[Donald_MacKenzie]_Material_Markets_How_Economic_(Bookos.org)

《[Donald_MacKenzie]_Material_Markets_How_Economic_(Bookos.org)》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

TheClarendonLecturesinManagementStudiesarejointlyorganizedbyOxfordUniversityPressandtheSaïdBusinessSchool.Everyyearaleadinginternationalacademicisinvitedtogiveaseriesoflecturesonatopicrelatedtomanagementeducationandresearch,broadlydefined.ThelecturesformthebasisofabooksubsequentlypublishedbyOxfordUniversityPress.ClarendonLecturesinManagementStudies:TheModernFirmOrganizationalDesignforPerformanceandGrowthJohnRobertsManagingIntellectualCapitalOrganizational,Strategic,andPolicyDimensionsDavidTeeceThePoliticalDeterminantsofCorporateGovernancePoliticalContext,CorporateImpactMarkRoeTheInternetGalaxyReflectionsontheInternet,Business,andSocietyManuelCastellsBrokerageandClosureAnIntroductiontoSocialCapitalRonBurtReassemblingtheSocialAnIntroductiontoActor-Network-TheoryBrunoLatourGatekeepersTheRoleoftheProfessionsinCorporateGovernanceJohnC.CoffeeScience,Innovation,andEconomicGrowth(forthcoming)WalterW.PowellTheLogicofPosition,TheMeasureofLeadershipPositionandInformationintheMarket(forthcoming)JoelPodolnyGlobalCompaniesinthe20thCentury(forthcoming)LeslieHannah MaterialMarketsHowEconomicAgentsAreConstructedDonaldMacKenzie1 3GreatClarendonStreet,Oxford26OxfordUniversityPressisadepartmentoftheUniversityofOxford.ItfurtherstheUniversity’sobjectiveofexcellenceinresearch,scholarship,andeducationbypublishingworldwideinOxfordNewYorkAucklandCapeTownDaresSalaamHongKongKarachiKualaLumpurMadridMelbourneMexicoCityNairobiNewDelhiShanghaiTaipeiTorontoWithofficesinArgentinaAustriaBrazilChileCzechRepublicFranceGreeceGuatemalaHungaryItalyJapanPolandPortugalSingaporeSouthKoreaSwitzerlandThailandTurkeyUkraineVietnamOxfordisaregisteredtrademarkofOxfordUniversityPressintheUKandincertainothercountriesPublishedintheUnitedStatesbyOxfordUniversityPressInc.,NewYork©OxfordUniversityPress2009ThemoralrightsoftheauthorhavebeenassertedDatabaserightOxfordUniversityPress(maker)Firstpublished2009Allrightsreserved.Nopartofthispublicationmaybereproduced,storedinaretrievalsystem,ortransmitted,inanyformorbyanymeans,withoutthepriorpermissioninwritingofOxfordUniversityPress,orasexpresslypermittedbylaw,orundertermsagreedwiththeappropriatereprographicsrightsorganization.EnquiriesconcerningreproductionoutsidethescopeoftheaboveshouldbesenttotheRightsDepartment,OxfordUniversityPress,attheaddressaboveYoumustnotcirculatethisbookinanyotherbindingorcoverandyoumustimposethesameconditiononanyacquirerBritishLibraryCataloguinginPublicationDataDataavailableLibraryofCongressCataloginginPublicationDataDataavailableTypesetbySPIPublisherServices,Pondicherry,IndiaPrintedinGreatBritainonacid-freepaperbyBiddlesLtd.,King’sLynn,NorfolkISBN978–0–19–927815–210987654321 ToBarbara,Moyra,andRobin CONTENTSAcknowledgementsviiiListofFiguresx1.Introduction12.TenPreceptsfortheSocialStudiesofFinance83.AssemblinganEconomicActor37withIainHardie4.Derivatives:TheProductionofVirtuality635.TheMaterialSociologyofArbitrage85withDanielBeunzaandIainHardie6.MeasuringProfit1097.ConstructingEmissionsMarkets1378.Conclusion:OpeningtheBlackBoxesofFinance177Glossary187Notes190References200Index221 ACKNOWLEDGEMENTSIamextremelygratefultotheSaïdBusinessSchool,UniversityofOxford,andtoOxfordUniversityPressfortheirinvitationtodelivertheClarendonLecturesinManagementStudies,onwhichthisbookisbased.IwouldalsoliketothanktheUKEconomicandSocialResearchCouncilfortheirawardofaProfessorialFellowshipon‘SocialStudiesofFinance’(RES-051-27-0062),whichmadepossibleboththewritingofthisbookandthebulkoftheresearchthatunderpinsit.MygratitudetoototheInstituteofAdvancedStudy,DurhamUniversity,foraterm’sfellowshipthathelpedmebringthebooktocomple-tion,andtoInnogen,theESRCCentreforSocialandEconomicResearchonInnovationinGenomics,forsupportingtheunderlyingworkontheconstruc-tionofeconomicagents.Ihavealsoincurredmanymorepersonaldebtsofgratitude:toIainHardieandDanielBeunza(co-authorsofChapters3and5),forallowingmetoreuseourjointworkhere;toMoyraForrest,whoprovidedmewithcopiesofthevastmajorityofthepublicationsthatformthisbook’sreferences;toBarbaraSilander,whoword-processeditstextandbibliography;andtoRobinWilliams,whohasbeenasteadfastsupporterofmyworkformanyyears.Robinhasalsobeenawonderfulfriend,andBarbaraandMoyradeservespecial,long-termthanks:theyhavenowworkedwithmeonfivebooks.Theinputofthemanyintervieweeswhosetestimonyformsthebook’smainempiricalunderpinningwasessential.Thenatureofthebook’ssub-jectmatter—and,inmanycases,interviewees’preferences—meanthattheircontributionsaregenerallyanonymous.Theywerevitallyimportantnever-theless.IamparticularlygratefultothosewhoallowedmetowatchLIBORbeingcalculatedandtothehedgefundthatpermittedtheobservationsbyIainHardieandmyselfthatformthebasisofChapter3.Threechaptersbuilddirectlyuponarticlespublishedelsewhere:Chapter3isarevisedversionofHardieandMacKenzie(2007);Chapter4ofMacKenzie(2007a);andChapter5ofBeunza,Hardie,andMacKenzie(2006).OtherworkIhavedrawnonincludesHatherly,Leung,andMacKenzie(forthcoming)and Acknowledgements/ixMacKenzie(2007b;2008).TheGlossarydrawsuponMacKenzie(2006).Iamgratefultothecopyrightholdersforpermissiontomakeuseofthismaterialhere.SectionsofChapter5writtenbyBeunzasimilarlydrawonmaterialinBeunzaandStark(2004).Asalways,though,Isavemymostheartfeltthanksofallformyfamily:Caroline,Alice,andIain. LISTOFFIGURES2.1.Extensionalsemanticsandfinitism272.2.Whitetoplayandmateinasinglemove283.1.Layoutofthehedge-fundtradingroom424.1.Totalamountsofexchange-tradedderivativesoutstandingatendofJuneofeachyearfrom1998to2006646.1.ExamplesofcodesfromtheUniversityofEdinburgh’schartofaccounts1136.2.FrequencydistributionofreportsofnetannualincomebyUScorporations,1976–19941267.1.Thepriceofallowancesforphase1oftheEuropeanUnionEmissionsTradingScheme167 1IntroductionJustafter11.00a.m.,everyweekdaythatisnotabankholiday,anapparentlymundanecalculationisperformedatacoupleofdesksinanunremarkableopen-planofficeinLondon’sDocklands.Smallsetsofnumbersarriveelectron-icallyorbytelephone.Thetwopeopleorchestratingthecalculationcorrectobvioustypingmistakesandchecklessclear-cutdiscrepanciesbytelephonecalls:‘Hello,it’s[X].Justwanttochecktheone-weekontheDanish[Krone].Youguysarequoting2.51.Youwanttokeepitaroundthat?’or‘Everyoneelseiscominginagoodbitunderthat.’Sometimestheytelephoneandremindthosewhoshouldhaveprovidedfigures.Oncehalfthenecessarynumbershavebeeninputintothecomputersystemthatperformsthecalculation,itbeginstoprocessthem.OnthedayIwaswatching,alltheinputshadbeenreceivedandchecked,andtheprocessingcompleted,by11.43,andoneofthetwostaffmembersinvolvedthensaidtohiscolleague:‘Youcanpublishaway.’Thisundramaticsequenceofeventsproduceswhatis,fromtheview-pointoftheamountofmoneyhingeingdirectlyonitsoutcome,oneoftheworld’smostconsequentialsetofnumbers:BritishBankers’AssociationLIBOR(LondonInterbankOfferedRate),thedominantglobalbenchmarkforinterestrates.OnthedayonwhichIwatchedLIBORbeingset,over$170tril-lionoftheworld’sfinancial‘derivatives’(theequivalentofaround$26,000foreveryhumanbeingonearth)wereindexedtoLIBOR,fluctuatinginvalueasitchangedfromdaytoday.1TheimportanceofthecalculationisreflectedinthearrangementsifaterroristincidentorothereventdisruptstheofficeinwhichIwitnessedit.Anearby,similarlyequippedofficebuildingiskeptinconstant 2/Introductionreadiness;dedicatedlineshavebeenlaidintothehomesofthoseresponsibleforthecalculation;apermanentlystaffedback-upsite,over250kmaway,canalsocalculateLIBOR.Ordinarily,LIBORisofinteresttomarketparticipantsalone,butasthe‘creditcrisis’beganin2007itsvaluesstartedtobediscussedinTVnewsbulletins,becausethemarketturmoilsuddenlyhighlightedthecrucialimportanceofwhetherbankswerepreparedtolendtoeachotherandatwhatratesofinterest.IbeginwithLIBORbecauseattentiontoitisaninstanceofwhatisdistinctiveabouttheapproachtomarketsexploredinthisbook.Thatapproachiscur-rentlybestexemplifiedin‘socialstudiesoffinance’.ThetermcameintousefirstamongstyoungscholarsinParisinthelate1990s.Initsbroadmeaning,itsignalstheapplicationtofinancialmarketsofsocialsciencedisciplinesbeyondeconomics(andalsowiderthanthoseapproachesto‘behaviouralfinance’thatarerootedinindividualpsychology),suchasanthropology,genderstudies,humangeography,politicalscience,andsociology.Smallnumbersofscholarsinthesefieldshavebeeninterestedinfinancialmarketsformanyyears,2butrecentlysuchinteresthasgrownandcoalescedasthoseinvolvedhavebecomeawareofwhattheirpeersinotherdisciplinesaredoing.3Amorespecificmeaningof‘socialstudiesoffinance’is,however,morepertinenthere.Inthismeaning,thetermreferstoapproachestomarketsthatareinspiredbysocialscienceresearchonscienceandtechnology.4Perhapsthemostprominentnameforthelatterresearchis‘socialstudiesofscience’,hencetheanalogousexpression‘socialstudiesoffinance’.Thosewhohaveworkedinthesocialstudiesofscienceandtechnologytendtoacquiresensitivities,interests,andintellectualresourcesthatdifferatleasttoadegreefromthoseofthewiderdisciplinestowhichwebelong.(Ofcourse,wealsoalwaysborrowfromthesocialsciencesmoregenerally,andIhavedonesohere.Sectarianismisneveravirtue.)Whatisperhapsmostcharacteristicofaperspectiverootedinthesocialstudiesofscienceandtechnologyisitsconcernwiththematerialityofmarkets:theirphysicality,corporeality,technicality.5Evenafinancialmarket,tradingasitdoestokensofrightsandobligationsratherthangoodsorservicesthatcanbeconsumeddirectly,ismadeupofphysicalartefactsandtechnologies.Forexample,asemphasizedinChapter5,apricemusttakephysicalform—spokenorwrittennumbers,electronicsignals,andsoon—ifitistobecon-veyedfromonehumanbeingorcomputersystemtoanother,andthephysicalformittakesisconsequential. Introduction/3Anemphasisonmaterialitypoints,however,tomorethantheimportanceofobjectsandtechnologies.Thehumanactorswhomakeupmarketsarenotdisembodiedagentsorabstractinformationprocessors,howeverconvenientitmaybeforeconomicstomodelthemassuch.Theyareembodiedhumanbeings,andbodiesarematerialentities.Thecapacitiesandlimitationsofthesematerialentities(includingthoseofhumanbrains)arehugelyimportanttohowmarketsareconstructed.Thefact,forexample,thatabrainisnotaninformationprocessorwithinfinitecapacitymeansthatconceptualtoolsthatsimplifythecognitivetaskofgraspingwhatisgoingoninamarketcanbeenormouslyimportant.Theinterbankmarket,forexample,isdiverseandhugelycomplicated,asonequicklylearnsbysittingbesidethetradersandbro-kerswhospendallday,everyday,seekingtounderstanditandactwithinit.YetLIBORcondensesallthatcomplicationintoasinglesetofnumbers,makingitpossible,forinstance,toindex$170trillionofderivativestoconditionsintheinterbankmarket.Suchatoolismorethanarepresentationofamarket:itisaconstitutivepartofeconomicaction,shapingsuchactionanditsconsequencesformarketprocesses.Aboveall,perhaps,whatonemightcallthe‘materialsociology’inspiredbythesocialstudiesofscienceandtechnologyemphasizesthetechnical-ityofmarkets.Thepropertiesofartefacts,technologicalsystems,concep-tualtools,andsoonarenot‘details’thatsociologicalanalysesshouldsetaside:fullyroundedanalysesneedtoincorporatethem.Take,forexample,WayneBaker’ssuperbstudyoftheChicagoBoardOptionsExchange(Baker1981;1984a;1984b),stillexemplarynearlythreedecadesafteritscompletion.6(An‘option’isasecuritythatgivesitsholderarightbutnotanobligation,forexampletobuyasetnumberofsharesorotherassetsatasetprice.)Bakerdemonstratesbeautifully,usinginterviews,networkanalysis,andpricedata,theconsequentialityofsocialrelationsamongstoptiontraders.Yetitisalsoimportantthattradingontheexchangewasinfluencedbyeconomicmodelsofoptionpricing,notablytheBlack–ScholesmodeldevelopedbyeconomistsFischerBlack,MyronScholes,andRobertC.Merton.Blackandotherssoldpapersheetsoftheoreticaloptionprices,whichmanytradersusedtoinformtheirtradingdecisions,andIhaveargued(inMacKenzie2006)thatamongsttheeffectswastoshiftpatternsofoptionpricestowardsthemodel.Thecaseofoptionstradingexemplifiesacentralaspectofwhatisdistinctiveaboutlookingatmarketsfromtheviewpointofthesocialstudiesofscience 4/Introductionandtechnology.EconomicagentssuchasChicagooptiontradersarenotjust‘naked’humanbeings,norsimplyhumanbeingsembeddedinsocialnetworks.Their‘equipment’matters.AtraderequippedwithBlack’ssheetswasadifferenteconomicagentfromonetradingonthebasisofintuitionandexperiencealone.TheBlack–Scholesmodelwasindeedtechnical,butitwasnota‘meretechnicality’:itwasaconsequentialpartofhoweconomicagentswereconstructed.(The‘agent’ofthisbook’stitleisanyeconomicactor,notmerelyonewhoactsonbehalfofanother.)Thechaptersthatfollowaimtoelaborateandillustratethe‘socialstudiesoffinance’,inthissenseofperspectivesinspiredbysocialscienceresearchonscienceandtechnology.Chapter2doesthisbyformulatingasetofpreceptsofworkofthiskind.Itisasetofprecepts.7Manyofthemare,Ithink,widelysharedbythosewhocometothestudyofmarketsfromabackgroundinthesocialstudiesofscienceandtechnology,butanysuchsetisboundtobeidiosyncratic,andIhavenowishtofoistthesepreceptsonmycolleagues(orevenontheco-authorsofChapters3and5).Furthermore,mychosensetofpreceptsisclearlyincomplete(especiallywhenoneshiftsfocustomarketsinspheresotherthanfinance),noraretheyallillustratedequallyinthechaptersthatfollow.Nonetheless,IhopethepreceptshelptofleshouttheapproachtotheunderstandingofmarketsIamadvocating.Chapters3to7areprimarilycasestudies,eachbroadlywithintheframe-workoutlinedinChapter2.Chapter3,writtenjointlywithIainHardie,isorga-nizedaroundoneofthecentralideasdiscussedinChapter2:MichelCallon’sideaoftreatingeconomicactorsasmadeupofagencements,ofcombinationsofhumanbeings,materialobjects,technicalsystems,texts,algorithms,andsoon.HardieandIapplythenotionofagencementtoacategoryofagentthatisofgrowingimportanceinfinancialmarkets:ahedgefund.(TheGlossaryexplainsfinancialmarkettermssuchas‘hedgefund’,andthemoreimportantsuchtermsarealsodiscussedinthechapters.)Ouraccountisbaseduponabriefperiodofobservationofoneparticularfund’strading,oninterviewsconductedwiththepartnersinthefund,andonawidersetofinterviewswithtradersinotherhedgefundsandininvestmentbanks,andalsowiththosewhosupplyvitalservicestohedgefunds,suchasthe‘fundsoffunds’thatchannelcapitaltothem.Likealmostallsophisticatedtraders,thefundHardieandIstudiedtradednotjustbasicassetssuchasbondsbutalso‘derivatives’ofthoseassets:contracts Introduction/5orsecuritiesthevalueofwhichdependsonthepriceoftheunderlyingasset,oronotherparameterssuchastheprobabilityofanissuerofbondsdefaulting.Chapter4examinesthedevelopmentofexchangesdevotedtothetradingoffinancialderivatives.Asrecentlyas1970,therewasnofinancial-derivativesexchangeanywhereintheworld.BytheendofJune2006,contractsout-standingonsuchexchangestotalled$84.4trillion,around$13,000foreverypersonontheplanet.Thelimitedliteraturebeyondeconomicsonderivativesoftenemphasizestheir‘virtual’natureas‘money’s“newimaginary”’(PrykeandAllen2000).DrawinguponoralhistoryinterviewswithkeyfiguresinthedevelopmentoforganizedderivativestradingintheUSAandUK,Chapter4examineshowthose‘virtual’productshavebeenbroughtintomaterialexis-tence.Itexploresthesimilaritiesanddifferencesbetweentechnologicalinno-vationandinnovationinderivatives,anddiscussestheroleofthe‘internal’culturesoffinancialmarketsandofthewiderculture(inparticular,thelegaltracesofhostilitytogambling).Thechapteralsoexamineshowthefact-generationmechanismsoffinancialmarketsarecrucialtoderivativestrading.Chapter5,co-authoredwithDanielBeunzaandIainHardie,examinesaspecificformoftradingthatiscentraltohowderivativesexchangesproducefacts,aswellashavingmanyotherrolesinfinancialmarkets:arbitrage,inotherwordstradingthatexploitsdiscrepanciesinrelativeprices,forexamplebetweenanassetandaderivativeofthatasset.DrawingonworkconductedbyHardieandmyself,andonanethnographyofaWallStreetarbitragetradingroombyBeunzaandDavidStark(2003;2004;2005),thechapteremphasizesboththematerialityofprices(asnotedabove,pricesarephysicalentities,andtheextentandspeedoftheirmobilityarecrucialtoarbitrage)andthedelicatesocialrelationsthatexistamongstarbitrageursandbetweenthemandothers.Chapter6shiftsfocusfromfinancialmarketsthemselvestotheprocessesthatprovidecrucialdataforthosemarkets:thebook-keepingandaccount-ingthatproducecorporatefinancialstatements,especiallycorporateearnings(‘profits’).Thechapterdrawsuponabodyofliteratureinaccounting,littleknownoutsidethefield,on‘earningsmanagement’,andonabriefcasestudyofaparticularaccountingclassificationbyafirmheavilyinvolvedinearningsmanagementtoemphasizetheapplicabilitytoaccountingofanotheroftheideasdiscussedinChapter2:finitism.Inthefinitistperspectiveexploredin 6/Introductionthatchapter,‘[w]ecouldtakeourconceptsorrulesanywhere’(Bloor1997:19)whenweperformclassifications,measureitems,andfollowrules,allofwhichareactivitiescentraltoaccounting.Giventhat,whatkeepsaccounting(relatively)orderly?Chapter6suggeststhattheanswertothisquestionisintrinsicallysociotechnical:itinvolvesbothpeople—notjustasindividuals,butasmembersoftheculturesofbook-keepingandofaccounting—andtechnicalsystems.Chapter7movesoutsidethesphereoffinanceinanarrowsensetoexaminemarketsinpollutionpermits,particularlygreenhouse-gasmarkets.Itempha-sizesseveraloftheissuesdiscussedinChapter2,suchastheactiveroleofeconomicsandofeconomistsinbringingmarketsintobeing(ratherthanstudyingthemasalready-existing,external‘things’)andespeciallythepoliti-calnatureofthedetaileddesignofmarkets.Indeed,thechaptersuggeststhatemissionsmarketsarepolitics.Inparticular,theyarewaysofundermininginertiaandachievingapparentconsensus;buttheyalsoseemtodisplacecon-flict,whichshiftsfocustotherulesgoverningtrading(especiallygoverningthecrucialmatteroftheallocationofpermits).The‘economicexperiment’(MuniesaandCallon2007)ofEuropeancarbontradinghassofarhadamixedoutcome:forexample,limitedenvironmentalimpact,butasurprisingdegreeofsuccessin‘technicizing’thepoliticsofallocation.Chapter8isthebook’sconclusion.Itemphasizesthemodest,exploratorynatureofthepreviouschapters,andreturnstotheoverallquestioninformingthebook:whatcananapproachrootedinthesocialstudiesofscienceandtechnologycontributetoanunderstandingofmarkets?Thechapterexaminestherelationsofsocialstudiesoffinancetoeconomicsociologymoregenerally,andmakesexplicitthepossibilityhintedatinChapter7:thatanapproachtomarketsthatdrawsuponthesocialstudiesofscienceandtechnologycanbecome‘publicsocialscience’,analogoustowhatBurawoy(2005)calls‘publicsociology’.ANotere.InterviewSourcesThemaindatasourcedrawnoninthechaptersthatfollowisasetof189interviewswithpeopleinvolvedinfinancialandemissionsmarketsastraders, Introduction/7managers,brokers,marketdesigners,providersofadministrativeandotherservices,accountants,andauditors.(Twenty-threeoftheseinterviewswereconductedbymycolleagueIainHardie,andsixty-sixhavealreadybeendrawnuponinMacKenzie2006.)Apartfromincasesinwhichtheidentityoftheintervieweeisimportanttoanhistoricalepisodebeingdiscussed,Ihavefol-lowedthepreferenceofmostofmyintervieweesandcitedinterviewdataanonymously. 2TenPreceptsfortheSocialStudiesofFinanceWhatmightbeinvolvedinanapproachtothestudyofmarkets,especiallyoffinancialmarkets,thatisinspiredbythesocialstudiesofscienceandtech-nology?ThischapterbuildsonChapter1’sbriefdiscussionofthatquestionbysketchingtenpreceptsthatmightunderpinresearchofthiskind,researchthatfocusesonthephysicality,thecorporeality,andthetechnicalityofmarkets.Precept1:FactsMatterLetmebeginwithacrucialfacetofthetechnicalityofmarkets:theroleinthemoftheproductionandcirculationoffacts.A‘fact’is‘Athingknown...tohaveoccurredortobetrue’,1andfactsareobviouslyacrucialtopicofthesocialstudiesofscience.Thatfielddoesnotaskwhetherclaimedscientificfactsare‘reallytrue’,whichisaquestiononlyscience,notsociology,cananswer.Instead,thesociologyofscientificknowledgeisinterestedinquestionssuchashowfactsareproducedandwhatsecurestheirfacticity,inotherwordstheirstatusasfacts.‘Produced’istherightword:scientificfactsarenotingeneralsimply‘outthere’,awaitingthefortunatediscovererwhostumblesoverthem.AstheLatinrootof‘fact’—facere,tomake,todo—remindsus,sci-entificfactsaremade:byexperiment,byintellectualwork,andbyobservation TheSocialStudiesofFinance/9thatisnormallytechnologicallymediatedandtypicallyisdisciplinedandgoal-orientedratherthanhaphazard.The‘career’ofafacttendstomakethehistoryofitsproductionvanish.Full-fledgedscientificfactsarepublic,collective,‘social’facts,notsimplyprivatefacts:theirstatusasfactsisassessedbyscientificcommunities,notjustbyindividuals.2Asoneoftheclassicethnographiesoftheproductionofscien-tificfactsemphasizes,tobecomeafactastatementmustlosethe‘traceofauthorship’inthesenseofbeing‘freedfromthecircumstancesofitsproduc-tion’(LatourandWoolgar1986:82and105).Ifitcannotbefreedfromthosecircumstances—if,forexample,aphenomenoncanbeexhibitedonlybyaspecificpersonoraparticularlaboratory,andnotbyothers—itisnormallyfatedtobeclassedinsteadasan‘artefact’:notatruthabouttheworld,butthespuriousresultoftheparticularproceduresusedinitsproduction.Thefactsthatcirculateinfinancialmarketsmightseemweakandvulner-ablecomparedtothoseproducedbyamaturescience.ButconsiderBritishBankers’AssociationLIBOR,LondonInterbankOfferedRate.Thelow-keydeclaration,‘Youcanpublishaway’,whichIheardwhenstandingintheLon-donofficeinwhichLIBORiscalculated,initiateditselectronicdisseminationviamultipledataproviders.Almostinstantaneously,itwaspossibleatmorethanhalfamillionterminalsacrosstheglobetoclickamouseafewtimesand/orstrikeahandfulofcomputerkeys,andpromptaflowofelectricalsignalsthatwouldcausethenumberscalculatedintheLondonofficetoappearonscreen.3AsfarasIcantell,thosewhodothishaveuntilveryrecentlytreatedthenumbersthatappearasunproblematic.Theyhavebeen,forexample,contenttohavehugesumsofmoneychangehandsonthebasisofwhatappearsonscreen.Likescientificfacts,LIBORisproducedratherthanstumbledupon—itsproductionisdescribedinChapter4—butlikethemithadbeenfreedalmostentirelyfromtheparticularcircumstancesofitsproduction.(Contro-versyoverLIBORdid,however,eruptin2007–8asaresultofthecreditcrisis,asdescribedinChapter4.)Surprisingly—giventheamountofmoneythatdependsonthem—norealcontroversynormallysurroundedthenumbersthatflowoutofthatLondonoffice.Ofcourse,notalleffortstogive‘marketnumbers’factualstatusaresuccess-ful.Consider,forexample,anotherdaily‘pricefixing’:the10.00a.m.CheddarcheeseauctionconductedontheflooroftheChicagoMercantileExchange.Acheeseauctionmayseemanesotericmatterofnowiderinterest,butthe 10/TheSocialStudiesofFinanceresultsoftheauctionareinputsintotheformulathathasbeenusedsincethe1930stosetthegovernment-mandatedminimummilkpricereceivedbyfarm-ersintheUnitedStates,andtheythushelpdeterminetheconsumerpricesofdairyproductsintheUSA.Inthesummerof2006,fiercecontroversyeruptedintheUSAovertheauction.Onefarmers’spokespersonclaimed:‘Thereareveryfewbuyerswhoaresettingthepriceinaverythinmarket.’Anothersaidtherewas‘deepscepticism’aboutthemechanism’s‘legitimacy...asatruereflectorofsupplyanddemand’,andinJuly2006sixsenators,includingHillaryRodhamClinton,demandedagovernmentinvestigation(Grant2006).Theaccusation,contestedvigorouslybytradersontheChicagoMercantileExchangeandbyothers,wasineffectthattheresultsoftheauctionwereartefacts,notfacts.4Whethermarketnumbersachievethestatusoffactsisconsequential.AsdiscussedinChapter4,itis,forexample,amajorinfluenceonwhetherderiv-ativesmarketscansuccessfullybeconstructed.Moregenerally,facticityisoftenapreconditionforliquidmarkets.AsCarruthersandStinchcombeputit,‘liquidityis...anissueinthesociologyofknowledge’(1999:393).Amarketis‘liquid’iftheitemstradedinitcanreadilybeboughtandsoldinsubstantialquantityatorclosetoprevailingmarketprices,withoutunduedelayorheavytransactioncosts,andachievingliquiditytypicallyrequiresstandardizationoftheitemstradedand,crucially,adegreeofconsensusontheircharacteristics.Themarketforinterest-ratederivatives,forexample,wouldnotbeliquidifeachcontractrequiredtheresolutionofadisagreementoverhowinterestratesaretobemeasured,ifeachsubsequentpaymentwasvulnerabletolitigationovertheadequacyofthemeasurementthatdetermineditssize,orifonefearedthatone’scounterpartiesmightbeabletomanipulatethemeasurementintheirfavour.ThecontributionofthefacticityofBritishBankers’AssociationLIBORtotheliquidityoftheinterest-ratederivativesmarketshasbeenthatallthesemessyobstaclesareavoided.Precept2:ActorsAreEmbodiedAllmarkets,whetherliquidornot,arecombinationsofhumanbeingsandphysicalobjects.Itmayseemtooobvioustoneedsaying(untilonerealizesthatfewanalysesofmarketsdevelopthepoint),buthumanbeingshavebodies,are TheSocialStudiesofFinance/11bodies.Corporeality—inthesenseofthematerialcapacitiesandlimitationsofthosebodiesandbrains—iscriticaltohowmarketsfunction.Insomemarkets,embodimentcanhardlypassunnoticed.Takean‘open-outcry’tradingpit,asteppedamphitheatreinwhichdealsaredonebyvoiceorbyeyecontactandhandsignals(see,e.g.,Zaloom2006).Crucialskillsforpittradersaretobeabletodetectthebodilysignsoffear,whichoftenindicateatraderwhoisdesperatetoexitaposition,andtosuppressthesignsofone’sownfear.Whereonestandsinsuchapitmatterseconomically:thetoprungisbest,becauselinesofsightarebetterthere,andbecauseitisadvantageoustobeclosetothebigbrokers,withtheirlargevolumesofcustomerorders,whostandthere.Forthatveryreason,bodilypositioniscontested.Onefrequentlygetsjostled,orhastojostle,andfistfightsaren’tuncommon.Physicalheightmatters,bothtoseeingandespeciallytobeingseen.Thepitisaplaceofmalebodies,withwomenformingnomorethanasmallminority.Itisofcoursetheexoticnatureofpittradingthatmakesitsembodimentstandout.Aspitshavedeclined,theyhavebeenreplacedbyscreen-basedtrad-ing,andtradingbytelephonealreadyhasalonghistory.It’sparticularlyeasyimplicitlytopositadisembodiedactorwhenstudyingsuchtrading,becausethebodilyactionsinvolvedaremostlyfamiliartoanyofficeworker.Muchofthetime,forinstance,onewouldbehardpushedtodistinguishthephysicalactionsofatraderinabank’sdealingroomfromthoseofanacademicathisorherdesk.There’salotofsittingandstaringatascreen,typingandmovingacomputermouseabout,talkingonthetelephone,reading,chatting,anddrinkingcoffee.Disappointinglyfromtheviewpointofexoticism,shouting,swearing,andraucousbehaviourindealingroomsarenowfarrarerinactual-itythaninfilmportrayals.Bodilycapacitiesstillmatter,however.LetmeagainusetheexampleofLIBOR,whichispertinentherebecauseitisanapparentlydisembodiedsetofnumbers.AsdescribedinChapter4,theinputstotheLIBORcalculationcomefrombankdealingrooms,buttheyareheavilyinfluencedbyinterdealerbrokers.Althoughbrokersincreasinglygivetheirclientsthecapacitytotradeelectronically,thecoreoftheirbusinesswas(andtoasignificantextentstillis)‘voicebroking’.Afirm’sbrokersinagivenmarket(forexample,thesterlinginterbankmarket)sitclosetogetherataclusterofdesks,withnearbyclustershandlingrelatedmarketssuchasininterest-rateswaps.Eachbrokerhasonhisdesk(itisanotherpredominantlymaleniche)a‘voicebox’—acombinationof 12/TheSocialStudiesofFinancemicrophone,speaker,andswitches—connectedbydedicatedtelephonelinestoeachofhisclients.Interdealerbrokersdonotthemselvestrade:ifaclientbankwantstoborrowmoney,thebroker’sjobistofindabankthatwilllend(orviceversa).Sothekeyskillisknowingwhowantstodowhatandwhoispreparedtodowhat.Itispartlyamatteroffosteringrelationshipswithclients.Iwasrepeatedlytoldbyinterdealerbrokersthattheirsis‘arelationshipbusiness’.Asoneofthemsaidtome,abrokerinthe‘moneymarket’(themarketaboutwhichLIBORisa‘fact’)might‘speaktohisbigclients...haveconversationswiththemmaybetwenty-fivetimesaday,whichistwenty-fivetimesasoftenastheyspeaktotheirwives’.There’salso,however,acrucialbodilyskillininterdealerbroking,askillthoseinvolvedcall‘broker’sear’:thecapacityaurallytomonitorwhatisbeingsaidbyalltheotherbrokersataclusterofdesks,whileoneselfholdingavoiceboxconversationwithaclient.Asanintervieweeputit:Whenyou’reonthedeskyou’reexpectedtoheareveryoneelse’sconversationsaswell,becausethey’reallrelevanttoyou,andifyou’reonthephonespeakingtosomeoneaboutwhat’sgoingoninthemarkettherecouldbeahotpieceofinformationcominginwithoneofyourcolleaguesthatyouwouldwanttotellyourclients,soyou’vegottobeabletohearitcominginasyou’respeakingtotheperson.Wheninterviewingbrokersattheirdesks,Isometimesfound‘broker’sear’disconcerting.Someonecouldapparentlybepayingfullattentiontohiscon-versationwithme,whenhewouldsuddenlyrespondtoacommentorques-tionfromfiveorsixdesksawaythatIsimplyhadn’theard.(Themultipleconversationsmakebrokers’officesnoisy,andthey’remorecloselypackedandmoreraucousplacesthanbanks’dealingrooms.)Broker’searisanacquiredskill:asonetoldme,‘youdon’tjustdoitfromdayone....Somepeoplenevermakeit.’Broker’searhelpsbrokersquicklytomatchborrowersandlenders(orinothermarkets,buyersandsellers),thuscontributingtotheliquidityoftheinterbankmarket,andtheinformationthatbroker’searaggregatesalsohelpsbrokers’clients—manyofwhommaketheinputstoLIBOR—tounderstandwhataresometimesrapidlychangingconditionsintheinterbankmarket.AsnotedinChapter4,there’snofullyalgorithmicwayofgeneratinganappropri-ateLIBORinput.Judgement,basedonanunderstandingofmarketconditions,isinvolved,andbroker’searisonebodilyfoundationofthatjudgement. TheSocialStudiesofFinance/13Precept3:EquipmentMattersAsemphasizedinChapter1,humanbodiesandbrainshavetheirlimitationsaswellastheirremarkablecapacitiessuchasbroker’sear,andthismakestheirsupplementationby‘equipment’(physicaland‘cognitive’)crucial.Acentralconjectureofthesocialstudiesoffinanceisthatequipmentmatters:itchangesthenatureoftheeconomicagent,ofeconomicaction,andofmarkets.Consider,forexample,physicalequipmentsuchasthestockticker(Preda2006)ortradingscreensconnectedinelectronicnetworks(seeKnorrCetinaandBruegger2000;2002a;2002b;KnorrCetina2005).Tickersandtradingscreenspartiallycircumventthemostbasicofallbodilylimitations—theinabilitytobeintwoplacesatonce—andtheirintroductiondidnotsimplymakeexistingformsofbehaviourinmarketsmoreefficient:itreshapedmar-kets.Tickersweretelegraphicprice-disseminationsystems,atthereceivingendofwhichtheabbreviatednamesofsecurities,theirprices,andthevolumesoftradeswereprintedoutontopapertape.Theymadefine-grainedknowledgeofpricemovementsavailableinclosetorealtimetogeographicallydispersedmarketparticipants(atleastifthoseparticipantswereabletocongregateintheofficesofbrokerswithtickers),whenpreviouslysuchknowledgerequiredonetobepresentphysicallyintheexchangeswheresecuritiesweretraded.Preda(2004b;2006)conjectures,forinstance,thatthetickerhelpedprompttheriseof‘chartism’or‘technicalanalysis’:thebelief—stillwidespread—thatpatternscanbefoundinpricegraphsthathavepredictivevalue.Theintro-ductionoftradingscreenswasatleastequallyconsequential:‘“themarket”nolongerresidedinanetworkofmanyplaces,butonlyinone,thescreen,whichcouldberepresentedidenticallyinallplaces’(KnorrCetina2005:51).Forexample,asbrieflydiscussedinChapter5,thisalteredhowthepervasivemarketactivityof‘arbitrage’(theexploitationofpricediscrepancies)hadtobeconducted.Actors’equipmentgoesbeyondphysicaltechnologies:their‘conceptualequipment’alsomatters,orsothesocialstudiesoffinanceposits.Financialmarketsare,asnotedinChapter1,complicatedplaces,especiallyinthecon-temporaryworldofbewilderingarraysofproductsandofpricedatathatchangesecondbysecond.Giventhelimitedmemoryandcomputationalcapacityofthehumanbrain,economicagentsmustdevelopandacquiresystematicwaysofmakingsenseofmarketsthatreducethiscomplication 14/TheSocialStudiesofFinancetoalevelthatismentallytractable.Organizationsmustdevelopproce-duresforinteractingwithmarkets,andtoanincreasingextentthosepro-ceduresareimplementedinalgorithmsinautomatedpricing,trading,andrisk-managementsystems.Sometimes,thewaysofthinking,procedures,andalgorithmsthatareemployedderivefromfinancialeconomics,suchasthetheoryofoptionstouchedoninChapter1.Probablymoreoften,however,practitioners’waysofthinkingandassociatedwaysofactinghavenodirectconnectiontoacademiceconomicsorindeedareregardedbyeconomistsasmistaken.Chartism—usinggraphsofpricemovementstoattempttopredictfuturemovements—isanexampleofthelatter:financialeconomistsregarditasonaparwithastrology,butmanytraderstakeitseriously,andactonthebasisofit.Thatconceptualequipmentmattersisaconjecture,notapresumption.Althoughindividualeconomists(forexample,MertonandBodie2005)havenoteditspotentialimportance,financialeconomicsdoesnottreatthisequip-mentinanysystematicway.Itmodelsaworldinwhich,forexample,marketprocessesforceoptionpricestotheir‘correct’values,withoutexplicitconsid-erationoftheeffectsonthoseprocessesofwhetheractorshaveoptiontheoryavailabletothem.Theevidenceconcerningtheeffectsofitsavailabilityisnotunequivocal,butinmyviewsupportstheconclusionthatactors’deploymentofoptiontheorydidaffectprices.5Italsoseemslikelythatchartistbeliefsandprocedures,iffollowedsufficientlywidely,willaffectpricemovements(thisisonereasonthateventradersscepticalofchartismcitefortakingitseriously),thoughthereisagainlimiteddirectevidenceonthepoint.6Actors’conceptualequipmentcanalsohaveeffectsthataresubtlerthaninfluencingpricesdirectly.Oneofthewaysinwhichfinancialmarketsarecomplicatedisthatthemembersofaclassofproductsoftenvaryinconse-quentialwaysintheirdetailedcharacteristics.Thusoptionsevenonthesameunderlyingassetwilldiffer:somewillbe‘calls’(optionstobuytheasset),andsome‘puts’(optionstosellit),andtheirexpirationdateswillvary,aswilltheirexerciseprice(thepriceatwhichtheygivetherighttobuyortoselltheassetinquestion).Animportantroleofoptiontheoryisinpermittingthereductionofthatcomplicationtoacommonunderlyingmetric:‘impliedvolatility’,thevolatility(extentoffluctuations)ofthepriceoftheunderlyingassetconsistentwiththepricesoftheoption,accordingtoanoption-pricingmodel.Translatingoptionpricesintolevelsofimpliedvolatilityallowstheeasycomparisonofoptionswithdifferentcharacteristics(theimpliedvolatility TheSocialStudiesofFinance/15ofthisoptionis17percent,ofthatoption19percent...),andindeedpricequotationsforoptionsaresometimesexpressedasimpliedvolatilitylevelsratherthanamountsindollarsorothercurrencies.Althoughitwouldbehardtodemonstrateitquantitatively,itseemsplausiblethattheavailabilityofthismetriccontributestotheliquidityoftheoptionsmarket.‘Impliedvolatility’isnotuniqueinitsroleasasimplifyingmetricandfacili-tatorofcommunicationaboutthepropertiesoffinancialproducts.Chapter3,forexample,conjecturesthatthe‘yield’ofabondplaysthesamerole,althoughthenotionof‘yield’andtheoriginaltechniquesforcalculatingitareoldandcomefrom‘mixed’(applied)mathematicsratherthaneconomics(HawawiniandVora2007).Anotherexamplecomesfromthemarketfor‘creditderivatives’suchascollateralizeddebtobligations(CDOs),contractsthevalueofwhichdependsontheriskofdefaultbyeachofalargesetofcorporationsorotherissuersofdebt(perhapsahundredormoreissuers);onthelikelyextentofthepay-out,ifany,todebtholdersafterdefault;andontheextenttowhichtheriskofdefaultbyoneissueriscorrelatedwiththeriskofdefaultbytheothers.Inthecredit-derivativesmarket—whichintheyearspreceding2007’screditcrisiswasthe‘hot’areaofderivatives—themetricof‘basecorrelation’7seemstohaveafunctionsimilartothatof‘impliedvolatil-ity’asacommunicativetool.Forexample,itfacilitatesnegotiationamongstsophisticatedparticipants.ThebasecorrelationimpliedbyaCDOquotationcanbeusedasthebasisforareasonedargumentwhythequotationneedstobealtered.Theavailabilityofconceptualequipmentcanmatterevenifthetheoryunderpinningtheequipmentisnotunderstood—softwaresystemsallowtraderswithonlyaroughgraspofthetheoryofoptionsorofCDOstocal-culateimpliedvolatilitiesorbasecorrelations—ornotbelieved.Thosewhodounderstandthemodelsthatareusedinsuchcalculationsfrequentlyviewthemasoversimplifications.Ihave,forexample,yettointerviewacredit-derivativestraderwhoregardsasadequatethe‘single-factorGaussiancop-ula’modelnormallyusedincreditcorrelationcalculations.Nevertheless,thesimplemodelsremaininwideuse.Morecomplexmodelsfaceformidablebarriersascommunicativetools,becauseforfullcommunicationbothpartiesmustbeusingthesamemodel,andthatisseldomthecaseonceonemovesbeyondsimplemodels.Furthermore,thesimplemodelstypicallyhavejustonefreeparameter—‘impliedvolatility’,forexample—withtheotherpara-metersbeingeitherfixedbymarketconvention(CDOpricing,forexample, 16/TheSocialStudiesofFinancewasoftendonebyassumingarecoveryrateafterdefaultof40percent,whateverthecorporationthathasissuedthedebtinquestion)orregardedasempiricallyobservablefacts.Whennumbersoffreeparametersarelarger,orparametersdonothaveintuitiveinterpretations—asisoftenthecasewithmorecomplexmodels—communicationandnegotiationbecomemuchharder.Precept4:CognitionandCalculationAreDistributedandMaterial‘Publicfacts’suchasLIBOR,technicalequipmentsuchasstocktickers,graphi-calrepresentationssuchasthoseusedbychartists,and‘conceptualequipment’suchas‘impliedvolatility’or‘basecorrelation’areallaspectsofthediversecognitiveandcalculativeprocessesthattakeplaceinfinancialmarkets.Theseprocessesare‘distributed’inthesensethatagiventaskisoftenperformednotbyasingleunaidedhumanbeingbutbymultiplehumanbeings,objects,andtechnicalsystems—anaspectthathasbeenexaminedinotherspheresintheliteratureondistributedcognition(see,aboveall,Hutchins1995a;1995b).8Thesocialstudiesoffinanceinheritsthebasicconjectureoftheliteratureondistributedcognition.Notonlycancombinationsofmultiplehumanbeingsandobjectsdothingsanunaidedindividualcannot,buttheperformanceofthesametaskbyanunaidedindividualcanbeexpectedtohavedifferentpropertiesfromitsperformancebyacombinationofthissort.Hutchinsputstheunderlyingpointeloquently,deployingashisprimeexamplenavigationasconductedinUSwarships.Humanbeings,heargues,‘createtheircognitivepowersbycreatingtheenvironmentsinwhichtheyexercisethosepowers’.Tounderstandcognitionthatinvolvesmultiplecollaboratinghumanbeingsand/orinteractionwithobjectsandtechnicalsystems,onemustgobeyondthepsychologicalorcognitive-scienceanalysis‘oftheindividualboundedbytheskin’.‘[L]ocalfunctionalsystemscomposedofapersonininteractionwithatoolhavecognitivepropertiesthatareradicallydifferentfromthecognitivepropertiesofthepersonalone’,anda‘groupperforming[a]cognitivetaskmayhavecognitivepropertiesthatdifferfromthecognitivepropertiesofanyindividual’(Hutchins1995a:xvi,176,and289). TheSocialStudiesofFinance/17Forexample,asinglehumanbeingcanproduceaLIBOR,butnotaLIBORthatwouldhavethestatusofafact.Tohaveasufficient‘feel’forintra-dayandday-to-daychangesinconditionsintheinterbankmarket,heorshewouldneedtobeanactiveparticipantinthatmarket—one,forexample,towhombrokerswouldregarditworththeirwhilerelayingup-to-dateinformation:notjustprices,butthenatureoftheunderlyingtransactions(andsome-timesevenwhichparticularbankshaddonewhat).Butthatparticipationwouldmakehimorheraninterestedparty.ThatmultiplehumanbeingsareinvolvedintheproductionofLIBOR—thatitsproductionisinthatsense‘distributed’—insuchawaythatthereisagoodchanceoftheirinterests‘cancellingout’isanessentialaspectofitbeingafact,asdiscussedinChapter4.Somethingsimilaristrueoftheproductionofcorporateaccounts,asoutlinedinChapter6.Notonlyisthatproduction(foranyotherthanthesmallestofcompanies)beyondthepowersofanindividual,butsuchlimited‘hardness’astheresultantnumberspossessisintrinsicallyboundupwiththeinvolvementofmultiplehumanbeingsinrolesthatarestructuredinparttechnologically.Allcognitionandallcalculationarephysicalprocesses(thebrainisabio-logicalorgan),butthematerialityofcalculationisperhapsmostclearlyseenwhenitinvolvesextensivenumericalcomputation.Forexample,theoreticalmodelsofthevalueofCDOsseldomyieldequationsthatcanbesolved,otherthanveryslowly,withthetraditionalmaterialtoolsofpencilandpaper.9ThemodelsusedforpricingCDOsrunfairlyquicklyonmoderncomputers,butcalculationsofhedgingratiosandrisk-managementparametersaremoredemanding.Evenusinggridsofseveralhundredinterconnectedcomputers,theriskcalculationscantakeseveralhours.Aphysicalconstraintbecomesrelevant:heat.Manybankswantcrucialcomputationstobeperformedinorclosetomainoffices,tradingfloors,andriskmanagers,becauseevenfibre-opticconnectionsarestilltooslowforsomepurposestoallowallcompu-tationtotakeplaceindistantsites.InLondon,though,thosemainofficesareoftenintheCity,whereexpansionisfrequentlyimpossibleorhugelyexpensive.Onecan’tkeeppackingmoreandmorecomputersintoanygivencomputerroom,becausetheheattheygeneratewilleventuallyexceedthecapacitytoremoveitbyair-conditioning.Sothosewhosellcomputerhard-waretoinvestmentbanksareawarethat‘performanceperwatt’isnowoneoftheparametersonwhichthathardwareisjudged.Arisk-analysiscomputerrunisacalculation‘internal’toafinancialmarket.Butallsuchmarketsareconnectedtoprocessesoutsidethemselves,andthis 18/TheSocialStudiesofFinancetooinvolvesdistributedcognitionandmaterialcalculation.Taketheemis-sionsmarketsdiscussedinChapter7,forexample.AstheUSEnvironmentalProtectionAgency(EPA),whichcoordinatesthemarketinsulphur-dioxideemissions,pointsout:‘Anessentialfeatureofsmoothlyoperatingmarketsisamethodformeasuringthecommoditybeingtraded.’10Unaidedhumansensescannottellhowmuchsulphurdioxideisbeingemittedbyasmokestack,yetforamarkettoworkthequantityofthesmokestack’semissionsmustbeafact.Hencetheimportancetomarkets(notjustinemissions,butinahugerangeofothercommoditiestoo)of‘metrology’:ofthescienceandtechnologyofmeasurement,inparticularofstandardizedunitsandmeasurementproce-duresthatensureanohm,acentimetre,oragrammeasuredinoneplaceandatonetimeissufficientlylikethesamequantitymeasuredatadifferentplaceandtime.11Eachsmokestackofeverycoal-firedoroil-firedelectricpowerplantintheUSA(otherthantheverysmallestsuchplants)nowhastocontainequipmenttomeasuretheflowofgasesthroughit(LevinandEspeland2002).Thisequipmentmustalsosamplethosegasesevery15minutes,determinetheconcentrationofsulphurdioxideinthesample,12andproduceanelectronicrecordofthemeasurementresults,whichiscapturedbyadataacquisitionandhandlingsystemandaveragedtoproducehourlyrecords.ThedatafilescontainingtheserecordsaretransmittedelectronicallytotheEPAeverythreemonths,andpublishedviaitswebsite.13Beneaththefirstlayerofmetrologyinthesulphur-dioxidemarketisanother:themetrologythatprovidesassurancethatthemetersarecorrectlycalibrated,forexamplebycheckingtheoutputofamonitorwhenfedasampleofgasnotfromitssmokestackbutfroma‘calibrationcylinder’containingaknownconcentrationofsulphurdioxide.Beneaththatlayerisyetanother:thetestsandstatisticalproceduresthatmeasurethecompositionofthecontentsofcalibrationcylindersbycomparingthemtostandardizedgassamplesfromthefederalagencywithoverallresponsibilityformetrologicalmatterssuchasstandardreferencematerials,weight,andmeasures,theNationalInstituteofStandardsandTechnology.14Thispyramidofmetrologymakessulphur-dioxidepollutionmeasurable:itproducesfactsaboutit,thusmakingthemarketinemissionspermitspossible.Metrologyisn’tjusta‘physical’matter;itisalsopolitical.Sulphur-dioxideemissionscould,forexample,havebeenmeasuredmorecheaplybythemethodof‘massbalance’:byrecordingtheamountsofcoaloroilconsumed TheSocialStudiesofFinance/19andtheirsulphurcontent(astandardcommercialdatum),andthenapplyingcorrectionfactors—factorsthatwere‘wellknownbasedonengineeringstud-ies’(Ellermanetal.2000:248)—totakeintoaccounttheretentionofsulphurinashanditsremovalby‘scrubbers’insmokestacks.Buttheresultantfactswouldnothavebeenhardenough.ManyenvironmentalistsintheUSAhaddeepconcernsaboutthewisdomandethicsofmarketsinpollutionpermits,andwouldhavefearedthatutilitiescouldmanipulateemissionsdatacalculatedinthisway.Thosefearsmightwellhavetippedthepoliticalbalanceagainsttheproposedmarket.Anexpensive,largelyautomaticmeasurementsystem,inwhichcorruptiblehumanbeingsplayedonlyalimiteddirectrole,‘wasthepriceofassuagingenvironmentalistconcernsaboutemissionstrading’(Ellermanetal.2000:249).Anothermoti-vationforthesystemwastoincreasemarketliquidity.EconomistRichardSandor,whohadlongexperienceoftheChicagoderivativesmarketsdescribedinChapter4,advisedtheEPAthatitwasessentialthatemissionsmeasure-mentsbemadepublicasquicklyandasfrequentlyaspossible.Otherwisethespeculatorswhowouldprovideliquiditywouldbedeterredfromenteringthemarketbythefearthattheutilitycompaniesknewtoomuchmorethantheydid.InSandor’sview,onlyautomatedmeasurement,notmass-balancecalculations,couldavoidasituationof‘asymmetricinformation’thatwoulddamageliquidity(Sandor,interviewedbyauthor,19February2007).Precept5:ActorsAreAgencementsOnewayofexpressingthehypothesisthatactors’equipmentisconsequentialandtheobservationthatcalculationisdistributedandmaterialistosaythatactorsareagencements.AlthoughthenotionofagencementisdrawnfromDeleuze(forexample,DeleuzeandGuattari2004;seeWise2005),thesenseinwhichIusethetermismoredirectlythatinwhichithasbeenusedbyMichelCallon,whoseworkonmarketshasbeenamajorresourceforthesocialstudiesoffinance.Callon’sapproachisrootedinthe‘actor-networktheory’thathedevel-opedwithBrunoLatour(CallonandLatour1981;Callon1986;Latour1987;2005)andwithotherssuchasMadeleineAkrich,JohnLaw,andVololonaRabeharisoa.Thetheoryiscurrentlyperhapsthesinglemostprominentform 20/TheSocialStudiesofFinanceofwhatinChapter1Icalled‘materialsociology’.Itsmostdistinctivefeatureisitsagnosticismastothenatureofagentsoractors,whicharetakenaspotentiallyincludingnon-humanentitiesaswellashumanbeings.(Withthegeneralreaderinmind,Ihaveusedtheword‘agent’,not‘actor’,inthetitleofthisbook,butdonotintendanysystematicdistinctionbetweenthetwo.)Theactor-networknotionof‘actor’isthusquitedifferentfromthestandardsociologicaluseoftheterm,whichrefersonlytohumanbeings.Inonesense,theactor-networknotionof‘actor’—alsotherelatednotionof‘actant’—simplyfollowshowthosetermsareusedinsemiotics,especiallythesemioticsofA.J.Greimas.15(AsLatour2005:53putsit,someentitiesthat‘modifyastateofaffairsbymakingadifference’are‘actants’,not‘actors’.Thedifferenceisthatthelatterhaveandtheformerhavenot‘beenprovidedintheaccountwithsome...featuresthatmakethemhavesomeformorshape’.)Theactorsandactantsinnarrativesareplainlynotallhumanbeings:‘Theconceptofactant...appliesnotonlytohumanbeingsbutalsotoanimals,objects,orcon-cepts’(GreimasandCourtés1982:5).Whereactor-networktheorydiffersfromtraditionalsemioticsisinapplyingsuchnotionsmorebroadlythantonarra-tivesandothertexts.AsJohnLawputsit,‘semiotics...tellsthatentitiestaketheirformandacquiretheirattributesasaresultoftheirrelationswithotherentities...actor-networktheorymaybeunderstoodasasemioticsofmateriality.Ittakesthesemioticinsight...andappliesthisruthlesslytoallmaterials—andnotsimplytothosethatarelinguistic’(Law1999:3–4,emphasesinoriginal).ThusCallonarguesthataction,‘includingitsreflexivedimensionthatpro-ducesmeaning,takesplaceinhybridcollectives’:combinationsof‘mater-ialandtechnicaldevices,texts,etc.’andhumanbeings.InCallon’sanalysis,therefore,aneconomicactorisnotanindividualhumanbeing,norevenahumanbeing‘embeddedininstitutions,conventions,personalrelationshipsorgroups’(aseconomicsociologyposits).ForCallon,anactoris‘madeupofhumanbodiesbutalsoofprostheses,tools,equipment,technicaldevices,algorithms,etc.’—inotherwordsismadeupofanagencement(Callon2005:4–5).Thereisadeliberateword-playinthenotionofagencement.Agenceristoarrangeortofittogether:inonesense,unagencementisthusanassemblage,arrangement,configuration,orlayout.ThereferentineverydayFrenchisoftenphysical,suchasthepartsofamachine;indeed,inordinaryparlance,lesagencementsarefixturesandfittings,andtobebienagencéistobewellequipped TheSocialStudiesofFinance/21(Collin,Knox,Ledésert,andLedésert1982).Theothersideoftheword-play,however,isagence:agency.(IretaintheFrenchagencementbecausethisword-playdoesnotcarryoverintotheterm’susualEnglishrenderingas‘assemblage’,whichthushassomewhattoopassiveaconnotation.)AsCallonandCaliskan(2005:24–5)putit:‘Agencementsdenotesocio-technicalarrangementswhentheyareconsideredfromthepoint[of]viewoftheircapacitytoactandtogivemeaningtoaction.’Thenotionthatactorsaresociotechnicalcombinations—thattheyareagencements—hasthevirtueofimplicitlyhighlightingthequestionoftheattri-butionofagency,inthesenseofthecapacityforintentionalaction.(Notethatthisisaweakersenseof‘agency’thanacommonsociologicaluseofthetermtomeanthecapacityforintentionalactionthatalterssocialstructures.)Ordinarily,wethinkofagencyinmarketsasresidinginindividualhumanbeings,butthatattributionisnotinevitable.Thelawofcontract,forexample,oftenattributesaspectsofagencynottoindividualsbuttotheorganizationstowhichtheybelong.Anintrinsicaspectofmakingadeal—acommonplaceintentionalactioninfinancialmarkets—isthetakingonofcommitments,forexampletodeliversecuritiesormoney(atleasttheelectronictracesthereof).Whiletradersmayspeakortypethewordsthatbringadealintobeing,thosewordscommitnotthosewhohavespokenortypedthembuttheorganiza-tionsofwhichtheyaremembers.Shouldtheindividualsinquestionleavethoseorganizations,theseobligationsdonotdepartwiththem.Muchofthesometimesfiercedebatearoundactor-networktheorythathastakenplacewithinthesocialstudiesofscienceandtechnologyhasconcernedtheattributionofagencytonon-humanentities(see,especially,CollinsandYearley1992).Atleastequallyinterestingfromtheviewpointofmarkets,however,ishowtheattributionofagencydistributesitacrosshumanbeings.Sometimes,forexample,theagencyoftradersisdenied.Forinstance,priortotheintroductionofsufficientlysmallandportablecomputers,tradersmakinguseofoptiontheoryusuallydidsobyconsultingpapersheetsoftheoreticaloptionvalues.InbothChicagoandLondon,thosewhodidthiswerefre-quentlyreferredtobytheirpeersas‘sheetmonkeys’:agency-lessslavesofthephysicalsheetsandofthemathematicalmodelembodiedinthem.Thepresent-dayequivalentepithetis‘F9monkey’.PressingtheF9keywhenusingaspreadsheetinstructstheprogramtoperformthecalculationsimplementedinthespreadsheet,andan‘F9monkey’issomeonewhoallowstheresultantnumberornumberstodeterminetheiractions. 22/TheSocialStudiesofFinanceTheattributionofagencyisofteninterwovenwithissuesofgender.Thusonesheet-usingoptionstradertoldmethatothersquestionedthemasculinityoftheuseofsheets.Fellowtraders‘wouldlaughatyouandtrytointimidateyououtofthe[trading]pit,saying,“You’renotamanifyou’reusingthosetheoreticalvaluesheets.”They’dtakeyoursheetsandthrowthemdownonthefloorandsay,“Beaman.Tradelikeaman....Youshouldn’tbehere.You’renotatrader.Youcan’ttradewithoutthose”’(MacKenzieandMillo2003:124).MycolleagueLuciaSiuhasdiscoveredthatinatleasttwoofChina’scommodityfuturesexchangesmosttradersarewomen.AlthoughtheirrolesastradingintermediariesaresimilartothoseofmanyoftheirpredominantlymaleWesterncounterparts,theirjobslacktheprestigeandhighrewardsofthelatter.Theyareseenmoreasthetoolsofotheractors,andtheskillandinitiativetheybringtotheirtradingarelargelydisregarded.Thenotionofagencementthushasavirtuethatisnotamongstthosestan-dardlyclaimedforactor-networktheory.Tracingtheagencementmakingupaneconomicactor,ratherthanfocusingexclusivelyonwhatonemightcallaction’sglamorousagentialpeaks,broadensthefieldofviewofthesocial-scienceinvestigationoffinance,notjusttowardsthingsbuttowardslesshigh-statushumanbeings.Therehas,forexample,beenconsiderableatten-tiontoaccountants,totheirprofessionalizationandregulation,andsoon.Thedetailedcontentoftheworkoftheirlower-statuscounterparts,book-keepers,hasincontrastalmostneverbeenstudiedbysocialscientists,yet—asemphasizedinChapter6—itisbook-keeperswhotypicallyperformtheprimaryclassificatoryactsthatarethefoundationofcorporateaccounts.16Again,genderisinvolved.Thebifurcationofthetaskofproducingaccountsthattookplacefromthelatenineteenthcenturyonwardsbecameagenderedbifurcation,aseparationbetweenthealmostexclusivelymaleprofessionalandtheclerk,whowasincreasinglylikelytobefemale(KirkhamandLoft1993).Anothervirtueofthenotionofagencementisthatitsuggeststhatactorsshouldnotbeseenashavingfixednaturesorfixedcharacteristics.Theequip-mentthatmakesanactorwhatitis,theparticularmaterialprocessesofcalculationitengagesin,thespecificsofthedistributionofcognition—alltheseshapethenatureofactors,orsothesocialstudiesoffinancepostulates.Again,thepointisfundamentallyanactor-networktheoryone:‘actorsarenetworkeffects.Theytaketheattributesoftheentitieswhichtheyinclude’(Law1999:5). TheSocialStudiesofFinance/23Considerrationality,forexample.Thecentraldivideinmodernfinancialeconomicsisbetweentheorthodoxviewofactorsasrationalandthe‘behav-iouralfinance’viewofactorsassubjecttosystematicpsychologicalbiasesofthekindidentifiedintheexperimentalworkofDanielKahnemanandAmosTversky(forexample,KahnemanandTversky1979).Implicitlyorexplicitly,bothviewsconceiveoftheactorasanindividualhumanbeingorakintosuchahuman.Typically,neitherviewtakesintoaccountinanysystematicfashiontheways,discussedabove,inwhichindividualhumanbeingsareembeddedinagencementsofmultipleotherhumanbeingsandtechnicalandconceptualequipment.Ratherthanendorsingeithertherational-actororbehavioural-financeviewpoints,thesocialstudiesoffinanceinvestigateshowthenatureofactorsisshapedbytheagencementsthatconstitutethem.Forexample,theincorpor-ationofconceptsandproceduresfromeconomicsintoactors’conceptualequipmentmaymakethebehaviourofactorsmorerational,inthesenseofbecomingclosertothepostulatesofeconomicmodels.Atitsmostbasic,ahumanbeingequippedwithafinancialcalculatorisadifferentactorfromonewithoutone.Iwasamused,forexample,todiscoversomeyearsagothatcolleagueswhowereusingclassroomexperimentstoillustratebehaviouralfinanceinsistedthattheirstudents(employeesinthefinancesector)notbringtheirfinancialcalculatorswiththem.Thecalculatormadethestudenttoorational!Theclaimthattheattributesofactorsarenotfixedcanbeextendedbeyondrationalitytotheotherstereotypedattributeoftheeconomicactor:selfish-ness.Itiseasytoassumethattheindividualsandorganizationsinfinancialmarketsaremotivateduniformlybyprivatemonetarygainalone.However,significantaspectsofthosemarketsaregifteconomies.Forexample,thosewhomakeinputsintotheLIBORcalculationarenotpaidtodoso,andthedataprovidersthatdisseminateLIBORpaytheBritishBankers’Associationonlyamodestfee.Furthermore,manyaspectsoffinancialmarketsrelyontrust,forexamplethatadealmadeverballybetweentraderswillbehonouredevenifsubsequentpricemovementsrenderitdisadvantageous.ItisnotaccidentalthatthecoatofarmsoftheLondonStockExchangebearsthemottodictummeumpactum:mywordismybond.Anypre-digitalfinancialmarketwouldhavebeenhamstrungifverbalcommitmentscouldnotinpracticebereliedupon.Evenintoday’sderivativesmarkets,dealsareoftenstruckinformallyandfullconfirmationsbyfaxorelectronicallyfollowonlydaysorweekslater. 24/TheSocialStudiesofFinanceSelfishnessisperhapsmostinterestinginrelationtocollective-actionprob-lems.Thesearesituationsinwhichthereisacourseofactionavailablethatwillbeofnetbenefittoallthoseconcerned,butinwhichthebenefitcannotberestrictedtothosewhohavetakenpartintheaction,andthebenefittoeachparticipantresultingfromhisorherindividualcontributionisoutweighedbythecostofthatcontribution.Insuchsituations,selfishlyrationalactorswillfree-ride,leavingittootherstotakethenecessaryaction,whichwillthereforenothappenifallthoseinvolvedarerationalegoists.Typicalexamplesarecasesinwhichtradersneedtoforgoprivateprofit-makingopportunitiesinordertobenefittheirmarketasawhole(forexample,bypreservingitsreputationfororderlinessorprobity),17orinordertopreserveatacitprice-fixingagreementfromwhichtheyallgain.Gifts,trust,andcollectiveactionareofcoursequestionsofstatusandofculture.Asallsocialscientistsknow,giftsshouldnotbethoughtofasarbitraryaltruism:tobeinvitedtojoinoneoftheBritishBankers’Association’sLIBORContributionPanelsincreasesabank’sstandinginthemarketinquestion.London’straditionof‘gentlemanlycapitalism’(Thompson1997)helpedmakedictummeumpactumnotanentirelyemptyboast:torenegeonaverbalpromisewoulddemonstratethatonewasnotagentleman.Yetphysicalsettingandtechnologymaymattertoo.Theexperimentalevidenceoncollectiveactionsuggeststhatitiseasiertosustainincontextsofface-to-faceinteraction,and‘open-outcry’tradingpitsprovidedjustsuchacontext.18Free-ridinganddefectionforprivatebenefitfromtacit‘sharing’agreementsmaybeharderforothertraderstodetectwhentradingisconductedbytelephoneorelectroni-cally,andtradersinthosecontextsmaynotbeabletoexpresstheirdisapprovalaseffectivelyastheycaninface-to-faceinteraction.Theissueofwhetherandhowactorscanbeconfiguredsothattheywillcontributetocollectiveaction,ratherthanfree-ride,isparticularlypressinginrelationtoenvironmentalquestions,manyofwhicharecollective-actionand/orinter-generationalproblems.Asnotedabove,sulphur-dioxidemetrol-ogywasmobilizedintheUSAasthefoundationofamarket.InEurope,itbecamepartofasociotechnicalsystemdesignedtopromoteinternationalcollectiveactionbymakingbothcontributionsandfree-ridingvisible:‘tote-boarddiplomacy’asithasbeencalledbythemainstudyofit.19(A‘tote-board’isapublicindicatorofthelevelofcontributionstoadrivetoraisefundsforcharity.Theroleofasimilar‘technologyofaltruism’in‘Téléthon’tele-visionfund-raisingisnotedbyCallonandLaw2005.)TheUnitedKingdom’s TheSocialStudiesofFinance/25reputationinthe1980sand1990sasthefree-riding‘dirtymanofEurope’—areputationthatwasdamagingtothethenConservativegovernment—arosefromthevisibilityofitsapparentfailureinthisrespect.20Itisalsoimportantthatthecourseofactionthatisinanactor’sinterestisoftennotself-evident,anddecisionmakersareatleastsometimesinfluencedbycalculationsperformedbyeconomistsandotherprofessionalcost–benefitanalysts.(Again,actorssometimesseemtoneedeconomiststohelpthemberational!)ThustheMontrealProtocol,theinternationalagreementphasingoutozone-depletingsubstances,wasfacilitatedbyacost–benefitanalysisbytheEnvironmentalProtectionAgencythatwasacceptedasineffectshowingthatfortheUnitedStatestheissuewasnotacollective-actionproblem:thecoststotheUSAofphasingoutozonedepleterswouldbeoutweighedbythebenefitsofitsownrestraint,evenifnoothercountryexercisedrestraint(Barrett2003:227–30).Asimilarcost–benefitanalysisfromthePresident’sCouncilofEconomicAdvisorshadasignificantinfluenceindebatesinWash-ington:ithelpeddefeatopponentsoftheprotocolandpersuadekeyofficials(Benedick1991).Incontrast,internationalactiontoslowglobalwarmingmaybebeinghamperedatleasttosomedegreebytheabsenceofasimilarlyauthoritativecalculation.TheUKTreasury’sreview,ledbytheeconomistSirNicholasStern,wasintendedtoprovideit,andthereview(Stern2007)indeedconcludedthattheglobalcostsofcurbingemissionsareoutweighedbytheglobalbenefitsofdoingso.However,thereview’smethodology—particularlyitsapproachtothenumericallydominantfactorinsuchcalculations,thechoiceofthe‘dis-countrate’usedtocalculatethepresentvalueoffuturecostsandbenefits—hasprovedcontroversial,andconsensushasnotemerged.Itwould,ofcourse,behopelesslynaivetoimaginethatpolitical(orindeedmanagerial)decisionsaresimplytheresultofcost–benefitanalyses.Thepoint,rather,isthis.Ifactionisshapedbyactors’interests(aplausiblepropositionsharedbysocialscienceperspectivesrangingfromMarxismtorational-choicetheory),thenattentionmustbepaidtothematerialcalculationofthoseinterests.EveryoneI’vemetinfinancialmarketswantstomakemoney,butithardlyadvancesthestudyofsuchmarketstonotethatthisinterestisper-vasive.Whatoneneedstoexamineishowbeliefsaboutmarkets,physicalandconceptualtools,thestructureoftheconnectionsbetweenactors(Burt2005),andsooninfluencecalculations.Interestsarenotgiven:theyarecalculatedwithinagencements. 26/TheSocialStudiesofFinancePrecept6:ClassificationandRuleFollowingAreFinitistProcessesAstheabovediscussionindicates,actor-networktheoryhasbeenanimportantinspirationforsocialstudiesofscience.Itwould,however,beamistakeforthefieldtorestrictitselftoasinglesetofintellectualresources.Adifferentintellectualtraditionfromthesocialstudiesofscienceandtechnology(andrelatedfields)hasmuchtooffertoo:finitism.21Fromtheviewpointofthisbook,finitismismosteasilyintroducedasatheoryoftheapplicationoftermstoinstancesorparticulars.Consideraterm‘A’,whichcouldbeaneverydaywordsuchas‘walk’or‘red’;amathematicaltermsuchas‘converge’or‘polyhedron’(Lakatos1976);or—asdiscussedinChapter6—anaccountingtermsuchas‘depreciate’,‘asset’,or‘financelease’.Oneviewoftheapplicationoftermstoinstancesistoconceiveoftermsashavingfixedmeanings.Oncewehavedecided,individuallyorcollectively,whatatermmeans,thentheinfiniteuniverseofitems,processes,activities,statesofaffairs,andotherparticularsisdividedupintoinstancesofAandofnot-A:reditemsanditemsthatarenotred;walkingandactivitiesthatarenotwalking,suchasrunning;polyhedraandentitiesthatarenotpolyhe-dra;purchasesofcapitalassetsandactivitiesthatarenotpurchasesofcapitalassets;‘financeleases’andstatesofaffairs,suchasoperatingleases,thatarenot‘financeleases’;andsoon.22Onthisviewofmeaning,whichissometimescalled‘extensionalsemantics’,aterm’s‘extension’—the‘setofthingsofwhichitistrue’(Barnes1982:31)—isfixedinadvanceofusageoftheterm.ItmaysometimesbedifficulttodeterminewhetheranewlyencounteredparticularisaninstanceofAornot,butifextensionalsemanticsiscorrectthedifficultyismerelyempirical.Incontrast,finitismdeniesthattheuniverseofalltheitemsandactivitiesthatmayeverbeencounteredshouldbethoughtofasdividedupinadvanceintoinstancesofAandofnot-A.Allweeverhave—asindividualsorasanentireculture—isafinitesetofpastapplicationsof‘A’toparticulars.Whenanewparticularisencountered,thedifficultyismorethantheempiricaloneofdeterminingitsproperties:weneedtodecidewhetheritissufficientlylikethepreviousparticularswehaveclassedasAtowarrantthatclassification.Notwodirectlyobservableentitiesoractivitiesareeverentirelyidentical;therearealwaysdifferencesbetweenthemthatcouldbepointedtoaswellassimilarities;‘everysituationisindetaildifferentfromeveryother’(Hesse1974:12). TheSocialStudiesofFinance/27ConceptAConceptAInstance1notAInstance1ofAInstance1notAInstance1ofAl2l2l2El2l•lX•33•••T•E•n•••N•always•SlnfinitelnIn=∞OOpenpossibleNDecisionunderdeterminedClosurebyexistingclusteringl1lUndetermined2linstances3Statusofallinstancesaseither•AornotAgivenpriorto•Statusofalldeterminedusage•instancesrevisableEXTENSIONALFINITISMSEMANTICSFigure2.1.ExtensionalsemanticsandfinitismSource:RedrawnfromfigureinBarryBarnes,T.S.KuhnandSocialScience,1982,TheMacmillanPressLimited,reproducedwithpermissionofPalgraveMacmillan.ExtensionalsemanticsandfinitismareneatlysummarizedbythesociologistofscienceBarryBarnesinthediagramIhavereproducedasFigure2.1.Notethatfinitismgoesbeyondtheassertionthatmeaningsaresocialconventions(Barnes,Bloor,andHenry1996).Thatassertionisentirelycompatiblewiththeextensional-semanticsviewthatoncethemeaningof‘A’ischosen,theinstancestowhichitcorrectlyappliesarethenfixed.Rather,inafinitistperspectiveeveryapplicationofatermtoaninstanceisimplicitlyadecision.NotonlyistheextentofsimilaritytopreviousparticularsclassedasAalwaysinprinciplecontestable,butthosepreviousinstancesarealwaysrevisable:wemaydecidethatoneormorepreviousapplicationsof‘A’weremistaken.Anareainwhichfinitismisofparticularimportanceisinunderstandingwhatitistofollowarule,theclassicdiscussionofwhichisinWittgenstein(1967).Rulefollowingmattersfortheunderstandingoffinancialmarketsbecauseoftheextenttowhichtheyarerule-governed.AsVogel(1996)suggests,althoughwethinkofthecurrentepochasoneof‘deregulation’ofmarkets,rulesareproliferating.Inapparentparadox,creating‘freermarkets’ 28/TheSocialStudiesofFinancedemands‘morerules’.Inparticular,accounting—thetopicofChapter6—isanactivitythatissubjecttoaveryextensivebodyofrules.Accordingtofinitism,‘nothingin[a]ruleitselffixesitsapplicationinagivencase’(Barnes1995:202).Consider,forexample,thesixthcommandment,‘Thoushaltnotkill’.Itseemsastraightforwardenoughprohibition,butcon-sideritsapplicationtoenemysoldiers,enemycivilians(as‘collateraldamage’),terminallyillpeopleingreatpainwhohaveexpressedawishtodie,humanfoetuses,animals(inexperiments),animals(forfood),andsoon.Whatthecommandmentimpliesforeachofthesehasbeentheobjectoffiercelyvaryinginstincts.Murdermayseemtooeasyanexamplewithwhichtoillustraterulefinitism,butconsiderahardercase:chess.Itlacksmurder’smoralsignificance,andisa‘microworld’:alimited,artificialdomain,deliberatelystrippedofambiguity.23(Chesshas,forinstance,beenrelativelyeasytoautomate.)Here,surely,weshouldfindrulestowhichextensionalsemanticsapplies.Consider,though,thepuzzlereproducedinFigure2.2.White,toplay,mustdeliverFigure2.2.WhitetoplayandmateinasinglemoveSource:PuzzlecomposedbyRichardHaddrell. TheSocialStudiesofFinance/29checkmateinasinglemove.Thesolutionisthatwhitemoveshisorhermostadvancedpawntotheeighthrankandreplacesitwiththerookthatcurrentlyblocksthediagonalfromthewhitebishoptotheblackking.Thelatteristhusexposedtocheck,andhasnoflightsquarebecausetherooknowcontrolstheeighthrank.Itischeckmate.Anychessplayerwillrespondthatthe‘solution’isnotalegalmove:itiscontrarytotherulesofchess.Consider,however,thepawn-promotionruleasitstoodintheLawsofChessinMay2005:‘Whenapawnreachestherankfurthestfromitsstartingpositionitmustbeexchangedaspartofthesamemoveforaqueen,rook,bishoporknightofthesamecolour’(FIDE2005a:rule3.7.e).Thesolutiontothepuzzlecould,inprinciple,surelybedefendedasareasonableinterpretationofwhatitisto‘exchange’thepawnforarook.Onecouldofcourseimaginearuleofinterpretationbeingaddedtoseektospecifywhatthepawn-promotionrulemeans,butiffinitismiscorrectthatissimplytoenteraregress,sincetheruleofinterpretationwillitselfcontaintermsthatneedtobeinterpreted.24(Oneexplanationofthesheervolumeoftherulesgoverningfinancialmarkets,especiallygoverningaccountingintheUSA,ispreciselythatthisregresshasbeenentered.)Thechiefadvantageoffinitismasaresourcefortheanalysisoffinancialmarketsislessthewayinwhichithighlightsthein-principleflexibilityofclassificationandrulefollowing(apointthatmanyinthefinancesectorarealreadywellawareof)thanthewayinwhich—preciselybyhighlightingthatin-principleflexibility—itchannelsattentiontothefactorsthatinpracticeconstrainit.AsthephilosopherandsociologistofscienceDavidBloorputsit,inapassageofwhichIhavealreadyquotedpartinChapter1:Accordingtomeaningfinitism,wecreatemeaningaswemovefromcasetocase.Wecouldtakeourconceptsorrulesanywhere,inanydirection...Wearenotpreventedby‘logic’orby‘meanings’fromdoingthis...Therealsourcesofconstraint[are]ourinstincts,ourbiologicalnature,oursenseexperience,ourinteractionswithotherpeople,ourimmediatepurposes,ourtraining,ouranticipationofandresponsetosanctions,andsoonthroughthegamutofcauses,startingwiththepsychologicalandendingwiththesociological.(Bloor1997:19–20,emphasisinoriginal)Anactor-networktheorist—attunedtothewaysinwhichtechnologies(walls,weapons,prisons,writing...)contributetotheproductionofsocialorder—wouldwanttoadd‘things’or‘technologies’toBloor’slist.Ithinkthat’sright.Considerthechesspuzzle.Ifasimilarpositionappearedinachess 30/TheSocialStudiesofFinancetournament,andaplayerovercamehisorherinstinctssufficientlytoplayamoveanalogoustothe‘solution’,whatwouldstopitbeingsuccessfulisindeed‘interactionswithotherpeople’:theopponentandtournamentcontrollerwouldnotacceptit.Butifonewereplayingachesscomputer,theconstraintistechnological:themovecouldnotbeinputintothemachine(atleastunlessonehadaccesstothesourcecodeofthechessprogramandtheskillstoalterit).AswillbeseeninChapter6,it’sthesamewithaccounting.Constraintcomesnotonlyfromotherpeople,butalsofrommachines.Precept7:EconomicsDoesThingsResearchersinthewidersocialscienceswhostudymarkets,forexampleeco-nomicsociologists,oftenseethemselvesasincompetitionwitheconomics.Thereisnoreasonwhythatshouldbethecaseforsocialstudiesoffinance.Modernfinancialeconomics,forexample,offersananalysisofmarketsthatisremarkablysuccessful—moresuccessfulthananysystematicalternative(behaviouralfinanceincluded)—andthesocialstudiesoffinanceshouldnot,inmyview,seeitselfascommittedtodisputingthatsuccess.Thatattitudecomesnaturallytosomeonewithabackgroundinthesocialstudiesofscience.Nophilosopher,historian,orsociologistofphysicsseeshim-orherselfasincompetitionwithphysics.(It’snotthattheknowledgeproducedbythenaturalsciencescan’torshouldn’tbecontested,butthattocontestitistodoscience,nottoanalyseitsociologically.)However,theattitudeshouldnotbecomesimplycomplacentacceptanceofahegemonicdiscipline.Ifamaterialsociologyofmarketsistotalkabouteconomics—anditneedsto—itmusthavesomethinginterestingtosay.Apromisinglineofenquiryinregardtofinancialeconomicsexploresthehypothesisthatthefield’ssuccessisinparta‘performative’success,notsimplyadescriptiveoranalyticalsuccess.Theterm‘performative’wascoinedbythephilosopherJ.L.Austin,andabasicusageistodistinguishutterancesthatdothings—‘Iapologize’,whichifonesaysitisanapology—fromutter-ancesthatreportonstatesofaffairsseparatefromtheutterance,suchas‘Itisraining’(seeAustin1962).Aneconomicmodel,forexample,issome-timesnotsimplyarepresentationofamarketasanentityentirelyseparate TheSocialStudiesofFinance/31fromthemodel,butatoolusedbymarketparticipants,andsometimesthewidespreaduseofsuchatoolcanchangemarketsinconsequentialways(MacKenzie2006).Applyingthenotionof‘performativity’toeconomics(asdoneinfluentiallyinCallon1998)thussuggeststhehypothesisthatfinancialeconomicssucceedsbecauseithasbeenabletoreshapetheworld.Bybeingincorporatedintoregulatorystructures,pricingsoftware,tradingstrategies,andsoon,ithascreatedconditionsofwhichitisareasonablygoodempiricaldescription.Notethatsuggestingthateconomicsisperformativeisnotquitethesameassayingthatideasfromeconomicsinfluencepeople.Theydo,inthesensethatthosewhohavetakenaneconomicsdegreeoraneconomics-influencedMBAdoseemoftentothinkdifferentlyfromthosewhohaven’t.ButasI’veemphasizedabove,aneconomicmodelembodiedinasystemforpricingandriskmanagementcanhaveeffectseveniftheusersofthesystemdon’tbelievethemodel,don’tunderstandit,orevendon’tknowthatitexists.Economicsisembodiedinproceduresandphysicalartefacts,notjustinideas.SinceIhavediscussedtheperformativityoffinancialeconomicsatlengthelsewhere,25Igiveadifferentexampleofthepreceptthat‘economicsdoesthings’inthisbook:thepartplayedbyeconomicsandbyeconomistsintheestablishmentofthemarketsforemissionspermits.It’sactuallyamoreclear-cutcaseoftheprecept,becauseunlikemostofthemarketsanalysedbyfinan-cialeconomistsemissionsmarketsdidnotpreviouslyexist:theywereinventedbyeconomists.They’rethusnotachallengingcasefromtheviewpointofestablishingthateconomicsdoesthings,buttheemergenceofthesemarketsisnonethelessworthattentionbecauseofthepivotalplacethattheynowoccupyinenvironmentalpolicyandpolitics.Precept8:InnovationIsn’tLinearPartofwhatmakestheroleofeconomistsininventingemissionsmarketsinterestingisthatinventionisnotthesimplethingenvisionedbya‘linear’viewoftechnologicalinnovation,and‘theperformativityofeconomics’ismisreadifseenthroughthelensofalinearview,asithasbeenbysomecriticsofthenotion.26Insuchaview,toputmattersverysimply,thefirststep 32/TheSocialStudiesofFinanceiswhenscientistsdiscoverfeaturesofthenaturalworld.Next,thepracticalimplicationsofthosediscoveriesarededucedbytechnologistsandbecomeinventions.Thoseinventionsthendiffuse,andhaveeffectsonsocietiesandeconomies.Allaspectsofalinearviewofinnovationhavebeencriticizedforseveraldecades—astill-usefulaccountofmanyofitsflawscanbefoundinBarnesandEdge(1982)—andindeedit’squestionablejusthowwidelyitwaseverbelieved,atleastintheabovesimple,starkform(Edgerton2004).27Ihavealreadytouchedupontheweaknessofthenotionofscientific‘discovery’whenitisinterpretedtomeanstumblingoverwhatisalreadythere.Noristechnologicalinnovationsimplythedeductionoftheimplicationsofscientificdiscovery.Technologistsmayormaynotdrawuponscience,andwhentheydotheyuseitcreativelyasaresourceratherthansimplydeducingitsimplications.Muchinnovationtakesplaceduringdiffusion:Fleck(1994)callsthis‘innofu-sion’.Theusersoftechnologiesareoftenanimportantsourceofinnovation(OudshoornandPinch2003).Technologiesareshapedtofittheir‘contexts’(MacKenzieandWajcman1999),and‘contexts’areoftendeliberatelyreshapedtofittechnologies(MacKenzie1990),rathersimplychanginginresponsetoadoptionofthosetechnologies.AsCallon(2007)suggests,alinearviewoftheperformativityofeconomicswouldbenomorevalidthanalinearviewoftechnologicalinnovation.Newfinancialproductsandtradingmechanismscanindeedbeviewedasinnova-tions(asChapter4discusses),andeconomicsisreasonablyoftendrawnuponintheseinnovations.However,itisnottheonlysourceofthem,andtheout-comeofprocessesofeconomicinnovationisaffectedbymanyotherfactors.Financialinnovationisshapedbymatterssuchaslegalstructures,politicalprocesses,andevenbybroaderculturaldifferences:the‘culturalgeographies’discussedinChapter4.Themarketsinemissionspermitsareindeedecono-mists’inventions,butasdiscussedinChapter7,theyareshapedbymuchmorethantheintentionsoftheirinventors.Precept9:MarketDesignIsaPoliticalMatterThatlinearviewsareinvalidhasanimportantconsequence.Iftheprocessbywhichtechnologiesormarketsdevelopwereasalinearmodelposits, TheSocialStudiesofFinance/33thepoliticsoftechnology(orthepoliticsofmarkets)wouldbereducedtoasimplebutunattractivesetofchoices:toembraceinnovationindis-criminately;toacquiescepassively;ortoresistinnovation.However,becauselinearviewsofinnovationarefalse,amorediscriminatingandnuancedpol-iticsoftechnology,seekingactivelytoshapetheinnovationprocessanditsoutcomes,ispossible,andthesameissurelytrueofthepoliticsofmarkets.Thedesiretoprovideastraightforwardillustrationofthewayinwhichmar-ketdesignisapoliticalmatteristhemainreasonwhythisbookgoesbeyondfinancialmarketstoexaminethenewmarketsincarbonemissions.Thedetailoftheirdesignisenormouslyconsequential,anditisbeingcontestedpolitically(althoughonnothinglikethescalethattheissue’simportancedeserves).Thecarbonmarketsare,however,nomorethananillustration.Thatthedesignofmarkets—forexample,theformalandinformalrulesthatgovernthem—isapoliticalmatteristruemorewidely.Apparentlyminormatters—‘technicalities’,oftentechnicalitieslittleunderstoodbynon-participants—canhavebigeffects,forexample,givingadvantagestosomeactorsandsomestrate-giesanddisadvantagingothers.(Myfavouriteexample,discussedinChap-ter5,isthe‘uptickrule’intheUSA,whichtranslatedintopoliticaltermsmadeithardertovote‘against’acorporationthan‘for’it.)Aneffectivepol-iticsofmarkets—whether‘left-wing’or‘right-wing’ininspiration—needstoengagewithsuchapparent‘technicalities’,notjustwiththeoverallvirtuesanddemeritsofmarkets.Precept10:ScalesAren’tStableAprejudicethatimpedesthedevelopmentofapoliticsofmarketsoftheabovekindisthetendencytodividephenomenaintosmall,‘micro’phenomena(details,technicalities,interpersonalinteractions,andsoon)andbig,‘macro’phenomena(globalization,neoliberalism,capitalism,theinternationalsystemofstates,andsoon)andtothinkofonlythelatteraspolitical.Amaterialsociologyofmarketsshouldbesuspiciousoftheassumptionthatscalesarefixed:that‘micro’phenomenawillremainsmall,and‘macro’phenomenastaybig.Inscienceandtechnologyitisfrequentlythe‘details’andthe‘technicalities’thatmatter,separatingasuccessfulexperimentormachine 34/TheSocialStudiesofFinancefromafailure.Animportantemphasisinscienceandtechnologystudiesistheneedto‘opentheblackbox’,toinvestigatethecontents,normallyhidden,ofsuccessfulproceduresandsuccessfulmachines.(Inengineers’terminology,a‘blackbox’isadevicetheinternalstructureofwhichisopaqueorcanbedisregarded,andwhichcanbetreatedsimplyastransforminggiveninputsintopredictable,appropriateoutputs.)Solongasthey‘work’,blackboxesareanimportantsourceofpower:indeed,onanactor-networkviewtheyaretheonlysourceofpower.‘Amacro-actor’,wroteCallonandLatourinthefirstexpositioninEnglishofactor-networktheory,‘isamicro-actorseatedonblackboxes’(1981:299).Themostvividinstancesofchangesinscalearethoseinwhichapparently‘micro’matters(patternsofsocialrelationsamongsmallnumbersofpeople,‘technicalities’,andsoon)becomelarge.I’vediscussedelsewhereavividexam-pleofthefirst—the‘imitative’tradingamongstarbitrageurs,manyofwhomkneweachotherpersonally,thatledtothedownfallofthehedgefundLong-TermCapitalManagementandtothenear-paralysisofsignificantpartsoftheglobalfinancialmarkets(MacKenzie2003)—soletmehereofferanexampleofthesecond.Itindeedconcernsasmalltechnicality:thearithmeticofsecurityprices.Until1997,stockpricesintheUSAweredenominatedineighthsofadollar,soastockcouldcost$453/($45.375),butwasnotallowedtohaveaprice8of,forexample,$45.37,$45.38,or$45.40.Researchbytwofinancialecono-mists,WilliamChristieandPaulSchultz(1994),foundthatbroker-dealersonNASDAQ(theNationalAssociationofSecurityDealersAutomatedQuo-tationsystem)rarelypostedpricequotationsthatendedinanoddeighth(1/,3/,5/,or7/).SubsequentinvestigationbytheDepartmentofJustice8888andSecuritiesandExchangeCommissionof4,500hoursofbroker-dealers’tape-recordedtelephonecallsfoundthatthiswasamarketnorm,notjustastatisticalquirk,withbroker-dealersphoningeachotheraboutodd-eighthquotes:Trader1:WhotradesCMCAF[ComcastUKCablePartnersLtd.]inyourplacewithoutyellingitout?Trader2:...SammyTrader1:Sammywho?Trader2:Itmaybetheforeigndepartment... TheSocialStudiesofFinance/35Trader1:What?Trader2:Theforeigndidn’trealizetheyhadtotradeit.Trader1:Well,he’stradingitinaneighthandhe’sembarrassing...Trader2:...foreigndepartmentTrader1:He’stradingitineighthsandhe’sembarrassingyourfirm.Trader2:Iunderstand.Trader1:Youknow.Iwouldtellhimtostraightenuphis[expletivedeleted]actandstopbeingamoron.(DepartmentofJustice1996:25–6)Itiscommoninmanymarketsforpricestoclusteratroundnumbers,andtherewerepossibleinnocentexplanations,todowiththelimitedprocessingcapacityofhumanbrains,fortheavoidanceofoddeighths.Roundnumbers(inthiscase,zero,aquarter,ahalf,three-quarters)arecognitivelysalienttohumanbeings(Yule1927)andtheuseofcoarseunits—quarters,noteighths—asamarketnormmayreducetheriskoferrorandminimizenegotiation(Harris1991).28However,ChristieandSchultzsuspectedimplicitcollusiontokeepthespreadbetweenthehighest‘bid’priceatwhichadealerwouldbuyastockandthelowest‘offer’priceatwhichadealerwouldsellitatleastat25cents,not12.5cents.29Aninformal‘nooddeighths’rulewouldmakedefectionfromatacitagreementofthiskindinstantlyvisiblewhenNASDAQbroker-dealersscrutinizedtheirpeers’on-screenprices.(Thenormisofinter-estinthelightoftheabovediscussionofcollectiveactioninthatNASDAQisnotaface-to-facemarket.Thecollectiveactionaspectarisesbecauseabroker-dealerusingonlyeven-eighthquoteswasforgoingmarketsharethatwouldhavebeenobtainedbypostingbidsoroffersthatwereaneighthbetterthanhisorhercompetitors’.)Theissue‘scaledup’inaliteralsense.Twelveandahalfcentspersharebecomesahugesumwhenaggregatedovertransactionsinvolvingbillionsofshares:aclass-actionlawsuitbyNASDAQinvestorsledtothepaymentofdamagesreportedtobe$910million,then‘thelargestcivilantitrustsettle-mentinhistory’(Ingebretsen2002:153).Italsoscaledupinawiderway.ThecrisisdestabilizedNASDAQ,arguablytheworld’ssecondmostimportantstockmarket(aftertheNewYorkStockExchange).TheSecuritiesandExchangeCommissionrespondedtotheodd-eighthsepisodebyinsistingonchangedaccessandorder-handlingrules,amongsttheeffectsofwhichwastocreate 36/TheSocialStudiesofFinanceopportunitiesforfast,low-coststocktradingbydirectcomputerizedmatch-ingofbuyandsellorders,withnomediationbybrokers.Thismadepossible‘daytrading’bylaypeople,whichbecamewidespreadintheUSAinthelate1990sandwas,forexample,acomponentofthedot.comboom.AnapparenttechnicalityofthearithmeticofstockpriceshelpedshiftthetechnologicalfoundationsandwiderculturaleconomyofstocktradingintheUnitedStates.Scaleswereindeednotstable. 3AssemblinganEconomicActorwithIainHardieLondon:Wednesday,5January2005.HardieandMacKenzieareobservingtrad-ingbyahedgefund.Withtraditionalinvestmentvehiclesproducingpoorreturns,hedgefundshavebeengrowingrapidlyinnumberandincapital.Theytendtoclusterinparticularplaces,notablyinNewYork(someinmid-townManhattan,butmainlyinthenorthernsuburbs,especiallyGreenwich,Connecticut)andintheareainwhichweare:inandaroundMayfairandStJames’sinLondon’sWestEnd.Theroomwearesittinginisneitherostentatiousnorlarge:evenhedgefundsmanaginghundredsorbillionsofdollarsaresmallorganizations,andthisoneismadeupofonlyfivepeople.Theroomfacesoffthestreet,andoftentheonlynoisestobeheardaretypingonkeyboardsandthehumofthefancoolingthefund’spowerfulcomputerserver.Yettheworldiscontinuouslybeingbroughtintothisquietroom.Atnoon,Europe’sthree-minutesilencetocommemoratethevictimsoftheAsiantsunamiisobserved.Electronicmailfromothermarketparticipants,oftenatmajorinvestmentbanks,arrivesalmostcontinuously:itbringselec-tronictracesofpricesandothernews;analysesofmarkets;confirmationoftrades;andsoon.Screensdisplaynumbersrepresentingindicativepricesinthemarketsinwhichthefundconcentrates:emergingmarketbonds,the‘creditdefaultswaps’thatprovideprotectionifabondissuerdefaults,andcurren-cies.Ascrollingon-screen‘ticker’listsmajortransactionsinemerging-market 38/AssemblinganEconomicActorbonds.Telephoneconversationstakeplace.Mostarebrief:often,thefund’smaintraderisseekingpricequotations,andifthenumbersheisquoteddownthetelephonelineareattractive,apurchaseorsaleworthasmuchas$5mil-lioniscompletedinafurtherfewquickwords.Atonepoint,thespeakerphoneisswitchedontolistentostaffataninvestmentbankansweringquestionsabouttheirviewsonparticularemergingmarkets.Themattersthatattractattention,intheformeitherofconversationsorofclosescrutinyofcomputerscreens,areheterogeneous:theminutesoftheUSFederalReserve’sOpenMarketCommittee,releasedthepreviouseveninginLondontime,whicharetakenasindicatingthatfurtherinterest-raterisesareontheway;thepricesofthegovernmentbondsofthePhilippines,whichhavedefiedasharpglobaldecreasefollowingthereleaseoftheminutes;thesoon-to-be-announcedfigureforUSnon-farmemployment;theexchangeratesoftheMexicanpesoandSouthAfricanrand;politicsinEcuador;pricequotationconventionsintheTurkishbondmarket;andmuchelse.Economicactorsareagencements,asdiscussedinChapter2:particulararrangementsofembodiedhumanbeings,physicalobjects,technicalsystems,procedures,andsoon.Whatweareobservingisamodestagencement—acom-binationofonlyfivepeople,somefamiliartechnologies(theserver,thekey-boards,thescreens,thenetworkconnections),andsomespecializedalgo-rithms,procedures,andformsofknowledge—yetonethathastospantheglobe.Ittradesbonds,currencies,andderivativesoriginatinginAfrica,Asia,Europe,NorthAmerica,andSouthAmerica.Theworlditspansisinfinitelycomplex,andeventheattenuatedinformationthatflowsintotheroomviathecomputernetworksisinpracticeunlimitedinquantity.Fortheirfundtobeaneconomicactor,thesefivepeople,alongwiththeirphysicalandconceptualequipment,havetoturnthiscomplexityintopatternsthataresimpleenoughtograsp,andthentakeappropriateaction.Howthishappensisthetopicofthischapter.StudyingaHedgeFundHedgefundsare,asnotedabove,actorsofgrowingimportanceintheglobalfinancialmarkets,buthavebeentheobjectofremarkablylittlesocialscienceattentionoutsideoffinancialeconomics.1Thedistinctionbetweena‘hedge AssemblinganEconomicActor/39fund’andmarketactorsofotherkindswasoriginallyacreationoflawandofregulation,especiallyofthewaveofsecuritiesregulationintheUSAthatfollowedtheWallStreetcrashof1929andsubsequentGreatDepression.Somelimitedexceptionsaside,theInvestmentCompanyActof1940madeitillegalforinvestmentcompaniestoshortsell(tosellsecuritiestheydonotown,forexamplebyborrowingthemintheexpectationthatbythetimetheyneedtobereturnedtheirpriceswillhavefallen)ortouseleverage(tobuysecuritiesusingborrowedfunds).Inconsequence,anyeconomicactorintheUSAdesiringthecapacitytoactinthosewayshadtoconfigureitselfsothatitwasnotan‘investmentcompany’withinthemeaningoftheact.2Restrictionsonshortsellingandleveragevarybetweencountries,andhavegenerallyeasedinrecentdecades.However,a‘hedgefund’canstillbeviewedasaneconomicactorsetupinsuchawaythatconstraintsofthiskindonitareminimized.Althoughthereissomevariationinternationallyinlegalrequire-ments,intheUK,USA,andmostothercountrieshedgefundsareallowedtoacceptonly‘qualifiedinvestors’—individualswhoarewealthyand/ordeemedsophisticated—anddirectinvestmentbymembersofthegeneralpublicisprohibited.Hedgefundsarealsonotnormallypermittedtoadvertise(thisisknownasthe‘non-solicitation’requirement).Thereisinadditionsometimesalimitationonthenumberofinvestorspermitted:forexample,undersection3ofthe1940InvestmentCompanyAct,nomorethan100.WhatisgenerallyregardedasthefirsthedgefundwasA.W.Jones&Co.,setupin1949.(JoneshadaPh.D.insociologyfromColumbiaUniversity,butthereseemstohavebeennoconnectionbetweenhisacademicworkandhishedgefund.)Jones’sstrikingsuccesswasmadepublicbyanarticleinFortune(Loomis1966),anditbegantoattractimitators,as,later,didtheQuantumFundledbythecelebratedGeorgeSoros.Thehedge-fundsectorhasnotenjoyedentirelysmoothgrowth—therehavebeenwell-publicizedsetbacks,suchasthenear-failureofLong-TermCapitalManagement(LTCM)inSeptember1998(MacKenzie2003)—butinrecentyearsithasexpandedsharply.In1990,therewerefewerthan1,000hedgefunds,managing$25billioninassets;by2004,thereweremorethan8,000funds,managingalmost$1,000billion.3Theflowofcapitalintothehedge-fundsectorsinceourobserva-tionsinJanuary2005hascontinuedtobesubstantial—forexample,around$50billionamonthinthelaterpartof2006—andbyJuly2007assetsundermanagementwereestimatetohaverisentoover$2,000billion(Mackintosh 40/AssemblinganEconomicActor2006;Thomas2007).Hedgefundsmaybeabouttomoveintotheretailinvest-mentmainstreamforthefirsttime.Atthetimeofwriting,theUKFinancialServicesAuthorityhasproposedrulesthat,ifimplemented,wouldadd‘fundsoffunds’(which,asthenamesuggests,investinportfoliosofhedgefunds)toitsauthorizedproductlistandallowthemtotakeinvestmentsfromthegeneralpublic.Hedgefunds’annualmanagementfeesof1to2percentareinlinewiththoseofotheractivelymanagedinvestments,buttheyalsochargeaperfor-mancefee,typically20percentofprofits—thatis,ofincreasesinnetassetvalue.(Normally,netassetvaluehastoriseaboveits‘high-watermark’inpreviousperiodsbeforethisfeeapplies.)Tocurbtheincentivetoexcessiverisk-takingcreatedbythisfeestructure,hedge-fundmanagersareconventionallyexpectedtohaveasmuchashalfoftheirownpersonalnetworthinvestedinthefundthattheymanage,sothattheysufferlossesaswellasbenefitfromgains.Attimes,hedgefundscanbecomeimportantownersofparticularclassesofsecurity:inearlySeptember2005,forexample,hedgefundswerereckonedtoholdbetweenaseventhandaquarterofthestockofGermany’sleadingcorporations,takeninaggregate(JenkinsandMilne2005).Becausenearlyallhedgefundsareactivetradersratherthanpassive‘buy-and-hold’investors,andbecausetheuseofleverageiscommon,theircontributiontooveralltradingvolumesismuchhigherthantheproportionsofinvestors’capitalthattheymanage.In2005,hedgefundswerebelievedresponsibleforbetweenaquarterandathirdoftradingontheNewYorkandLondonStockExchanges(anon.2005b),andforaroundhalfoftotaltradinginthemainmarketinwhichthefundwestudiedoperates,emerging-marketgovernmentbonds(anon.2005a).By2007,asinglefund,Citadel,wassaid‘toaccountformorethan5percentofalldailysharetradingvolumeontheNYSEandTokyoStockExchanges’(Gangahar2007)and‘formorethan10percentoftradinginthemostliquidTreasuries’:thesovereignbondsoftheUnitedStates(BealesandTett2007).Researchaccesstohedgefundsishard.Thesectorisadiscreetone,partlythroughnecessity(thenon-solicitationrequirement)andpartlythroughchoice,withmanyhedge-fundmanagerstraditionallyshunningpersonalpub-licity,especiallypublishedphotographs.Asnoted,theindustryishighlycon-centratedgeographically.AlthoughNewYorkhasalwaysdominatedintermsofassetsundermanagement(andstilldoes),London—byfartheleading AssemblinganEconomicActor/41hedge-fundsiteinEurope—hasbeengrowingfast.In2002,28oftheworld’s50largesthedgefundswerebasedinNewYorkandonly3inLondon.ByJanuary2007,however,NewYork’sshareofthetop50hadfallento18,whileLondon’shadgrownto12(Willman2007).Thefundtowhichwe(HardieandMacKenzie)gainedaccesswasbasedphysicallyinLondon,althoughitsprimaryregistration,likethatofmanyotherhedgefunds,wasintheCaymanIslands.Itwasofroughlyaveragesizeintermsofassetsmanaged.Thecategoryintowhichitfellatthetimeofourobservationsin2005—$25–100million—coveredapproximatelyathirdofallhedgefunds,withslightlylessthanathirdbeinglargerandslightlymorethanathirdsmaller.4Thefundconsistedinearly2005ofthetwopeople,whomwecallpartnersAandB,whosetitup;a‘strategist’(partnerC,atrainedeconomist);anoperatingofficer(partnerD,whothoughbasedinthetradingroomisresponsibleforaspectsofwhatinfinancialmarketsiscalled‘back-office’work);andatrader’sassistant.(Aninternwasalsopresentononeofthedaysofourobservations.)Inagroupassmallasfive,thepresenceofevenasingleresearcherisintrusive,sowefeltitunreasonabletoaskforprolongedaccess,andourobservationsofthefund’stradingarerestrictedtothefirstweekofJanuary2005.5BecauseMonday,3Januarywasamarketholiday,thoseobservationscoverfourdays.Hardiewaspresentthroughout;MacKenzie(becauseofothercommitments)forpartof4Januaryandallof5January.Mostly,wesimplytooknotes,butthefundallowedustotape-recordthe‘strategy’meetingsitholdsat9.00a.m.eachmorning,andaswebegantodevelopasenseofwhichtrading-roomverbalinteractionswereinterestinganalytically,wesoughtandweregrantedpermissiontotape-recordthosetoo.Whilenotingasbestwecouldwhatothersdidandsaid,weconcentratedourobservationsonpartnerA,thetrader.(AlthoughpartnerBsometimestrades,partnerAwasresponsibleforallthetradingduringourobservations,andreferencesbelowto‘thetrader’arealwaystopartnerA.)Heallowedustositbehindhim,slightlytooneside;wecouldobserveallhisactionsandallthevisibleobjectsofhisattention(seeFigure3.1).Wecouldhearhispartinalltelephoneconversations,andwewerealsoabletolistentotelephoneconferencecalls.Weoccasionallyaskedthetradertoexplainactionshehadjusttaken,tryingtotimesuchenquiriessoasnottodisturbtheflowofhisactions.(Heath,Jirotka,Luff,andHindmarsh1993outlinethecuesparticipantsindealingroomsusetoavoiddisruptingothers’actionsequences.) 42/AssemblinganEconomicActorpartnerC‘strategist’sometimestrader’sinternassistantcomputerscreensHardiepartnerApartnerB‘trader’MacKenziepartnerD‘operatingofficer’Figure3.1.Layoutofthehedge-fundtradingroomThetraderseemedremarkablyunperturbedbythiscloseobservation,butplainlyourpresencecouldhaveaffectedwhatheorhiscolleaguesdid.How-ever,itisworthnotingthatthefocusofourobservationswasbehaviourthatiscentraltoeffectivetrading.Investorsinhedgefundsoftenjudgethemontheirperformancemonthtomonth,especiallyinthecaseofarelativelynewfund(asthisonewas),andinthatcontextfourtradingdaysareconsequential.Todepartfromsuccessfulroutineswouldhavehadahighcost,andtheimpres-sionwehaveisthatthetraderandhiscolleaguesdidnotallowourpresencetodisturbtheiractionsinthisrespect. AssemblinganEconomicActor/43Wesupplementedthoseobservationswithfollow-upinterviewswithpart-nersA,B,C,andD,andwithseveralinformalmeetingsbetweenHardieandpartnerA.PartnerAalsopermittedustoforwardtoourselvesnearlyalltheelectronicmailmessageshereceivedandsentduringtheperiodofourobservation.(Noselectivityonhispartwasinvolved:wesimplyranoutoftimetoforwardthecompletesetofmessages.)Printedout,thesee-mailsfilleightlever-archfiles.Ourobservationsareofcourseverybrief,andanyonehedgefundislikelytohaveidiosyncraticfeatures,sotoplaceourstudyofthisparticularfundincon-textwehavealsodrawnuponawidersnowballsample(sofarinvolvingfifty-oneinterviewees)oftradersinotherhedgefundsandininvestmentbanks,ofthosewhomanagesuchtradersandprovidethemwithotherservices,andofthe‘fundsoffunds’thatarenowthedominantcategoryofinvestorinhedgefunds.Whenaninterviewisquotedwithoutattributiontoourfund’straderoroneofhiscolleagues,thequotationcomesfromthiswidersetofinterviews.TheArrangementofTradingSinceanagencementisanarrangement,letusbeginwithlayout.InJanuary2005,thehedgefundwestudiedleasedtwomodestroomsinasharedofficebuild-ing.Oneroomwasusedforthe9.00a.m.andothermeetingsandforsomeconferencecalls.Theother,wheremostoftheactionweobservedtookplace,couldhavebeenmistakenfornormalaccommodationforclericalworkers,exceptfortwofeatures(seeFigure3.1).First,ratherthanbeingdistributedforprivacy,desksformedasinglerectangleinthemiddleoftheroom,andtheoccupantsofchairsallnormallyfacedinwards.Second,thereweremorecomputerscreensthanoccupantsoftheroom:infrontofthetrader,forexample,werefourscreens.Theplethoraofscreensinterferedsomewhatwithlinesofsightwhenseated—partnerBwouldoftenstandtotalktothetrader—butthecentripetallayoutoftheroomsuggestsadesiretofacilitatecommunicationandmutualvisibility.Thefundspecializesin‘emergingmarkets’:countriessuchasTurkey,Lebanon,thePhilippines,SouthAfrica,Russia,Hungary,andthenationsofLatinAmericathatareoutwiththeheartlandsoftheglobalfinancialsystembutneverthelesshavesignificantcapitalmarkets.(Sometimescountriessuch 44/AssemblinganEconomicActorasIceland,whicharedevelopedbutontheperipheryofthemetropolitanheartlands,arealsoconsideredasemergingmarkets,althoughnotbythishedgefund.)Thegovernmentsofallthecountriesinwhichthefundspecializesissuebondsintheirownorforeigncurrency.Bondsaretradeabledebtsecuritiesthattypicallycommittheirissuertorepaythecapitalsum(theprincipal)onagiven‘maturity’dateandtopay‘coupons’(periodic,normallyfixed,interestpayments)untilthatdate.Theyarethemainmeansbywhichbothdevelopedandemerging-marketgovernmentsbridgetheshortfallbetweenrevenuesandexpenditures,ashortfalltheynearlyallencounteralmostcontinuously.Thecapacityforsuccessfulbondissuanceenhancesagovernment’sfreedomofaction:moneycanbespentnow—oninfrastructure,education,health,warfighting,andsoon—andrepaidonlyinthefuture,andgovernmentsfre-quentlypaytheprincipalonexistingbondsthathavereachedtheirmaturitybyissuingnewbonds.Thepricesandcouponratesatwhichinvestorsarepreparedtobuybondshaveadirecteffectonagovernment’sdebt-servicecosts,andthusonitsbudgetbalanceandultimatelyonthepolicychoicesopentoit(Hardie2007).Thefundwestudiedalsotradescurrencies,butthecoreofitstradingisofbondsandbond-derivativessuchasbondfutures(a‘future’isastandardizedexchange-tradedcontractwhichisclosetoequivalenteconomicallytoacom-mitmenttoafuturepurchaseorsaleoftheassetinquestionatasetprice)andcreditdefaultswaps(which,asnoted,arecontractsroughlyanalogoustoinsuranceagainstabondissuerdefaulting).Thebondsthefundtradesareidentifiedbycountryandmaturitydate(andsometimesalsobycouponrate).Whenthetradertelephonesasalespersonataninvestmentbankandasks,‘CanyougetmealevelonBrazil14s—onefour?’,whatarebeingsoughtare‘bid’(purchase)and‘offer’(sale)pricequotationsfortheBraziliangovernmentUSdollarbondsmaturingin2014.(‘Onefour’isawiseprecautionbecausetheBraziliangovernmentbondsmaturingin2040arealsoactivelytraded,andamistakebetweenthetwowouldbeserious.)Thefurtherqualification—‘Canyouaskhim[thebank’strader]toshowmeabidfor[$]5[million]?’—indicatessomethingofthescaleonwhichthefundtrades.Ifthepricequota-tionisattractive,afewfurtherwordsonthetelephoneandabriefe-mailorBloombergmessageconfirmthedeal.Suchtransactionscanbeconductedwithapparenteaseandinformalitybecausetheymobilizeentitiesnotallofwhicharerenderedevidentsimplyby AssemblinganEconomicActor/45observingandlisteningtothetrader.Hereweencounteranaspectofagencementthatthesocialstudiesoffinance(withtheexceptionofLépinay2004)hassofarlargelyignored:the‘back-office’infrastructureoftrading.Whenthetraderhasstruckadeal,hewritesdownitsparametersonpaperona‘tradeblotter’inafolderthatliesonthedeskbetweenhimandpartnerD.Ononeofherscreensistheelectronic‘blotter’ofatrade-captureandportfolio-managementsystemthefundleases.Likeothersuchsystems,itcontains‘securitymasters’,acces-sibleon-screenviapull-downmenus,whichcontainautomaticallyupdatedelectroniccharacterizationsofallthesecuritiestheirusersarelikelytotrade.(AnintervieweeatafirmthatprovidestechnicalsystemstohedgefundstoldMacKenziethatasofJanuary2006hisfirm’ssystemcontained220,000securitymasters,updateddailyascouponsorprincipalarepaid,andsoon.)AsmallbutcriticalpartofpartnerD’sroleinensuringthefund’ssmoothoperationistousethemenustocalluptheappropriatesecuritymasterandenterintotheelectronicblotterwhetherthetransactionwasasaleorapurchase,thequantityandprice,andsomeotherdetailssuchastheidentityofthecounterparty.PartnerD’sworkaside,thefund’s‘backoffice’isnotphysicallypresent.Thetrade-capturesystemtransmitstherecordofthefund’stradingtoits‘administrator’,whichisaseparatefirm,therelevantofficeofwhichisinDublin.Amongsttheservicestheadminstratorprovidesis‘reconciliation’:ensuringthatthefundanditscounterpartyhaveindeedmadethesametrade.Inconsistencies—‘breaks’astheyarecalled—arecommonintheworldoftrading.Sometimesthepartiestoatrademanuallyenterdetailsthatdonotmatch;sometimestheirtwosecuritymasters,supposedlycharacterizingthesamesecurity,infactdiffer.(Ofthe1,300employeesofonefirmthatprovideshigh-techadministrationservices,680arebasedinMumbai,workingthroughtheLondonandNewYorknights,manyofthemidentifyingand,asfaraspos-sible,resolvingbreaks.)Crucially,too,theadministrator’sstaffandtechnicalsystemsemploythetrade-capturedatatocheckthetrader’sandhisassistant’scalculationsofthechangingdailyvaluesofthefund’sassets,whicharecriticalfiguresbecausetheydetermineperformance.Thefundisalsolinkedelectronicallytoits‘primebroker’,aleadinginter-nationalinvestmentbank.Whenthefundagreesatrade,theprimebrokermakesthenecessarytransfersoftheelectronictracesofmoneyoroftitletosecurities.(Abond,forexample,isnowalmostneverapapercertificate:itisaniteminanelectronicdatabase.)Thebankcommitsitselftomakethese 46/AssemblinganEconomicActortransfersevenifthefundisunabletopayforthem,thusfacilitatingthefund’stradinginanadditionalsense:thefund’scounterpartiesknowthatnotjustitscreditworthiness,butthebank’s,standsbehindthetrades.Datatransfersfromthefund’stechnicalsystemstothebank’smakeitpossibleforthelattertomonitorthefund’srisk-takinganditscashflow.Attheendofeverytradingday,thebank’ssystem‘sweeps’thedozensoftradingpositionsthatmakeupthefund’saccount,andplacesexcesscashonovernightinterest-bearingdeposit.Whenthefundisshort(hassoldsecuritiesitdoesnotown,thecapacitytodowhichis,asnotedabove,almostadefiningfeatureofahedgefundasaneconomicactor),thebankwilltryhardtolendittherequisitesecurities,eitherfromitsowninventoryorelsewhere,eveniftheyare‘hard-to-borrow’.DistributingCognitionThefund’scapacitytoenacttradesthusdependsonpeopleandtechnicalsystemsnotphysicallypresentinitstradingroom.Sotooitscapacitytoknowwhichtradestoenact.Thefunddealsinthecurrenciesandgovernmentsecuritiesoffar-flungcountrieswithcomplexeconomiesandintricatepolitics,securitiesthatareentitlementstopaymentsthataresometimesfarinthefuture.Who,forexample,canconfidentlyknowwhetherthefiscalsituationofacountrythathasissuedathirty-yearbondwillbegoodenoughforittorepaytheprincipalwhenitfinallyfallsdueorwhetheritsgovernmentthenwillbewillingtomakethepayment?Thedecisionstobemadearedifficultones,andtheamountofpoten-tiallyrelevantinformationisvast.Muchofitisavailableinthetradingroomdirectlyon-screen.Sittingattheirdesks,theirattentionontheirscreensinthemannerdescribedbyKnorrCetinaandBruegger(2002a),thetraderandhiscolleaguesalmostcontinuallysiftthisincomingflowofinformation.Itcomesfrommanycountries.Somepartsofitarequantitative:aboveall,dataonpricemovementsinthemanymarketsinwhichthefundtrades.Otherpartsarequalitative.Ononeofthetrader’sscreens,forexample,weretitlesofReutersnewsstories.Ifhechooses(heseldomdoes,forreasonssuggestedbelow),thetradercouldclickhismousetoopenupReuters’accountofconditionsinthericemarketinManila,orlearnthat‘afterlonglanguorEgyptianpoliticswakesup’. AssemblinganEconomicActor/47Fivepeople(theinternpresentononedayplayednoactivepart)thuscon-frontedmultipleappresentationsofmarketsandinnumerablerepresentationsofeventsinmuchoftheglobe.Ifhumanbeingshadunlimitedpowersofinformationprocessing,calculation,andmemory,asingleunaidedhumancouldperhapsturntheinformationflowingintotheroomintoanoptimaltradingportfolio.Sincehumancapacitiesarelimited,asnotedinChapter2andasHerbertSimonemphasizedlongago(Simon1955),thenecessarytasksaredistributedacrosstechnicalsystemsandmultiplehumanbeings:whatgoesoninthetradingroomisindeed‘distributedcognition’inHutchins’ssense(1995a;1995b).Forexample,atechnicaltoolonwhichallbondtradersdependisayieldcalculator.Thecomplicateddiversityofbondpricesfrequentlyneedsconvert-ingquicklytoandfromamoreuniformmetric.Thinkingintermsofyieldsenablesdifferentbondstobecompared,andindeeditiscommonforbondpricestobequoted,orbond-auctionbidstobepriced,notassumsindollarsorothercurrencies,butasyields.Intoday’sfinancialmarkets,thecalculationofyieldshasbecomearoutine,‘black-box’softwarefeature.However,anincidentontheseconddayofourobservationssuddenlyrendereditvisible.Thetraderaskedhisassistanttoproduceasoftware-implementedcalcula-tortoenablepricequotationsforTurkishgovernmentbondsintheformofyieldstobeconvertedtoandfromliraprices.Theassistantdidso,employingthestandarddefinitionof‘yield’astheaverageannualrateofreturnofferedbyabondoveritsentireremaininglifespanatitscurrentmarketprice,whichiscalculatedbyfinding,byiteration,thediscountrateatwhichthesumofthepresentvaluesofthebond’scouponsandprincipalequalsitsmarketprice.Thetrader,however,quicklysawthathisassistant’scalculatorwaswrong.Theassistantdidnothaveacrucialpieceof‘localknowledge’—theconventionintheTurkishbondmarketistoemploynotthestandarddefinitionofyieldbuttheannualcouponpaymentsexpressedasapercentageofmarketprice:Trader:TurkishT-billsworkonasimpleyieldandnotacompoundyield.Didyouknowthat?Assistant:No.Oncecorrected,theTurkishyieldcalculatorbecomespartofthesociotech-nicalagencementthatconstitutesthehedgefund.Thecalculator’sconstructionisitselfheterogeneous.Itmixesprogrammingexpertise,knowledgeofmarket 48/AssemblinganEconomicActorconvention,andspecializedfactualknowledge:thecouponratesofTurkishgovernmentbondswithspecificmaturities.Twopeopleproduceit:thetradercouldinprinciplehavewrittenithimself,butinpracticeheneedstodelegatethetask,andalsotoassesswhetherithasbeencarriedouttohissatisfaction.Intheseaspects,theTurkishyieldcalculatorisunusualonlyinthatwewerepresentasitwasbeingconstructed.WhatHutchins(1995a:374)saysofnav-igationistruealsoofourhedgefund:‘Thesettingof...workevolvesovertimeaspartialsolutionstofrequentlyencounteredproblemsarecrystallizedandsavedinthematerialandconceptualtoolsofthetradeandinthesocialorganizationofthework.’The‘socialorganizationofthework’,inthesenseofthedistributionofcognitionandactionacrossthepeopleintheroom,wasevidentduringourobservationsinmanyways.Forexample,thetraderfrequentlyaskscolleaguesquestionsregardinginformationhehas(possiblytemporarily)forgotten:‘AtwhatpricedidIdothattrade?’or‘Whatwasthe[US]unemploymentratelastmonth?’(AsHutchins1995a:134putsit:‘rememberingisjointlyundertaken’.)Whenthetraderisoutoftheofficeduringofficehours(whichtypicallyhappensonlybriefly),hereliesonhiscolleaguestoobservemarketactivity.Whenhereturns,hisfirstwordsoftenare‘What’shappening?’or‘Howisthemarket?’PartnerCalsofrequentlytakestheinitiativeinorientingthetrader’satten-tiontoforthcomingdatareleases,andpartnerBoftenpointshimverydirectlytorelevantmarketdevelopments:‘Hey...youcanputthetradeonagainat110’,or‘Wow.Phil[Philippinegovernmentbonds]istradingdown.Don’tyouseethesemessages?’Implicitinpointersofthelatterkindisoftenaviewofanappropriatetrade.Ifthatviewconflictswiththetrader’s,abriefdiscussionwilloftentakeplace:PartnerB:...haveyouseentheZAR[SouthAfricanrand]?Trader:Yes,it’sgoingmyway.Whatisyourproblem?Doyouwantmetotakeitoffnow?Onotheroccasions,however,thetraderwilldonomorethanacknowledgethecomment(‘Yeah,Isawit’),orwillnotreplyatall.Althoughthetraderhasevidentconfidenceinhisviews,heacknowledgesthatothershaveexpertisethathedoesnot:inparticularthe‘strategist’,part-nerC.Thelatter’sroleistofolloweconomicandpoliticaldevelopmentsin AssemblinganEconomicActor/49emerging-marketcountriesbutalso(forreasonsweexplorebelow)economicdevelopmentsinothercountries:mainly,butnotexclusively,theUnitedStates.Thefollowingexchange,forexample,tookplaceafterthemonthlyreleaseoftheUSemploymentfigures,thedataeventduringtheperiodofourobservationstowhichbyfarthemostattentionwasdevoted.ThetraderandpartnerCarelookingatthesameinformationscreensduringtheexchange,andthetraderissimultaneouslytryingtocompletethepurchaseofsomeBrazilianbonds(thebreaksinthetextaremainlywhenheistalkingonthetelephoneaboutthis):PartnerC:Christmassaleshavebeenkindofsluggish.Byallaccountstherewasalotofdiscountingandgoingoutandorderingnewmer-chandiseandalsothisemploymentreport,thereasonwhyitisbelowexpectationswasbecauseretailjobscutby20,000.Trader:...Sotheeconomyisweak,yeah...PartnerC:Sotheretailersarehavingatoughtime.They’renothiringliketheyusuallydoinDecemberandareprobablydiscounting.Trader:...Sothenumber.Basedonthisnumber,whatdoyouthinkTreasuries[USgovernmentbonds]shoulddo,overall?Ifyouhadtocloseeverythingelseoffand,basedonthisnumber,justonthesenumbers,what?PartnerC:JustTreasuries?Trader:Yeah,whatwouldyousayTreasuriesshoulddo?Up,down,unchanged.PartnerC:Inaword,Iwouldsayunchanged.Trader:Right,thankyou.Butdon’tyouthinkthatmarketprofessionalswilllookthroughthenumbersandtheywillimplysomethingforinflation?Orthat’snotgoingtohappen?PartnerC:YouaskedforonewordsayIgaveyouthe,yeah,I’llembellishabitmore.Yeah,it’sasIsaid.ItappearsthatChristmaswaskindof,overallitwasokayforretailers.Itwasokaybecausetheywerediscountingtomovethemerchandise.Sothat’spositiveforinflation,forDecemberanyway...Butyouhavetoweighthatupagainstthefactthattheaveragehourlyearningswasabithigherthanexpected.Trader:Right.PartnerC:So? 50/AssemblinganEconomicActorTrader:That’swhatImeantaboutthe,readingthroughthose...PartnerC:Yeah,that’swhyIsayunchangedbecausetherearethesecross-currentsgoingthrough.ThatwasanexchangeaboutthebearingofeconomicconditionsontheUSbondmarket.OtherexchangesbetweenthetraderandpartnersBandCdebatespecificfeaturesoftrades.Theyoftenbeginquitecasuallyandmovegraduallytowardsacollectivedecision:Trader:Shouldwedo,ImeanI’dliketodothetrade,theTaiwandollartradeversusthe[US]dollar.Youdon’tthink...PartnerB:He’ll[partnerCwill]tellyoutodoitversustheeuro.Trader:Butthisisabigchange.Imeanyouknow,youcan’tchangelikethat,likeovernight.Let’sthinkaboutit.Maybewechangeit,okay,but...PartnerB:Changewhat?Trader:Imean,inaway,it’satradethatitsaysgolongthedollarandshorttheeuro,right?Imean...thistrade,ifyoudon’tthinkthedollar/euroisgoingtowards,let’ssay,inat120[anexchangerateof1euro=$1.20].Ifyouthoughtthedollar/eurowasgoingto135,youwouldn’tproposethistrade.PartnerC:Well,Iwouldinthesensethat,there’sthreescenarios,twoofwhich,thisTaiwanthingwillwork...inEurope.OneisthatUSdoestherightthing.Trader:Right.PartnerC:Andtheeuro...moreagainsttheeuro.Trader:Myproposal,whichI’vemadeontheMexicanpesoaswell,isthatwedothesetradesagainstabasketofdollarsandeuros,atthispoint,ratherthanjustgoallshorteuros.PartnerC:Yeah,ImeanI’mnotproposingthatI’vechangedmyviewonthedollargenerally,just,IwasthinkingjustthroughtheTaiwanandtheAsian,Taiwanbutallthe,it’sonewayortheother,ifyoubelievethatstoryisgoingtohappenthisyear...PartnerB:Ibelieveit.PartnerC:AndIdotoo.PartnerB:Buttheissuewe’rediscussingisnotwhethertobuytheAsia.Whatwe’rediscussingiswhattoshortagainstit.... AssemblinganEconomicActor/51PartnerC:...ifyou’reambiguousabout,ifyou’realittlebitambiguousaboutdollar’sdirectionfortheyear,thenitwouldbe,Ithink,stillakindofawin/winagainstTaiwanversustheeuro...Cognitionisseldomentirelyseparatefromemotion(Damasio1995).Afterdiscussionssuchastheabovehavestabilizedaninterpretationandgeneratedadecision,orevenwhenthetraderhastakenadecisionwithoutconsultinghiscolleagues,theyfrequentlyprovidehimwithemotionalsupport.Hisworkisstressful,involvingactionsinwhichlargeamountsofmoney(hisownandhiscolleagues’,aswellasthefund’sinvestors’)areatstake.Supportfordeci-sionsthathavealreadybeentakenwasoftenrestatedexplicitly:‘Ireallylikethattrade’or‘Yes,Iwouldbeprettycomfortablewiththat.’Ifpricesdonotmoveaspredicted,colleagues’commentsbothsupportthetraderandencour-agehimtomaintainfocus:‘Yeah,don’tletitaffectyou’;‘You’regoingtomakenomoneythinkingaboutit.Justforgetaboutitandmoveon.’Sometimestheencouragementtodothislastisveryexplicit:‘Whatelsewouldyoubuy?Whatelseistheretosell?’Multi-SiteCognitionThecognitiveprocessesthatinformthefund’sdecision-makingaredistrib-utedmorewidelythanoverthepeopleandtechnicalsystemsinitstradingroom.Thenecessarysiftingofpotentiallyrelevantinformationisalsocon-ductedelsewhere,oftenindifferentcountriesorcontinents.Sometimes,theresultsofthissiftingarriveviatelephonecallsorviatelephoneorweb-castteleconferencesorganizedbyinvestmentbanks.Mostcommonly,however,theresultsofothers’siftingarrivesintheformofelectronicmailmessages.Intothiscategoryfallthevastmajorityofthosee-mailmessagesreceivedbythetraderduringourfourdaysofobservationsthatdidnothaveaspecificpurposesuchastoconfirmadealortogiveapricequotation—andevenmessagesgivingpricequotationsoftenalsocontainabriefcommentaryonmarketdevelopments.Usually,thesee-mailsarenotthebilateralmessagesonwhichKnorrCetinaandBruegger(2002a)focus,butmessagestomultiplerecipients.Asthetraderputit:‘Inawaythee-mailsthatyougetarelikebeing...inanareawhere,youknow,therearetwentydifferentpeoplesharinginformation.’Asampleofthe 52/AssemblinganEconomicActore-mailsfollows(itemsmarkedwithanasteriskarethetitlesofelectronicallyattachedpagesfromservicessuchasBloombergNews):(sender1)4.1.05,14:58:‘,’∗Atlast...(sender2)5.1.05,00:30:Today’shighlightsBrazil:Intermsofdatareleases,watchtodayfortheDecemberC[onsumer]P[rice]I[ndex]-Fipe([Sender’sbank’sprediction]:0.6%)andfx[foreignexchange]flowsforDecember.Mexico:Thepesonearlyreachedour11.45recommendationtargetandweadvocateclosinglongUSD/MXDpositionswhenthepesogetsclosertothatlevel.(sender3)5.1.05,02:34:ROP[RepublicofthePhilippines]flyingdespiteEM[emergingmarket]selloffandrates....!!!!technicals...(sender4)5.1.05,07:19:[Philippines’governmentbonds]holdinginverywellvs-restofemg[emergingmarket]spreadstighterby8–10[basispoints](sender3)5.1.05,07:38:‘Philippine10-YearDollarBondsRiseonNarrowerBudg’∗—marketonfiredespiteovernightactioninLatam[LatinAmerican]credit.IfthisstoryisthesoledriverthemarketparticipantsaremuchmorenaivethanevenIgavethemcreditfor....after11monthsbudgetdeficitwasat160,anannualisedamount175bnpesos....ofcoursetheannualdeficitwaslikelytocomeinatsomewherebetween170–190...thisstoryissurelynosurprise.(sender5)5.1.05,08:22:StillthinkPhillisellsoffmoreasitisonlydown1/4point(sender3)5.1.05,08:24:[Philippines’governmentbonds]Justonfire.(sender6)5.1.05,11:26:,27’S,40,34[bondmaturities]Theseincomingelectronicmailmessagesgenerallycontaininformationalreadyavailabletothetraderviathescreensinfrontofhim.HehasaccesstoReuters,BloombergNews,andotherservices.HecaneasilyfindoutwhentheleveloftheConsumerPriceIndexforBrazilisduetobeannounced,theexchangerateoftheMexicanpesoagainstthedollar,thepricesofthebondsofthegovernmentsofBrazilortheRepublicofthePhilippines,andtheextentoftheirreportedbudgetdeficits.Theseelectronicmailmessagesthusgenerallyservetodrawthetrader’sattentiontosomeofthedataitemsavailabletohim,andnottoothersofthoseitems,andoftenexplicitlyorimplicitlysuggest‘framings’:waysofinterpretingdataitems(seeBeunzaandGarud2004).Forthetraderandhiscolleaguesto AssemblinganEconomicActor/53monitorallavailabledataitemswouldbeinfeasible.Theconstantarrivalof‘pointers’reducestheneedtoattempttodoso,andsometimesfeedsdirectlyintoaction.Afteraflurryofe-mailonthemorningof5January(includingthemessagesfromsenders3,4,and5quotedabove),at8.30thetraderconcludesthatthepricesofthegovernmentbondsoftheRepublicofthePhilippinesareabouttofall,andshortsells$5millionofsuchbonds(denominatedinUSdollars)tosender3’sbank,e-mailingtohiscounterparty(whoislocatedinHongKong):‘Youhaven’tmoved[yourprices].Londonwillsellit.’Clearly,themessagesquotedabovediffer.Sender1isdrawingattentiontoanewsitem,andofferingapersonalopinion.Sender2drawsattentiontoaforthcomingdataannouncement(manyincomingmessagesdothis),and,inrespecttoMexico,offershisinvestmentbank’sexplicittradingadvice.Themessagesfromsenders3,4,5,and6taketheformofreportsonoranalysesofmarketdevelopments:noadviceisexplicit,butonlyabraverecipientoronewithalong-termorientationwouldreceivesender6’smessageandpromptlybuyBraziliangovernmentbonds.Thesourcesofmessagessuchastheseareofcourseeconomicactorsintheirownright,whocanbepresumedoftentohaveaninterestinwhatotherswilldowhentheyreceivethem.Itisineffectexpectedthatmarketparticipantswill‘talktheirbook’whencirculatingideasfortrading—thattheywillalreadyholdaposition,thevirtuesofwhichtheyarepropounding—andacertainamountof‘gilding’orexaggerationinsodoingisdiscounted(lesspardonableiswhatBiggs2006referstoas‘sandbagging’:advocatingapositionwhileoneselfunwindingit).Itisimportant,however,thatmanyoftheexchangeswearediscussing(suchasthee-mailsaboutPhilippinebonds)aremultilateral,sothatopportunismthatleadstoanidiosyncraticviewpointmaybedetectable.Asthetradersays,‘it’sfairlyobvious...becauseyoucanseeifonepersonissayingA,andeverybodyelseissayingtheopposite’.Furthermore,somesourcesofideasorsiftingaremorecredibleandmoreauthoritativethanothers.Asthetraderputsit:‘somepeoplearemoreinformed;...somepeoplearemorethoughtful,sophisticated;somepeoplearesimpler,youknowtheyhavesortofbasedtheirdecisionsonhunchesandsoforth...sothere’sadifferenceinstyle,andIdon’tliketopigeonholethisguyisalwaysright,thisguyisalwayswrong...butofcoursetheyhave,youhavesomesortofabiaswhether,youknow,howtheythink,andtheirstyle...andyoufactorthatintoyourdecision.’PartnerBlikewisenotes:‘asyoutalktoallthosepeopledayafterdayafterday,youkindofdevelopafeelforwhohasthe 54/AssemblinganEconomicActorrightmindsetandwhodoesn’t...youwillfeelthat,yeah,thisguyhasagoodcallonthemarketandthatguynotsomuch.’SelectivityinInformationSiftingFullytocharacterizethesiftingofinformationthatgoesonwithinthefund’stradingroomandinotherlocationsconnectedelectronicallytoitisataskbeyondthischapter.Here,wediscussonlytwoaspects.Thefirstisgeograph-ical:thesurprisingextenttowhichtheattentionofourfund,forexampleinthe9.00a.m.strategymeetings,wasactuallydirectedtotheUSA.Asalreadyseen,althoughthefundtradesthebondsofcountriessuchasBrazilandthePhilippinesitneverthelesspaiddetailedattentiontomatterssuchas,forexam-ple,thepre-ChristmasretailmarketintheUSA.Ourfollow-upinterviewsconfirmedthatthatwasnotunusual.PartnerCreckonedthattheweightgiventointernationalfactors—‘usuallytheUS,really’—inthefund’sdeci-sionswasaround30–40percent,withconsiderationsspecifictotheemerging-marketcountryinquestionaccountingforaround60–70percent.Themetricof‘yield’allowsthevastrangeofbondsissuedworldwidequicklyandeasilytobecompared.Forexample,theyieldofthedollar-denominatedbondsissuedbyBrazilorthePhilippinescanbecomparedwiththeyieldofsimilarUSTreasurybonds,andtheperceivedprobabilitiesofdefaultbyBrazilorbythePhilippinesarecondensedinto‘creditspreads’oftheyieldsoftheirbondsoverTreasuries.Indeed,apricequotationforanemerging-marketbondwilloftentaketheformofaspreadofitsyieldovercomparableTreasuries.Thevaluationofemerging-marketbondsasspreadsoverUSoreurogov-ernmentbondsmeansthat,ceterisparibus,thepriceoftheformerwillmoveinlinewithmovementsinthelatter.However,ourfund‘hedgedout’thisdirectconnection,forexamplebytakingoffsettingpositionsinUSandeurobondfutures.Alessdirectlinkneverthelessremains.IfdomesticinvestmentsintheUSAearnonlylowyields,emerging-marketbonds(withtheaddi-tional‘spreads’theyoffer)seemtobecomemoreattractive.Amongsttheconsequencescanbeanimprovementintheperceivedcreditworthinessofemerging-marketgovernments:moreattractivebondsmeanlowereddebt-servicecostsandthusimprovedbudgetbalances,andanincreasedpossibilityofsellingbondswithlongermaturities,whichhastheeffectofreducingthe AssemblinganEconomicActor/55risksintrinsictofrequentrefinancing.Incontrast,ifUSyieldsrise,emerging-marketbondslosesomeoftheirattractiveness,andthisvirtuouscirclecanreverse,withperceivedgovernmentcreditworthinessdecliningandcreditspreadswidening.6Inconsequence,USinterestratesandbondyieldsaffectnotjusttheoveralllevelsofemerging-marketbondyieldsbutalsothespreadofthoseyieldsoverTreasuries,afactortowhichourfundwasexposed.Muchoftheactioninemerging-marketbondsthatweobservedhadtodowiththereleaseonthenightof4–5JanuaryoftheminutesoftheDecembermeetingoftheUSFederalReserve’sinterest-rate-settingOpenMarketCommittee,whichasnotedaboveindicatedaclearlyincreasedprobabilityofinterest-raterisestocome.Despitetheprotectionofferedbythetrader’shedges,itisthusnotsurprisingthatonthemorningof5Januaryweobservedhimreadingthoseminuteswithgreatcare.Theywerethekeyinterpretativecontextfortheemerging-market‘selloff’,includingthepuzzlinginitialfailureofthebondsofthePhilippinestofallinconcert,andtheanomalyinthepricingofBrazilianbondstobediscussedinChapter5.Suchphenomenaare,ofcourse,partofthemeaningofthatmostfamil-iarofnotions:‘globalization’.Itisworthnoting,however,thatinthiscaseglobalizationactsinpartthroughanalgorithm.Withoutthemetricof‘yield’,comparingthebondsissuedbydifferentgovernments,withalltheirpartic-ularities,wouldbeslowerandmuchharder.Anagencementthatincludesayieldcalculator(oritslessmobilepredecessors,the‘yieldbooks’thatbanksusedlaboriouslytoproduce)differsfromonewithoutsucharesource(justasChapter1suggestsisthecaseforanagencementincludinganoptionpricingmodel).Inparticular,themetricof‘yield’helpstoconstructaglobalbondmarket.Asecondissueconcerningselectivityincognitionistheextenttowhichattentionispaidtothepoliticsoftheemerging-marketcountriesinwhosebondsthefundinvests.Atonepoint,wenoticedthetradercarefullyreadinganewsstoryaboutAbdaláBucaram,themaverick,populistformerPresidentofEcuador,styled(byhimself,aswellasbyhisenemies)elloco,themadman.Thisstoryhadnotarrivedviaane-mailmessage,anditpromptedustoaskthetraderhowimportantaconsiderationwasthepoliticsofthecountrieswhosebondshetraded.Hisanswerwassuccinctandgeneral:‘Theweakerthecredit,themoreimportantthepolitics.’Thehighertheprobabilityofagovernmentdefaulting 56/AssemblinganEconomicActoronitsbonds,themoresalientisinformationonthatcountry’spolitics.ThatPinochetmightfinallystandtrialwas,asfaraswecouldtell,simplypoliticalnewsaboutChile,andasourceofpersonalsatisfaction(orotherwise).ThatellocomightreturntoEcuadorfromhisexileinPanamawas,incontrast,newsofadifferentkind.ThecreditofEcuador,saidthetrader,‘isoneoftheweakestthereis’(Ecuadordefaultedonitsbondsasrecentlyas1999).Insuchacase,‘oneortwoguyscanchangethewaythingsare’.Asingletraderspendingafewminutesreadinganewsstoryisofcourseaweakdatum,butthetrader’sexplanationisconsistentwiththeextensivestudybyMosley(2003).Akeybond-marketdivideisbetweengovernmentsthatarereckonedreasonablylikelytodefault,andthosewhosedefaultisregardedaseffectivelyinconceivable.Ecuadorisinthefirstcamp.TheUKandUSAareinthesecondcamp,andChilehasmadepartial,sometimespainfulprogresstowardsjoiningthem.Bondinvestorsmonitorbothcampsinrespecttogovernmentdeficits,inflation,andinterest-ratedecisions,but‘politics’incountriesinthesecondcamptendstobeofinterestonlytotheextenttowhichitislikelytoaffectthesefactors.Only58percentofMosley’sintervieweesmentionedelectionsinsuchcountriesasafactortheytookintoaccount,andofthosewhomentionedthemalmostnine-tenthssaidtheywerenotimportant(Mosley2003:56).Incontrast,asoneintervieweetoldher:‘Politicsishugeforemergingmarkets’(Mosley2003:129).Ourtrader’sexplanationimpliedafurtherdifferentiation—forexample,thepoliticsofEcuadorisseenasdemandingattentionofadifferentkindfromthatofChile—andinhisattentiontoEcuadoreanpoliticsthetraderwasnotunique.InApril2005,Ecuadorhadtoabandonanattemptedbondissue‘becauseofrisingpoliticaltension’(Weitzman2005).ConclusionThenotionofagencementdoesnotdisplacetheclassicconcernsofeconomicsociology.Socialnetworks,forexample,plainlystillmatter(indeed,arepartofagencements),forexamplewhensecuritiesarehardtoborrow,butaprimebrokerwithgoodconnectionstocustodybankscanstilldoso.Nevertheless,agence-mentisapotentiallyusefulbroadeningofeconomicsociology’sintellectual AssemblinganEconomicActor/57resources,inparticularinitsemphasison‘technical’linkagesaswellason‘social’ones.Theriskofbroadening,however,isthatitbecomesindiscriminate.Thetaskoftracinganagencementinaninterconnectedworldisformallyendless,andthenotioncouldbecomesimplyajargonintowhichtotranslatebanaldescriptionandnarrative,ashappenedtosomedegreewhenactor-networkconceptsfirstbecamefashionableinEnglish-languagescienceandtechnol-ogystudiessometwentyyearsago.Itisthusessentialtobeselective:tofocus,forexample,onaspectsofagencementsthatarenotobviousandonwaysinwhichthecompositionandconfigurationofagencementsaffecteconomicaction.Atthemostbasiclevel,thenotionofagencementhelpfullydirectsustotheconditionsofpossibilityofeconomicactors:theoften-ignoredinfrastruc-turethatenablesthemtobetheactorstheyare.Whyaretherenowover8,000hedgefunds?Thereasonsofcourseincludeeconomicandpoliticalchanges,butitisalsoimportantthatsettingupahedgefundismucheas-ierthanitwastwentyyearsago.Thereal-timeinterconnectionoftrade-captureandothersystemsmakesitpossibletostandardize,automate,andrisk-manageadministrativeandprime-brokerageservices,whichcanthusbesuppliedonanindustrial(ratherthan‘cottageindustry’)scaleandrelativelycheaply.Itisworthnotingthatitisintheinfrastructureofeconomicaction—ratherthaninwhatChapter2calledaction’sglamorousagentialpeaks,suchastrading—thatemploymentislargelytobefound.Whileweknowofnoprecisebreakdownoffinance-sectoremploymentinthisrespect,itisclearthattradersareonlyasmallminority.Thevastbulkofjobsconcernotherrolesinagencements.Theirgenderbalanceisdifferent:tradingisstillmainlyamalepre-serve,butmorewomenaretobefoundintheinfrastructurethatunderpinsit.Thegeographicallocationoftheinfrastructuraljobsalsodiffers,atleastpoten-tially,fromthatoftheglamorousones.TheRepublicofIreland,forinstance,isnotaprominentsiteoftradingbuthasbecomeperhapstheworld’sleadingsiteofhedge-fundadministration,providingnotjust‘offshore’legalstatusandafavourabletaxregime,butalsoarobustcommunicationsinfrastructure.(‘Locatingcomputersinaplacewithhurricanes,it’sjustnot...agoodplan’,notedoneintervieweefromtheworldofadministration,explainingwhytheCaymanIslandswereunattractiveinthisrespectdespitetheirtaxadvantages.) 58/AssemblinganEconomicActorIrelandalsoofferstrainedEnglish-speakingstaffandresponsiveregulation:‘inDublin,youcangoandvisittheregulator...haveacupofcoffee’,saidthesameinterviewee.Asmoothlyfunctioninginfrastructureisnormallyinvisible:wehaddeliber-atelytoseekouttheinfrastructureofourfund’seconomicaction,ratherthanitbeingdrawntoourattentionbyeventsinthetradingroom.However,inthewiderhedge-fundworldtherearewaysinwhichapparentlyinfrastructuralissuescansuddenlyimpingeonfreedomofaction.Thusonemainmeansbywhichriskiscontrolledinthatworldis‘mark-to-market’collateralizationofcontracts.Asmarketpricesmoveinfavourofoneorotherpartytoacontract,collateralassetsaretransferredbetweenthem.Suchtransfersnowoftentakeplacedaily.Inthewordsofoneinterviewee,itislogicaltomeetsuch‘mark-to-marketcalls’bypledgingout‘themostilliquidcollateralthatyouhavethatfitsthecollateralrequirements’.Sowhatremainsinahedgefund’seasilygrasped‘box’(theunlentandunpledgedsecuritiesthatitowns)willoftentendtobeitsmostliquidassets.Undernormalcircumstancesthisisunimportant,butwhenafundsuddenlyneedscash(forexample,becauseofinvestorwith-drawalsorofmark-to-marketrequirementsthatmustbemetincash)itcanbeconsequential,especiallyifcombinedwithtechnicalsystemsthathavenotbeensetuptoincludefieldsthatallowthefundreadilytodeterminematterssuchashowquicklyassetspledgedascollateralcanbereclaimedandmadeavailableforsale.Financialassetsmayseemabstract—theyare,asnotedinChapter1,tokensofrightsandobligations,normallyinelectronicform,notobjectsthatcanbeconsumeddirectly—butwhetherornotahedgefundhasatechnicalsystemthatallowsitquicklytoascertainthestatusoftheseassetscanaffectakeyaspectofitasaneconomicagent,itsfreedomofaction:Themarketstartstogodown,nowyougottosellsomething,becauseyou’regettingcallsallovertheplaceonmark-to-market.Soyoujust,youlookandsay,‘oh,whatcanwesell?’Inaperfectworldyou’dsellabalanceofyourportfolioofliquiditiessothatyoukeepsomesenseofcontroloverthebalanceofwhat’sthere.Butifyoudon’tknowwherethatcollateralisandwhenit’scomingbackyouarejustrelyingonthefaithofpeoplethatyoupledgeitforaweekanditwillcomebackaweeklater,andyoudon’tcare’cosit’sgoingtocomeback.Inanormalmarketthatwouldbefinebutinafastmarketyou’vegottosellnow.So,yousay,‘ohshit...it’sgoingtotakemeamonthtofigurethatout[whichilliquidassetscanbesold]soI’mjustgoingtosell AssemblinganEconomicActor/59this[highlyliquid]UStwo-yearT-bill’,orwhateveritis.So...a...hedgefund...ifthey’renotpreparedforit,willbeleftwiththeirmostilliquidcollateralonly,whichisalsothestuffthattheStreetalreadyknowsyouown.Cosyoucreatedalotofattentionwhenyouboughtit...Andthenyoustartsellingit,andtheygo,‘shit,thatguyowns30percentofthatissue,webetterstartsellingittoo’.Aswellasagencementconstitutingtheconditionsofpossibilityofeconomicaction,thedistributionofcognitionandofactionitinvolvesmayshapethepropertiesofactors.Forexample,asnotedabove,aneconomicactorequippedwithayieldcalculatorisdifferentfromonewithoutanyequivalentlyeasywayofcomparingbonds.Moregenerally,asChapter2pointsout,orthodoxfinancetheoryposits(forthepurposesofmodelling)aninvestorwhoisacompletelyrationalindividualwithunlimitedcognitivecapacities,butthishasbeenchallengedby‘behaviouralfinance’.Thelatterfieldalsoviewsinvestorsasindividuals,but,asnotedinChapter2,seesthemashamperedbythesystematiccognitivebiasesrevealedbyexperimentsofthekindconductedbyKahnemanandTversky(1979)andsummarizedintheir‘prospecttheory’.Onesuchbiasisasystematictendencytobehavedifferentlyinsituationsofperceivedgain(inwhichmanysubjectsbecomeriskaverse,unwillingtotakethechanceoflosingwhattheyhavewon)andperceivedloss,inwhichthepropensityistogambletorecouptheloss.Fortraders,thetemptationisthustoavoidmakingaloss‘real’byliquidatingaloss-bearingposition,buttocontinuedoggedlytoholdit—inthejargonoftrading,tobecome‘married’toit—inthehopeitrecovers(Fenton-O’Creevy,Nicholson,Soane,andWillman2005).Hencetheapparentlycommonplacemattersofemotionalsupportandassistanceinfocusingafteratradinglossareoftheoreticalsignificance:amongsttheireffectsmaybetodiminishthe‘prospecttheory’biasreferredtoabove.Traders’cultureiscertainlyreflexivelyawareofthatbias.InthepitsoftheChicagoBoardofTrade,forexample,traderssometimeshummedMendelssohn’sweddingmarchtosignalthatacolleagueappearedtohavebecome‘married’toaposition(Zaloom2006),andourwiderinterviewscon-firmthattraders(andespeciallythosewhomanagetraders)arealerttothepossibilityof‘marriage’.Notetheformofthegeneralpointthissuggests.Thatindividualtradersareaffectedbytheircolleaguesandmanagers,thattheircultureisreflexive,andthatcognitionandactionaredistributedacrosspeopleandtechnicalsystemsmayhavetheeffectofmakingtheeconomicactormorelikethefullyrationalagentpositedbyorthodoxfinancetheory. 60/AssemblinganEconomicActorHowever,alsonotethateconomicallyrationalactionmaynotalwayspro-motestability.Anotherpossibleeffectofthecompositionandconfigurationofagencementsisontheriskofcontagion:thespreadofafinancialcrisisinonecountrytoothers,includingcountrieswithfewtradeorotherconnectionstotheoriginalsite.Withthetechnicalinfrastructuresofmodernmarketsmakingiteasyforahedgefundorothereconomicactortoinvestinmanycountriessimultaneouslyevenifithasonlymodestassets,highlyselectiveinformation-processingroutinesarelikelytobeoptimal(CalvoandMendoza2000),andwecertainlyobservedsuchselectivity.Itwould,forexample,bemostunlikelytobecost-effectiveforourfundtohireaMagyar-speakingeconomisttodeepenitsunderstandingofHungary,onemarketamongstmanyinwhichitoperates.Theremayinconsequencebesituationsinwhichtheoptimalstrategyforactorswhospreadtheirinvestmentsovermanycountriesinvolvesimitation:ifanactorobservesotheractors—especiallythosejudgedtohaveexpertunder-standingofthecountryinquestion—buyingorselling,itmaybesensibletodothesameasquicklyaspossible.Weshouldemphasizethatwedidnotwitnessourfundbehavinginthisway,butourobservations(andthecorpusofe-mailstothetrader)containampleevidenceofthecirculationofinformationaboutparticularclassesofactorbuyingandselling.Whatis,however,hardertoobserveiswhyanactorisbuyingorselling.Asaleofassetsinonecountrymay,forinstance,arisesimplybecauselosseshavebeenincurredinothercountrieswithminimaleconomiclinkstoit.Suchsalesmaynotconveyanyinformation,superiororotherwise,aboutthecountryinquestion,butmaybemisinterpretedasconveying‘badnews’(CalvoandMendoza2000).Withtechnicalsystemsfacilitatingultra-rapidreactionstosalesandpricefalls,suchprocessescancreatesurprisinginterconnections.On22February2006,forexample,apessimisticanalysisofIceland’sprospectsbythebond-ratingagencyFitchtriggeredfallsincurrenciesrangingfromtheSouthAfricanrandtotheIndonesianrupiah:theBrazilianreal,forexample,tem-porarilyfellalmost3percent(JohnsonandSimensen2006).ParticularlyyokedtogetherinFebruaryandMarch2006wasthegeographicallydiversetrioofIceland,Hungary,andNewZealand.Thecruciallinkagewasthe‘carrytrade’,ahedge-fundstaple(thoughnotastrategyemployedbyourfund).Inthis,afundborrowsinalow-interest-ratecurrency,typicallytheyen,andinvestsinthebondsorotherassetsofahigh-interest-ratecountrysuchasIceland.Asignificantdepreciationofthecurrencyofthehigh-interest-rate AssemblinganEconomicActor/61countrycancauseacarrytradetobecomeloss-bearing,andatonepointon22FebruarytheIcelandickronahadfallenby9percentfromitsdollarexchangerateon20February.Whatappearstohavehappenedon22Februaryisthat‘theemergingmarketcontagion[was]causedbyinvestorscuttingprof-itablepositionsinordertoplugtheirIcelandiclosses’(JohnsonandSimensen2006).Finally,whatofagency?Anactor-networkeconomicsociologydoesnotitselfattributeagency,butinsteadfollowsthewayinwhichsuchattributionsareshapedandchannelledbyfactorsincludingthecompositionandconfig-urationofagencements.Letussetasidethenotoriouscontroversyconcerningactor-networktheoryandtheattributionofagencytonon-humanentitiessuchasphysicalobjects,andconsideronlyitsattributiontohumanbeings.7Agencyisofcoursecommonlyattributedtoindividualssuchasthetrader,butisalsooftenattributedto‘higher-level’entities.Ourhedgefund,forexample,isalegalentity,and,asnotedinChapter2,thelawofcontractattributesagencytoit,nottotheindividualswhocompriseit.Undersomecircumstances,too,marketconfigurationscanbesuchthatagencycanseemtohaveleftparticulareconomicactorsandtoresideinamarketasawhole:forexample,inthecreditcrisisthatbeganinsummer2007,hedgefundsandothereconomicactorsoftenhadlittleornochoiceastotheircoursesofaction,andhadtosellassetsevenwhenitwasabadtimetotrytodoso.Theattributionofagencyto‘themarket’isindeedcommonwhenitisanentityinvokedinpoliticaldiscourse,bothright-andleft-wing.Moreexotically,theeconomicagencyofhumanbeingsisalsosometimesattributedtoentities‘inside’them,suchasspecificbrainstructures.Success-fullydoingsorequiresanagencementincludingspecialisttechnicalequipment,inparticularamagnetic-resonancebrainscanner,andistheterrainofthefascinatingnewfieldof‘neuroeconomics’(see,forexample,Sanfeyetal.2003).Theattributionofagencymayseemanesoteric,academicissue,butforfinancial-marketpractitionersitisinfactapervasiveconcern.Manyoftherewardstotradersandthoseimmediatelyaroundthemcomeintheformofbonusesthataresupposedtoreflectindividualcontributionstoafirm’sprof-its.Becauseagencementiscollective,this‘singularization’—theattributionofagencytospecificcomponents—isproblematic,8and,unsurprisingly,isoftentheobjectofbitterjealousyandintenseconflict.ItsrichnessasasociologicaltopichasbeendemonstratedbrilliantlybyGodechot(2004;2007). 62/AssemblinganEconomicActorTheexampleoftheattributionofagencyreflectstheoverallmeritsofthenotionofagencement.Usedinaselectiveway,itcanhelptracelinkagesthatarecrucialinconstructingeconomicagentsandframingcontemporaryeconomiclife,thusthrowingfamiliarphenomenaintonewlightanduncoveringsur-prisingconnectionsandunderpinnings.Agencementsconstitutemarkets,andinvestigatinghowtheyaremadeupisacrucialtaskforthesocialstudiesoffinanceandthematerialsociologyofmarketsmoregenerally. 4Derivatives:TheProductionofVirtualityInChapter3,HardieandInotedthatthehedgefundwestudiedsoughttohedge—reduceorcancelout—itsdirectexposuretochangesinUSdollarandeurointerestrates.Thefund’stradercoulddothisbyopeningupatradingscreenonhiscomputerandclickinghismousetobuyortosellthebondfuturestradedontheChicagoBoardofTradeoritsEuropeancounterparts.Itisasimpleoperationthattakesnomorethanafewseconds,butonethathasacomplexhistory,whichisthetopicofthischapter.Fortyyearsago,interest-rateriskcouldbehedged,ifatall,onlywithdiffi-culty.Notonlydidtradingscreenslieinthefuture:theexchangesviawhichfinancialfuturesandotherfinancialderivativessuchasoptionsareboughtandsoldtodayeitherdidnotexist,ortraded,asdidtheChicagoBoardofTrade,onlyfuturesonphysicalproductssuchasgrain.AsrecentlyasJanuary1970,noorganizedfinancial-derivativesexchangeexistedanywhereintheworld.Suchfinancial-derivativestradingastookplaceatthestartofthe1970swastinyinvolumebytoday’sstandards,andwasadhoc.Itwasconductedeitherintheintersticesand/oronthefringesofstockmarkets,orelse‘overthecounter’:bydirectnegotiation,especiallybetweenbanks.Theburgeoningsincethenoffinancial-derivativesexchangeshasbeenoneofthedefiningfeaturesofthemassivetransformationoftheworld’sfinancialmarkets.AttheendofJune2006,exchange-tradedderivativestotalling$84.4trillion(theequivalentofaround$13,000foreveryhumanbeingonearth)wereoutstandingworldwide, 64/Derivatives908070605040totalamounts($trillion)3020100199819992000200120022003200420052006yearFigure4.1.Totalamountsofexchange-tradedderivativesoutstandingatendofJuneofeachyearfrom1998to2006Source:Half-yearlystatisticsfromtheBankforInternationalSettlements(),incor-poratinglaterdataadjustments.andthetotalwasgrowingfast,havingincreasedroughlysixfoldoverthepre-viouseightyears(Figure4.1).Manyofthesecontractswillhavebeenenteredintotooffsettherisksofotherderivatives(soanunknownproportionofthetotalisthusinasenseself-cancelling),butthechangefromtheearly1970sisnonethelessstriking.Mostacademicwritingonderivativesisofcoursebyeconomists,whohavefocusedprimarilyonthepricingofderivatives(althoughthereisamore‘institutional’literatureontheeconomicsoffinancialinnovationandalsousefulworkonwhysomederivativessucceedandothersfail,bothofwhichIwilldrawonbelow).1Recently,however,theattentionpaidtoderiva-tivesbyhumangeographers,anthropologists,andsociologistshasincreasedsharply.2Acommonthemeinthismoresociologicalliteratureis‘thestrangelyimaginary...orvirtualcharacterofderivatives’(Arnoldi2004:23).Allfinancialsecuritiesare‘virtual’inthesensethattheirvalueliesnotintheirphysical Derivatives/65substanceaspapercertificatesorentriesinanelectronicdatabasebutintheclaimsonfuturestatesoftheworldthattheyembody:rightstodividendsfromacorporation,tointerestpaymentsfromagovernment,andsoon.Aderivativeofsuchasecurityisthusanentitythatderivesitsvaluefromwhatisalreadyanabstractclaim,andsothedevelopmentofderivativesmarketscanbeseenasafurtherstageoftheabstractionofmonetaryforms.Derivativesare‘money’s“newimaginary”’,notePrykeandAllen(2000).Buthoware‘abstract’or‘virtual’assetsbroughtintobeingandmadetradeable?Asthedevelopmentsincomputertechnology(aboveall,‘virtualmemory’)3thathavegivenusthemodernnotionof‘virtuality’remindus,virtualityisalwaysamaterialeffect,indeedanelaborate,sophisticated,andexpensiveone.Inthischapter,Iexplorethreeaspectsofmaterialproductionofvirtuality.Thefirstistheparallels(andalsothedissimilarities)betweenfinancialandtechnologicalinnovation.Thisthemeisintheliteratureineconomics(e.g.Silber1981),buttheviewoftechnologicalinnovationtobefoundthereistoonarrow,excludingforexampleitspoliticaldimension.Viewingderivativesasinnovationsoffersaperspectiveonhow‘economicsdoesthings’(Chapter2’sseventhprecept)andhighlightsissuessuchastheintellectualpropertyregimewithinwhichinnovationtakesplaceandtheneed,ifinnovationistobesuccessful,totakeintoaccounttheinterestsofintended‘users’andtofindaworkablecompromisebetweenincompatibleinterests.Thesecondissuetobeexploredisthe‘culturalgeographies’(Thrift2000)ofderivatives.WiththeexceptionofMaurer’sworkon‘Islamic’derivatives(Maurer2001;seealsoMaurer2005),eventhegeographicandanthropologicalliteratureonderivativesissurprisinglyhomogenizing,seemingimplicitlytopositaworldinwhich,atleastwithinitsmetropolitancore,‘place’nolongermattersgreatly.Spatialandculturallocationisstillsignificant,however,eveninthemetropolitanheartlands.Thediffering‘internal’culturesoffinancialmarketshavelefttheirstamp,ashavethedifferentwaysinwhichtherela-tionshipbetweeninvestingandgambling(onwhichseedeGoede2005)isarticulatedlegally.Therelationshipbetweenaderivativeandawagerisnotjustageneralculturalissue.Theexigenciesofkeepingderivativesseparatefromwagers(or,insomecircumstances,makingsurethattheyarewagers)havebeencriticalaspectsoftheirmaterialproduction.Theyhaveaffectedwhichderivativescanbetradedandwhichcannot,andtheextenttowhichaderivativecanbeabstractandvirtual. 66/DerivativesThethirdissuetobediscussedisfacticity.Thevirtualcharacterofaderiva-tivecontractisenhancedif,asisincreasinglythecase,itcanbesettledonlybythetransferofcash,withneitherpartyabletodemandorimposedeliveryofanunderlyingasset.Themeasureusedtodeterminetheamountstobepaidmustthereforebea‘fact’:itmustbeanacceptablerepresentationoftherealityofwhichitspeaks,andnotbesubjecttomanipulation.Ifocusonthemostimportantsetoffactsofthiskind(BritishBankers’AssociationLIBOR,LondonInterbankOfferedRate,alreadytouchedoninChapters1and2),andnotethefiercecontroversyoverLIBORthateruptedin2007–8.Thechapter’smainempiricalfocusisthedevelopmentoffinancial-derivativesexchangesintheUSAandUKsince1970andtheemergenceoftheBritishfinancialspread-bettingindustry.Thespecificityofexchangesandofspreadbettingbuildsintothestudyabiastowardsthediscoveryofhet-erogeneity:nationaldifferencesinover-the-countertradingalmostcertainlyexist,butwouldbehardertoidentify.Thatbias,however,isbalancedbythechoiceoftheUSAandUKascomparatorcountries.Theiroverallfinancialsystemshaveverysimilarcontours,andintheliteratureon‘varietiesofcapi-talism’(e.g.HallandSoskice2001)theyarenormallylumpedtogether.ThedifferencesfoundbetweenderivativestradingintheUSAandUKarethusdifferencesbetweentwootherwisesimilarcases.Greaterheterogeneitywouldprobablyhavebeenfoundhadthestudyencompassedthefinancial-derivativesexchangesnowthrivinginlocationsasdiverseasFrankfurt,Stockholm,SãoPaulo,andSingapore,orthoseemerginginmanyothercountries,suchasRussiaandthePeople’sRepublicofChina.Beforeturningtothemainbodyofthechapter,itmaybehelpfultohaveabriefintroductiontothemainorganizationscovered,therelevantchronol-ogy,andthesourcesofinformationIhavedrawnon.Theinitialmodernefforttobeginorganizedfinancial-derivativestradingwasthecurrencyfutureslaunchedbytheNewYorkInternationalCommercialExchangein1970.(A‘future’isastandardized,exchange-tradedcontractthatisequivalenteco-nomicallytoonepartycommittingitselftobuy,andtheothertosell,asetquantityofagivenassetatagivenpriceatasetfuturetime.)Thateffortfailed,butcurrencyfutureslaunchedbytheChicagoMercantileExchange’sInternationalMonetaryMarketin1972succeeded,asdidtheChicagoBoardOptionsExchange,spunoffbytheChicagoBoardofTradein1973.(Anoptiongivestheright,butunlikeafuturedoesnotimposetheobligation,tobuy—orinanalternativeformofthecontract,tosell—anassetatasetpriceon,orup Derivatives/67to,agivendate.)TheChicagoBoardofTradeitself,andanumberofotherUSexchanges,alsobegantradingfinancialderivativesinthemid-1970s.TheInternationalCommercialExchange,ChicagoMercantileExchange,andChicagoBoardofTradehadalloriginallybeenagriculturalcommoditiesexchanges.IntheUK,theeventuallymostsuccessfulfinancial-derivativesexchange—LIFFE,theLondonInternationalFinancialFuturesExchange—wasanentirelynewdevelopment,establishedin1982.In1978,theLondonStockExchangesetupaTradedOptionsMarket,whichmergedintoLIFFEin1992.In1991,theLondonFuturesandOptionsExchange,astheLondonCom-modityExchangewasthenknown,launchedpropertyandhousingderiva-tives;ittoomergedintoLIFFE,initscasein1996.(Propertyderivativesareofparticularinterestbecausehousing,commercialproperty,andlandaccountforwealthcomparableinmagnitudetothetotalityofstocksorbonds,yetthemarketinderivativesofthemhasbeensmall,atleastuntilveryrecently.4Thefailureofderivativesinthesphereofpropertythrowstheirsuccessinotherspheresintoananalyticallyinterestinglight.)FinancialspreadbettingbeganwithbetsontheFT(FinancialTimes)-30shareindexofferedbythebookmakerJoeCoral,andgainedmomentumwiththeestablishmentin1974byStuartWheelerofIG(InvestorsGold)Index.In1981,IGIndexbegantoofferspreadbetsontheFTSE(FinancialTimes-StockExchange)-100andDowJonesindexes.Anotherfirm,CityIndex,beganoffer-ingfinancialspreadbetsin1983,andotherssuchasCantorIndexhavejoinedtheindustrymorerecently.Spreadbetsarederivatives—theyarecontractsthevalueofwhichdependsuponquantitiessuchasstock-marketindexesorexchangerates,justasanindexfutureorcurrencyfuturedoes—buttheyaredeliberatelyconstructedaswagers,forreasonsdiscussedbelow.Thechapterdrawsuponfoursetsofsources.ThefirstisexistinghistoriesoftheChicagoBoardofTrade(Falloon1998),ChicagoMercantileExchange(Tamarkin1993;MelamedandTamarkin1996),andLIFFE(Kynaston1997).Thesecondisthetradepress,whichisoftenvaluableinparticularforrevealingfailedinitiativesinderivativestrading.Thethirdisasetoftwenty-sevenoral-historyinterviewsconductedbytheauthorwithpeoplecentraltothedevel-opmentoffinancialderivativesexchangesintheUSAandUKandoffinancialspreadbettingintheUK.ThefourthisafurthersetoftwelveinterviewswhichfocusedontheLondoninterbankmarket,ontheroleofbrokersinthatmarket,andonLIBOR,andwereaccompaniedbybriefobservationofbrokers’offices,bankdealingrooms,andtheprocessbywhichLIBORisconstructed.As 68/Derivativeselsewhereinthisbook,intervieweesareanonymousexceptinthecaseofthosewhoplayedthemostimportantpersonalrolesinthedevelopmentsunderdiscussion.InnovationThevastbulkoftoday’sfinancial-derivativestradingisinproductsthatdidnotexistin1970.Theseproducts,especiallythosetradedonorganizedexchanges,didnotsimply‘evolve’.Theywereinvented.Indeed,today’sfinancial-derivativesexchanges,especiallythefreshlyestablishedonessuchasLIFFE,aretheresultofconscious,deliberateprocessesofdesign.Innovationinfinanceandinphys-icaltechnologyisnotthesame—threekeydifferencesarediscussedbelow—butthecomparisonisanalyticallyproductive.Letmebeginwithsimilaritiesbetweenfinancialandtechnologicalinno-vation.Priortothenineteenthcentury,whatwenowthinkofas‘science’playedlittleroleintechnologicalinnovation,butthatrolehasnowgrownconsiderably.Sotoowithfinance.Theacademicdisciplineofeconomicshadlittleeffectonderivativestradingbefore1970,butsincethenitsrolehasbeenmajor(MacKenzie2006).AsnotedinChapter2,awidespreadunderstandingoftechnologicalinno-vationisthe‘linearmodel’inwhichscience‘discovers’truths,technologists‘apply’sciencebyworkingoutitspracticalimplications,andtheresultantproducts‘diffuse’unchangedtousers.Thoughstillinfluentialinpublicdis-cussion,thelinearmodelhasbeendiscreditedbythemodernliteratureontechnologicalinnovation(e.g.,BarnesandEdge1982;Fleck1994;SørensenandWilliams2002;OudshoornandPinch2003).Instead,thatliteraturesug-geststhatscienceandtechnologyinteractnotasdisembodiedknowledgebutasembodiedexpertise(oftenviathecirculationofpeople);thatscienceisaresourcethatengineersdrawoncreatively,ratherthansimplyapplying;thatcarefulattentiontousers’needsandto‘localpracticalknowledge’(Fleck1994)isnecessaryforsuccessfulinnovation;andthatmuchinnovation—Fleckcallsit‘innofusion’—takesplaceinwhatisconventionallyregardedas‘diffusion’.Alloftheseaspectsoftechnologicalinnovationalsocharacterizeinno-vationinderivatives:asCallonnowputsit,the‘performativityofeco-nomics’isa‘coperformance’involving‘economistsinthewild’—layaswellasprofessional—ratherthanjust‘confined’(laboratoryoruniversity)econo-mists(Callon2007).Academiceconomicshasunderpinnedderivativestrading Derivatives/69bothtechnicallyandbyprovidinglegitimacy,especiallyagainstthechargeofgambling(MacKenzie2006).However,keyinnovationsinexchange-tradedderivativeshaveinvolvedeconomistswholeftacademiatoworkinthemar-kets,suchasRichardSandor,wholefttheUniversityofCaliforniaatBerkeleyfortheChicagoBoardofTrade,andhisMercantileExchangecounterpartsFredArdittiandRickKilcollin.Theseeconomistsinthewilddidnotsimply‘apply’economics.Theyfoundthemselvesinvolvedinprocessesofinnovationthatinvolvedcloseinterac-tionwiththethreemaincategoriesofusersofderivatives:hedgers,whoareconcernedtoprotecttheirorganizationsagainstarisksuchascurrencyorinterest-ratefluctuations;speculators,whohopetoprofitbycorrectlyanti-cipatingthosefluctuations;andmarketmakers,whostandreadybothtobuyandtoselltheproductinquestion,earningthedifferencebetweenthe‘bid’andthe‘ask’(thepricesatwhichtheyarepreparedtobuyandtosell).Hedging,speculating,andmarketmakingarecategoriesofactivityratherthanofpeopleandorganizations:marketmakers,forexample,oftenhedgetheirpositionsordeliberatelytakespeculativepositions,whilesomewell-publicizedderivativesfiascoshaveresultedfromorganizationsstartingoutbyhedgingbutslippingintospeculating.Nevertheless,thecategoriesof‘hedger’and‘speculator’arepartofthe‘laysociology’thatparticipantsinderivativesexchangesdeploy,and‘marketmaker’isadesignatedrolewithspecificrespon-sibilities.Innovativeexchange-tradedderivativesneedtobeshapedinsuchawayastobeattractivetoallthreecategoriesofuser.Forexample,theInternationalCommercialExchange’scurrency-futurestradingoverlappedwiththestartofthebreak-upoftheBrettonWoodssystemoffixedexchangerates,afavourabletimebecausevolatilityencouragesderivativestradingbygivinghedgers,spec-ulators,andmarketmakersincentivestoparticipate.However,insufficienteffortseemstohavebeendevotedtodesigningandmarketingcontractsthatsatisfiedtheneedsofhedgers.‘[T]hecontractspecificationshadtobeattractivetobanktradersandcorporatetreasurers....Successfulfuturescontractsneed,ataminimum,20to25percentcommercialparticipation.Youcannothaveamarketjustforspeculators’(MelamedandTamarkin1996:174;seealsoBlack1986).Exchange-tradedderivativesarestandardizedproducts,sotheirspecifica-tionsneeddecidinginadvance.Theseincludehowbigasinglecontractistobe;the‘ticksize’(theminimumincrementinprice);thelimits(ifany)ondailypricemovesandonthesizeofpositionanyonetradercanaccumulate; 70/Derivativestherequirementsfor‘margin’(thesumsparticipantsinanexchangehavetodepositwiththeexchangeclearinghousewhentheyfirstbuyorselladerivative,andthenhavetoadjustaspricesfluctuate);theexpirationdatesofcontracts;andtheproceduresfordeliveryoftheunderlyingassetorforcashsettlement.5Successfulchoiceofthespecificationsofderivativescontractsinvolvescare-fulattentiontosometimesconflictinginterests:ofhedgersandspeculators;ofexchangemembersandexternalcustomers;andofthe‘longs’whohaveboughtaderivativeandthe‘shorts’whohavesoldit.Theseinterestsareneithereasytodetermine—extensiveresearchoftenseemstobenecessarytoelicitthem,givingcontractdesignsomethingoftheflavourofeconomicexperimentation(seeMuniesaandCallon2007)—norfixed.Indeed,amajorentrepreneurialactivityoffinancial-derivativesexchangesistopersuadebothexternalcustomersandexchangemembersthatitisintheirintereststotradeanewderivative(seeMacKenzie2006:154–5and173–4forexamplesfromthehistoryoftheChicagoMercantileExchangeandChicagoBoardOptionsExchange).Thepotentialforintereststoconflict,evenaftertheyhavebeenelicitedand‘translated’(Latour1987)inthisway,makescontractdesign—liketechno-logicaldesign(Winner1980)—aninherentlypoliticalproblem.Itisonethatcannotbesolvedsimplybyfiat(overlyfavouringtheinterestsofonegroupwillprobablybefatal,becauseotherswillthennotparticipateintrading),butrequiresbalanceandcompromise.RichardSandor,forexample,notedthatthedeliveryprocedurehedesignedfortheChicagoBoardofTrade’sfirstfinancialderivative,futuresonmortgage-backedbonds,‘iscomplicatedandcumbersome.Itappearstocausedifficultiesforboththelongsandtheshorts.Itisinthatsensefair,andmaybethereasonithasbeensuccessful’(SandorandSosin1983:267).Design,marketing,andtheencouragement(oftenviaface-to-facemeet-ings)ofparticipationintradingarepressingmatters,especiallyintheearlydaysofanewcontract,becauseexchange-tradedderivativesaresubjectto‘virtuous’and‘vicious’circlesakintothoseidentifiedintechnologicalinnovationbyArthur(1984)andDavid(1992).ThearchetypalexampleoftheeffectsofthesecirclesisthedominanceoftheQWERTYkeyboard.Itisnotdemonstrablyoptimalforelectronicword-processing—itsoriginalmotivationwastoreducethechancesoftheleversofamechanicaltypewriterstickingtogetherbyminimizingthefrequencywithwhichadjacentkeyswere Derivatives/71struckinsuccession—butQWERTYis‘lockedin’totheEnglish-speakingworld’skeyboards,anditsrivals‘lockedout’:nonehasarealisticchanceofdisplacingit.Lock-inresultsfromtheadvantagesthatsometimesflowtoanincumbenttechnologyorderivativesexchangesimplybyvirtueofbeingincumbent.QWERTY’sadvantagesarethefamiliarityofmillionsofuserswiththatkey-layandthedifficultiestheywouldfaceinthefirstfewweeksofusingadifferentlayout.Theinternalcombustionengine’sadvantagesincludethecenturyofintensiveresearchanddevelopmenteffortthathasbeendevotedtoit(andnottoitsrivals),andthehugeinfrastructureoffuelsupplyandmaintenancethatarivalwouldhavetocreateafresh.Inthecaseofderivativesexchanges,businesstendstoflowtowhereexistingvolumesoftradingarehigh,becausehighvolumesmeanliquidity(evenlargetransactionscanbeconductedquickly,easily,andwithoutalargeimpactonprice),lowtransactioncosts,andarobustmarketprice.Conversely,lowvol-umesmeanilliquidity,highcosts,andunreliableprices.Soanexchangethatgainsanestablishedpositioninaparticularderivativebecomes,likeQWERTY,hardtochallenge(Silber1981:132).LIFFE,forexample,foundthatthecur-rencyfuturesitlaunchedincompetitionwiththoseofChicagoMercantileExchangewerenotsuccessful,despiteLondon’soverallprominentroleinforeignexchange(Kynaston1997:95–6and126–7;LeslieandWyatt1992:91).Instead,LIFFE’ssurvivalandsuccesscametorestonderivativesthathadnowell-establishedrivals,notablyFTSE-100futuresandUKandGermanbondfutures.Thereare,however,alsodifferencesbetweenfinancialandmosttechno-logicalinnovation.Thetaxtreatmentofderivativesismorecriticaltotheirsuccessthaninthecaseofmostphysicaltechnologies.Forexample,theappealoftheLondonTradedOptionsMarkettocustomerswasinitiallylimitedbythewayinwhichoptionsweretreateduntilSeptember1980inUKtaxlawas‘wastingassets’,whichhadtheconsequencethatcapitalgainstaxliabilitiescouldbeincurredonloss-makingaswellasonprofitabletrading(Steen1982).Incontrast,alargepartoftheappealoffinancialspreadbettingisthatintheUKcustomers’winningsarefreefromtax.Spread-bettingfirmsincurtaxliabilitiesasbookmakers,butthesearemodestandabsorbedintothespreadbetweenthepricesatwhichthefirmsbuyandsellcontracts.Financialinnovationsareeasierto‘reverseengineer’thanmosttechnolo-gies(Tufano1989:230;AllenandGale1994:53).Tominimizetheriskofdispute 72/Derivativesandlitigation,thespecificationofderivativeshastobemadeasexplicitaspos-sible.Tradingderivatives,pricingthem,andhedgingtheirrisksmayrequiretacitknowledge,buttheirdesigniseasytocopy.Innovativetechnologies(especiallythosethatareeasilycopied,suchaspharmaceuticals)areprotectedfromimitationbyintellectualpropertylaw,particularlypatenting.Incontrast,thelegalprotectionofinnovativefinancialproducts(and,forexample,ofderivativespricingmodels)hasbeenlimited,atleastuntilveryrecently.IntheUSA,forexample,financialproductsandmodelswerepresumedtofallwithinthe‘businessmethod’and/or‘mathematicalalgorithm’exemptionsfromthepossibilityofpatenting.Thegeneralshiftofintellectualpropertylawfromapresumptionofopenaccess—towhichpatentsweretheexception—towardsapresumptioninfavourofprivateproperty(Merges2000)hasonlyquiterecentlyencompassedfinancialinnovations.ApivotalcasewasStateStreetBank&TrustCo.v.SignatureFinancialGroup.6ItconcernedUSPatent5,193,056(9March1993),assignedtoSignature,whichcoveredadata-processingsystemforcalculatingassetvaluesandallocatingexpensesina‘HubandSpoke’™systeminwhichmutualfundssharetheownershipofacommoninvestmentportfolio.StateStreethadsoughttohavethepatentruledinvalid,butinJuly1998theCourtofAppealsfortheFederalCircuit,whichenjoys‘nationwidejurisdiction’overpatentcases,7foundinfavourofSignature.StateStreetsoughttoappealtotheSupremeCourt,butinJanuary1999thelatterdenieditleavetodoso(Lerner2002:903).Itisremarkablethat,atleastuntilStateStreet,financialderivatives,centralastheyaretotheglobalcapitalistsystem,developedinalegalregimewithonlylimitedintellectualpropertyrights.Didthatregime(a)slowinnovationbyreducingincentives,or(b)enhanceinnovationbyfacilitatingcopyingandadaptationinacontextinwhichQWERTY-like‘first-mover’advantageswereanadequateincentive?Thatquestionpointstoafamiliardebateaboutpatent-ingthatcannotbeenteredintohere,buttheextraordinarypaceofderivativesinnovationmightinclineoneto(b).Whatis,however,clearisthatcopyingwasindeedeasy.Specificderivativeshavefrequentlybeenimitatedwithout,atleastuntilrecently,fearoflitigation.IGIndexwould,likewise,havebeenunabletopreventotherfirmsofferinganalogousspread-bettingcontracts.Indeed,thereisasenseinwhichentireexchangeshavebeenimitated.LIFFE,forexample,wasmorecloselymodelledontheChicagoexchanges,particularly Derivatives/73theMercantileExchange,thanonanyBritishprecedent(Kynaston1997;LeslieandWyatt1992:91).CulturalGeographyTheestablishmentofLIFFEhighlightsathemeprominentinethnographiessuchasAbolafia(1996):tradingisaculturalaswellasaneconomicactivity.TheChicagofinancial-derivativesmarketsinheritedfromtheirparentagriculturalfuturesexchangesatraditionofoftenfrenzied,open-outcry,face-to-facetrad-ingin‘pits’.Chicago’swasatradingculturequitedifferentfromthatoftheNewYorkStockExchange.TherewasnoequivalentamongstChicago’scom-petingmarketmakersofNewYork’s‘specialists’,whoenjoyedwhatinChicagowasoftenperceivedtobeunfairlyprivilegedaccesstothe‘book’ofunfilledorders(inreturnforanobligationtomaintainanorderlymarket,inparticulartotradewiththeirowncapitaliftherewasatemporaryimbalancebetweenorderstobuyandtosell).TherewasanevengreatergulfbetweenChicago’sroughandtumbleandthe‘gentlemanlycapitalism’(Thompson1997)thatplayedadominantroleinLondonuntiltheearly1980s(the‘BigBang’deregulationof1986wasakeymomentinitsdemise).Itiseasytostereotype—toforgetthatanurbaneself-presentationisperfectlycompatiblewithdedication,financialacumen,andevenhard-edgeddealing—butneverthelesstheeliteofLondon’sfinancialsectorformedsomethingofa‘statusgroup’inWeberianterms(seeWeber1970;2000a;2000b).DavidSteen,akeyfigureinthedevelopmentoftheLondonTradedOptionsMarket,nicelyexpressedinmyinterviewwithhim(21June2001)thedifferencethathesawbetweenLondon’sethosandthatprevalentintheUSA:They[Americans]aremuchkeenertomakemoneythan[we]arehere...WhenIwasyoung,ifyou’dbeentoapublicschool,andparticularlyifyou’dbeentoOxfordorCambridge,youreallydidn’tneedtoworrymuchmoreaboutanythingelseasfarassocialstatuswasconcerned.Youcouldgoanywhereandyou’dbeacceptedanywhere.Youknewwhereyouwere.Establishedsocialstandingmadeitpossibletodisdainsmall-mindedpursuitofpecuniaryadvantage,whichwassometimescalled‘tizzysnatching’(‘tizzy’was 74/Derivativesnineteenth-centuryEnglishslangforsixpence):asSteenputit,‘peopletradingandtakingasnatchatprofitofsixpenceashare’.InChicago,incontrast,theequivalentofatizzywasconsideredwellworthsnatchingenergetically.LIFFEplumpedunequivocallyforChicagocultureovergentlemanlycap-italism,optingsymbolicallyforChicago’sbrightlycolouredtradingjacketsratherthanthedarksuitsandblackshoestraditionalintheCity.(LIFFEdrewthelineonlyatUnionJackjackets,fearingthey‘wouldbeseenontelevisionsellingthepounddowntheriver’(Kynaston1997:73).)LIFFE’straderswereoftendefiantlyEastEndor‘Essexboys’(Zaloom2003;2006)ratherthangen-tlemen.TheLondonTradedOptionsMarket(LTOM)wasfarmoreambivalentlyplacedthanLIFFE.ItsinspirationtoowasChicago(initscase,thesuccessoftheChicagoBoardOptionsExchange,althoughtheimmediatespurtoitsestablishmentwasthethreatthatoptionsonLondonsharesmightbetradedinAmsterdam),butLTOM’sLondonStockExchangeparentagewastoostrongforitfullytoembracethemoreflamboyantaspectsofChicagotradingculture.OnemarketmakerwhomovedfromChicagotoLTOMin1986recallsthathe‘wasbooedoffthefloorfirstdaybecauseIhadbrownshoeson’.Hiscolleagueswerenodoubtteasing,buthefoundtheattempttotranslateChicagoattitudesandpracticestoLondonsometimesuncomfortable.Tracingtheeconomicconsequencesofdifferencesofthiskindisdifficult.ThespreadsbetweenLTOM’smarketmakers’‘bid’and‘ask’pricesinthelate1970sandearly1980swerelarge(GemmillandDickins1986),farbiggerthanthoseinChicago,anditistemptingtoattributethistothewayinwhichChicago’sethosoffiercecompetitionbetweenmarketmakersfailedtosur-vivethetranslationtowhatwasinsomerespectsstillagentlemanlyworld.However,thatmaynotbecorrect,forthereareotherpossibleexplanationsoflargespreads.8Forexample,therewereeconomicallyconsequentialtensionsbetweenLTOManditsparent,theLondonStockExchange.Inparticular,stock-exchange‘jobbers’(marketmakers)valuedtheirrightunderexchangerulesnottodiscloselargetransactionsforninetyminutes,becauseitmadeiteasiertohandlebigblocksofshares.DelayeddisclosurecauseddifficultiestoLondon’soptionsmarketmakers(noequivalentrighttodelayeddisclosureexistedintheUSA),becauseitmeanttheycouldneverbeentirelyconfidentofthepriceatwhichtheycouldhedgeanoptionsposition.Widebid-askspreadscanthusbeseenashelpinginsulatethemfromtherisksattendantonthedifficultyofhedging. Derivatives/75Moreclear-cutistheeffectuponderivativesmarketsofoneaspectofthewidercultureinwhichtheyareembedded:thetraceleftinthelegalsystemofhostilitytogambling.Section18oftheUKGamingActof1845laiddown‘ThatallContractsorAgreements...bywayofgamingorwagering,shallbenullandvoid’,renderinggamblingdebtsunrecoverableinlaw.TheUSAwentfurther,withmoststates(including,cruciallyfromtheviewpointoftheChicagoderivativesmarkets,Illinois)outlawinggambling.Althoughorganizedexchangesdedicatedtothetradingofderivativesoffinancialassetsarerecent,suchderivativeshavelongbeentradedinadhocways,andexchangesdedicatedtoderivativesofagriculturalcommodities(grainfutures,forexample)haveexistedsincethenineteenthcentury.Theissueofhowtodrawthelegaldistinctionbetweenalegitimatederivativescontractandawageristhuslong-standing,anditisnotstraightforward:aderivativecanindeedseemtoresembleabetonthemovementofthepriceoftheunderlyingasset.Ifitwereruledthataderivativewasawager,aderivativecontractwouldhavebeenillegalintheUSAandunenforceableintheUK.Ineighteenth-centuryEnglishlegaldoctrine,theoveralldistinctionbetweenalegitimatecontractandawagerwasinformedbywhatO’Malleycallsa‘materialisttheoryofexchange’,inwhich‘theactofexchangemustincludesomeelementofmaterialvalueortitleto[material]value’(2003:239–40).The‘abstract’or‘virtual’natureofderivatives—which,asnoted,isamainthemeofrecenttheoreticaldiscussionofthem—isthusinfacttheirmostlong-standinglegaldrawback.Adoctrineaccordingtowhichlegitimateexchangehastoinvolvethetransferoftitletomaterialvalueendangersthelegalityofoptionsonsecurities,whichareattworemovesfrommaterialvalue,beingatbestaclaimonatitleofownershiporotherright.Norwastheproblemrestrictedtoderivativesofsecurities.Agrainfuturemightseemunequivocallytoinvolveeventualtransferofownershipofamaterialasset,butinpracticefuturescontractsongrainorothercommoditieswerenormallysettledbycashpayments.Deliveryofgrain(orevenoftheelevatorreceiptsthatwerethemainformoftokenofownershipofgrain:seeCronon1991)wasrare.Forreasonsthatscholarshaveyettoexploreindetail,nineteenth-centurylegaldoctrine,inbothEnglandandtheUSA,becameless‘materialist’andmorefavourabletoderivatives.Thedistinctionbetweenalegitimatecon-tractandawagerwasredrawnaroundwhatbecameknownasthe‘intent 76/Derivativestest’(Swan2000:212–13):ifthepartiestoacontractintendedthedeliveryoftheassetinquestion,thenthecontractwasnotawagerandwaslegalandenforceable,evenifdeliverydidnotactuallytakeplace.Toagrariancriticsofagriculturalfuturesexchanges,‘intent’couldseem‘anemptylegalfiction’(O’Malley2003:243),sinceitwaseasyforfuturestraderstoclaimthattheyhadintendedtodeliverthecommodityinvolved,andhadfailedtodososimplybecausecircumstanceshadchanged.Nevertheless,critics’effortstorestore‘theeighteenth-centuryprincipleofmaterialexchange’failed(O’Malley2003:244).Tradingoffuturesonphysicalcommoditiesandofstockoptionspassedtheintenttest(stockoptionscouldbesettledbyhandingoversharecertificates,andthathadcometocountasdelivery).However,thetestcreatedproblemsformoresophisticatedfinancialderivativespreciselybecauseoftheirmoreabstractnature.Astockindex,forexample,isamathematicalabstraction(itisnotthepriceofanysingleentity,butisanaverageofprices),sobyfarthesimplestwaytoconstructafutureonanindexistomakeitsettleablebycashpaymentalone.Butclaimingintenttodeliverwouldthenbeimpossible,andthecontractwouldasaresultbeliabletoberuledtobeawager.Inconse-quence,althoughtheChicagoexchangeshadwishedtointroducefuturesonstockindexesfromthelate1960sonwards,theywereunabletodosountil1982.(HowthenecessarylegalandregulatorychangeswerebroughtaboutisdiscussedinMacKenzie2006.)InEngland,LIFFEfacedtheintenttestandthe1845GamingAct(stillonthestatutebook),initiallyinregardtointerest-ratefuturesbaseduponLIBOR.Again,theissuewasthatLIBOR,beinganaverageinterestrate,wasnotdeliverable.LIFFEdevisedwhatithopedwasalegallyadequatehybrid:cashsettlement,butwiththe‘long’havingtherighttodemanddeliveryofadepositsimilartoaloanintheinterbankmarket(themarketthatLIBOR‘summarizes’inthewaydiscussedbelow).InJuly1982,LIFFEobtainedCounsel’sopinionthat‘suchacontractisnotawagerinlaw’(Kynaston1997:58).In1984,asimilarhybridwasdevisedforLIFFE’snewFTSE-100futures,with‘buyersandsellers[ableto]nominatesharestheymightwishtoreceiveordeliver’,againbecauseofthefearthat‘GamingActimplicationsmightprecludecash-onlysettlement’(Kynaston1997:131).TheissueofgamblingwasresolveddecisivelyintheUKonlyin1986,when,Kynastonreports,LIFFE’s‘traditionallygoodrelationship’(1997:155)withtheDepartmentofTradeandIndustryledtotheinclusionintheFinancial Derivatives/77ServicesActofaprovision(section63oftheAct)layingdownthatnocontractthatconstitutedinvestmentbusinesswithinthemeaningoftheActcouldberendered‘voidorunenforceable’onthegroundsthatitwasawager.Theprovisionremovedthebarriertoderivativesthatcouldbesettledonlyincash.Theymightstillfailtheintenttestandthusbeclassedasbets,buttheywerenowlegallyenforceable.The1986provisionhad,however,aninadvertentconsequence:itrescuedthenascentBritishspread-bettingindustry(whichhassubsequentlygrowntocompetewithLIFFEforthebusinessofindividualcustomers)fromtheconsequencesofthe1987stock-marketcrash.BecausegamblingremainedlegalintheUK,IGIndexhadbeenabletoturntheanalogybetweenderivativestradingandgamblingfromaproblem(asithadbeenintheUSAandforLIFFE)intoaresource:makingaderivativeintoabetconfersthetaxadvantagenotedabove.ThemostcommonofthestandardizedcontractsthatIGIndexanditscom-petitorsoffertheircustomersareanalogoustofutures(themaindifferenceisthatthecontractsaredirectlywiththespread-bettingfirm,ratherthanbetweencustomers).InthecaseofFTSE-100contracts,forexample,firmsquoteapriceatwhichcustomerscan‘buy’theindexfromthefirm,andalowerpriceatwhichtheycan‘sell’ittothefirm.(Aswithmarketmakersonexchanges,thefirms’profitscomemainlyfromthespreadbetweenthetwoprices.)Acustomerwhobelievestheindexwillrisewillbuytheindex,stakingacertainamount(typicallyoftheorderof£5)perindexpoint,hopingthattheindexwillhaverisenbymorethanthespreadbythetimeheorshesellstheindexbacktothefirm.Acustomerwhobelievestheindexwillfallwillbeginbysellingtheindex,andclosethebetbybuying(see,e.g.,Vintcent2002).Aswithexchange-tradedfutures,spreadbetsthusofferthepotentialthatalimitedinitial‘margin’depositcanbecomeamuchlargergainorloss.Spread-bettingfirmshedgeanylargeresultantexposuretomarketmovementsbytakingapositionsimilartothattakenbytheaggregateoftheircustomers(oftenusingfuturesonLIFFEorotherexchanges).Until1986,however,itwasimpossiblelegallytorecoversumscustomersowedthefirm.IGIndexcontrolledthatriskbyrequiringadepositlargeenoughtocoverlikelylosses,butcalculatingthatdepositinvolvedestimatingthesizeofplausiblemarketmovements;demandingtoobigadepositwouldputcustomersoff.AsnotedinMacKenzie(2004),the1987crashinvolvedamarketmovefargreaterthanseemedlikely,anditleftmanyofIGIndex’scustomerswith 78/Derivativesliabilitiesexceedingtheirdeposits.Atthattime,thefirmwasnowherenearaswellfinancedasitistodayanditcouldeasilyhavebecomeinsolvent.Becauseitscustomerswereinaggregate‘long’—hadbetthatpriceswouldrise—itshedgingmeantthatittoowaslong,andhugepricedeclinesmeantitowedlargesumstoitsbrokers,whichhadtobepaidimmediately.However,manyofitscustomers(whothoughttheyknewgamblingdebtstobeunenforceable)refused,orwereunable,topaywhattheyowedIGIndex.Fortunatelyfromitsviewpoint,IGIndexwasabletopointthemtosection63oftheFinancialServicesAct,whichmeanttheyhadtopay(interviewwithStuartWheeler,1March2005).Facticity‘Culture’isthusnotsimply‘thecontext’withinwhichderivativestradingtakesplace.Viamatterssuchasthelawofgambling,itshapesandisinter-meshedwiththedetailedmechanicsofthistrading.Anothercrucialaspectofthosemechanicsisthenatureoftheasset,rate,orotherquantityunderlyingaderivative.Foragriculturalfuturesexchanges—whichwere,asnoted,thesitesfromwhichmodernfinancial-derivativesexchangessprang—themostpressingissueinthisrespectwasstandardizingtheunderlyingassettoanextentsuf-ficientforclaimsonittobetradeablewithoutreferencetoanyspecificphysi-calentities.InChicagograintrading,standardizationseemstohavebeenanemergentproperty,co-evolvingwithfuturestrading(Cronon1991).Later,standardizationwasanexplicitpartoftheplanningforanewcontract,suchastheChicagoBoardofTrade’sfuturesonmortgage-backedbondsoritsfuturesonTreasurybonds,introducedinAugust1977and‘theexchange’smostsuc-cessfulcontractever’(Falloon1998:251;thefineethnographybyZaloom2006isoftheBoardofTrade’sbond-futurestrading).Bondsthemselvescouldnotplausiblybestandardized,soinbothcasesthetrickyproblemofmakingdif-ferentissuesofbondscommensurablehadtobetackled.Thesolutionsfoundwerealittleelaboratebutrobust,thoughsuddenshortagesofthe‘cheapest-to-deliver’bond(sometimestheresultofadeliberate‘squeeze’)arearecurrentproblemofwhichallbond-derivativestradersmustbewary.Aderivativescontractthatcanbesettledonlyincashavoidssuchproblems,andcashsettlementalsofacilitatesthedevelopmentofderivativesonentities Derivatives/79thatcannotstraightforwardlybedelivered:firstofallstockindexes,andnowamuchwiderrangeofquantitiesincluding,forexample,weatherandhumanlongevity.9However,cashsettlementraisesadifficultyofadifferentsort(onequitedistinctfromthelegalvulnerabilityarisingfromthe‘intenttest’).Themeasureusedtodeterminecashsettlementsums—whetheritbeaprice,anindexlevel,aninterestrate,orameasureofweather,longevity,orotherquantity—mustbeafact.Oneaspectoffacticity—thatis,ofstatusasafact—isadequacyofrepre-sentation.Themeasureusedforcashsettlementmustbebelievedgenuinelytoexpressconditionsinthemarketorprocessunderlyingthederivative,sothatsomeoneusingthederivativetohedgeriskcanbesurethat(ifconditionsareunfavourable)thegaintheywillmakefromthederivativewillcanceloutthelossestheywillincurintheunderlyingmarketorfromtheunderlyingprocess.Problemsofadequacyofrepresentationmay,forexample,havebeenonefactorinthefailureoftheLondonFuturesandOptionsExchange’spropertyfutures(Patel1994).ThemeasureusedforitshousingfutureswastheNation-wideAngliahousepriceindex,butthatwasbasedonlyontransactionsinwhichNationwideAngliawasthelender.ItwasonlyoneofseveralcandidatemeasuresoftheoverallstateoftheUKhousingmarket(eventoday,differ-entindexesoftenoffermarkedlydifferentestimatesoftherateofchangeinhouseprices).Furthermore,theaveragecountrywideconditionofthehousingmarketwaslessrelevanttohedgers—suchasdevelopersconcernedthatthehousestheywerebuildingwouldnotfetchtheanticipatedprices—thanlocalconditions,whichinthehousingmarketcanoftenvarymarkedly.(In2006,theChicagoMercantileExchangelaunchedhousingfuturesandoptions,usingindexesthatarespecifictoparticularcitiesandbasedonidentifyingrepeatsalesofthesameproperties,butitistooearlytotellhowsuccessfulthesecontractswillbeinthelongterm.)Incontrast,LIBORisanexampleofameasurethathasusuallybeentakenasanadequaterepresentationoftheunderlyingmarket.Itisthebasisbothofimportantexchange-tradedderivativescontracts,suchastheChicagoMer-cantileExchange’sEurodollarcontractorLIFFE’sShortSterlinginterest-ratecontract,andalsoplaysthecentralroleintheover-the-counterinterest-rateswapsmarket.(Inatypicalswap,partyApayspartyBarateofinterestthatisfixedforthecontract’sduration,whileBpaysAafloatingrate,mostcom-monlyLIBOR.)Theswapsmarketisthelargestofallderivativesmarkets—it 80/Derivativesdwarfseventhehugemarketsinexchange-tradedderivatives—withthecon-sequencethat,asnotedinChapter1,by2006contractstotallingover$170trillion(around$26,000foreveryhumanbeingonearth)wereindexedtoLIBOR.Giventhat,itissurprisingthatLIBORhasnever,tomyknowledge,beentheobjectofasocialsciencestudy(theonlydetailedsourceonitisunpublished:Mason1999).LIBORistheinterestrateatwhichmajorbankscanborrowfundsfromotherbanksintheLondoninterbankmarketinaparticularcurrencyforagivenperiodoftime.(Becausearangeofcurrenciesandtimeperiodsareinvolved,LIBORisasetofnumbers—six-monthUSdollarLIBOR,forinstance—notasinglenumber.Whythemostcrucialfactsareasetof‘London’rates,not‘NewYork’rates,isanintriguingquestionintheculturalandpoliticalgeographyoffinancialmarketsthatunfortunatelycannotbediscussedhere.)TounderstandhowLIBORisconstructedrequiresabriefdiscussionoftheinterbankmarket.Akeyroleinitisplayedbythe‘voicebrokers’discussedinChapter2.Suchbrokerssitatdesksinroomsthatresemblebanks’tradingrooms,butaremorecrowded,noisier,andmoreraucous.Oneachbroker’sdeskisa‘voicebox’(consistingofamicrophone,loudspeaker,andswitches),whichisconnectedbydedicatedtelephonelinestosimilarvoiceboxesatthedesksofeachofthebroker’sclientsinbankdealingrooms.Sometimesinterbankdealsarestruckdirectly,butmoreoftenabank’sdealerwhowishestoplaceortoreceiveaninterbankdepositwillusehisorhervoiceboxtotellabroker,whowillthendooneofthreethings:(a)usehisorhervoiceboxtotrytofindacounterparty;(b)shoutouttheordertohisorhercolleagues;or(c)aska‘boardboy’(astheyarestillcalled)towritetheorderononeofthelargewhiteboardsthatsurroundthebrokers’desks.Brokerssupplytheirbankdealing-roomclientswithscreensthatindicatecurrentbidandofferratesforinterbankdeposits,andthosescreensarethemostimportantminute-by-minuterepresentationoftheinterbankmarket.Thereis,however,anelementofjudgementintheratesthatbrokersdisplayonthescreens.Notallbanksareequal:theircreditratingsdiffer,andabank’screditriskdepartmentwilltypicallyimposealimitontheamountofmoneythatcanbeondepositwithanyparticularcounterpartybank.Soabrokermay,forexample,choosenottodisplaythemostattractiveinterestratethatheorsheknowsof,ifitssourceisabankwithapoorcreditratingtowhichmanyofhisorherclientswouldbeunabletolend. Derivatives/81Dealersinbanksalsoexercisejudgementininterpretingtheratesthescreensdisplay.AskedhowheestimatesLIBOR,onedealertoldme:withinsaythepoolofsixteen[banksonaLIBORPanel:seebelow]...you’llprobablyhavethreeaggressivelenders,sotherun-throughyougetfromthebrokeriswhereyou’regoingtogetthefirstthreelotsofmoney.Afterthatyouhavetomoveyourpriceupuntilitbecomesattractiveenoughforthepeoplethatdon’twanttolendtosuddenlythink,‘well,thisisbecomingattractiveenoughtodoit’,andthat’swherethisspread...comesfrom[ALIBORestimateis]notgoingtobeamid-marketrate,it’sgoingtobethepointatwhichyouarelikelytogetthemoney.ThejudgementthusinvolvedinestimatingLIBORraisesanotheraspectofthefacticityofthemeasureusedtocash-settleaderivative:itsrobustnessinrespecttoattemptstomanipulateit.ThosewithnodirectinvolvementinthemarketofwhichLIBORisarepresentationmightbeguaranteedtobeunbiased,buttheywouldlackthedetailedknowledgeneededtoexerciseinformedjudgement,sothereisnopracticalalternativetorelianceonthosewhoseinvolvementmeanstheymayhave‘interests’intheoutcome.WhentheChicagoMercantileExchangelaunchedLIBOR-settledEurodol-larfuturesinDecember1981,itthussetupitsowndailypoll,designedbyitschiefeconomist,FredArditti.EachofadesignatedsetofbankswasaskedtogiveanestimateofLIBOR,butbeforetheaveragewastakenthehighestandlowestestimateswereeliminated,sonoonebankcouldinfluencetheresultbygivingaveryhighoraverylowestimate.‘[I]nthebeginningthere[was]someminorgrousing’,LeoMelamed,thenchairoftheexchange,toldme,but‘thebeautyofthe[LIBOR]“fixing”wasthatitwassooverwhelminglyacceptedasthe“true”priceforinterestrates’.10In1985,theBritishBankers’Association,membershipofwhichisopentointernationalbankstradinginBritainaswellasBritish-domiciledbanks,introducedacentralizeddailyLIBOR‘fixing’(similarinoutlinetoArditti’s)thateventuallyreplacedallotherfixings,althoughother‘LIBORs’arestillsometimesquoted.TheForeignExchangeandMoneyMarketsAdvisoryPaneloftheAssociationselectspanelsofeight,twelve,orsixteenbanksforeachcurrencyonthebasisofthosebanks’‘reputation’,‘scaleofactivityintheLondonmarket’,and‘perceivedexpertiseinthecurrencyconcerned’,while‘givingdueconsiderationtocreditstanding’(Mason1999:3–4).By11.10a.m.eachbusinessday,eachbankonaLIBORpanelreportstoTelerate(nowpartofReuters)‘therateatwhichitcouldborrowfunds 82/Derivatives[“unsecured”,and“governedbythelawsofEnglandandWales”]wereittodosobyaskingforandthenacceptinginter-bankoffersinreasonablemarketsizejustpriorto11.00’(Mason1999:slides8and9)inthecurrencyandforthetimeperiodinquestion.Theratesarethenrankedinorder,thetopandbottomquartilesareignored,andthemeanofthesecondandthirdquartilesiscalculated.ThatmeanisBritishBankers’AssociationLIBOR,andbyaround11.45a.m.itisdisseminatedworldwideviaallthemainmarketnetworks.Thefixingtakesinputsthatmayseemimprecise—‘weaskthem[thebanksonthepanel]totelluswhatotherpeopleareoffering’;thereisnorequirementthatanyloanactuallybetakenoutatthatrate;andwhatconstitutes‘reasonablemarketsize’isdeliberatelynotdefinedexactly(Mason1999:4–5,emphasesinoriginal)—andfromthoseinputsitproducesfacts.Thefixingisdesignedtobesociologicallyrobust,sotospeak.ThebanksthatproducetheinputswillverylikelyhavelargederivativesportfoliosindexedtoBritishBankers’AssociationLIBOR,thevalueofwhichwillbeaffectedbythefinalfigure,butaswellasthelattertheirinputsarealsodisseminated.Oneintervieweeshowedmethatday’sinputsintothree-monthsterlingLIBOR,pointingwithsuspiciontoabankthathadreduceditsinput—byasinglebasispoint(thatis,ahundredthofapercentagepoint)—fromthepreviousday’s,whileallothershadeitherincreasedtheirsorleftthemunchanged.Anidiosyncratic,manipulativeinputisthusonpublicdisplaytothemarket.Furthermore,theexclusionofthetopandbottomquartilesmeansthatanoverlyidiosyncraticinputwouldinanycasebethrownoutofthecalculation.Finally,anadvisorypaneltotheBritishBankers’AssociationselectsthemembersofContributorPanels,andabankthatregularlymadedubiousinputscouldfaceembarrassingremoval.Thissanction,oneintervieweetoldme,hasbeenused,albeitnotrecently.During2007–8,however,confidenceinLIBORcameunderunprecedentedstrain.TheexclusionofthetopandbottomquartilescannotbyitselfprotectLIBORfromissuesthataffectallbanks.WithfirstNorthernRockandthenBearStearnsneedingtoberescued,almostanybankcouldseematrisk.SuspicionsemergedthatbanksonLIBORpanelswerereluctanttoreportthattheycouldborrowonlyathighrates,becauseoffearofrumoursthatthebankinquestionwasintrouble.Thisledtoconcernsthatbankswere‘herding’—keepingtheirinputsclosetoothers’anticipatedinputs—andthattherewasadownwardbiastoLIBOR.TheWallStreetJournalbecameaprominentcritic,forexamplereportingon16April2008aclaimthatthree-monthdollarLIBORwasthirtybasispointsbelowwhatit‘should’be.TheBritishBankers’ Derivatives/83AssociationvigorouslydefendedLIBOR,andthoughthereweresuggestionsforalterationstoit(suchasaskingeachbanktoreportnottheratesatwhichitcanborrowbutthoseatwhichotherscanborrow)’theAssociationrightlyrejectedknee-jerkchanges.Nevertheless,therewasgrowinginterestinbasingnewderivativesonotherindexes:inJune2008,forexample,LIFFElaunchedafuturescontractbasedonEONIA,aweightedaverageoftheratesonovernighteuroloandealsbetweenbanks.ConclusionThemoretheoreticallyorientedofthecontributionstothegeographical,anthropological,andsociologicalliteratureonfinancialderivativeshavehadatendencyinadvertentlytoreplicatetheappearanceoftheproductstheydiscuss:theyhaveformedaratherabstractliteratureonapparentlyabstractproducts.However,amarketintheseproducts‘ismorethanabrightidea’,saysLeoMelamed,wholedtheChicagoMercantileExchange’smoveintofinancialderivatives.‘Ittakesplanning,calculation,arm-twisting,andtenacitytogetamarketupandgoing.Evenwhenit’schuggingalong,ithastobecrankedandpushed’(MelamedandTamarkin1996:295).Thatcalculation,cranking,andpushing(forsomeinsightintoarm-twisting,seeMacKenzie2006)isakeypartofthematerialproductionofvirtuality,asareissuessuchasthedesignofderivativescontracts,thelegalboundarybetweenlegitimatetradingandgambling,localculturesoftrading,andhowthefacticityessentialtocash-settledderivativesisgenerated.Suchissuesshouldmattertothosewhoareinterestedinthe‘bigquestions’ofthetheoreticalliterature—suchastheextentanddistributionofriskorthescopeofglobalizationandcommodification—becausewell-groundedanswerstothosebigquestionsinevitablymustinvolvetheapparentdetailofmaterialsociology.Thequestionofspatiality,forexample,demandsanuancedanswer.Yes,globalfinancialintegrationisaveryrealphenomenon;butno,ithasnotbroughtabout‘theendofgeography’(O’Brien1992).LIBOR,forinstance,isaglobalfact;butitisalsoLondonInterbankOfferedRate.Spreadbetting,totakeanotherexample,permitsrapidlyincreasingnumbersofresidentsoftheUKtousethescreensandkey-padsoftheirmobilephonestoenterinto 84/Derivativesinexpensivederivativescontractsonthousandsofglobalassets:theNikkeiindex,Brentcrude,gold,carbon-emissionspermits,andsoon.11Thesimpleoperationinvolvedtakeslessthantensecondsandthecontractsareusuallyconfirmedinaslittleasfiveseconds.NofullyequivalentexperiencehasbeenavailabletoresidentsoftheUSA,andsuchlivedexperiencesofmarketsaresurelyconsequential:by2006,some400,000peopleintheUKhadspread-bettingaccounts(BradyandRamyar2006).Currently,thereisanattempttorepeatintheUSAthesuccessoffinancialspreadbettingintheUKintheformof‘hedgelets’—similartospreadbets,butformulatedinsuchawayastostopthembeingclassedaswagers—buthowsuccessfulthatwillberemainstobeseen.Thematerialsociologyofderivativeshasmanyfacetsnotdiscussedinthischapter.OnethatwillbediscussedinChapter5ishowarbitragetiesthepricesofderivativestotheirunderlyingasset(butalsohowitsometimesfails).Alsoimportant,buteithernotdiscussedinthisbookortouchedononlyverybriefly,aretherolesinderivativestradingofbodies,whicharematerialentitiesparexcellence,andoftechnologies(thesometimestraumaticshiftfromopen-outcrytoelectronictradingisdiscussedbyZaloom2006);thecrucialfunctionsofclearinghouses(Millo,Muniesa,Panourgias,andScott2005);thestructur-ingroleofsystemsofregulation;andsoon.ItisclearlyimportanttoextendtheanalysisbeyondtheUSAandUKtotheworld’smanyotherderivativesexchanges,whetherestablishedornascent.Theissuesofinnovation,culturalgeography,andfacticityneedaddressingalsointhecontextoftheover-the-counter(directinstitution-to-institution)market,notjustexchange-tradedderivatives.Nevertheless,Ihopethatthispreliminaryanalysisindicatesatleastthatweshouldnotsimplybefascinatedbythevirtualcharacterofderivatives,butneedtoinvestigateindepthhowthatvirtualityismateriallyproduced. 5TheMaterialSociologyofArbitragewithDanielBeunzaandIainHardieThederivativesdiscussedinthepreviouschapter,whichnowplaysuchacentralroleinglobalfinance,arecontractsorsecuritieswhosevaluedependsonthepriceofan‘underlying’asset,onthelevelofanindex,oronanexchangerate,interestrate,orotherquantity.Here,aselsewhereinfinance,factualstatusisimportant.Foraderivativesmarkettobecredible,thepriceofderivativesmustnotbeseenasarbitrary,orassubjecttomanipulation,butmustadequatelyreflectconditionsinthemarketfortheunderlyingasset,inparticularthepriceoftheassetortheleveloftheindex,rate,orotherquantityinvolved.Crucialtothefunctioningofderivativesmarketsisthataspecificmater-ialprocesscantiethepriceofafuture,option,orotherderivativetotheunderlyingassetorrate:arbitrage.Inthischapter,weseektodevelopamaterialsociologyofarbitrage.Asdiscussedbelow,‘arbitrage’isatermwithdifferentmeanings,butthischapterfollowsmarketpractitionersindefiningitastradingthataimstomakelow-riskprofitsbyexploitingdiscrepanciesinthepriceofthesameassetorintherelativepricesofsimilarassets.Forexample,ifthepriceofaderivativedriftstoofarawayfromthatimpliedbyconditionsinthemarketfortheunderlyingassetorrate,thenarbitrageurscanstepinandexploitthepricediscrepancy.Theybuythe‘cheap’instrument 86/TheMaterialSociologyofArbitrageandsellthe‘expensive’one,solimitingorreducingthediscrepancy,recon-nectingthemarketforthederivativewiththatfortheunderlyingassetorrate.Importantthoughtheroleofarbitrageinderivativesmarketsis,itisfoundinmanyothercontextsaswell.Aclassicexamplehistoricallywasgoldarbi-trage.IfthepriceofgoldinSaudiArabiaexceedsitspriceinNewYorkbymorethanthecostoftransportation,arbitrageurscanprofitbybuyinggoldinNewYorkandsellingitinSaudiArabia(orviceversaifgoldischeaperinSaudiArabia).Bybuyingandsellingasclosetosimultaneouslyaspossible,arbitrageursavoidtherisksof‘directional’trading:theyprofitirrespectiveofwhetherthepriceofgoldgoesontoriseortofall.Arbitragerequirestechnologicalresources,sustainedeffort,andexpertisebeyondthecapacityofnearlyalllayinvestorsinfinancialmarkets.Itisthepreserveofmarketprofessionals,andisacrucialformoftrading.(TobringthetopicclosetohomeforacademicreadersintheUSA,wewouldnotethatthemostsuccessfuloflargeuniversityendowmentfunds,suchasHarvard’s,areoftenskilledpractitionersofarbitrage.)Indeed,arbitrageconstitutesmarkets,forexamplehelpingtodeterminetheirscopeandtheextenttowhichtheyareglobal:thatinternationalgoldarbitrageispossiblecreatesaworldmarketingoldwitha‘worldprice’,ratherthangeographicallyseparatemarketswithdifferentprices.Inconstitutingmarkets,arbitragehaswiderconsequencesforeconomiesandpoliticalsystems.Forexample,inthelate1990sarbitrageursinhedgefundsandinvestmentbanksbegantoperceivegrowingsimilaritybetweenthebondsissuedbythegovernmentofItalyandthoseissuedbyotherEuropeancountries,notablyGermany.Foravarietyofreasons(includingdistrustofthefiscalefficiencyoftheItalianstateandconsequentfearsofitdefaultingonitsbonds),thepricesofItaliangovernmentbondshadtraditionallybeenlowrelativetothoseofcountriessuchasGermany,thusimposinghighdebt-servicechargesonItaly.AsarbitrageursbegantobuyItalianbonds,theirrela-tivepricesroseandtheproportionofItaly’sgovernmentexpendituredevotedtodebtservicefell.Theprocess—whichwasassistedbytheliquiditycreatedbytheMTSelectronicbond-tradingsystem,setupbytheItaliantreasuryin1988anddiscussedbelow—helpedItalymeettheMaastrichtcriteriaforEuro-peanEconomicandMonetaryUnion(EMU).Arbitrageurs’beliefsthushadaself-validatingaspect—theypromptedtradingthatmademorelikelytheevent,Italy’squalificationforEMU,onwhichthebeliefswerepredicated—and TheMaterialSociologyofArbitrage/87arbitragehelpedtocreateaEuropeangovernmentbondmarket,ratherthanseparatenationalmarkets.Thefailuresofarbitragecanbeasconsequentialasitssuccesses.Suchfailureswereattheheartoftwoofthemostseriouscrisesofthepost-warfinan-cialsystem:the1987stock-marketcrashandthe1998turmoilsurroundingthehedgefundLong-TermCapitalManagement(LTCM).Acrucialaspectoftheformerwasthebreakdownofthelink—normallyimposedbyarbitrageinthewaysketchedabove—betweenthestockmarketandakeyderivativesmarket:stock-indexfutures.InthecaseofLTCM,theforcedunwindingofarbitragepositionscausedhuge,sudden,highlycorrelatedpricemovementsacrosstheglobeinapparentlyunrelatedassets,bringingsomemarketsclosetoparalysis.Thereisanenormousdisciplinaryimbalanceinregardtoarbitrage.Ithasreceivedalmostnosustainedattentionineconomicsociology,ineconomicanthropology,ineconomicgeography,orinthestrandofpoliticalscienceknownasinternationalpoliticaleconomy,eveninthesubsetsofthosespecial-itiesthatdealwithfinancialmarkets(thelimitedexceptionsincludeMiyazaki2003;Robottin.d.;BeunzaandStark2004;Hardie2004;andMacKenzie2003).Incontrast,thecentraltheoreticalmechanisminvokedbymodernfinancialeconomicsis‘arbitrageproof’.Thefieldpositsthattheonlypatternsofpricesthatcanbestablearethosethatpermitnoopportunitiesforarbitrage.Partic-ularpatternsofpricesarethenshowntobenecessarybydemonstratingthatifpricesdeviatefromthatpattern,arbitrageispossible.Theentiremoderntheoryofassetpricing—especiallythetheoryofthepricingofderivativessuchasoptions—relieson‘arbitrageproof’ofthiskind.ParadigmaticistheNobel-Prize-winningtheoryofoptionsdevelopedbyFischerBlack,MyronScholes,andRobertC.Merton(BlackandScholes1973;Merton1973).Intheirmodel,thepriceofanoptionisdeterminedbythefactthatitcanbereplicatedexactly:itispossibletoconstructacontinuouslyadjustedportfolioofholdingsorborrowingsoftheunderlyingassetandcashthatwillhavethesamepay-offastheoptioninallstatesoftheworld.Thepriceoftheoptionmustequalthecostofthisreplicatingportfolio,forotherwisearbitrageispossible.Theconceptualizationof‘arbitrage’intheworkofBlack,Scholes,andMerton,andinmainstreamfinancialeconomicsmoregenerally,differsfromthearbitrageasmarketpracticethatisthefocusofthischapter.Orthodoxeconomistsdefinearbitrageasdemandingnocapitalandinvolvingnorisk,whileinmarketpracticearbitrageseemsalwaystorequiresomecapitaland 88/TheMaterialSociologyofArbitrageinvolvessomerisk,eveniftheriskisonlythatacounterpartytoatransactionwillnotfulfilitsobligations(Hardie2004).Indeed,apuristwouldarguethatthetradingweconsiderinthischaptershouldnotbeconsidered‘arbitrage’.1Purism,however,hasitscosts—apuristdefinitionof‘arbitrage’excludesthereal-worldcounterpartsofthecanonicalarbitragesoffinancetheory,suchasthearbitragethatimposesBlack–Scholes–Mertonoptionpricing—andpurismisnottheonlypossibleresponse.Financialeconomists—especially‘behavioural’economistssuchasAndreiShleifer—havebeguntoinvestigatetheconsequencesofmakingthedefinitionofarbitragemorerealistic(see,forexample,ShleiferandVishny1997).Theseeconomistsrightlyseethetopicasacrucialone.Since,inorthodoxviews,itisaboveallarbitragethatmakesmarketsefficient,theexistenceoflimitstoarbitragecastsintodoubtthefullvalidityofthecentraltenetofmodernfinancialeconomics:theefficientmarkethypothesis,accordingtowhichpricesinmaturecapitalmarketsfullyreflect,effectivelyinstantaneously,allavailableprice-relevantinformation.Weshallsuggestbelowthattherearepotentiallyproductivelinkagesbetweentheemergingliteratureineconomicsonthelimitsofarbitrageandthe‘materialsociology’ofarbitragethatweadvocate.AsnotedinChapter2,materialsociologypaysattentionto,amongstotherthings,theroleplayedinsocialrelationsbytechnologicalsystemsandotherphysicalobjectsandentities(includinghumanbodiesviewedasmaterialentities).Sincethatroleisofcoursepervasive,allsociologyshouldbematerialsociology,yetsocialtheoryfrequentlyabstractsawayfromphysicalobjectsandempiricalenquiryoftendoesnotfocusonthem.Asweshallargue,aproperunderstandingofarbitragerequiresustotakeintoaccountbothits‘physical’and‘social’aspects,andthetwoareultimatelyinseparable.Arbitrageissimultaneouslya‘physical’anda‘social’process.Indevelopingamaterialsociologyofarbitrage,thischapterdrawsuponthreedatasources.OneisfieldworkconductedbyBeunza,incollabora-tionwithDavidStark,andreportedpreviouslyinBeunzaandStark(2003;2004;2005).BeunzaandStarkconductedaparticipant-observationstudyoftheLowerManhattanarbitragetradingroomofaglobal,non-Americaninvestmentbankwith128officesin26countriesacrosstheUSA,Europe,andAsia.Thestudycomprised65half-dayvisitstothetradingroomoverthecourseof34months.Duringthosemonths,BeunzaandStarkunder-tookdetailedobservationatfourofthetentradingdesksintheroom(mergerarbitragedesk,statisticalarbitrage,specialsituations,andcustomer TheMaterialSociologyofArbitrage/89arbitrage),sittinginthetightspacebetweentraders,followingtradesastheyunfolded,andsharinglunchandjokeswiththetraders.Theycomplementedthisdatawithin-depthinterviewsofaselectedgroupoftradersinamoreprivatesetting,typicallyasmallconferenceroomjustoffthetradingroom.Inthefinalyearoftheinvestigation,BeunzaandStarkwereintegratedmoreformallyintothetradingroom—providedwithaplaceatatradingdesk,acomputer,andatelephone.ThesecondsourceisastudyofarbitragebyMacKenzie,whichfocused(a)onLTCM,wherethemainpurposewastounderstandthefund’scri-sisandwhatitimpliedforthelimitstoarbitrage,and(b)onformsofarbitrage,especiallyoptionsarbitrage,thatdrawheavilyuponfinancethe-ory,wherethegoalwastounderstandingtheextenttowhichsucharbi-trage‘performs’financetheoryinthefashionpositedbyCallon(1998).Thestudywasbasedonsemi-structuredinterviewswitha‘snowball’sampleof26arbitrageurswhosetradingwasrelevanttothesequestions,onananaly-sisofthepricemovementsinthecrisismonthsinthemarketsinwhichLTCMoperated,andonexistingeconometricanalysesofoptionspricing(seeMacKenzie2003;2006).ThethirdsourceisastudybyHardieandMacKenziewhichfocusesonseveraloftheissuesdiscussedinthischapter,includinginformationflowbetweencompetingtraders,relationsbetweentradersandmanagers,andcir-cumstancesinwhichtradesthatwillalmostcertainlybeprofitableifheldforlongenoughneverthelesshavetobeunwound.Itssnowballsample(oftradersinhedgefundsandinvestmentbanks,ofthosewhomanagethesetraders,ofthosewhoprovideadministrativeservices,andofthe‘fundsoffunds’thatchannelcapitaltohedgefunds)comprises51interviews,2andthestudyalsoinvolvesthebriefperiodofobservationofahedgefunddrawnoninChapter3.Itwasduringthisperiodofobservationthatwewitnessedthearbitragetradethatwedescribeinthenextsectionofthechapter.However,theexampleischosenforitssimplicity,notbecauseitencapsulatesalltheissueswewishtodiscuss—nosingleexampledoes.Indeed,ourgoalhereisnottoreportsys-tematicallyonthethreeunderlyingstudies,whichhavemanyaspectsbeyondthosediscussedhere.Rather,ouraimistooutlineamaterialsociologyofarbitrageconsistentwithourobservations,interviews,andothersourcesofdata,andalsowiththeresultsofothersinthesocialstudiesoffinancewhohavetoucheduponarbitrage. 90/TheMaterialSociologyofArbitrageBrazil14sand40s5January2005:HardieandMacKenzieareobservingtradinginthehedgefunddescribedinChapter3.Justafterthefund’smorningmeeting,part-nerBnoticesanoddityintheBraziliangovernmentbondmarket.TheminutesoftheUSFederalReserve’sOpenMarketCommittee,releasedthepreviouseveninginLondontime,havebeentakenbymarketparticipantsasindicatingthatfurtherinterest-raterisesareontheway,andhaveledtogeneralpricefallsintheBrazilianbondmarket.However,the‘14s’(anissueofdollar-denominatedbondsthatmaturein2014)are‘tradingup’:theirpriceishighrelativetootherbonds.‘Hitthebid’(sellthem),suggestspartnerB.Thetraderdoesnotrespondimmediately,buthegoesontoaskhisassistanttoproduceachartofthepricesoverthelastthreemonthsofthe‘14s’andthe‘40s’(Braziliangovernmentdollar-denominatedbondsthatmatureinNovember2040).Asthedayproceeds,thetradertakesapositioninthe14sandthe40s,shortsellingtheformerandbuyingthelatter.HealsosendsacontactinaninvestmentbanktheExcelfilecontainingthepricechartproducedbyhisassistant,encouraginghiscontacttocirculateittoothers.(Laterinthischapterwediscusswhyhedoesthis.)Abondmaturingin2040seemsverydifferentfromonematuringin2014:muchcouldhappeninthequarter-centurybetweenthetwodates.Butthebondmaturingin2040is‘callable’:theBraziliangovernmenthastherighttorecallthebondbyrepayingtheprincipalearly,in2015.IfBrazilianbondscontinuetotradeatanythingliketheircurrentprices,itwillbeinthegov-ernment’sinteresttodoso,sinceitwillbeabletoreplacetheborrowingmorecheaply.The‘40s’thusineffectmaturein2015andso,despiteappearances,a‘14’anda‘40’arequitesimilar.Onthemorningof5January,noneofthisissaidexplicitly:itispartofwhatthetraderandpartnerB,likeallsophisticatedparticipantsintheBrazilianbondmarket,simply‘know’.(Hardiewasaninvestmentbankerbeforereturn-ingtoacademia,andwasinvolvedintheinitialsaleofthe‘40s’onbehalfofthegovernmentofBrazil,soheknowsittoo,thoughheneedstowhis-peranexplanationtoMacKenzie.)Nevertheless,thechartproducedbythetrader’sassistantisamaterialrepresentationthatmakesvisiblethereasoningunderpinningthetrade.Oncehehasconfiguredthechartaccordingtothetrader’swishes—initially,itshowsthepricesofthe14sandthepricesofthe TheMaterialSociologyofArbitrage/9140s,whenthetraderwantsittodisplaythedifferenceinprices—thepricesofthetwobondscanbeseentofolloweachotherclosely,aswouldbeexpected,butwiththe40salmostalwaysslightlymoreexpensivethanthe14s.Again,thereasoniscommonknowledgeamongstaficionados.The40sarethemostliquidofBraziliangovernmentbonds,theonesmostreadilyboughtandsold,andthusthemostattractive,forexampletothosewhowishtocreateandtoexitpositionsquickly.Indeed,inalaterinterviewpartnerCdescribestheBrazil40sasthe‘asset-classbenchmark’foremerging-marketgovernmentbondsasawhole:‘ifpeoplearenegative/positiveonemergingmarkets,theybuyorsellthatspecificissue.’Inordertoreadthetrader’sassistant’schartcorrectly,oneneedstoreal-izethatonit‘time’flowsfromrighttoleft(theearliestdatesareontheright).Oncethatisgrasped,however,someoneviewingitcanclearlyseewhatthetraderhasseen:inrecenttradingdaysthe14shavebecomemoreexpensivethanthe40s,withthedifferenceincreasingsharplythepreviousday(4January).Thetraderknowshismarketwellenoughtoinferacausethatisconfirmedonlylaterinthedayinatelephoneconversationwiththeabove-mentionedinvestmentbankcontact.Thesell-offtriggeredbyUSFed-eralReserve’sminuteshasconcentratedinBrazil’sliquid40s.Indeed,asthecontacttellsthetrader,unusually‘therealmoneyguys[tradersnotinhedgefundsbutinbiggerinstitutions]shorted40s’.Thetraderthusconfidentlyassumes—andmakesexplicitinatelephoneconversationwithhiscontact—thatthefactthat14saremoreexpensivethan40sisapricediscrepancythatwillbetemporary.Byshortselling14sandbuying40s,he—andindeedothers—canperformanarbitrage(inmarketpractitioners’senseoftheterm).Thediscrepancywouldbeexpectedtovanishinthenormalcourseofevents,butespeciallyifotherschoosealsotoexploitit(perhapsbecausetheinvestmentbankcontactcirculatestheassistant’scharttothem),theprocesswillbehastened,maybeconsiderably.Byearlyafternoon,thetraderhasaccumulatedsome$13millionofshortsalesof14sandanother$13millionofpurchasesof40s.Bymid-afternoon,heisabletosay‘it’smovedinmyfavour’—thediscrepancyhasstartedtoreduce—‘butnotenoughtounwind’:hekeepsthepositionon,expectingfurtherreductionsinthedis-crepancy.Onlyattheendoftheweekdoesheliquidatehisposition,earningahealthyprofit.Notewhatthetraderisnotdoinginthistrade.Likethegoldarbitrageur,heisnottakinga‘directional’view.Heisnotattemptingtopredictthepoliciesof 92/TheMaterialSociologyofArbitragetheBraziliangovernment,toestimatetheprobabilityofbonddefaultbyBrazil,ortoanticipatethefuturecoursesofinterestratesorinflation:becausethe14sandthe40saresosimilar,changesinfactorssuchasthesewillaffectthepricesofeachbondroughlyequally,andwiththetrader’smatched‘long’and‘short’positionstheeffectswillcancelout.Asthetraderputsitinatelephonecalltohiscontactintheinvestmentbank,‘thereiszeromarketrisk’inthetrade:itsprofitability(‘thereisatleasthalfapointinthattrade’)shouldnotbeaffectedbyoverallrisesandfallsinthepricesofBraziliangovernmentbonds.Infact,asheacknowledgestous,thetrader’spositionisnotentirelyfreefromrisk—seebelow—butinitsinsulationfromthemajorriskfactorsinhismarket,itislowrisk.Askedbythethirdauthorabouttherationaleofthetrade,thetradersays(justasafinancialeconomistwould)thatthefact‘thatthistradehaspresented[itself]indicates[an]inefficiency’.Temporarily,pricesarereflectingsomethingotherthanmerelyavailableinformationsuchastherelativeliquidityofthetwobonds.Althoughthetrader’smotivationmaysimplybetoearnmoneyforhishedgefund,hisactionsarehelpingtoeliminateadiscrepancyandcorrecttheeffectsofan‘inefficiency’.Inthatrespect,histrading,evenifnotfreeofrisk,resemblesarbitrageasconceivedbyfinancialeconomics.TheMaterialityofArbitrageApriceisathing.Likeallprices,thosetowhichthetraderwasresponding(andcirculatingintheformofthechartpreparedbyhisassistant)werephysicalentities—patternsoncomputerscreensandspokennumberstransmittedbytelephone.Theformsofembodimentofpricesarevarious—thesoundwavesthatconstitutespeech;penorpencilmarksonpaper;theelectricalimpulsesthatrepresentbinarydigitsinacomputerizedsystemorencodesoundoveratelephoneline;handsignalsin‘open-outcry’tradingpitsthataretoonoisyforvoicestobeheard;andsoon—butarealwaysmaterial.Ifapriceistobecommunicatedfromonehumanbeingtoanother,orfromonecomputerizedtradingsystemtoanother,itmusttakeaphysicalform.Thematerialityofpricesmatterstoarbitragebecausetheirphysicalembod-imentaffectstheextentandspeedoftheirtransmission.Classicalformsofarbitrageexploitedthedifferencesbetweenpricesindifferentplaces.The TheMaterialSociologyofArbitrage/93commoditiesandcurrencyarbitrageurJ.Aron&Company,forexample,usedtokeeptelephonelinestoSaudiArabiaopenconstantlysoitcouldasquicklyaspossibledetectandexploittheemergenceofdiscrepanciesingoldorsilverprices(RubinandWeisberg2003:90–1).Thedevelopmentofelectronicpricedisseminationsystems(especiallythe‘Monitor’system,introducedbyReutersin1973)largelyunderminedthetime-spaceadvantagesthatfirmssuchasAronhadachievedbytheuseofsocialnetworksandoldercommunicationtechnologies,notablythetelegraphandthenthetelephone.Itwasmuchlesslikely,forexample,thatatradercouldperformarbitragebyhavingtwotelephonelinesopen,sellinganassettoonecounterpartywhilebuyingitatalowerpricefromanother.‘AftertheintroductionofMonitor,prices[initiallycurrencyexchangerates,andlatermanyotherpricesaswell]suddenlybecameavailablegloballytoeveryoneconnectedtothesystem’(KnorrCetinaandBruegger2002b:395).Electronicpricedisseminationdoesnot,however,entirelyeliminatediffer-encesinthespeedoftransmissionofprices,andthosedifferencesremaincon-sequential,eveniftheyarenowmeasuredinsecondsorevenmicroseconds,nottheminutesorhoursoftraditionalarbitrage.Forinstance,apricedelayofaslittleastwosecondsmadetheindexarbitragedeskofabankthatcompetedwiththebankstudiedbyBeunzalosemillionsofdollars.Indexarbitrage(theformofarbitrageattheheartofthe1987crash,forexample)exploitsdiffer-encesbetweenthepricesofindexfutures,forinstanceontheS&P500index,andthepricesofthestocksmakinguptheindex.Onthedayinquestion—describedinBeunzaandStark(2004)—thenumbersrepresentingstockpriceswerebeingtransmittedbythecompetitor’sReutersserverwithsmalldelays,whilefuturespriceswerearrivingnormally,givingtheappearance(onadayonwhichthemarketroseconsistently)ofpersistent,attractivemispricing.Seeingapparentopportunities,thecompetitor’sarbitrageurstradedinhugevolumesandincurredlargelosses.Asatraderexplained,‘whiletheywerebuying,wewereselling...thetradersherewerewritingticketsuntiltheirfingerswerebleeding.Wemade$2millioninanhour,untiltheyrealizedwhatwashappening.’ThewayinwhichoneofthetradersinthebankstudiedbyBeunzaconfig-uredhistwoUnixworkstationsandBloombergterminalindicatestheimpor-tanceofspeedsoftransmissionandprecisetimes.Everyday,theclocksinhisUnixworkstationsaresynchronizedtoanatomicclock.Acrossthetopofoneofhisthreescreens,hehasaslashsignthatrotatesandmovesfromsidetoside. 94/TheMaterialSociologyofArbitrageItisa‘pulsemeter’,usedtogaugethe‘pricefeed’—thespeedwithwhichinfor-mationonpricesisarriving—whichstopsmovingwhenpricesstoparriving.OnhisotherUnixworkstation,thetraderhasfivecolouredsquaresthatworkas‘speedometers’,indicatinghowquicklyordersaregettingthroughnetworkservers:iftheyaregreen,everythingisfine;iftheyareyellow,thenetworkiscongestedanddealsaredelayed;iftheyarered,serversareclogged.Two‘CPU-meters’alsomeasurecongestioninthebank’sorderflow(BeunzaandStark2004).Thespeedwithwhichthephysicalentitiesthatembodypricesmoveisnotdictatedbyphysicalandtechnologicalconsiderationsalone.Forinstance,intheearlydaysofconnectionbymodemtotheNewYorkStockExchange,memberfirmstriedtoobtainanadvantagebyinvestinginfastermodems.Topreventcostlycompetitioninever-fasterhardware,a‘speedlimit’of9.6Kbaudwasputinplace.DuringBeunza’sobservations,thislimitbecameaproblemforsomeofthebanksaffectedbytheterroristattackof11September2001.Thosebanks’existing9.6Kbaudcommunicationsystemswereinoperable,butsufficientlyslowmodemswerenolongercommerciallyavailable(BeunzaandStark2003).Suchdeliberatebarrierstocompetition,however,havebeenerodingfast,andan‘armsrace’hasbeenunderwayforsometimeamongstarbitrageurs,andalsothoseusingautomatedorder-placingsystemstooptimizetheirtrad-inginotherways,inrespecttotransmissiondelaysincomputernetworks(Bear,Hod,Enness,andGraham2006).Inthisarmsrace,thephysicalplace-mentoftherelevanthardwarematters.Inparticular,itiscrucialtominimizeevenmicrosecondsofdelaybetweenatraderinitiatingthepurchaseorsaleofsecuritiesandthetradebeingimplementedonanexchange’sorder-matchingsystem.Firmsarethuspreparedtopayapremiumtohavetheircomputersys-temsphysicallyclosetotheexchange’ssystemandconnectedtoitasdirectlyaspossible.Theendofface-to-facetradingonexchangefloorshasmeantthatthehumanbodiesparticipatinginsuchtradingneednolongerbelocatedinoneplace,butarecentralizationoftechnologicalsystemsisrunningalongsidethedecentralizationofbodies.3Alsosometimesimportanttotheconductofarbitrage,andsimultaneouslybothphysicalandsocial,aretheallowableformsthatpricescantake.Ineconomictheory,pricesaretypicallyrepresentedasrealnumbers—theequiv-alentofthepointsonacontinuousline—whilethephysicalembodimentofpricesinmarketpracticerequirespricestobeasubsetoftherationalnumbers, TheMaterialSociologyofArbitrage/95thenumbersexpressibleasaratioofintegers(seeMirowski2002:543).Further-more,theallowablesubsethasoftendifferedfromthatintheeverydayusageofmoney.Forexample,asnotedinChapter2,untilJune1997stockpricesintheUSAweredenominatedineighthsofadollar.Thesizeofminimumunitsofpricesuchasthisaffectsthemagnitudeofthediscrepancyatwhicharbi-tragebecomesfeasible.AfterJune1997,whentheNewYorkStockExchangereducedtheunitofpricetoasixteenthofadollar,thetypicaldiscrepancythattriggeredindexarbitragedecreased,andtherewerelargernumbersofindexarbitragetrades(HenkerandMartens2005).Whenin2001theSecuritiesandExchangeCommissionrequiredthefurthershiftfrombinaryfractionstodecimals(dollarsandcents),theadditionalfacilitationofarbitragemeantthatindexarbitrageursfacedenhancedcompetition—asoneputit,‘Weusedtomakeourmoneybygettinginbetweenthesixteenthincrements’—andsomeincurredlosses.Amongstthematerialentitiesinvolvedintheperformanceofarbitragearearbitrageurs’bodies.Concludingatransactionoverthetelephonewithonepartytobuygold,acurrency,orotherasset,whileatthesametimetellingacolleaguetosellittoanotherpartyatahigherprice,isunlikelytosucceedifone’sconversationwiththecolleaguecanbeheard.Itisthusimportantinthis(andinmanyotherusesofthetelephoneinfinancialmarkets)thatoneswitchesoffthetelephoneorvoiceboxmicrophonewhentalkingtocol-leagues.Thetelephonesusedindealingroomsoftenhavethumb-operatedswitchesbehindtheearpiecethatmakeiteasytodothis,andmanypeoplealwaysusethemtoswitchoffthemicrophonewhenthepartyattheotherendofthelineisspeaking,evenifthereisnoparallelconversationforthemtooverhear.Thatway,itbecomesabodilyhabitthatwillnotdesertoneinsituationsofexcitementorstress.Electronicallyconductedarbitragecanalsoinvolvebodilyskills.Suchtrad-inginvolvesplacing‘bids’(offerstobuy)or‘asks’(offerstosell)fortheassetinquestion.Thisisgenerallydonebyusingacomputermousetoclickonascreenthat,atleastinthecaseofelectronicallytradedfutures,showsforeachpricelevelthenumbersofbids(ofteninblue)andofasks(ofteninred).Atbusytimes,thesenumbersandlevelschangefromsecondtosec-ond,withblueandredbarsseemingtodanceupanddown.Ifanarbitrageopportunitypersistsonlyforseconds(asisoftenthecase),constantattentionandrapidphysicalexecutionareneeded.TheanthropologistCaitlinZaloom 96/TheMaterialSociologyofArbitragereportsthatastraineefuturestraderssheandhercolleaguesweremaderepeatedlytopractisewithacomputerizedgold-pricearbitragesimulation,sothatthedisciplinedattentionandfast,accurateactiontheywouldneedbecamebodilyhabits.Theywereencouraged‘toplaycommercialvideogamesonourowntimetoincreaseourreactionspeedsandhand-eyecoordination’.Aparticulardangertheyweretrainedtoavoidwas‘fatfingering’,inwhich,forexample,insteadofleft-clickingthemouseto‘jointhebid’(puttinginanoffertobuyatasetprice)theyaccidentallyright-clicked,inadvertentlybuyingtheassetinquestionatitscurrentmarketprice.Themanagers’aimwasto‘trainourbodiestooperateasuninterruptedconduitsbetweenthedealingroomandtheon-lineworld,allowingourfingerstobecomeseamlessextensionsofoureconomicintentions’(Zaloom,personalcommunication;seeZaloom2006).Thebodilyaspectsofarbitragearemostprominentwhenitisperformedinopen-outcrytrading‘pits’:steppedamphitheatres,traditionallyoctagonal.Dozensorhundredsoftradersstandontherungsofapit,makingdealsbyvoiceorbyeyecontactandanelaboratesystemofhandsignals.InChicago(theprimesiteofopen-outcrytrading),thehand-signallanguagethatisusediscalled‘arb’becauseitsspeedwasessentialtoarbitrage.Forexample,whenatradingfirmspottedanarbitrageopportunity,canonicallybetweenthepricesofgoldfuturestradedinChicagoandinNewYork,itwasquickerto‘arb’(hand-signal)instructionsfromthefirm’sboothtothetradingpitthantosendaclerkrunningtothepitwithawrittenorder(Lynn2004:57–9;seealsoZaloom2006).Wherebodiesarepositionedwithrespecttoeachothercanbeofconsid-erablesignificancetoarbitrageinopen-outcrytrading.Forexample,thetwomainformsofoptionarecalls(optionstobuyataset‘exerciseprice’)andputs(optionstosellatasetprice),anddiscrepanciesbetweencallandputpricescanbeexploitedbyarbitragessuchas‘conversion’.(Inconversion,atradersellsacalloptionandsimultaneouslybuysaputoptionwiththesameexercisepriceandexpirationplusthestockorotherunderlyingassetinquestion.)OptionsarbitrageursontheAmericanStockExchangefounditadvantageoustostandinbetweenthe‘specialist’(designatedmaintrader)responsibleforcallsandthespecialistresponsibleforputsonthesamestock.Thatwastheoptimumbodilypositionfordetectingandexploitingopportunitiesforconversionandsimilararbitrages. TheMaterialSociologyofArbitrage/97Thatthephysicallocationofbodiesisofsuchimportancetoopen-outcrytradingisonereasonwhy,asnotedinChapter2,pitsandtradingfloorsarebodilyplacesinanespeciallystrongsense:theextensivejostlingandoccasionalfistfightsareoftenovertherighttooccupyaparticularspatiallocationinapit.Notonlyareheightandaloudvoiceadvantages,butevenmundaneartefactssuchasshoesareconsequential:Chicagotradersoftenwearplatformheelstoincreasetheirheight.Overthelasttenyears,however,open-outcrytradinghasdeclinedrapidly,anditmaysoonbecomeextinct.Nevertheless,bodilylocationstillmattersinthescreen-basedtradingthathasreplacedit.Becausearbitrageinvolvestradingpairs(orlargersets)ofsecurities,theexpertiseitdemandsisoftendistributedacrossmorethanoneperson.Thephysicalplacementoftradersandothermembersofstaffintheinvestment-banktradingroomstudiedbyBeunzawasdesignedquiteconsciouslytofacil-itatethenecessaryinteraction,andalsotoinhibitinteractionsthatmightbedetrimental(BeunzaandStark2004).Forexample,thebank’scustomerdeskexecutesordersforclientswhilethe‘specialsituations’desktradescomplex,hybridstrategiesforthebankandisnotindirectcontactwithclients.Thetwosetsoftraderssitfacingeachother,separatedonlybytheirkeyboardsandcomputermonitors.Onewayinwhichthecustomerdeskgeneratesideasisbyconsideringwhycustomersmightbedoingwhattheyaredoing:forexample,astheheadofthecustomerdeskputit,ifsomeonepaysalotforastock,itcanbeinterpretedas‘whatdoesheknowthatIdon’t?’Thephysicalproximityofthesituationsdeskmeantthatpossibleinterpretationsofpuzzlingclientorderscouldquicklyandinformallybediscussedbetweendesks:I[theheadofthecustomerdesk]lookedatitandsaid,‘whydoeshedothat?’ItalkedtoJosh[aproprietarytrader]anditdidn’tmakeanysense.‘Thatguy’scrazy,’wethought.Thatwasthetip-off.Westructuredwhatwethoughtwasabettertrade...‘Itislikebrainstorming’,theheadofthedesksaid:‘wereallydon’tknowwhatwe’regonnathinkintheend.Icouldhavetold“buy”tothoseguys,andconcludefiveminuteslaterthatitwas“sell”.’Incontrast,thefour‘statisticalarbitrageurs’inthebank’stradingroomweredeliberatelykeptatsomedistancefromeachother.‘Statisticalarbi-trage’involvesthedetectionofpatternsintheever-changingfluxofsecurityprices.(Itisoftheoreticalsignificancebecausebydetecting,exploiting,and 98/TheMaterialSociologyofArbitrageeliminatingpredictablestatisticalstructureinpricemovements,itmayleavebehindonlyunpredictablerandomness,makingmarkets,intheterminologyoffinancialeconomics,‘weakform’efficient.Statisticalarbitrageurs,however,denythatstructureisentirelyeliminatedinthisway.)Thestructureexploitedbystatisticalarbitrageursisseldomorneverdeterministic:theyhopeonlytodobetterthanpurechance.Inconsequence,therisksoftheactivityaremin-imizedifdifferentstatisticalarbitrageursareexploitingdifferentpatterns,andincreasedif—perhapsbecausetheyhavesharedideas—theirtradingbecomestoosimilar.Asaseniortraderputit:‘Wedon’tencourage[statisticalarbi-trageurs]totalktoeachother.Theysitapart’(seeBeunzaandStark2004).TheSocialityofArbitrageApriceisathing,butitisalsosocial.Allformsofarbitragedependfortheirsuccessonwhatotherswilldo.Evenintheclassicformsofarbitragethatexploitdifferencesinthepricesofthe‘same’assetindifferentplaces,othersmustbedependedupontofulfiltheirobligations:forexample,todelivergoldifthearbitrageurhasstruckadealtobuyit,ortodelivermoneyifthearbitrageurhassoldgold.ProcedurescarriedoutbyothersmustalsobereliedupontoensurethatgoldinRiyadhis‘thesame’asgoldinManhattan.Othersyetagainmaybeneededtotransportgoldfromoneplacetoanother.(Whensecuritieswerepapercertificates,theirtransportationfromplacetoplaceandtheriskoflossofthemduringsuchtransportationwereissuesthatarbitrageurshadtoconsider.)The‘sameness’ofgoldisestablishedbyassayprocedures‘external’tothemarketthatcanbetreatedbymarketpractitionersasa‘blackbox’—areliableprocesswhosedetailstheydonotneedtoconsider—andnowadays‘trans-portation’ofsecuritiesisalsousuallytreatedbytradersasablack-boxmatter.However,many—probablymost—currentformsofarbitrageexploitdiscrep-anciesinthepricesnotofthe‘same’assetbutof‘similar’assets:Brazil14sand40s;stocksandstock-indexfutures;stocksandoptionsonthosestocks;ItalianandGermangovernmentbonds;newlyissued(‘on-the-run’)govern-mentbondsandpreviouslyissued(‘off-the-run’)bonds;governmentbondsandbondscarryingimplicitgovernmentguaranteesbutbackedbypoolsofmortgages;thesharesofthetwolegallydistinctbuteconomicallyintegrated TheMaterialSociologyofArbitrage/99corporationsthatuntil2005madeuptheRoyalDutch-Shellgroup;andsoon.However,thesimilarityofassetssuchastheBrazil14sand40s,orofsharesinRoyalDutchandinShell,depends,atleastovertheshortandmediumterm,onotherswithinthemarkettreatingthemassimilar,andthearbitrageurcanseldomaffordtotreatthisasablackbox.The‘similarity’offinancialassetsisalwaysinasensetheory-dependent.Sometimes,thetheoryinquestionisasophisticatedmathematicalmodel,suchasBlack–Scholes–Mertonoptionpricing.Atothertimes,thetheoryisvernacularanddowntoearth:forexample,thatthe40swillremainBrazil’smostliquidgovernmentbonds,orthattheintendedEurozonewouldcon-verge,makingItalianbondssimilartoGermanbonds.Toembarkuponarbitrage,tradersthushavetoconvincethemselvesthatthetheoryonwhichthearbitragerestsiscorrect,oratleastplausibleenoughtobethebasisofpracticalaction.Theywilloftenalsowanttoorneedtoconvinceothers.Inourobservationsbothoftheinvestmentbankandthehedgefund,therewasmuchdiscussionofpossibletradesandofthetheoriesunderlyingthem,bothinsidetheorganizationandintheformofanalysescominginfromoutside(andoccasionallyflowingintheoppositedirection).Criticalrolesinthesediscussionsareoftenplayedbymaterialrepresentationsofvalue,suchasthechartshowingtherecenthistoryofthedifferenceinpricesbetweenthe40sandthe14s,ora‘spreadplot’,showingtherelativepricesofHewlettPackardandCompaq,whichBeunzaobservedbeingcloselyfol-lowedin2001–2by‘riskarbitrageurs’hopingtoexploittheprobable—butnotcertain—mergerbetweenthetwocorporations(BeunzaandMuniesa2005).Butmaterialrepresentationsareoftennotontheirownconclusive:informa-tionaboutwhatothertradersaredoing—forinstance,aboutthebehaviourof‘realmoney’intheBrazilianbondmarket—canalsobeimportantinallowingtheplausibilityoftheoriestobejudged.Theneedtoconvinceothersdoesnotnecessarilyceaseonceatradertakesonanarbitrageposition.Often,thepricediscrepancythatisbeingexploitedwillincreasefurtherbeforeitdecreases,whichmeansthatthearbitrageurwillincurapparentlosses.(Forexample,inSeptember1998,theHarvardendow-ment,whosearbitrageactivitieswerecaughtupinthecrisissurroundingLTCM,hadrackeduptemporarylossesthattheWallStreetJournal(Sandler1998)reportedtobeinexcessof$1billion.)Sometimes,apparentlossesareactualoutflowsofmoneyorsecurities(or,atleast,theelectronictracesthereof),forexampleasaresultofthedailyprocessinwhichexchangeclearinghouses 100/TheMaterialSociologyofArbitrageadjustthe‘margin’depositsthatparticipantsmustmaintaininordertobeallowedtocontinuetoholdtheirpositions.Atothertimes,therearenoactualoutflows,butasbanksandhedgefunds‘marktomarket’(revaluetheirtradingpositions,whichisnowalsonormallydoneatleastdaily),apositionshowsaloss.Ineithercase,thelosseswillbetemporary(theoutflowwillbereplacedbyaninflow,a‘paper’losswillturnintoarealizableprofit)ifthetheoryunderpinningthearbitrageiscorrect,butothersmayneedtobeconvincedofthistoallowthearbitrageurtocontinueholdingtheposition.Inalargeinstitutionsuchasabank,theimmediatelyimportantaudienceforarbitrageisanarbitrageur’smanagerormanagers,whowillnormallybecloselyattentivetothe‘P&L’(profitandloss)figuresofthosetheysupervise.‘There’sasayingintradingcircles’,onetraderandmanagertoldus:‘thewhitesheet[P&Lsheet]doesn’tlie’—lossesarereal,andshouldbeacteduponasiftheyarereal.Thearbitrageur’sproblem,however,isthatfromhisorherviewpointthewhitesheetdoessometimeslie,atleasttemporarily.Acommoncomplaintamongstarbitrageursisofbeinginstructedbymanagerstoliquidateloss-bearingpositionsthattheywerecertainwouldbecomeprofitable.Even‘textbook’arbitragescanbesubjecttothisrisk:thetradersintheJapanesesecuritiesfirmstudiedbyMiyazaki(2003)reportedbeingforcedtoabandonindex-arbitragepositionsbecauseoftheapparentlossesincurredwhentheyhadtodepositadditionalfuturesmargin.Suchmanagementbehaviourmayseemincomprehensibleuntilonerealizesthattheboundarybetweenarbitrageandspeculationisporous,anditcanbehardformanagerstobecertainthatarbitrageurshavenotinfactstartedtospeculateontheriseorfallofprices.Twoofthemostcelebrated‘roguetraders’—NickLeesonofBaringsBankandJérômeKervielofSociétéGénérale—werearbitrageurswhocovertlybecameverylarge-scalespeculators.Inhedgefunds,themanager/arbitrageurdivideseemstypicallytobemuchlessmarked:eveninlargefundssuchasLTCMthetworolesarenotdistinct.Investors,however,formamoreimmediateaudiencethantheydointhecaseofbanks.Hedgefundsreportchangesinnetassetvaluestotheirinvestorsmonthly,whilebanksreportquarterlyorlessfrequently(dependingonthejurisdictioninwhichtheyareincorporated),andlossesinahedgefund’stradingarenotmaskedbytheprofitabilityofotherlinesofbusinessastheyoftenareinbanks.Theincreasinglyimportant‘fundsoffunds’,whichallocatetheirinvestors’capitaltohedgefundsthattheyselect(andalsofrequentlywithdrawit),areabletodemandreportsmoreoftenthanmonthly:sometimes TheMaterialSociologyofArbitrage/101evendaily.Soalargelossbyahedgefundconductingarbitrage—evena‘paper’loss—quicklybecomesvisible.Onehedge-fundmanager(andformerinvest-mentbanker)toldusthatinabank‘youcanjustifywhyyouwanttoholdontothosepositions’,whilehedgefundinvestors‘don’tcare.Theyjustlookatthenumber[changeinnetassetvalue]’.Thethreatofinvestorswithdrawingtheircapitalfromthefundisthusalmostcontinuous:‘thereisverysmalltolerancetolosingmoney....[W]ecannothavealosingmonth.’Theriskofarbitrageursinabankhavingtoabandontheirpositionsbecauseoftemporarylossesisreducedifmanagersunderstandandacceptthethe-oryunderpinningatrade,andthusbelievethatlosseswillindeedbetempo-rary.OneadvantageofinvestmentbankswithlongexperienceofarbitrageovernewcomerssuchastheJapanesefirmstudiedbyMiyazakiisthatthisunderstandingismuchmorelikely.Often,though,thetechnicaldetailsofarbitragetradingaredauntingeventothosewithextensivemarketexperi-ence.Forexample,arbitragebetweenUSgovernmentbondsandmortgage-backedbondsissuedbybodieswithimplicitgovernmentguaranteesinvolves(a)adjustingthespreadbetweentheiryieldstotakeaccountoftheconse-quencesofthefactthatmostmortgagecontractsallowborrowerstotakeadvantageoffallsininterestratesbyrepayingamortgagewithfundsborrowedmorecheaplyfromanotherlender,and(b)offsettingthoseconsequences,forexamplebyappropriatepurchasesofinterest-rateoptions.Neithertheadjustmentnortheoffsettingisanelementarymatter.Insuchcases,trustinarbitrageofteninpracticehastobetrustinthearbitrageurorarbitrageursasparticularpeople,justasinmanycasestrustinsciencecomesdowntotrustinthescientist(seeShapin1994).Ahedgefund,auniversityendowmentmanager,oranindividualtraderortradingdeskatabankwhoorwhichhasbuiltupagoodreputationismorelikelytobetrusted.LTCM’sfounderJohnW.MeriwetherhadledWallStreet’spremierarbitragedesk(atSalomonBrothers),andhiscolleaguesincludedothertraderswithhighpersonalreputations.TheywereabletohaveLTCM’sinvestorsacceptathree-year‘lock-in’inwhichtheywerenotallowedtowithdrawcapital,andevenafterthenearbankruptcyin1998theysuccessfullyrecruitedinvestorstoasuccessorfund,JWMPartners.Duringthe1998crisis,theoverseersoftheHarvardendowmentappeartohavetrusteditsmanagement,andratherthaninsistingthatloss-bearingpositionsbeliquidated,theytoleratedtheapparenthugeloss—thusmakingitpossibleforittobetemporary(asindeeditwas). 102/TheMaterialSociologyofArbitrageLosses,eventemporary,caninadditionbeavoidedifotherarbitrageursandprofessionaltradersalsocometoviewthepricedifferencethatanarbitrageurisexploitingasadiscrepancy.Inourobservationsandinterviews,wewerestruckbytheextentofthecirculationamongsttradersindifferentfundsandbanks,mainlybyelectronicmail,ofideasfortradingandbytheattentionmostprofessionaltraderspaytowhatothersseemtobedoing.Ifthatdiscussionandattentionleadsothersalsotoseektoexploitadiscrepancy,thentheirpurchasesandsaleswillnarrowthediscrepancy,oratleastreducetheriskofitwidening.That,forexample,waswhythetraderdiscussedinthischapter’ssecondsectionwantedthechartdisplayingthe14s/40sanomalycirculatedtoothers.‘AllIwantispeopleeventotalkaboutit,’thetradertoldus.Ifothersalsotookactiononthepricinganomaly,theywouldpreventitwidening.Shoulditwiden,thetraderexplained,hemightevencometodoubthisbelief(the‘theory’behindthetrade)thattheanomalywasadiscrepancythatwouldclose.‘Theremightbeareason[fortheanomaly]Idon’tunderstand.Imighthavetoreconsiderthedecision[toconstructatradingpositionpredicatedonitnarrowing].’Anotherwayofminimizingtheriskofprematurecapitalwithdrawalisdiversification.Ifafund,tradingdesk,orbankholdsawidevarietyofarbi-tragepositions—forexample,indifferentpartsoftheworldandindifferentassetclasses—then,onthefaceofit,thereislittlelikelihoodofenoughofthosepositionslosingmoneysimultaneouslytocreateaseriousoverallloss.(Thematched‘long’and‘short’positionscharacteristicofarbitragemeanthatcommonfactorssuchasglobaleconomicconditions,thelevelsofinterestrates,andthebuoyancyofstockmarketsshouldhavelittleornoeffect.)Diversificationofthiskindwas,forinstance,acoreaspectofLTCM’sstrategy.However,theconstantattentionofmanyprofessionaltraderstowhatothersaredoingmayundercutthebenefitsofdiversification.Iflargenumbersoftradersareallledtotakesimilarpositions,thenarbitragesthat‘ought’tobeuncorrelatedcansuddenlybecomelinked.This,forexample,waswhatcausedLTCM’sdiversificationtofail.LTCMtriedhardtokeepitspositionsprivate:asaverylargemarketparticipantwithalargelylocked-incapitalbase,itwasconcernedlesswiththebenefitsofotherspreventingdiscrepancieswideningthanwiththeirtradingcausingtheopportunitiesitwasexploitingtodiminishorvanish.However,othersdidfrequentlytakeonsimilarpositions,eitherbecausetheywerefollowingthesamegeneralstrategy(inpart,inemulationofLTCM’ssuccess)orbecausetheylearnedspecificsofLTCM’stradingdirectly TheMaterialSociologyofArbitrage/103orindirectlyfromthosewhotooktheothersideofthosetrades.‘Ican’tbelievehowmanytimesIwastoldtodoatradebecausetheboysatLong-Termdeemeditawinner’,saysonehedge-fundmanager(Cramer2002:179).TheresultantoverlappingsetofarbitragepositionsmadeitpossibleforaneventtowhichLTCMitselfhadonlyalimitedexposure—theRussiangovernment’sdefaultonitsrouble-denominatedbondson17August1998—suddenlytocausehighlycorrelatedadversepricemovementsacrosstheglobeandinapparentlyunrelatedassetclasses.ArbitrageurswhoincurredlossesinRussiahadtoliquidatepositions(eveninapparentlyunrelatedassets)tomeetmargincalls,withdrawalsbyinvestors,andotherdemandsontheircapital.Inaggregate,thepositionstheysoughttoliquidateoverlappedconsiderablywitheachotherandwithLTCM’sportfolio.Theseliquidationsinturncausedmorelosses,leadingtofurtherliquidations,andsooninadisastrous,market-paralysingspiral.Thesocialityofarbitragegoesbeyondrelationstoparticularotherssuchasmanagers,hedge-fundinvestors,andotherarbitrageurs:theconductofarbi-trageisaffecteddeeplybytheformsofactioninfinancialmarketsthatareseenaspermissibleandtobeencouragedorasimpermissibleandtobediscouraged.Onepersistentissueisthedifferenceinthisrespectbetweenthetwostandard‘legs’ofanarbitragetrade.Typically,apricediscrepancyisexploitedbybuying(orinotherwaystakinga‘long’position)inanundervaluedasset,andshortsellingasimilarovervaluedcounterpart.Longpositionsarealmostalwaysregardedasunproblematic,butshortpositionshavehistoricallyoftenbeentheobjectofsuspicion.Shortsellersarefrequentlyblamedforfallsinprice,andtheactivityisseenasmorallyreprehensibleforotherreasons:forinstance,incurrentinterpretations,thesecuritiesborrowinginvolvedinshortsellingiscontrarytoSharia,creatingaproblemforthosewhowishtosetup‘Islamic’hedgefunds.Insomemarkets(forexample,Mexicangovernmentbonds),onlyspecific,trustedmarketpar-ticipantsareallowedbyregulatorstosellshort.Inothermarkets,shortsellingbyawiderangeofparticipantsispermittedbutisconstrainedinotherways.Until2007,forexample,shortsalesofstockintheUSAweresubjecttothe‘uptickrule’(see,forexample,Robottin.d.)—theywereprohibitedunlessthelastpricechangehadbeenupwards—whichcouldcausesubstantialdelaysinshortsellingifpricesarefallingconsistently.Notalltheproblemsofshortsellingaretheresultofdeliberatepolicy(otherconstraintsincludetheavail-abilityofsecuritiestoborrow,thecostofsuchborrowing,andsometimesthe 104/TheMaterialSociologyofArbitragevulnerabilityofshortsellerstopredatorytradingbythosewhohopetoprofitbyforcingthemtounwindtheirshortpositions),buttheresultantdifficultiescanbecrucial.Accordingly,forexample,the‘stockloan’desk(whicharrangesstockborrowing)occupiesapivotalpositiononthebanktradingfloorstudiedbyBeunza.Becausetheextentoftheproblemsofshortsellingvariesfromassettoasset,systematiceffectsoftheseproblemscanbedetected.ThusDowJonesfuturesandotherstock-indexfuturesseemtotendmoreoftentobebelowthevalueimpliedbythelevelofunderlyingindexthanaboveit(Shalenn.d.).Thetradingrequiredtoexploit‘overpricing’offuturesisstraightforward:thearbi-trageurhastoestablishashortpositioninfutures(whichmeanssimplysellingfutures,andinvolvesnoparticulardifficulties),whilebuyingthestocksthatmakeuptheindex(alsostraightforward).Incontrast,exploiting‘underpric-ing’offuturesrequiresthearbitrageurtobuyfutures(againstraightforward),butitalsoinvolvesshortsellingtheunderlyingstocks,whichis,asnoted,oftenmoreproblematic.Arbitragecansometimesalsoraisemorespecificquestionsofpropercon-ductinmarkets.AninterestingcaseofthiswashugetradesinEurozonegov-ernmentbondsandbondfuturesundertakenbyCitigroupGlobalMarketsLtd.onthemorningof2August2004.Likeconventionalarbitragebetweenfuturesandtheunderlyingasset,Citigroup’stradingintendedtoexploitadiscrepancy,butinthiscaseitwasnotapre-existingpricediscrepancybutoneresultingfromdifferencesintheliquidityofthebondandbondfuturesmarkets.InthewordsofaCitigroupinternalmemo(quotedinSkoreckiandMunter2005),Citigroup’stradershadnoticedthat‘theliquiditybeingofferedinthebondsisfargreaterthanthatofferedinthebund[Germangovernmentbond]future’.(BundfuturesarethebenchmarkEuropeangovernmentbondfuture.)Inconsequence,astandard‘marketneutral’positionofthekindoftenconstructedbyarbitrageurs(inthiscase,shortfuturesandlongbonds)couldbeunwoundprofitably,evenintheabsenceofapre-existingpricediscrepancy,ifthepositionwasverylarge.Buyingfuturesonalargescaletounwindtheshortfuturespositionwouldleadtolossesthrough‘slippage’(thepriceoffutureswouldriseasthepurchaseswerebeingmade),butthearbitrage-imposedlinkbetweenfuturesandbondpriceswouldthencausethelatteralsotorisecorrespondingly.Atthatpoint,thegreaterliquidityofthebondmarketmeantthatthelongpositioninbondscouldbeunwoundbysellingbondsatthesefavourableprices,andcruciallythesaleswouldnotleadtoslippage: TheMaterialSociologyofArbitrage/105theycouldbecompletedbeforepriceswereforceddown.Theasymmetryinliquiditywouldmeanthattheprofitfrombeingabletosellbondsattheseelevatedprices,withoutslippage,wouldmorethancompensateforthelossescausedbyslippageinthefuturespurchases.Thebond-marketliquidityonwhichtheprofitabilityofCitigroup’stradedependedhadnotarisenspontaneously,butwastheresultofaconscious‘liquiditypact’betweenthebanksandcontinentalEuropeangovernmentsusingtheMTSbond-tradingsystem,whichhasnowexpandedfarbeyondItaly.Thegoalofthepactistoensurethattheeuro-denominatedgovernmentbondsofeventhesmallerEuropeancountriesremainliquid.Onmostprice-disseminationsystems,pricesareindicative:onecanconcludeadealonlybydirectlycontactingtheparticipantthathaspostedaprice,anditisnotobligedtotradeatthatprice.OnMTS,incontrast,thebanksusingthesystemhavetocommitthemselvestotradeuptoasetquantityofbondsatpostedprices.Thisexposesthemtotradinglosses,butitmakestheEurozonegovernmentbondmarket,despiteitsfragmentation,moreliquidandthusmoreattractivetoinvestors.TheFinancialTimesclaimsthat‘bankswerepreparedtosubsidisetheirMTSbusiness’becausegovernmentsoftenselectbanksatthe‘topofthelistintermsofMTStradingvolumes’whenawarding‘lucrativebusinesssuchasderivativestransactionsorsyndicatedbondsales’(vanDuynandMunter2004).CriticaltoCitigroup’stradewasitsmateriality:itrequired‘hitting’allthebids(offerstobuybonds)ontheMTSsystemnearlysimultaneously,ataskthatwasimpossiblemanually.SoCitigroup’stradersdevelopedsoftware—whichtheyreferredtoasthe‘spreadsheet’—todothis.From9.12to10.29a.m.on2August2004,theymadetheplannedbondfuturespurchases,andthepriceofthesefuturesandofEuropeangovernmentbondsroseasanticipated.At10.28a.m.,theylaunchedthe‘spreadsheet’,withtheintentionofsellingbondstothevalueofC8billiontoC9billion(FinancialServicesAuthority2005).Althoughthe‘spreadsheet’hadbeentestedinsimulatedtrading,andhadbeenoperatedonasmallscaleintheactualmarket,ithadbeenimpossibletotryitoutonanythinglikethescalerequiredon2August,sohowitwouldoperateinmaterialrealitywasnotknownwithfullcertainty.Sometwentysecondsafterlaunchingthe‘spreadsheet’,thetradersbecameconcernedthatithadnotfunctionedproperly,sotheyactivateditasecondtime.Infact,ithadworkedevenbetterthananticipated.Inconsequenceofthesecondactivationandoftheunanticipatedsuccessofthefirstactivation,insteadofthebondsales 106/TheMaterialSociologyofArbitragecancellingoutearlierpurchases,CitigroupwasleftwithanetshortpositioninEuropeangovernmentbondsofC3.8billion.The‘spreadsheet’(designedtosellbonds)washurriedlyreconfiguredtobuythem,andreactivatedat11.25a.m.(FinancialServicesAuthority2005).AlthoughCitigroup’stradinghadnotgoneasplanned,itmadeaprofitofalmost£10million,muchmorethanthetradewasexpectedtoyield,becauseafterCitigroup’shugebondsalesotherparticipantsinMTSsharplyreducedthepricestheywerequoting,andthe‘spreadsheet’wasthusabletobuybondsbackatpricessignificantlylowerthanthoseatwhichithadsoldthem.Whatisofinterest,however,isthereactiontothetrade.Citigrouphadnottradedoninsideinformation,norhaditspreadfalserumours.Yet‘[m]anytradersonthedayCitigroupdiditsdealthoughtthebankwasbreakinga“gentleman’sagreement”’bytakingadvantageonsuchahugescaleofthe‘forcedliquidity’requiredofparticipantsinMTS.Aleadingbankersaidthat:‘BysomeEuro-peangovernmenttreasuries,thistradewasperceivedasopenwarfare.’Thesentimentwasnotuniversal—anotherseniorbankersaid,‘Citigroupspottedawaytomakeaquickbuck.Iguesswejusthavetosaywelldonetothem’(vanDuynandMunter2004)—butCitigroupwaswidelycondemned,andtheUKFinancialServicesAuthority(FSA)forcedittorelinquishtheprofitsfromthetradeandtopayafurtherpenaltyof£4million.TheFSAdidnotaccuseCitigroupofhavingbrokenthelaw,butneverthe-lessheldthatitstradinghadviolatedtwooftheauthority’s‘PrinciplesforBusiness’.Inparticular,theFSAruledthatCitigroupGlobalMarketsLtd.‘didnothavedueregardto...thelikelyconsequencestheexecutionofthetradingstrategycouldhavefortheefficientandorderlyoperationoftheMTSplat-form’(FinancialServicesAuthority2005:2).PreciselyasAbolafia(1996)positsinhisclassicethnographyoffinancialmarkets,behaviourinsuchmarketsisinpracticegovernedbymorethanthepursuitofself-interestandtheconstraintsofthelaw:lessexplicitnormsmattertoo.ConclusionOurargumentinthischapterhasbeenthatarbitrage—howitispractised,itsrisks,itsuncertainties,itslimits,anditscapacitiestoweldmarketstogetherintoafinancialsystem—canproperlybeunderstoodonlyifitisgraspedinits TheMaterialSociologyofArbitrage/107fullmaterialityandsociality.Thatkindofrich,qualitativeunderstandingisofcoursedifferentfromthemoreabstractbutquantitativelymorepreciseunder-standingtypicallysoughtbyeconomists,even‘behaviouralfinance’specialists.Nevertheless,thereareareasofoverlapbetweena‘socialstudiesoffinance’perspectiveandfinancialeconomists’recentinvestigationoftheconsequencesofrelaxingtheirdiscipline’straditionalpuristdefinitionofarbitrage.Forexample,ShleiferandVishny(1997)modeltheriskthatthosewhoprovidearbitrageurswithcapitalwillwithdrawitprematurelyinthefaceoftemporarilyadversepricemovements.BravandHeaton(2002)addresswhatinourtermsisthedifficultythatarbitrageurscanhaveconvincingthemselvesandtheiraudiencesthatapricepatternisindeedadiscrepancythatcanbetheobjectofarbitrage.IncirculatingthechartofthepricehistoryoftheBrazil14sand40s,thetraderweobservedwasseekingtosolveinpracticetheproblemmodelledbyAbreuandBrunnermeier:thelimittoarbitragethatcanarisewhen‘rationaltradersfaceuncertaintyaboutwhentheirpeerswillexploitacommonarbitrageopportunity’(2002:341).Attari,Mello,andRuckes(2005)modelariskthatbecameverypertinentforLTCMafterthefund’sdifficultiesbecameknowntoothersatthestartofSeptember1998,butofwhichalllargearbitrageursneedtobewary:thatthecombinationofcapitalconstraintsandpositionsknowntoothertraderscanmakearbitrageurs’actionspredictableandexploitable.ShleiferandVishny,BravandHeaton,AbreuandBrunnermeier,andAttari,Mello,andRuckesputforwardfourseparatemodels,eachcapturingoneoftheaspectsthatwepositasintrinsictoarbitrageasmarketpractice.Nointegratedmodelhasyetemergedfromtheliteratureineconomicsonthelimitsofarbitrage,butourfieldworksuggeststhatitisintheinteractionofarbitrage’saspectsthatitscruciallimitsmayreside.ThusthecrisissurroundingLTCMarosefromthewayinwhichcapitalconstraintsakintothosemodelledbyShleiferandVishnyinteractedwiththeconsequencesofothersimitatingasingleprominentarbitrageur,andLTCM’scrisiswasworsened(toadegreethatishardtodetermine)byothertraders‘arbitragingthearbitrageur’inthemannermodelledbyAttari,Mello,andRuckes.Wewouldthereforebehopefulthatthestudyofarbitragecouldbeaproductiveareaofcollaborationbetweenfinancialeconomistsandthoseinthewidersocialsciencespreparedtotacklefinancialmarketsintheirfullmaterialityandsociality.Weareinadditioncertainthatarbitrageisapivotaltopicforsocialstudiesoffinance.Thedetailsofarbitragemayseemtobelittle 108/TheMaterialSociologyofArbitragethings,buttheyarelittlethingsconnectedtobigissues:as,forinstance,thematerialityofCitigroup’s‘spreadsheet’connectstothe‘forcedliquidity’ofMTSandthustogovernmentbudgets,EconomicandMonetaryUnion,andeventheoverallprojectofEuropeanunification.Thepowersandlimitsofarbitragearecriticaltoglobalfinancialmarkets,andthematerialsociologyweadvocateisneededtounderstandthem. 6MeasuringProfitClinton,Mississippi,Friday,20April2001.TheDeepSouth:duringtheCivilWar,ShermanbasedhimselfherewhenbesiegingthecityofJackson.Clinton’sstreetsarelinedwithhistoricbuildingsandshadedwitholdtrees.YetClintonisalsotheNewSouth,thesiteoftheultramoderncorporateheadquartersofaleadingtelecomcompany—abuildingthatinthewordsofonereporterrises‘likeadarksteelmothership’(Ripley2002).Inthemothership,anaccountingmanagerisatwork,andpartofherjobistoproducefacts.Thetelecomcompanyisbuildingthephysicalinfrastructureoftheinformationage:theoptical-fibrenetworksforsynchronized,high-speed,high-capacitytransmissionofdata.Butitalsoleasescapacityonothercompanies’networks,primarilysoastolinkitsownnetworktothoseoflocalcarriers.Althoughallofthiscapacityhashadtobepaidforunderitsleaseagreements,notallofithasbeenused.Withafewkeystrokesandmouseclicks,sheallocatesthecostsoftheunusedportiontotwoaccounts:‘OtherLongTermAssets’and‘ConstructioninProgress’.Whatshehastherebydoneistoclassifyasetoftransactionsascreationofassets,andtheclassificationwillultimatelyfeedthroughintohercompany’spublishedaccounts.Acommonprejudiceaboutaccountingseesactsofclassificationofthiskindasboring;certainly,fewsocialscientistsoutsideofacademicaccountancyhaveeverstudiedthemindetail.Yetscalesaren’tstable.ThecompanyisWorldCom,with20millioncustomersand60,000employees.ItcarriesmoreInternettrafficandmoreinternationaltelephonecallsthananyothercom-panyintheworld(Jeter2003:xx–xxii).Overthemonthstocome,WorldComwillunravel,andthatFriday’sclassificatoryact—dictatedbytheaccountant’s 110/MeasuringProfitsuperiors—isattheheartoftheprocess.Asitunravels,itundoesmuchofwhatlittleconfidenceremainsintheUnitedStatesincorporatefinancialreporting.ThatconfidencehasalreadybeendamagedbadlybythebankruptcyofEnroninDecember2001.WorldCom’sannouncementon25June2002thatithasidentifiedaccountingirregularitiesamountingto$3.9billionaddstogrowingfearsthatEnronisnotunique.On21July2002,WorldCom’sunravellingculminatesintheworld’slargestevercorporatebankruptcy.WhateverdoubtspoliticalelitesintheUnitedStateshavefeltabouttheSarbanes–OxleyBill,withitsdraconianmeasurestoregulatefinancialreportingandpunishdubiouspractices,vanish.Fourdayslater,theHouseapprovesSarbanes–Oxley423:3,andtheSenatevotes99:0initsfavour.On30July,GeorgeW.BushsignstheActontothestatutebook.Overtheyearsthatfollow,internationalcompaniesseekingstock-marketlistingsstarttochooseLondon,inparttoavoidSarbanes–Oxley’scostsofcomplianceandheavypenalties.AfewkeystrokesonaFridayinMississippihavecontributedtoashiftinthecentreofgravityoftheworld’sfinancialmarkets.TheMeasurementofFinancialPerformanceTheconsequencesofthatFriday’sactionswereunusualintheirscale,buttheactionsthemselveswerenot.Today’ssocietiesaresaturatedwithmeasure-ment,tosuchanextentthatithasoftenbecomeinvisible.(If,forexample,apartinyourbicycleorcarfails,itsreplacementwillfitexactly,withnofilingorreshapingneeded.It’samundaneachievement,butonethatrestsonahuge,hiddeninfrastructureofmeasurementandprecisionengineering.1)Amongstthethingswemeasureisfinancialperformance,especiallyprofit,anditisperfectlyconceivablethattherearemorespecialists(book-keepers,accountants)inthatformofmeasurementthaninanyother.Themeasurementofprofithasbecomeagiganticenterprisebecauseitiscentraltoeconomicgovernance.Asenterprisesbecomebigger,whethertheyareprofitableorloss-makingceasestobeapparenttoeventhemostexperi-encedproprietorialeye:recordkeepingandcalculationarerequired.Whentheybecomepublic,listedcompanies,theirsharescanbeboughtandsoldbyinvestorswithnodirectpersonalcontactwiththecompaniestheyown. MeasuringProfit/111Anaudiencethusemergesforpublicfinancialfactsaboutcompanies,notjustinternalmeasuresofperformance.Thedesirethatthosefactsbetrustworthy—andrepeatedscandalsinwhichtheyturnoutnottobe—meanthattheirproductionisnotjustaprivatematter.Auditors,regula-tors,taxauthorities,governments—allthosenowhaveaparttoplay.Thetopfirmsspecializinginthemeasurementofprofit—the‘bigfour’:DeloitteToucheTohmatsu,Ernst&Young,KPMG,andPricewaterhouseCoopers—areamongsttheleadingglobalcompanies.Financialperformancegenerally,andprofitspecifically,arecrucialsignalsinafree-enterprise,capitalistsystem.Ifanactivityisprofitable,itattractsresources,andintermsofthelogicofsuchasystemitshoulddoso,becauseprofitindicatesthatthevalueofanactivity’soutputisgreaterthanthatofitsinputs.Conversely,ifanactivitymakesaloss,resourcesflowawayfromit.Acompany’sfinancialreportingnowtypicallytakesthreemainforms.Its‘incomestatement’(intheUK,its‘profitandlossaccount’)records,foragiventimeperiodsuchasthreemonthsorayear,thecompany’srevenuesandtheexpensesincurredinearningthem.Its‘balancesheet’recordstheamountsofitsassetsandliabilitiesattheendofthetimeperiodbeingreportedon.Finally,its‘cashflowstatement’recordsthecashitpaidandreceivedoverthetimeperiod.Theneedforthreeformsofreportarisesbecausemeasuringprofitisnotsimplycountingcash.Modernfinancialreportingis‘accrualsbased’,whichmeansthatthereisnosimplecorrespondencebetweenacorporation’scashflowinagiventimeperiodanditsrevenuesandexpensesinthatperiod.Forexample,thecostoftheelectricityconsumedbyacompanyinthecurrentquarteriscountedasanexpenseinthatquarter,evenifthebillhasyettoarrive.Ifanitemissoldtoacustomerinthecurrentquarter,thatisrecognizedasrevenue,evenifthecustomerhasnotyetpaid.Asanytextbookofaccount-ing(e.g.Perks2004:173–4)notes,acompany’sprofitsunder‘accruals-based’accountingisthusnotthedifferencebetweenthecashithasreceivedandpaidout.Instead,profitisthedifferencebetweenrevenuesearnedinagivenperiodandtheexpensesincurredinearningthem,evenifthecorrespondingcashreceiptsorpaymentstookplaceinanearlierperiodorwillnottakeplaceuntilalaterone.Ifamanufacturingcompanybuysamachine,oranairlineanaeroplane,onlyaportion—oftenquiteasmallportion—ofthecostsinvolvedwillnor-mallybeclassifiedasanexpenseintheperiodofthepurchase,becausethe 112/MeasuringProfitmachineoraeroplanecanbeexpectedtocontinuetogeneraterevenueyearsintothefuture.Instead,thecostsofthepurchasewillappearnotonthecompany’sprofitandlossaccount,butonitsbalancesheet—asan‘asset’.Astheassetages,itsbalance-sheetvaluewilltypicallybereduced,anditisthelossofvalueinagiventimeperiod—the‘depreciation’—thatistreatedastheexpenseandsetagainstrevenueforthatperiod.ThatisthesignificanceofwhattheWorldComaccountantdid.Thecostsofleasingunusednetworkcapacitycouldhavebeenclassifiedasanexpenseanalogoustotheconsumptionofelectricity.Alongwithsomeothersmalleradjustments,‘capitalizing’thosecosts—classifyingthemnotasanexpensebutaspurchaseofanasset,morelikeamachineoranaeroplanethanthecostofelectricity—reducedWorldCom’sexpensesinthesecondquarterof2001by$610million(Beresford,Katzenbach,andRogers2003:108),enoughtoturnwhatwouldotherwisehaveanoveralllossintoaprofit.AccountingClassificationActsofclassificationarefundamentaltothemeasurementofprofit.Allbook-keepersandaccountants,andmanylaypeople,havetodecide—oftendozensoftimesaday—whichaccountingcategorybestfitsatransactionorothereconomicitem.Inanageinwhichmostaccountingindevelopedcountriesiscomputerized,thistypicallytakesthematerialformofassigningtotherecordofthetransactionanumericalcodefromtheorganization’s‘chartofaccounts’,sopermittingtheautomatedprocessingandaggregationofitemsbytheorganization’saccountingsystem.Forinstance,whenIincurredacostchargeabletothemaingrantthatsupportedtheresearchreportedinthisbook,IhadtodecidewhichofthecodeslistedinFigure6.1(asubsetofEdinburghUniversity’schartofaccounts)toassigntoit.SometimesIfoundthetaskstraightforward.IfIwentbytrainwithintheUKtointerviewsomeone,thatstruckmeasunequivocally‘3202’.ButIconfessthatasapurelyamateurbook-keeperIwasoftenalittlepuzzled.IfIpaidsomeone’sexpensestocometotalktoaworkshopIwasrunning,wasthatstill3202?Wasitmaybe3102?WhenIboughtaconferencephonetodotelephoneinterviews,wasthat3102or3512?Givenabitofimagination,Icouldevenenvisageclassifyingthephonepurchaseas3202or3206,becauseitwasasubstituteforphysicaltravel. MeasuringProfit/1133004ResearchTeaching(At)3006ResearchAssistants(Ar)3032ResearchFellows(Af)3102ResearchOtherCosts3138ResearchConsumables3202Travel/Subs–WithinUKonly3206Travel/Subs–OutsideUK/EC3438InsuranceCover3512ExceptionalItems3712ResearchUoEO’headsRecoveredFigure6.1.ExamplesofcodesfromtheUniversityofEdinburgh’schartofaccountsSuchdecisionsarepervasive.Theyincludethefollowing:21.Isanitemanaccountingitem?Consider,forexample,abrand.Therighttocallyourcompany’ssweetfizzydrinks‘CocaCola™’isclearlyvaluable.Inabroadsenseoftheterm,thatrightisasmuchan‘asset’asthemachineryusedtoproduceandtobottlethedrinks;indeed,itmaybeworthvastlymorethanthatmachinery.Therightseemsclearlytomeetformaldefinitionsof‘asset’suchas‘Assetsareprobablefutureeconomicbenefitsobtainedorcontrolledbyaparticularentityasaresultofpasttransactionsorevents’(FASB1985:16).Accountantshave,however,wor-riedthatbrandslackacharacteristicthattheybelieveaccountingitemsmusthave:measurability.AsInternationalAccountingStandard38putsit,‘Anintangibleassetshallberecognised...onlyif...thecostoftheassetcanbemeasuredreliably’(IASB2005:1599).HowcanoneputafigureonthevalueoftherighttocallaliquidCocaCola?Howcanonetellwhetherthatrighthasgotmorevaluableoverthepastyear,orless?Achangeinvaluewouldaffectnotjustacorporation’sbalancesheetbutitsincomestatement,fromwhichitsearningsor‘profits’arecalculated.Ifabrandistreatedasanaccountingitem,woulditsowner’sprofitsbecomeamereopinion?Therearemethodologiesformeasuringthevalueofabrand,andfirmsthatspecializeinthetask,butthecredibilityoftheresultsiscontested(Power1992).2.Ifanitemisanaccountingitem,whatkindofanitemisit?Isitan‘expense’,an‘asset’,a‘liability’,‘income’,‘reserve’,or‘capital’?Asalreadynoted,suchclassificationscanhavesubstantialconsequences,forexam-pleforacorporation’searnings.3.Whatoverall‘scale’shallweusetomeasureanitem?Intermsofthemeasurementofassets,forexample,akeyissueiswhethertouse‘historic 114/MeasuringProfitcost’or‘fairvalue’.3Theformeristhetraditionalapproach:anassetisrecordedatwhatitoriginallycost,minustheextentofitssubsequentdepreciation.‘Fairvalue’,incontrast,is‘Theamountforwhichanassetcouldbeexchanged,oraliabilitysettled,betweenknowledgeable,willingpartiesinanarm’slengthtransaction’(IASB2005:2217).Bothapproacheshavetheiradvocates,oftenpassionateadvocates(thesupportersoffairvaluearecurrentlyintheascendancyinstandardsetting).Fairvaluehasthevirtuethatwhatafirm’sassetsarecurrentlyworthwillusuallybemorerelevantthanwhattheyoncecost.Historiccostmeasurementiscomplicatedbyinflationandbytheneedtoestimatedepreciation,anditcanseeminappropriateinthecaseofassetssuchaslandandbuildingswhosevalueoftenrisesthroughtime.Historiccost,though,hastheadvantagethatitwillusuallybeeasiertoestablishandtodocumentwhatanitemoriginallycostthanwhatitiscurrentlyworth.Toitscritics,‘fairvalue’achievesrelevanceatanunduepriceintermsofobjectivity.4.Howwillmeasurementactuallybedone?Takecosts,forexample.Ifacorporationmanufacturesitemsorsellsservices,howistheircosttobedetermined?Fewproductionorservice-provisionactivitiesareentirelyboundedofffromothersuchactivities.Afactory,forexample,willusu-allyproducemultipleitems,oftenofdifferentkinds.Howarethecostsoflabour,machinery,heat,light,andothercommonservicestobeallo-catedamongstthem?Nordoallproblemsvanishevenifafirmbuysinready-madealltheitemsitsells.Supposesuchafirmsellsanitemfromitsinventoryofsimilaritems.Whatisthecorrespondingcost?Isitwhatitpaidforthemostrecentlyacquiredsuchitem(‘lastin,firstout’),orfortheoldestsuchitem(‘firstin,firstout’),orsomekindofweightedaverage?Ifpricesarechanging,thedifferencebetweentheanswersmaybeconsequential.Similarly,howisrevenuetobemeasured?Inaccruals-basedaccount-ing,afirm’srevenueinagivenperiodis,asalreadynoted,notidenticaltothecashitreceivesinthatperiod.Therevenuefromasale,forexample,is‘recognized’whenanitemissold,notwhenitispaidfor.Itemscan,however,subsequentlybereturned,andsometimescustomerswillfailtopay.Becausereturnsandtheextentof‘baddebts’lieinthefuture,theyneedtobeestimatedifappropriate‘provisions’forthemaretobemadewhenmeasuringrevenues. MeasuringProfit/1155.Towhichtimeperioddoesanitembelong?When,forinstance,isanitem‘sold’?4Whenacustomersaysverballythatheorshewillbuyit?Whenacontractissignedorapurchaseorderreceived?Whentheitemleavesthewarehouse?Whentheinvoiceisissued?Whenpaymentisreceived?Furthermore,manyitemsaremanufacturedoveratimeperiodthatbeginsonlyaftertheyareordered,andmanyservicesareprovidednotatasinglepointintimebutoveranextendedtimeperiod.Isthecorre-spondingrevenueearnedatthestartofthetimeperiod,atitsend,atanintermediatepointorpoints,orgraduallythroughouttheperiod?Ifthelastofthese,whatproportionoftherevenueistoberegardedasearnedatanygivenpointintime?Theallocationofcostsacrosstimeperiodspresentssimilarissues.Takeadvertising,forexample.Isitgeneratingsalesinthecurrentperiod(inwhichcase,itshouldberecordedasanexpenseinthatperiod),orwillitleadtosalesinfuturetimeperiods(inwhichcase,itsrecognitionasanexpenseshouldbedeferredtothosetimeperiods)?6.Isanitem‘current’or‘long-term’?Acompany’sdebts,forexample,areconventionallydividedbetween‘currentliabilities’,whichareexpectedtofallduewithinayearorless,and‘long-termliabilities’,whicharenotexpectedtoneedhonouringwithinthatperiod.Similarly,‘currentassets’—thosethatcanberealizedwithinayearorless—alsoformadistinctsubcategory.Theclassificationisconsequentialbecausecurrentassetsandliabilitiesareprominentcomponentsoftheratiosusedtoassessacompany’sfinancialhealth.5Traditionally,a‘currentratio’(cur-rentassets:currentliabilities)of2:1wasregardedasanindicatorofhealth,whilearatioof1:1orlowercouldbereadasasignalthatacom-panywasclosetoinsolvency.Theclassificationofanitemas‘current’ornotisthusconsequential.7.Wherearetheboundariesofanorganization?6Whichtransactionsfallwithinthoseboundaries,andwhichoutsidethem?AprominentfeatureofEnron’spractices,forexample,wasthecreationof‘specialpurposeentities’thatcould,thefirm’sseniormanagersandauditorsbelieved,legitimatelybetreatedinaccountingtermsasseparatefromit.Thereareseveralentirelyproperreasonsforcreatingsuchentities(theyare,forexample,afoundationoftheentirefieldof‘structuredfinance’,includ-ingmuchofthecredit-derivativesmarkettouchedoninChapter2), 116/MeasuringProfitbuttheycanalsobeusedtoavoidclassifyingitemsasliabilitiesofacorporationonthegroundsthattheyareliabilitiesofthespecialpurposeentity,notofitsparent.How,then,isthe‘separateness’ofanentityfromitsparenttobedefined?Outsideinvestment?Ifso,howmuch(atthetimeofEnron,3percentwasregardedassufficient)?Outsidecontrol?Ifso,howiscontroltobemeasured?8.Isacorporationa‘goingconcern’:thatis,isitssurvivalreasonablyassured?Theanswermakesahugedifferencetohowitsassetsarevalued.Machinery,forinstance,willnormallybeworthalotmoreifthefirmthatownsitcanbeexpectedtocontinueusingittogeneraterevenues,bycomparisonwithavaluationofwhatitwouldfetchifsoldsecond-handorforscrap.Afirm’sdirectorsthushavetocertifyexplicitlywhetherornottheyareemployinga‘goingconcern’assumptionandauditorsmustconsiderwhethertheassumptionisappropriate.Theirdecisiontocon-testthatassumptionisnormallycatastrophicforthefirminquestion.Withitsassetsvaluedonlyattheproceedsofliquidation,theywillusuallybedwarfedbyitsliabilities,andbankruptcyensues.AccountingforEconomicRealityWhatfactorsmightstructurethemultipleexplicitandimplicitchoicesinvolvedinaccountingclassificationandmeasurement?Intheremainderofthischapter,Iexplorevariouspossibleanswerstothisquestion:first,‘eco-nomicreality’;second,therulesofaccounting;andthird,arangeofothersocialandtechnologicalfactors.Letmebeginwith‘economicreality’.Thepurposeoffinancialreporting,itwouldwidelybeagreed,istorepresentaccu-ratelytheeconomicsituationofacorporation,sothatitsexistinginvestors,itscreditors,andotherstakeholderscanassessmatterssuchaswhethertheirmoneyisbeingwellusedorwhethertheywillbepaid,andpotentialinvestorscandecidewhetherornottoentrustthecorporationwiththeircapital.Anobviouscomplicationisthestrongfeedbackfrom‘report’to‘reality’.Financialreportingdirectlyaffectstheeconomichealthofcorporations.Acorporationthatappearssoundandprofitableisattractivetoinvestorsandtolenders,whileabankthatappearsunsoundisvulnerabletoabankrun.7Towithdrawthe‘goingconcern’assumptionwillnormallycauseafirmto MeasuringProfit/117ceasetobeagoingconcern(Hines1988:256).Nevertheless,economichealthisnotsimplyamatterofaccountants’reports.Allreaderswillbeawarethattherearesomepurchasestheycannotmake,andsomepatternsofexpendi-turetheycouldnotsustainwithoutincreasingtheirincome;andsomethingsimilarholdsforcorporations.Acorporationcanbecomeunabletomeetitsfinancialobligationsjustasanindividualcan,andfinancialreporting—howeveroptimistic—maynotpreventthishappening.Indeed,accountingscandalsfrequentlytaketheformofthesuddeninsolvencyofanapparentlyprofitablecorporation.However,thegoalofcapturingeconomicrealityisinsufficienttodeterminethepracticeofaccounting,evenifnootherconsiderationsintrude.Forexam-ple,accountinghasitslocalculturesandtraditions.Notonlyhaveformalrulesandstandardsvariedconsiderablybetweencountries,butpracticesometimesvariesevenwhenrulesseemsimilar.ThusevenpriortorecentEuropeanharmonization,therulesgoverningthedepreciationoffixedassetsweresim-ilarintheUKandFrance,butthetypicalimplementationoftheruleswasquitedifferent(Walton,Haller,andRaffournier2003:23).Asaconsequenceofnationaldifferences,whentheassetsandprofitsofacorporationarecalculatedaccordingtothepracticesofmorethanonecountry,theresultantfigurescandifferconsiderably.In1993,forexample,Daimler-BenzAGlistedontheNewYorkStockExchange,anduntil1996(whenitstartedusingexclusivelyUSrules)itpreparedtwosetsofaccounts,USandGerman.Thevalueofitsshareholders’equity(thedifferencebetweenthevaluationsofDaimler-Benz’sassetsanditsliabilities)was40to45percenthigherinitsUSaccounts.Itsearn-ingsalsodiffered,andthemostdramaticdifference(in1993,Daimler-Benz’sGermanaccountsshowedprofitsof615millionDM,whenitsUSaccountsrecordedalossequivalentto1,839millionDM)seemstohavebeencausedmainlybyrevaluationsdesignedtoreducediscrepanciesinassetvalues(BayandBruns2003:397–9).Membersofdifferentlocalculturesoffinancialreportingmaybelievestronglythattheirpracticesbestcapturereality.Forexample,acontinentalEuropeancorporation’saccountantsmayhavefelttheyweretakingproperaccountofarapidlychanginganduncertainworld,whentotheirBritishandAmericancounterpartstheyseemedtobesaltingawaylargeundeclaredprofits.ThatsuchconvictionscanbepassionatemeansthattheharmonizationofinternationalaccountingstandardsacrosstheEuropeanUnionandtheongoingharmonizationbetweenEuropeandtheUSAhavebeenfraught.Such 118/MeasuringProfitharmonizationisintendedtomakeiteasierforglobalinvestorstocomparecorporationsthatreportindifferentjurisdictions,butthekeyfigureintheseeffortsforecast‘bloodalloverthestreets’astheycametofruition(SirDavidTweedie,quotedbyTricksandHargreaves2004).ParticularlycontroversialwasInternationalAccountingStandard39(IAS39),governingthevaluationoffinancialinstrumentssuchasderivatives.8Thekeyissuewasthebearingofthestandarduponsituationsinwhichderiva-tivesareusedtohedgearisk,forexamplewhenabankoffersitscustomersfixed-ratemortgagesorguaranteedinterestratesontheirdeposits,andusesderivatives(oftenLIBORbased)tooffsetitsconsequentexposuretochangesininterestrates.Bankstypicallytaketheviewthatinsuchsituationstheeco-nomicallyrealisticaccountingtreatmentiswhatiscalled‘hedgeaccounting’,inwhichfluctuationsinthemarketvalueofhedginginstrumentsarenotrecognizedintheirbalancesheetsandincomestatements,onthegroundsthatthosegainsandlossesareoffsetbyfluctuationsinthevalueoftheitemsbeinghedged.OpponentsofIAS39arguethatitsrulesgoverningthepermissibilityofhedgeaccountingaretoorestrictive,forexampleinfailingadequatelytotakeintoaccountthewayinwhichbankshedgeriskssuchasinterest-rateexposureinaggregate,notitembyitem.Thedanger,theyargue,isthatwhatineconomicrealityarerisk-reducinghedgingtransactionswillbemadetoappearriskybyinjectingspurious,artificialvolatilityintotheirearnings.(Forexample,anintervieweewhowasanaccountantforabankcomplainedthattheitemsbeinghedged—forexample,aportfoliooffixed-ratemortgages—areoftennot‘markedtomarket’—revaluedasmarketpriceschange—buttheinstrumentusedtohedgethemhastobe.)In2004,concertedlobbyingbybanksledtheEuropeanCommissiontoendorsethestandardonlyinpart,adecisioncondemnedsharplybytheUKAccountingStandardsBoard,whichwasreportedassuggestingthatUKcompanies‘shouldignoreit’(TricksandBuck2004).Itmightbeimaginedthatdisputesoverwhetheraccountingrulesreflect‘economicreality’couldbesettledbyturningtotheacknowledgedexpertsonthelatter:economists.Infact,thesmallminorityofeconomistswhohavetakenresearchinaccountingseriouslyhaverarelybeenabledecisivelytosettletheissuesatstake.Asalreadysuggested,perhapsthesinglemostimportantoverallquestioninfinancialreportingishowtodefineandmeasure‘income’(or‘earnings’or‘profits’).ThegreatBritisheconomistJohnHicksprovidedwhathasbecomeperhapsthecanonicaldefinitionof‘income’,9but MeasuringProfit/119headmitteditwasnotprecise.Makingitprecise—inparticular,separatingincomeunequivocallyfromcapital—mightbe‘chasingawill-o’-thewisp’,saidHicks.Economists,hewrote,‘shallbewelladvisedtoeschewincome’.Theconceptwasa‘badtool...whichbreak[s]inourhands’(Hicks1946:176–7,emphasisinoriginal).Accountants,however,haveneverbeeninapositiontoduckwhatisperhapstheircentralclassificatoryresponsibility.AsDennisRobertsonputit,‘Thejailsandworkhousesoftheworldarefilledwiththosewhogaveupasabadjobtheadmittedlydifficulttaskofdistinguishingbetweencapitalandincome’(quotedbyKay2004).TheRulesofAccountingIf‘reality’isnotsufficienttostructurethepracticeofaccounting,perhapsrulesare?Theextentofformal,writtenrulesofaccountinghasvariedwithtimeandplace,andtherehassometimesbeenstrongoppositionamongstaccountantstosuchrules.InBritain(andperhapsespeciallyinScotland,theoriginalhomeofanorganizedprofessionofaccounting),therehasoftenbeenaconvictionthattherequirementtocaptureeconomicreality—togive‘trueandcorrectview’or‘atrueandfairview’ofthefinancialsituationofcompanies,assuccessiveUKCompaniesActshaverequired—necessitates‘acustom-builtdocument’requiring‘theexerciseofaninformedjudgment’withwhichothers,evenaccountants’ownorganizations,shouldnot‘inter-fere’(Slimmings1981:14).Suchaperspectiveemphasizesprofessionalstatus:oneofthehallmarksofa‘professional’—asdistinct,say,froma‘bookkeeper’orother‘clerk’—hasbeentakentobetheexerciseof‘judgement’(see,forexample,Porter1995).Nevertheless,thedirectionofhistoricalchange,drivenaboveallbyaccountingscandals,hasbeentowardsrulesandprinciplesthatarespelledoutratherthanimplicit.TheGreatCrashof1929raisedhugequestionmarksoverwhetherthefinancialreportingofmanyUScorporationsreflectedtheireconomicsituations.Inparttowardoffgovernmentintervention(possiblyevencompulsorygovernmentauditingofcorporateaccounts),theAmericanInstituteofAccountantsmadeatleastasymbolicsacrificeofsomeoftheaccountant’sindividualdiscretion,andbegantopromulgateformalaccount-ingstandards(Zeff1984).Theeffortdidindeedhelptokeepaccountants 120/MeasuringProfitinchargeofformulatingstandards—in1938,theSecuritiesandExchangeCommissiondelegateditsstandard-settingpowerstotheInstitute’sCommit-teeonAccountingProcedure—butitmarkedthebeginningofaproliferationofformalstandards.Theepisodeservesasoneoftheprimeexamplesoftheriseofthe‘idealofmechanicalobjectivity,knowledgebasedcompletelyonexplicitrules’(Porter1995:7).Thesixbrief‘rulesorprinciples’formulatedbytheAmericanInstituteofAccountantsin1934hadby2008becometheFinancialAccountingStandardsBoard’s163standards,someofwhichexceedahundredpages.10TheequivalentsetofstandardspromulgatedbytheInternationalAccountingStandardsBoard—whichgovernaccountingintheEuropeanUnion,andhavealsobeenadoptedwidelyoutsideofEurope—isslimmer,butevenit(e.g.IASB2005)isalreadyover2,000pageslongintotal,andgrowing.Here,thefinitistperspectivesketchedinChapter2leadstoaclear-cutpre-diction:thateventhemostdetailedrulebookwillonitsownbeinsufficienttodeterminethepracticeofbookkeepingandaccounting.Indeed,theverysizeoftherulebooks—especiallyoftherulebookofAmericanaccounting,whichhastendedtobemoreexplicitlyprescriptivethanothernationaltraditions,espe-ciallythatoftheUK—andtheirtendencytogrowinexorablyinlengthyearafteryearcouldbeseenasanexemplificationoftheWittgensteinianregressoutlinedinChapter2.Ifoneattemptstoremedytheunderdeterminationofactionsbyrulesbyaddingrulesofinterpretation,oneisembarkedonwhatisinprincipleanendlesstask.Indeed,theInternationalAccountingStandardsBoardhasspawnedabodyspecificallyentrustedwiththetaskofinterpretation,theInternationalFinan-cialReportingInterpretationsCommittee(IFRIC).Itstaskis,however,notstraightforward.Determiningwhatstandards‘imply’isnotasimpleexerciseindeductivelogic.AsnotedinChapter7,forexample,IFRIC’sproposedInter-pretation(‘IFRIC3’)ofhowstandardsshouldbeappliedto‘emissionrights’provedimmenselycontroversial,andhadtobewithdrawn.An‘emissionright’isanovelaccountingitem,butitshouldnotbeimaginedthattheunderdeterminationofactionsbyrulesistobefoundonlyinthecaseofnewitems.Consider,forexample,asetofaccountingitemsofaveryoldkind,‘inventories’:companies’stocksofunsolditems.Thereis,tomyknowl-edge,nodisagreementwhatsoeverovertheprinciplethattheInternationalAccountingStandardsBoardlaysdownforvaluinginventories:‘Inventoriesshallbemeasuredatthelowerofcostandnetrealisablevalue’(IASB2005: MeasuringProfit/121662).11Implementingthisagreedprinciple,however,involvesmeasuringboth‘cost’and‘netrealisablevalue’.Assuggestedabove,thecostofamanufactureditemisnotsimplythecostoftherawmaterialsdirectlyusedinit.Workers’wagesarealsopartofthecost,butinmostsituationsthoseworkerswillbeproducingmanyitems,sometimesofdifferentkinds,notjusttheiteminquestion,soadecisionneedstobetakenaboutwhatproportionoftheirwages(andassociatedemployer’scosts)shouldbecountedinthecostoftheiteminquestion.Machinery,too,isnormallyusedtoproducemanyitems,thusraisingsimilarissuesconcerninghowtoapportionitsdepreciation.Becauseoftheseandtheothercontingenciesalludedtoabove—apportionmentofthecostofheat,light,buildings,administrativeservices,etc.—manyfirmsdonotattempttomeasurethecostsofindividualitemsofinventory,butinsteadestimatea‘standardcost’.Althoughtheaggregateofsuchstandardcostscanbecheckedagainsttherecordofactualexpenditures—aswasdone,forexample,inthefirmstudiedinMacKenzie(2008)—theyarecertainlynotself-evidentempiricalfacts.Similarlywithmeasuring‘netrealisablevalue’.Thatinvolvesdeterminingwhethertheitemoritemscanbesold,andifsoforhowmuch.Thatinturninvolvesjudgementsaboutbothmarketconditionsandtheitemoritems:havethey,forexample,beensupersededbysuperioralternatives?Unsurprisingly,suchquestionscansometimesbehardtoanswerunequivocally,anditisdif-ficulttoimagineeventhemostdetailedsetofrulesmakingthesatisfactorydeterminationof‘netrealisablevalue’entirelyalgorithmic.Measuringthevalueofinventoriesisthusfarlessstraightforwardthanmightbesuggestedbytheapparentlysimpleprinciple—‘thelowerofcostandnetrealisablevalue’—laiddowninthe‘rulebook’ofinternationalaccounting:indeed,inventoriesareoftenseenbyauditorsasamatterthatdemandstheirspecialattentionbecauseofthelargeeffectsdifferentvaluationscanhaveonafirm’sbalance-sheetsituationandprofits.Totakeanotherexampleofthewayinwhichagreementoveraruleordefinitiondoesnotpreventdisputeoverhowitistobeapplied,considertheaccountingclassificationwithwhichthischapterbegan:WorldCom’sclassifi-cationofthecostsofunusedportionsoflineleasesas‘assets’.BecauseWorld-Comhascometobeseenasastraightforwardcaseofaccountingfraud,itiseasytoimaginethattheclassificationwasunequivocally‘againsttherules’.Mattersare,however,notquiteasclear-cutasthat.WhenWorldCom’sChiefFinancialOfficer,ScottSullivan,wascalleduponbyitsBoardtojustifythe 122/MeasuringProfitclassificationhedidso,citingUSaccounting’scanonicaldefinition(alreadyquotedabove,andasfarasI’mawareentirelyuncontested):‘Assetsareproba-blefutureeconomicbenefitsobtainedorcontrolledbyaparticularentityasaresultofpasttransactionsorevents’(FASB1985:16).InhispapertoWorldCom’sBoard,Sullivanarguedthatthecostsofleasinglargeamountsofnetworkcapacity—largerthanmightactuallybeneeded—werecostsofacquiringcustomers(whichcanlegitimatelybeclassedasanasset);thattheclassificationofthosecostsasanasset‘doesnotcontradict’theabovedefinitionof‘asset’;andthatitmettherequirementsoftherelevantmoredetailedrules:Theleasecommitmentswereenteredintotoobtainaccesstolargeamountsofcapac-ityunderthetheorythatrevenuewouldfollowandfullyabsorbthesecostsandtoexpedite‘timetomarket’.Webelievethatthisprovidedanadvantageoverourcom-petitorsandcreatedtheleaderinInternetbackboneatOC192-c.12Thecommitmentswereenteredintowiththeknowledgethatwewouldincuranexpenseprematurelyandtherevenueswouldbeearnedsubsequenttothatdate.TheCompanywaswillingtoabsorbthiscostpriortorecognizingtherevenuestreambecauseitbelievedthatthefuturerevenueswouldbematchedupwiththesecosts.ThesecommitmentswereenteredintoastheresultofcustomersforwhichserviceswouldberenderedandtheleasecommitmentswereenteredintotoexpeditethecustomerprovisioningandrevenuestreaminaccordancewithSAB10113andasfurthersupplementedbyFASB91,14directandindirectcostsassociatedwithobtainingacustomermaybedeferredandamortizedovertherevenuestreamassociatedwiththatcontract.(Sullivan2002:1–2)Sullivan’sargumentwasnotapriorioutrageous.WorldComhopedthatthecustomersitattractedwouldstaywiththefirmforyears,andthecostsofacquiringthemwouldthusindeedgiverisetothe‘probablefutureeconomicbenefits’crucialtothedefinitionofan‘asset’.Thepracticeoftreating‘cus-tomerrelationships’asanasset—andthus‘capitalizing’thecostsofacquiringthoserelationships,ratherthantreatingthosecostsasexpenses—seemsrea-sonablywidespread:acolleagueofmineinaccounting,YannisTsalavoutas,wasablequicklytoprovidemewithasamplelistofUKcompanieswhichdojustthat.15Often,forexample,companiessuchaslawyersorconsultanciesareboughtspecificallytogainaccesstotheirclientele,andthecostsofdoingsotreatedasanasset.16 MeasuringProfit/123Sullivancanthusbeseenasattemptingtoconstructananalogy:thatleasinglargeamountsofnetworkcapacity(evenifitturnedoutnottobeneeded)waslikeotherwaysofacquiringcustomers,thecostsofwhichcouldlegit-imatelybetreatedasassets.Asweshallseebelow,hisargumentpersuadedoneaudiencebutfailedtopersuadeanother,morecrucialone.Thepoint,however,isthat‘persuade’istherightterm.Thecostsinvolveddidnotself-evidentlymeettheformaldefinitionof‘assets’,butnordidtheyself-evidentlyfailtomeetit.Thedefinitiondidnotcarrywithitinstructionssufficienttodeterminehowitshouldcorrectlybeappliedinaspecificcase.Unfortunatelyfromtheviewpointofthischapter,Sullivan’sargumentwasnevertestedincourt.Sullivanpleadedguiltytosecuritiesfraud,testifiedincourtagainstWorldCom’schiefexecutive,BernardEbbers,andreceivedajailsentenceofonlyfouryears,prosecutorshavingenteredapleainmitigationforhimonthegroundsthathehadbeenacooperativewitness.Ebbers’s—unsuccessful—defencewasthathewasignorantof‘detailed’matterssuchastheclassificationofthecostsofleasesasassets,notthattheclassificationwaslegitimate.Hewasjailedfortwenty-fiveyears.EarningsManagementWhatWorldComhadbeenseekingtoachievewas,ofcourse,afavourableportrayalofitsfinancialperformance.Justhowwidespreadareeffortsofthiskind?Clearly,thereisnounequivocal,directwaytoanswerthisquestion.Researchersandotheroutsidershavenoprivilegedaccesstotherealityofcompanies’economicsituationsthatcouldserveasabenchmark.Fortunately,however,thereisaconsiderablebodyofquantitativeresearchwithinacademicaccountingthatenablesusindirectlytoaddresstheques-tionoftheextentofeffortsatfavourableportrayal.Thisresearchconcerns‘earningsmanagement’,thecanonicaldefinitionofwhichisgivenbySchipper(1989:92):‘purposefulinterventionintheexternalfinancialreportingprocess,withtheintentofobtainingsomeprivategain(asopposedto,say,merelyfacilitatingtheneutraloperationoftheprocess)’.Schipper’sdefinitiondoesnotsaysoexplicitly,butearningsmanagementisusuallytakentobepermis-sible,legalformsofthisintervention.(‘Fabricatinginvoicestocreatefictitious 124/MeasuringProfitsalesrevenue’—Schipper1989:93—isfraud,notearningsmanagement.)Theprevalenceofearningsmanagementisthusofinterest,notjustasanindicatoroftheextentofeffortsatfavourableportrayal,butpreciselybecauseitisbehav-iour‘withintherules’.Theextentofearningsmanagementisthusameasure,albeitaverycrudeone,ofthedegreetowhichdiscretioncaninpracticebeexercisedevenwhen,asintheUSA(thesiteofmostresearchonthetopic),thefinancialreportingprocessisgovernedbyextensive,formalrules.What‘privategain’mightinducemanagementstoengageinearningsman-agement?Probablymostimportantiswhatappearstobeawidespreadbeliefamongstcorporatemanagersthatstockanalystsandinvestorsprefercorpor-ationswhoseearningsrisepredictablytocorporationswhoseearningsfluc-tuatesubstantially,evenifaroundthesameunderlyingtrend.(Thebeliefinthepervasivenessofthispreferenceseemstohaveinformedtheopposition,discussedabove,tothestandardgoverningaccountingforderivatives,IAS39.)Iftherewardsenjoyedbycorporateseniormanagersreflectstockprices,asinrecentdecadestheyincreasinglyhave,thereisthusanincentivefor‘incomesmoothing’,inotherwordsforexploitingpermissiblediscretiontoreducethevolatilityofearnings.Clearly,too,thereisusually,thoughnotalways,anincentivetoavoidreportinglosses,anditisoftenveryimportanttomeetortosurpassstockanalysts’predictionsofcorporateearnings.(Themainexceptiontotheneedtoavoidreportinglossesisthatoneofthetechniquesofearningsmanagementisthe‘bigbath’,inwhichacorporationreportsalargeloss,portrayingittoanalystsandinvestorsas‘one-off’andtheresult,forexam-ple,ofrestructuring.Apessimisticallycalculatedbig-bathloss—involving,forexample,largeprovisionsforliabilitiesandforbaddebts—enhancesfutureprofits,whichareboostedastheprovisionsareunwound.)Oneapproachtakeninresearchonearningsmanagementistoidentifysituationsinwhichthereisaclear,temporaryincentivetomanageearnings;toscrutinizecorporateaccountsforpatternsconsistentwithearningsman-agement;andtoexaminewhetherthosepatternscorrelatewithincentivesituations.ApioneeringstudyofthiskindwasJenniferJones’s(1991)examina-tionofthefinancialreportingoffirmsinindustriesthatwerepetitioningtheUSInternationalTradeCommissiontorecommendtariffsandimportrestric-tions.Suchpetitionsstoodachanceofbeinggrantedonlyiftherewasevidencethatdomesticindustrywasbeing‘hurt’byoverseascompetition.Jonesfocusedonaccruals:balance-sheetchangesforwhichthereisnoimmediatecash-flowcounterpartsuchasdepreciation,changesinthevaluationofproperty,plant, MeasuringProfit/125andinventory,andestimatesofaccountspayableandreceivable.Sheesti-matedthediscretionarycomponentofsuchaccrualsbysubtractingfromtotalaccrualsaregression-basedestimateof‘normal’,‘non-discretionary’accruals.Aggregatingresultsforfiveindustries(automobiles,carbonsteel,stainlessandalloytoolsteel,copper,andfootwear),sheshowedstatisticallysignificantnegativediscretionaryaccrualsintheyearsofInternationalTradeCommissioninvestigations.Initialpublicofferingsofstock(IPOs)areanothersituationinwhichthereisatemporarilystrongincentiveto‘window-dress’accounts(inthiscasetopor-trayfinancialstrength).Acomparisonofthe‘unexpected’accrualsofcompa-niesengagedinIPOswithamatchedcontrolgroupofsimilarcompaniesfoundthat62percentoftheIPOfirmshadhigheraccrualsthanthecorrespondingcontrol(Teoh,Wong,andRao1998:187,table3).Sincechanceprocesseswouldsuggesta50percentfigure,‘thisimpliesthatroughly12percentoftheissuingfirmsmanageearnings’(HealyandWahlen1999:373).Adifferentapproachtothedetectionofearningsmanagementistoexaminethestatisticaldistributionofearnings,lookingfordiscontinuitiesor‘kinks’atearningslevelsthatcorrespondtoparticularlystrongincentivestoearningsmanagement:zeroearnings(andthusthedividebetweenmakingaprofitandregisteringaloss);earningsinthepreviousyearorpreviousquarter;andcorporations’oranalysts’earningspredictions.Suchkinksturnouttobesubstantial(see,forexample,Figure6.2).Forinstance,analysisofUScorpor-ateearningsfor1976–94suggeststhat‘30%to44%ofthefirmswithslightlynegativepre-managedearningsexercisediscretiontoreportpositiveearnings’(BurgstahlerandDichev1997:124).Thedetectionofearningsmanagementaboundswithconceptualandmethodologicaldifficulties(see,forexample,McNichols2000).Analysesbasedon‘discretionary’or‘unexpected’accrualsareextremelysensitivetothemodelof‘normal’,non-discretionaryaccrualsthatisemployed(if,forexam-ple,earningsmanagementiswidespread,‘normal’levelsofaccrualsmayalreadyreflectsuchmanagement)andtheycannotdetecttechniquesofearn-ingsmanagementthatdonotinvolveaccruals.Noraredistributionalanaly-sesunequivocal.Adistributional‘kink’isnotinitselfevidenceofearningsmanagement.Itmaybe,forexample,thatanticipatedsmalllossesareturnedintosmallprofitsnotbychangesinaccountingclassificationsbutby‘real’interventions(salesdrives,cutsinexpenditureonmaintenanceoronresearchanddevelopment,andsoon).BurgstahlerandDichev(1997)attemptto 126/MeasuringProfit200015001000Frequency5000−.25−.20−.15−.10−.05.00.05.10.15.20.25.30.35EarningsIntervalFigure6.2.FrequencydistributionofreportsofnetannualincomebyUScorpora-tions,1976–1994Source:ReprintedfromJournalofAccountingandEconomics,24(DavidBurgstahlerandIliaDichev,’EarningsManagementtoAvoidEarningsDecreasesandLosses’,99–126),copyright1997,withpermissionfromElsevier.BurgstahlerandDichevdrawtheirdatafromComputstat,andthepopulationofcorporationsexcludesthefinancialsectorandregulatedindustries.Incomeisscaledbydivisionbythecorporation’smarketvalueatthestartoftheyear;intervalwidthis0.005.Thedashedlineisthezero-earningspoint.overcomethisproblembyinvestigatingcashflowfromoperationsandlevelsofaccrualsaroundreferencepointssuchaszeroearnings,butthiskindofanalysismaynotbeentirelyrobust(Dechow,Richardson,andTuna2003).Thereareevenpotentialissuesofreflexivity.Somesophisticatedinvestorsarealreadyemployingacademicearnings-managementdetectionmodels(Henry2004),anditwouldbesurprisingifregulatorswerenotdoingsotoo,sothereisnowanincentivetomanageearningsinwaysthatthemodelscannotdetect.Nevertheless,theoverallthrustoftheliteratureonearningsmanagementisconsistentwithpervasiveanecdotalevidence(mostauthoritativeisLevitt1998)ofextensiveearningsmanagementbyUScorporations,atleastinthe1980sand1990s.(Empiricalstudyofperiodspriortothe1980sistoosparsetoallowanydefinitiveconclusions,anditisnotyetclearwhethertheprevalenceofthepracticehasdeclinedsincethescandalsoftheearly2000sandthesubsequenthigh-profilecriminaltrials.)Attemptstoachievefavourableportrayalsappeartohavebeenwidespread,and—justasafinitistanalysiswouldpredict—theworld’smostrule-intensivesystemoffinancialreportingdidnoteliminatediscretionfromcorporateaccounting. MeasuringProfit/127ContestingClassificationItwould,however,bequitemistakentoconcludethataccountingispractisedfreefromconstraint.Forexample,againconsiderthecaseoftheWorldComaccountingclassificationwithwhichthischapterbegan.ThemostinterestingcharacterintheWorldComepisodeis,ultimately,notthechiefexecutiveEbbers(anunderstandablefigure,giventhetemptationsandpressuresofthemonthsinwhichthedot.comandtelecomboomturnedtobust),norSullivan(despitethefactthathisactionsserveasanalmosttextbookexampleofthelogicalgapbetweenrulesanddefinitionsandtheirapplications),butWorld-CommanagerCynthiaCooper,whoheadedaninternalauditteaminclud-ingtwokeycolleagues,GeneMorseandGlynSmith.Cooperwasacareeraccountant,broughtupinClinton,Mississippi,siteofWorldCom’sheadquar-ters.ShefirstlearnedaccountingatClintonHighSchool(itisasmall-townstory:shewastaughtthesubjectbyGlynSmith’smother),andwentontotakeanundergraduatedegreeinitatMississippiStateUniversityandapostgraduatequalificationfromtheUniversityofAlabama.HerroommateatMississippiStateremembershersittinginaccountingclasses‘inthefrontrow,deadcenter...shewouldpeppertheprofessorwithquestions’(Ripley2002).Cooperandherdivisionwereresponsibleforinternalmatterssuchasbud-gets,performanceevaluation,andidentifyinginefficiencies,notforcheckingWorldCom’sfinancialreporting:thatwastheroleofitsexternalauditors,ArthurAndersen.InMarch2002,however,herinterestinthedataunder-pinningWorldCom’sincomestatementandbalancesheetwaspiquedbyavisitfromthemanagerofWorldCom’swirelessdivision,JohnStupka.In2001,Stupkahadmadeprovisionforcustomerbaddebtstotalling$400million.Sullivanhadunwoundtheprovision,thusboostingWorldCom’sreportedincome,andStupkawasunhappy,becausethebaddebts,nolongerprovidedfor,wereabouttodamagehisdivision’sfinancialperformancefigures(PulliamandSolomon2002).ToCooper,Stupkahadbeenrighttocreateareservewhenhecouldreason-ablyforeseeaproportionofcustomersfailingtopay,andSullivan’sreversaloftheprovision‘smelledfunny’.ShequerieditwithanAndersenauditorresponsibleforWorldCom,who‘brushedheroff’(PulliamandSolomon2002).CooperthenturnedtoGeneMorse:‘Shecamebacktomeandsaid,“Godig”’(PulliamandSolomon2002).Shealsotooktheissueofthereversal 128/MeasuringProfitoftheprovisiontoWorldCom’sauditcommittee,whereshepresentedtheargumentsagainstit.Sullivanjustifiedit,buteventuallyconceded.Cooper’schoiceofMorseto‘dig’wasastute:hehadbothdeterminationandcomputerskills.Blockinghisenquirieswastheaccess-controlmatrixofWorldCom’scomputerizedaccountingsystem.Hewouldneedhigh-levelapprovalforthe‘read’permissionsheneeded,andthatrequestwouldmakeclearthattheinternalauditteamwasstrayingbeyonditsremit.Instead,MorseofferedtohelptestaprogramthatamemberoftheWorldCominformation-technologystaffwaschecking.TheprogramgaveMorseaccesstotherecordsofhugenumbersoftransactions,indeedsuchlargenumbersthathehadtobegintoworklateatnightafterhisdownloadingofgiantfiles—suchasthe350,000monthlyrecordsthatmadeup‘intercompanyaccountsreceivables’—begantobeblockedbecauseoftheireffectsonsystemperformance.(Suchrecordstypicallyincludescannedversionsofdocumentationsuchasinvoices,hencethesizeofsuchfiles.)KnowingCooper,Morse,andotherswereworkinglate,Cooper’sfatherwouldsometimestakeinsandwichesforthemtohaveatmidnight(Jeter2003:169).Morsebegantofocuson‘journalentries’suchastheonedescribedatthestartofthischapter,viawhichcostswereclassifiedaspurchasesofassets.Hisinstinctualresponsetosomeofthesewasthattheywerewrong.‘Thisstinks’,heremembersthinkinginonecase(PulliamandSolomon2002).BythestartofJune2002,Morsehadfoundaccountingclassificationstotalling$2billionthatseemeddubioustohimandtoCooper.SheandSmithbegantoraisethemwithMaxBobbitt,whochairedWorldCom’sauditcommittee,withFerrellMalone,theWorldCom‘engagementpartner’ofKPMG(whichhadtakenoverfromAndersenasthecompany’sauditors),andwithseniorcolleaguesinWorldCom’sClintonheadquarters,suchasControllerDavidMyers.Eventually,thedisputedclassificationsreachedWorldCom’sfullauditcommitteeon20June.Sullivanjustifiedthemverballytothecommittee,andoverthefollowingweekendhepreparedthewrittenjustificationquotedabove.ThecasemadebySullivanindefenceofclassifyingthecostsofunusedportionsofnetworkleasesasassetsdividedaccountantsfromlaypeople.Atleastsomeofthelatterfounditconvincing:‘Tosomenon-accountants[onWorldCom’sBoard],Sullivan’sjustificationsseemedreasonable,andsomethoughtKPMGdidnotsufficientlyunderstandtheCompanyortheindustry’(Beresford,Katzenbach,andRogers2003:128).Asthatquotationindicates, MeasuringProfit/129however,KPMG’saccountantssidedwithCooper’scase.EvenMyers,whoasWorldCom’sControllerhadbeenheavilyinvolvedinitshigh-levelaccount-ingprocessesanddecisions,concededtoCooperandSmiththat‘theentriesshouldn’thavebeenmade’(PulliamandSolomon2002).Indeed,heseemsinitiallytohaveresistedmakingthem,tellingSullivan‘thetransfercouldn’tbejustified’(Pulliam2003:M6).ItwasthelackofsupportfromSullivan’sfellowaccountantsthatwasfatalforhim.Whentheauditcommitteereassembledon24June2002,hecontinuedtodefendthedisputedclassification.However:KPMGtoldSullivanthatthistheoryforcapitalizingoperatinglinecostsdidnotholdwater.RepresentativesofAndersen,participatingbytelephone,saidtheaccountingwasnotinaccordancewithGAAP[GenerallyAcceptedAccountingPrinciples,theoverallframeworkoffinancialreportingintheUSA]andthatAndersenwaswith-drawingitsauditopinionsfor2001anditsreviewofthefirstquarterof2002.Afterthemeeting,Salsbury[WorldComGeneralCounselMichaelSalsbury]askedSullivanforhisresignation.Myers,whowasnotpresent,wasseparatelycontactedandaskedtoresign.Myerslaterresigned.Sullivandidnot,andwasfired.(Beresford,Katzenbach,andRogers2003:128–9)Ofcourse,contextmatters.Twoweeksbeforetheauditcommitteemeeting,thehugeauditingfirmArthurAndersenhadbeenconvictedofdestroyingevidenceinconnectionwithitsclient,Enron,anditwasalreadyclearthatthefirmhadeffectivelycollapsed:the‘bigfive’inauditinghadbecomethe‘bigfour’.Soallthosepresentattheauditcommitteemeetingwillhaveknownthatbothcriminalprosecutionandcorporatedemisecouldfollowfromaccountingdecisionsthatothersjudgedindefensible.Inothercircum-stances,Sullivan’sjustificationofthecontestedaccountingclassificationmighthavebeenseenasallowable,ashisearliercontentiousreversalofthebad-debtprovisionseemsinitiallytohavebeen,butbyJune2002therisksofacceptingitmusthavebeenviewedbythoseinvolvedastoohigh.ConstrainingAccountingAcrucialvirtueoffinitismasatheoreticalperspectiveonfinancialreportingisthatbysuggestingthataccountingcannotentirelybeconstrainedbyrules 130/MeasuringProfitorby‘reality’itsharpensempiricalcuriosityastothenatureoftheothersourcesofconstraint.Unfortunately,existingresearchisinsufficientfullytosatisfythatempiricalcuriosity.Certainly,Idonotpretendthatthemater-ialpresentedinthischapterorinasmallcasestudypresentedelsewhere(MacKenzie2008)hasdoneso.Forreasonstobetouchedonbelow,themostpromisingapproachtoinvestigatingtheconstraintsonaccountingislikelytobeethnographic,butresearchofthatkindonfinancialreportinghassofarbeenrare.Thesociologicallyorientedliteratureinaccounting,whichisrichontopicssuchasprofessionalization,managementaccounting,andtheregulationofaccounting,issparserinregardtothepracticesunderpinningfinancialreporting:‘Oneofthedisappointingcharacteristicsoffieldstudiesinorganizationsishowfewhaveexaminedhowaccountingandauditdecisionsaremade’(CooperandRobson2006:435).Somydiscussionoflikelysourcesofconstraintisspeculative.Althoughnotintendedasacommentonaccounting,thepassagefromDavidBlooralreadyquotedinChapter2isworthrevisitinghereasawayoforientingthediscussion:Wecouldtakeourconceptsorrulesanywhere,inanydirection...Wearenotpreventedby‘logic’orby‘meanings’fromdoingthis...Therealsourcesofconstraint[are]ourinstincts,ourbiologicalnature,oursenseexperience,ourinteractionswithotherpeople,ourimmediatepurposes,ourtraining,ouranticipationofandresponsetosanctions,andsoonthroughthegamutofcauses,startingwiththepsychologicalandendingwiththesociological.(Bloor1997:19–20,emphasisinoriginal)LetmebeginwithafactorprominentintheunravellingofWorldComandinanyfinitistanalysis:‘interactionswithotherpeople’.Financialreportingisnotanactivityconductedbyindividualsinisolation.Often,othersneedtobepersuadedthataclassificationorameasurementisappropriate,andthiscanbeanimportantsourceofconstraint.Themostobvioussuchothersareacorporation’sauditors,andthis,fortunately,isanareainwhichexistingresearchisrelativelystrong.AparticularlystrikingstudyisBeattie,Fearnley,andBrandt(2001),inwhichtheauthorstooksixUKcompaniesandinter-viewedboththefirms’financedirectorandthemostseniorexternalauditor.Theyshowedthatauditorsareindeedasourceofconstraint,butaflexibleone.Whenauditorsdisagreewithaclassificationorameasurement,whatensuesisoftennegotiation,ratherthanauditorsdictatingwhatmustbedoneforthemto‘signoff’thecompany’saccounts. MeasuringProfit/131Inonecase,forexample,boththefirm’sfinancedirectoranditsauditorsuspectedthatitsinventorywasovervalued.Tobeginwith,theauditorswereunabletodeterminebyhowmuch.However,installationofanewcom-putersysteminthefirmenabled‘slow-moving’andobsolescentstocktobeidentified.Astheleadauditorputit:‘Wejusthadtheinformationfromastockobsolescenceprintout.Butthey[thefirm]couldn’tdenyitbecausethey...generatedthatinformationthemselvesandthatcameupwithabignumber’(Beattie,Fearnley,andBrandt2001:85).Instead,however,ofinsistingthatthevalueofthestockshouldbewrittendownimmediatelybythe‘bignumber’,theauditorsagreedwithcompanymanagementthatitshouldbewrittendownonlygraduallyoverthreeyears.Manyaccountantsmighthavefeltthatthiswaswrong:inhisforewordtoBeattie,Fearnley,andBrandt’sbook,theleadingaccountingregulatorSirDavidTweediecommentsthat‘thecase-studiesareanonymous—whichisaswellfortheauditorwhoallowedover-valuedstocktobewrittendownoverthreeyears—forgoodnesssake!’(Beattie,Fearnley,andBrandt2001:xxi).Theseniorauditor,however,hadfelttheneedtoreachacompromise:‘I’mafraidItakeapragmaticapproachtoauditingandyoucan’teatanelephant...soIhadtodissectthethingand...getsomuchinoneyearandsomuchthenext’(quotedinBeattie,Fearnley,andBrandt2001:86).Interactionswithinacompanyarealsolikelytobeimportantasasourceofconstraintontheconstructionofaccounts,buthavebeenmuchlessstud-iedfromthisviewpointthaninteractionswithauditors.Seniormanagersarelikelytobeacrucialaudiencefortheworkofaccountants,andthelitera-tureonearningsmanagementpointsstronglytotheirdesireforfavourableportrayalsoffinancialperformanceoftenbeinganimportantfactorshapingfinancialreporting.Alessobvioussetofothers,however,arestafflowerincorporatehierarchiesthanaccountants,notablybook-keepers.Likeseniormanagers,theytoowillhavewhatBloorcallstheir‘immediatepurposes’,butthesearelikelytobe‘gettingthejobdone’,17ratherthanportrayingtheeconomicconditionoftheircompanyinonewayratherthananother.Theirremuneration,unlikethatofseniormanagers,isunlikelytobeaffectedatalldirectlybythatportrayal,andtheyarelikelytobeclassifyingamyriadofoftensmalltransactions,fewornoneofwhichwillontheirownbeconsequentialtotheportrayal.Theissueoftherelationshipbetweenthesituationsandworkofbook-keepersandofmoreseniorstaffhighlightsaconstrainingfactorthatisnot 132/MeasuringProfitexplicitinBloor’slist:technology.Thetechnicalsystemsofaccounting—whichforsmallentitiescanbestandalonesystems,butforthelargerfirmswillnownormallybeaspectsofcontinuouslyevolvingEnterpriseResourcePlanning(ERP)systemssuchasOracleorSAP(see,forexample,QuattroneandHopper2006)—areneithermerelyneutralmedianorsimplymeansofincreasingtheefficiencyofwhatunaidedhumanbeingsmightdo.Tech-nicalsystemslinktheworkofthemultiplepeopleneededtodotheaccountsofanylargeentityinstructuredways.Inparttoreduceopportu-nitiesforfraud,accountingsystemsaredesignedtoconstrainthewaysinwhichanygivenpersoncanaltertheresultsoftheworkofanother(andtopreventanyoneotherthanaspecific,limitedsetofindividualsmakingsuchalterations),andtoleaveanineliminabletracewhenanalterationismade.Suchdeliberate‘technical’barriersthenbecome‘social’constraints.Thevastbulkofaccountingclassificationsare,asalreadynoted,madebystafflowincorporatehierarchiessuchasbook-keepers.Technicalsystemsthen‘solid-ify’thoseclassificationsbyrestrictingwhatmoreseniorstaffcansubsequentlydotoalterthem.Isaseniormanagergoingtoattempttoinfluenceinadvancethousandsormillionsof‘primary’classificationsmadebydozensorhundredsofbook-keepers,orsubsequentlytoattempttoalterthoseclassifications(espe-ciallyifheorshedoesnothavetheaccesspermissionstodosoandifthealterationsleaveavisibleaudittrail)?TherecordofcasessuchasEnronorWorldComsuggeststhatthesearenotthemostattractivepaths.WhenGeneMorsewastoldbyCooperto‘godig’intoWorldCom’sfinancialreporting,hedidnotdiscovereffortsbySullivanorotherstoaltertheseprimaryrecords.Instead,earningsmanagementwasbeingaccomplishedbyleavingtheprimaryclassificationsintact,andperforminghigher-levelreclassifications.Thelatter,forexample,iswhattheWorldComaccountantwithwhichIbeganthischap-terdidontheFridayafternoonin2001.Thematerialformtakenbythistechnicalconstraintispassword-controlledaccesstoafirm’sERPsystem,togetherwithamatrixspecifyingthe‘read’and‘write’permissionsofeachuser.18Whilesenioraccountantsandperhapssomeothertopmanagerswillhaveextensiveorcomplete‘read’accesstoaccountingdata,thematrixnormallyrestrictstightlytheir‘write’access:indeed,thoseatthetopofcorporatehierarchiesmayhaveno‘write’accessatall.Processingsalesordersandinvoices,forexample,isnotpartoftheirjobs—itis‘book-keepers’work’—andwhiletheywouldusuallybeabletoviewtheresultant MeasuringProfit/133recordstheywouldnotnormallyhavethe‘write’permissionneededtocreatesucharecordortoalterit.Ofcourse,technicalconstraintisnotabsolute.Anysystem’scontrolscanbesubvertedifthetechnicalstaffinchargeofasystem’saccess-controlmatrixcanbepersuadedtoalterit,ifthosewhohavetherelevantpermissionscanbepersuadedtodisclosetheirpasswordstothosewhodonot,orifotherways(suchasthatusedbyGeneMorse)canbefoundtocircumventthecontrols.Thecreationofanaudittrailisadeterrentonlyifanauditorislikelytoscrutinizeitandtocontestthereclassificationsitreveals,whichreturnsustothesphereofhumanconstraints:ofBloor’s‘interactionswithotherpeople’and‘anticipationofandresponsetosanctions’.Indeed,onecan’tentirelyruleoutthepossibilitythattheapparentrarityofearningsmanagementperformedbytheenmassealterationofprimaryrecordsreflectssuccessindoingsouncontested.Nevertheless,suchconsiderationsdosuggestthatethnographiesoffinancialreportingwillneedtopaycloseattentiontoitstechnologicalbases.Theyalsopointtotheimportanceofstudyingtheworkofbook-keepersaswellasofprofessionalaccountants.Ifmyconjecturesarecorrect,itisbook-keepers—notaccountants—whoproducemuchofaccounting’sequivalentofscience’sobservationalbase.Book-keepers’classificationsarejustasopentofinitistanalysisasthoseofaccountants(andthereisno‘theory-independentobser-vationlanguage’inaccountingorinscience),buttheirclassificatoryworkhasalmostneverbeenexaminedinethnographicdetail.19AnotherissueworthattentionispresentexplicitlyonBloor’slist:training.Finitismsuggeststhatclassificationandconceptapplicationarebasedonrela-tionsofsimilarityanddifferencethat,ultimately,arelearnedostensively—thatis,byexposuretoauthoritativeexamplesof‘correct’classificationand‘appropriate’conceptapplication.Forinstance,ascientificparadigmisatrootasetofconcrete,exemplarysolutionstoscientificproblems.Scientifictrainingconsistsingoodpartoflearningofhowtoperformtheseexemplarysolutionsandhowtoextendthemtosimilarcases(Kuhn1970;Barnes1982).Socializa-tionintothe‘paradigm’inthebroadersenseofanoverarchingdisciplinaryframeworkisnotachievedsolelybytheframeworkbeinglearnedexplicitly(iffinitismisright,itcouldnotbelearnedinitsentiretyinthisway),butbyrepeated,authoritativeostension.Accountingandbook-keeping,Iconjecture,arealsolearnedingoodpartostensively.Theliteratureontheeducationofaccountants—suchasPower 134/MeasuringProfit(1991)andAnderson-Gough,Grey,andRobson(1998)—unfortunatelydoesnotofferaclear-cutanswertothequestionoftheextentofostensivelearning,butitseemsclearthatwhileaccountantsandbook-keepersdolearnmanyexplicitrules,theyalsohavetolearnhowtoapplytheserulestoparticulars.Someofthistrainingtakesplaceinformaleducationalsettings;muchofittakesplace‘onthejob’.Iftraininginaccountingorbook-keepingislikesci-entifictraining,wewouldexpectittoconsistlargelyinrepeatedexperienceofsolvingproblemsforwhichthereareauthoritative‘right’and‘wrong’answers.Iftheanalogywithscienceholds,theresultofsuchtrainingwillgobeyondtechnicalcompetenceinanynarrowsense.Itwillbefoundtobesocializationintoawayofviewingtheworldthatisnotwhollyexplicit,butisnotforthatreasonanylesspowerful.Thatitisunlikelytobeentirelyexplicitisonereasonwhyethnographic(ratherthan,forexample,interview-based)researchisneeded.Thosewhopractisethetechnicalculturesofbook-keepingandaccountingmaysimplybeunabletogiveafullverbalaccountofwhattheydoandwhytheydoit.Inparticular,prolongedostensivelybasedsocializationcan‘achievereal-ism’.Toneophytes,Iconjecture,classifyingitems(inaccounting,science,orelsewhere)willfrequently‘feel’likeclassification:hereisanitem;herearepossibleclassifications(X1,X2,X3,...,Xn);whichshallIchoose?(That,forexample,iswhatIoftenfeelwhen‘coding’researchgrantexpenses.)Theexperiencedpractitioner,incontrast,willoftenfeel‘ThisitemisanX3,’justastheexperiencedbird-watcherglancesatabirdandthinks‘Thatisanoyster-catcher,’not‘Iamclassifyingthatbirdasanoystercatcher.’Theclassificationcanstillbeanalysedasachoice(orsofinitisminsists),buttothoseinvolveditnolongerfeelslikeachoice,orindeedevenaclassification.Fromafinitistviewpoint,whathappensinsuchcasesisthattrainingandhabithaveformedastrongsenseofanalogy:thisbirdresemblespreviousoystercatcherssufficientlystronglythatitmustbeanoystercatcher.ThereactiontoSullivan’scontestedclassificationofthecostsofunusedportionsoflineleasescanalsobereadinthisway.Sullivanmaywellhavebeencorrectthat,asamatteroflogic,thisdidnotcontradicttheofficialdefinitionof‘asset’,buttotheotheraccountantsinvolved,suchasCooper,itdidnotresembleanasset,butseemedmuchmorelikeanexpense.WhatSullivanwascontendingwith,inotherwords,waspreciselythisdeeplyinternalizedsenseofanalogy.(Laypeoplewillhavelackedthatinternalizedsense,whichmayhavebeenwhyhehadgreatersuccesspersuadingthem.) MeasuringProfit/135‘WhenIobeyarule’,wroteWittgenstein(1967:85e),‘Idonotchoose.Iobeytheruleblindly’(emphasisinoriginal).Ostensivelybasedsocializationoffersonepossiblereasonwhytherulesandprinciplesofaccountingdoinpracticehaveforce.(Participantscertainlytreatthemashavingforce:thefiercedebatesocca-sionedbystandardssuchasIAS39wouldhavelittlepointifruleshadnoforce.)Whataruleorprincipleimpliesforanyconcretesituationisundeterminedlogically(onecan’t,forexample,calculatethevalueofaninventorysimplybylogicaldeductionfromtherelevantstandard,IAS2),buttothosesocializedintoaparticularcultureofaccounting,theremaywellbemanycasesinwhichtheclassificationorotheractionthatisdemandedseemsclear.Theextenttowhichaccounting(orbook-keeping)isinpracticeconductedinnominalist‘choiceamongstclassifications’modeorin‘realist’modeisanempiricalquestion.Myconjectureisthatbothmodeswillnormallybepresent.Routine,familiaritems,forexample,mayevoke‘realist’mode(‘thatisanX3’);unfamiliaritemsprovokeexplicitchoices(‘isthisanX1oranX2?’).But,clearly,otherfactorswillalsobepresent.Anaccountantengagedinearningsmanagementcanbeexpectedtooperatein‘choice’mode,andwillbea‘rulesceptic’,viewingrulesasopentointerpretationandseekingfavourableinterpretations.Heorshe,however,willalsoneedtotakeintoaccounttheclassificatoryimpulsesofthosein‘realist’mode,suchasCooper’sreactiontoSullivan’sreversalofthebad-debtprovision,which,asnotedabove,‘smelledfunny’toher.ConclusionThesetoffactorsdiscussedsofar—interactionswithotherpeople,techno-logicalsystems,training,andsoon—iscertainlynotexhaustive.WhatIhope,however,isthatthischapterbeginstosuggesttherichnessoftheissuesthatafinitistperspectivehighlights.Financialreportingandtheaccountingpracticesunderpinningitarecomplexsociotechnicalmattersthatare,asnoted,centraltotheeconomicgovernanceoftoday’ssocieties.Inaprofit-orientedeconomy,themeasurementofprofitisacentraltask,andonethatisfarfromsimple.Thefinitistargumentsthatclassificationsarealwaysinprincipledecisionsandthatrulesinthemselvesdonotdeterminetheirapplicationtospecificcasesmakefinitism,Iwouldargue,ausefulperspectivefromwhichtoanalyse 136/MeasuringProfitaccountingandfinancialreporting.Eventhemostmundaneofaccountingclassificationsbecomesofanalyticalinterestfromthisperspective.Giventhat,iffinitismiscorrect,anyitemcouldalwaysinprinciplebeclassifieddifferently,whatdeterminestheclassificationthatisactuallymade?Theanswer,thischap-terhassuggested,islikelytobeacombinationoffactorssuchastrainingandhabit,thegoalsthoseinvolvedaretryingtoachieve,thecontextwithinwhichtheyaremakingtheclassification,andthepropertiesofthetechnologicalsystemswithwhichtheyhavetointeract.Thischapterhasdonelittlemorethansimplysketchquestionssuchasthisandspeculateastolikelyanswers.Nevertheless,Iamconfidentthatresearchaddressingtheseissueswillbeofgreatinterest.Thefinitistaccountantandfinitistbook-keeperareattheheartofeconomiclife,andwebadlyneedtoknowwhatdeterminestheiractions. 7ConstructingEmissionsMarketsUniversitiescontainroomsandbuildingsthatacademicsneverenter:boilerhouses,forexample.Amongsttheircontentsaremetersthatmeasuretheconsumptionofelectricity,naturalgas,andoil.Thosereadingsdeterminethesubstantialsumsthatuniversitiesspendonenergy:around£9millionperyearformyuniversity.SomeofthemetersattheUniversityofEdinburghnowhaveasecondfunctiontoo,indirectlymeasuringthecarbondioxideemittedbytheboilersandcombinedheatandpowerunitstowhichtheyareconnected.Largelyun-noticed,themetershavebecomepartofwhatisineffectameasurementnet-workstretchingacrossEurope.Althoughnotasextensiveasthesociotechnicalsystemsviawhich,asdiscussedinthepreviouschapter,profitismeasured,thisnetworkisconsequential.ItisacrucialpartofthematerialinfrastructureoftheEuropeanUnion’sEmissionsTradingScheme,anewmarket—itcameintoexistenceonlyinJanuary2005—whichistheUnion’smaintoolforcombatingglobalwarming.Amongsttheactivitiestowhichthenewtradingschemeappliesare‘[c]ombustioninstallationswitharatedthermalinputexceeding20MW’(EuropeanParliament,Council2003:42).Onedoesn’tusuallythinkofauniversityasalargecombustioninstallation—asidefromininevitablejokesaboutthegenerationofhotair—butthecapacityoftwoofmyuniversity’sthreehighlyefficientcombinedheatandpowerplantspushesthemwithinthe 138/ConstructingEmissionsMarketsscopeofemissionstrading.Itisperhapsnotsurprisingthatoursciencecampusconsumesalotofenergy,buteventheplantintheareainwhichhumanitiesandsocialsciencesaretaughttotals20.5MW,andthusisoverthethreshold.LiketheotheroperatorsofEurope’slargecombustioninstallations,theUniversityofEdinburghhasinconsequencetoholdpermitstoemitcarbondioxide.Itreceivesanallocationofallowances,eachpermittingittoemitatonne(1,000kg)ofCO2.Itsemissionsaremeasured,andithas‘tosurrenderallowancesequaltothetotalemissionsoftheinstallationineachcalendaryear’(EuropeanParliament,Council2003:35).IfitemitsmoreCO2thanithasallowances,itmustbuyfurtherallowancesonthemarket,orelseitwillbefined.Ifitemitsless,itcansellitsexcessallowances.Allowancesaretraded‘overthecounter’(bydirectinstitution-to-institutionnegotiation,orviabrokers)andonorganizedexchangessuchastheNordicpowerexchange,NordPool.CO2futuresarealsotraded,intheircaseontheEuropeanClimateExchange,whichusestheelectronictrad-ingplatformofLondon’sInternationalPetroleumExchange.Asmallover-the-countermarketinCO2optionsalreadyexists.ThenetworkoftradingstretchesbeyondtheEuropeanUniontoencompass‘certifiedemissionreduc-tions’fromprojectsindevelopingcountriesand‘emissionreductionunits’fromprojectsintheformerSovietbloc.IfaEuropeanoperatorsuchasEdinburghUniversityhasashortfallofallowances,itcanmeetitsobligationsbybuyingthesereductionsandunits.PlansforemissionstradingaregatheringmomentuminNorthAmerica,inthenorth-east,midwest,andinCalifornia;politicalpressureisgrowingforaUSnationalcarbonmarket.Thoseinvolvedincarbontradingindeedenvisagetheemergenceofaglobalmarket,inwhichCO2willhavea‘worldprice’,justasgoldhasaworldprice.Thischapterbeginsbyexaminingtheintellectualrootsofemissionstradingwithineconomics.Itdiscussesthefirstlarge-scaleexperimentwithemissionstrading,onethathashadamajorinfluenceonlaterdebate:thesulphur-dioxidetradingintroducedbytheUSCleanAirActAmendmentsof1990.Inextturnbrieflytothe1997KyotoProtocol,inwhichcarbontradingplaysacentralrole,andtoexperimentsincarbontradinginthepost-KyotoyearsbyoilcompanyBPandbytheUKgovernment.Drawingonasetoftwenty-fourinterviewswithcarbon-marketparticipants,IthendiscussinmoredetailtheEuropeanUnionEmissionsTradingScheme,describingtheprocessbywhichitcametobeestablished,thefactorsthatexplainsalientfeaturesofitsdesign,theevolutionoftheschemesinceitsestablishmentin ConstructingEmissionsMarkets/1392005,andthelessonstobelearnedfromtheexperience.Amajorfocusofthechapteristheprocessesoftheallocationofemissionsallowances.Whilethemaininitialproposalwithineconomicsforanemissionsmarketassumedthatallowanceswouldbesoldtoemitters,politicalconsiderationshavegen-erallydictatedthatinpracticetheyhavebeendistributedfree.Thismakesthemechanismofallocationandthetotalamountsallocatedcrucialtechnicalandpoliticalmatters,centraltowhetherornotanemissionsmarketiseffectiveenvironmentally.EconomicsandEmissionsMarketsToafargreaterextentthananyoftheothermarketsdiscussedinthisbook,marketsinemissionspermitsareacreationofeconomics.TheyareaquintessentialexampleofastrongformofthekindofprocessdiscussedinChapter2,inwhicheconomicshasdonesomething:itsrolehasnotbeentoanalyseanalready-existingmarket,buttohelpbringanewmarketintoexis-tence.Whilemeteorologistsandothernaturalscientistshavebeentheexpertsontheextenttowhichhumanactivitiesareincreasingthe‘greenhouse’warm-ingoftheplanet,economistshaveplayedaleadingroleindiscussingwhatshouldbedoneinresponse.Fromaneconomicpointofview,theeffectsoftheproductionofcarbondioxideandothergreenhousegasesbyhumanactivitiesareaninstanceofafamiliarissue.Theyarean‘externality’,a‘costorbenefitarisingfromanyactivitywhichdoesnotaccruetothepersonororganizationcarryingontheactivity’(Black2002:167).Researchanddevelopmentisanexampleofanactivitythathaspositiveexternalities:itsbenefitstypicallyaccruemorewidelythantothefirmconductingit.Pollution,incontrast,isthetextbookexampleofanegativeexternality:Thecosttosocietyofhavingsomeofitslaborandsteelusedupinagivenfactoryis‘internalized’bythefirm,becauseithastopayforthoseinputs.Butthefirmdoesnothaveaneconomicincentivetominimizethe‘external’costsofpollution.(Jaffe,Newell,andStavins2005:165)Thetraditionalresponsebyeconomiststonegativeexternalitiessuchaspollutionwastoproposeataxsetatalevelthatwould‘internalize’the 140/ConstructingEmissionsMarketsexternality,inotherwordsatalevelcorrespondingtothewidersocialcostsoftheactivityinquestion,thusforcingeconomicactorstotakethosewidercostsintoaccount.Theargument,whichgoesbackatleasttoPigou(1920:168,193–4,andpassim),isthat‘whenthemarketdoesnotreflectthetruesocialcostsofsomeactivity,beitswitchingontheairconditioningordrivingintoLondonatrushhour,thegovernmentcanmakesocietybetteroffbyimposingataxthatreflectsthatsocialcost’(Harford2006).Governmentscouldofcourseseektocontrolharmfulactivitiesdirectly,forexamplebybanningthementirely,orbylayingdownhowtheyshouldbeconducted(forexample,imposingfixedlimitsonthequantitiesorconcentrationsofpollutants).Economistshave,however,generallyheldthatenvironmentalbenefitscanmoreefficientlybeachievedbymanipulatingthepricesystemviataxationthan,forinstance,byinsistingonparticulartechnologicalstandards.Thatway,marketprocesses—ratherthancentralizeddecisionmaking—woulddeterminethetechnologicalorothermeansbywhichreductionswereachieved.Theideathattherewasathirdsolutiontotheproblemofnegativeexternalities—beyondtaxesanddirectcontrolmeasures—isgenerallytracedtoeconomistRonaldCoase.A‘mutuallysatisfactorybargain’(Coase1960:4)couldbestruckbetweenthosewhoseactivitiesgenerateanexternalityandthoseaffectedbyit,eitherbytheformercompensatingthelatterorthelatterpayingtheformertoreduceorceasethedamagingactivities;whichwouldoccurwoulddependuponwhethertherewaslegalliabilityforthedamagecaused.Ifrationalagentscouldstrikesuchbargainswithoutincurringtrans-actioncosts,theoutcomewould‘maximise...thevalueofproduction’(Coase1960:8).Taxeswereunnecessary;bargainingbetweenprivateagentscouldachievethedesiredoutcome.WhileCoase’sattackonPigouandhissketchofanalternativetowhathecalled‘thePigoviantradition’(1960:39)wasinfluential,itdidnotdirectlyinspiretheideaoftradeablepermits.ItsmostimmediateprogenitorwastheUniversityofTorontoeconomistandeconomichistorianJ.H.Dales,whowasconsideringhowtocontrolpollutionoftheGreatLakes(Dales1968a;1968b).1Inhisanalysis,Dalesassumeditwaspossibletoestablishan‘equiva-lence...betweendifferentwasteproducts’,sothatquantitiesofdifferentprod-uctscouldbetranslatedintostandardized‘equivalenttons’.Hethensupposedthatithadbeendecidedthat‘nomorethanxequivalenttonsofwasteper ConstructingEmissionsMarkets/141yeararetobedumpedintothewatersofregionA,andthatxrepresentsa10percentreductionfromtheamountofwastethatiscurrentlybeingdischargedintotheregion’swaters’(Dales1968a:800).How,askedDales,mightadecisiontoreducepollutiontoafixed‘cap’beimplementedatleasttotalcost?Forgovernmentsimplytorulethateachdischargermustreduceitsdischargesby10percentwouldnotbeoptimal:somedischargersmightbeablecheaplytoreducetheirdischargesbymuchmorethanatenth,whileothersmightfacehighcostsinachievingthe10percentreduction.Norcouldgovernmentplausiblyminimizetotalcostsbytryingtoworkoutanoptimalallocationinwhichthosewhocouldcutwastecheaplywouldhavetomakebiggercutsthanthoseforwhomitwasexpen-sive.Mobilizingastandardargumentagainstcentralplanning(anargumentassociatedaboveallwiththeeconomistsLudwigvonMisesandFriedrichvonHayek),2Daleswrote:Tosupposethatoptimalityinthissenseispossibleistosupposethattheadminis-trativeauthorityisabletosolveasetofthousandsofsimultaneousequations,whentheinformationrequiredtowritetheequationsinnumericalformisnotonlynotavailable,butalsooftenunobtainable.(Dales1968a:800)Abetterwaytosolvetheproblemofachievingagivenoverallreductionofpollutionatminimumtotalcost,arguedDales,wasatax:an‘across-the-board’feeof‘somuchpertonofwastedischarged’(Dales1968a:800).Withataxorfeeofthissort:eachpolluterdecidesforhimselfbyhowmuch,ifatall,heshouldreducehiswastes.Theburdenofpollutioncontrolisthussharedinexactlytherightway,withoutthe[WaterControlBoard]havingtoagonizeoverthequestionofhowtofindajustandreasonablesharingofthecostofthescheme.Sinceeverypolluteradjuststothechargesinwhateverwayminimizeshiscost,thesocialcostofachievingthetargetamountofwastedischarge—whichisthesumofthecostsbornebyallpolluters(and,ofcourse,bytheconsumersoftheirproducts,inthecaseofindustries)—willalsobeautomaticallyminimized.(Dales1968b:92,emphasisinoriginal)Twoquitedemandingcalculationsneverthelessremained,Dalesargued:findingtheleveloffeethatwouldachievethenecessaryoverallreductionindischarges;andtakingaccountofthe‘new-comers(peopleorfactories)’whowouldalsodischargewastes.Theformerwouldrequire‘trialanderror’, 142/ConstructingEmissionsMarketsandthelatterwasa‘guesstimate’.Thosecalculationscould,however,beper-formedimplicitlybysettingup‘a“market”in“pollutionrights”’:Suchmarketswouldautomaticallysetthecorrectlevelofthepollutioncharge(insteadofitshavingtobesetbysomecommittee,afterlongandlearneddiscussion)andwouldalsoautomatically,andcontinuously,adjustthelevelofthechargetotakeaccountofeconomicgrowth.Asimplemarketthatcanbeoperatedbythreeorfourpeopleandasmallstaffofstenographerstoregisterpurchasesandsalesisverymuchcheaper,andjustasefficient,asalargebureaucracyrepletewithcomputerstogiveanswerstocomplicatedpricingproblems.Ifitisfeasibletoestablishamarkettoimplementapolicy,nopolicy-makercanaffordtodowithoutone.(Dales1968b:93,100)WhatDalesproposedwaswhathasbecomeknownasa‘cap-and-trade’scheme.AWaterControlBoardorothergovernmentauthoritywoulddecideonthecap,thetotalquantityofallowabledischarges(forexample,a10percentreductionfromcurrentlevels).Itwouldthenmakethecorrespondingnumberofpermitsavailableforinitialsale,andalsosetupamechanismforasecondarymarketinwhichpermitscouldbesoldandboughtatanypointuptothedateatwhichdischargersofwastehavetosurrenderthemtotheauthoritytoshowtheyhadtherighttodischargethequantitiestheyhad.Thefactthatdischargeswerenowcostlywouldcausesomefirmsandmunicipal-itiestoreducetheiremissionsandthenumberofpermitstheythusneededtopurchase,and‘whentheprice[ofpermits]hasrisenenoughtoreducethedemand’totherequisitelevel‘themarketwillbeinequilibrium’.Thosewhoseemissionswerelowerthantheyhadanticipated‘wouldhaverightstosell,andthoseinthecontrarysituationwouldbeinthemarketasbuyers’.‘Allofthesebuyersandsellers,throughtheirbidsandoffers,willestablishthepriceoftheRights.’Thatpricewould,forexample,beexpectedtoriseifpopulationorindustrialactivitygrows,soincreasing‘theincentiveforwastedischargerstotreat,orreduce,theirwastes’(Dales1968a:801;1968b:93–4).Dales’sproposalwasnotputintopracticeintheGreatLakes.Inthe1970sand1980s,however,asmallnumberoftradeablepollutionpermitschemeswereintroduced,mainlybytheUSEnvironmentalProtectionAgency.Forexample,astheuseofleadadditivesingasolinewasphasedoutintheUSAinthe1980s,refinerieswereallowedtemporarilytotrade‘leadcredits’,anddidsotoareasonablylargeextent:‘Approximately15percentofthetotalleadrightsusedweretraded’(Hahn1989:102). ConstructingEmissionsMarkets/143TradingSulphurDioxideMoresignificant,however,thanthelimitedandoftenclumsilyimplementedtradingschemesofthe1970sand1980swasthesettingupintheUSAinthe1990softhemarketdiscussedbrieflyinChapter2:themarketinpermitstoemitsulphurdioxide(SO)fromelectricpowerstations.3Therehadbeen2considerableconcernsincethe1970saboutthedamagingeffectsof‘acidrain’andotheraciddepositions.Sulphurdioxideemissions,especiallyfromcoal-firedpowerstations,hadbeenidentifiedasthemaincause.Numerousbillstoaddresstheproblemwereproposedinthe1980s,butallfailedinthefaceofoppositionfromboththeReaganadministrationandDemocratsfromstatesthatmightsuffereconomically,suchasthoseinAppalachiaandthemidwestthatproducedhigh-sulphurcoal.Finally,however,theCleanAirActAmend-mentsof1990setthegoalofreducingannualsulphurdioxideemissionsintheUSAdramatically—by‘tenmilliontonsfrom1980emissionlevels’—acutlargelytobeachievedbyapproximatelyhalvingSO2emissionsfromcoal-firedpowerstations.4Ratherthanlaydownhowpowerstationsweretocutsulphurdioxide(forexample,bymandatingtheinstallationof‘scrubbers’toremoveSO2fromfluegases),Congressestablishedatwo-phasemarketinemissionspermits.PhaseI,from1995to1999,covered‘the263dirtiestlargegeneratingunits’;inphaseII,from2000onwards,theschemewasextendedtoinclude‘virtuallyallfossil-fueledelectricgeneratingplants’inthecontinentalUSA(Ellermanetal.2000:6).Operatorsoftheseplantswereallocatedallowances,andtheownersofcoal-firedplantswerethusineffectleftfreetochoosebetweendoingnothingtoreduceemissions(andthusmostlikelyhavingtobuyadditionalallowancesonthemarket),payingthesubstantialcostsofinstallingandoperatingscrub-bers,orshiftingtolower-sulphurcoal.Theultimateinspirationofthesulphur-dioxidemarketwastheworkwithineconomicsonemissionsmarketstouchedonabove.However,econo-mistsdidnotsimplyproposethemarketandthensitbackandseeitadopted:theywerepoliticallyactiveadvocatesofit.ParticularlyimportantinthisrespectwereRobertStavinsofHarvardUniversity’sJohnF.KennedySchoolofGovernment;MITeconomistRichardSchmalensee,whowasamemberofPresidentGeorgeH.W.Bush’sCouncilofEconomicAdvisors;andRobertHahn,oneoftheCouncil’sprofessionalstaff.5Stavins,forexample,hadbeenaneconomistattheEnvironmentalDefenseFund,anadvocacygroupwhich 144/ConstructingEmissionsMarketshired‘economists,engineersandcomputeranalyststofindwaystohelptheenvironmentwithoutharmingtheeconomy’,6andwhichwassuccessfullypositioningitself‘asthemostpro-businessofthemajorenvironmentalNGOsintheUnitedStates’(VictorandHouse2006:2102).FromEnvironmentalDefense,StavinsmovedtoHarvard,wherehedirected‘Project88’,aninves-tigationof‘HarnessingMarketForcestoProtecttheEnvironment’,jointlychairedbytheRepublicanSenatorJohnHeinzandDemocratTimothyWirth.Project88’sreportpublicizedanestimatebyconsultantstotheEnvironmentalProtectionAgencythat‘amarket-basedapproachtoacidrainreductioncouldsaveus$3billionperyear,comparedwiththecostofadictatedtechnologicalsolution’(Stavins1988:5).Project88’sbipartisansponsorshipgaveitpotentialforinfluencewho-everwonthe1988presidentialelection.Thevictor,GeorgeH.W.Bush,‘hadpromisedtobe“theEnvironmentalPresident”andhadadvocatedlook-ing“tothemarketplaceforinnovativesolutions”toenvironmentalprob-lems’(Ellermanetal.2000:21–2),andsomeimportantmembersofhisnewadministrationwerefavourablyinclinedtoturningthosewordsintoaction.Forexample,RobertGrady,whohadbeenthechiefspeechwriterfortheBush–Quaylecampaign,wasgiventheresponsibilityfornaturalresourcesandenergyinthepowerfulOfficeofManagementandBudget,wherehewasanenthusiastfortheproposalforamarketinsulphurdioxideemissions(Stavins1998:78).Despitechannelsofpoliticalinfluenceofthissort,economistsontheirowncouldnothavecreatedthemarketinsulphurdioxide.Otherswereneeded:notjusttechnologists,whowereneededinordertocreatethesystemofmea-surementdescribedinChapter2,butalsolawyers.Forexample,anespeciallyimportantrole,bothaslobbyistandasdrafterofTitleIVoftheCleanAirActAmendments(thesectionthatcoveredsulphurdioxidetrading),wasplayedbyJoeGoffman.LikeStavins,Goffmanworkedoriginallyfor(andinhiscaseeventuallyreturnedto)theEnvironmentalDefenseFund,buthealsoservedasassociatecounselfortheSenatecommitteeworkingontheCleanAirAct,andthenoversawthedetaileddevelopmentofthetradingprogramme’srulesasanactingsectionchiefattheEnvironmentalProtectionAgency.7Anemissionsallowanceisasubtlelegalentity,soitisnotsurprisingthatlawyersaswellaseconomistshavebeenheavilyinvolvedinthedevelopmentofemissionstradingschemesmoregenerally.Acrucialissueiswhetheranallowanceisapropertyright.Wereittobesucharight,theconstitutional ConstructingEmissionsMarkets/145protectionawardedtoprivatepropertyintheUnitedStateswouldhavemadesubsequentchangestotheSO2tradingprogrammedifficult.Section403(f)oftheCleanAirActAmendmentslaiddownthatanallowancewas‘alimitedauthorizationtoemitsulfurdioxide...Suchallowancedoesnotconstituteapropertyright.’8AdecadelatertherewastobeasimilarassertionaboutthelegalstatusoftheunitstobetradedundertheKyotoProtocol,whenthe2001MarrakeshAccordsdecreedthattheprotocol‘hasnotcreatedorbestowedanyright,titleorentitlementtoemissions’.9Instead,Kyotounitsare‘unitizedanddivisibleembodimentsofpromisesacceptedbysovereignstatesinthecontextofamultilateralagreementwhichforthatreasoncanberevoked,revisedandalteredthroughfurthernegotiation’(Yamin2005:16).10Aswellasrequiringtheskillsoflawyersandothers,economistsfounditnecessarytoaccommodatetothedemandsofthepoliticalprocess,sometimessettingasidewhattheirdisciplinaryinclinationsmighthaveledthemtopre-fer.AsStavinsputit:‘PolicyinstrumentsthatappearimpeccablefromthevantagepointofCambridge,Massachusetts’—thehomeofHarvard,MIT,andmuchoftheeliteofacademiceconomics—‘butconsistentlyproveinfeasibleinWashington,D.C.,canhardlybeconsidered“optimal”’(Stavins1998:83).Twoformsofaccommodationstandout.First,animpulseofmanyeconomistswouldhavebeentoperformacost–benefitanalysis,workingouttheoptimumlevelofcutsinsulphurdioxideemissionsbyquantifyingthecostsofcutsandthebenefitsintermsofreduceddamagetotheenvironmentandhumanhealth.Assigningdollarvaluestosuchdamagewould,however,havebeenintenselycontroversial,andtheeconomistsinvolvedinpromotingsulphurdioxidetradingavoidedit,simply‘accepting—implicitlyorotherwise’(Stavins1998:77)theround-numberreductionof10milliontonsthatquicklybecamedominantinpoliticaldebate.Thatreductionwasinnosensedemonstrablyoptimal—itwasessentiallyacompromisebetweenenvironmentalistdemandsforacutof12milliontonsandindustryproposalsfor8million(Burtrawetal.2005:259)—butinsteadof‘debatingthecostsandbenefitsof[thetenmillionton]goal,they[economists]simplyfocusedonthecost-effectivemeansofachievingit’(Stavins1998:77).Second,inDales’sproposedmarket,permitswouldhavebeensoldtowastedischargers.Awell-designedauctionwouldquicklyhaveestablishedtheirmarketprice,withouthavingtowaitforsubsequenttradingtodothis.Auctioningsulphurdioxidepermits,however,wouldhavehadahugepoliticaldisadvantage:itwouldhaverequiredtheutilitycompaniesthat 146/ConstructingEmissionsMarketsownedpowerstationstopaylargesumsinordertocontinuetoemit.TherewouldhavebeenasubstantialtransferofresourcesfromthosecompaniestotheFederalgovernment.Asthemainstudyofthetradingprogrammeputsit,‘thisalternativewassimplynotonthetable’(Ellermanetal.2000:24).Inconsequence,nearlyallthesulphurallowancesweresimplydistrib-utedfreeofchargetothecompaniesthatranthegeneratingunitsthattheschemewastocover.(Asmallproportionofallowances—around3percent—isheldbackandauctionedannuallyonbehalfoftheEnvironmentalProtectionAgencybytheChicagoBoardofTrade,inanauctioninwhichprivatesellerscanalsotakepart.11)Atthepricespredictedatthestartoftheprogramme,phaseIallowanceswouldhavebeenworthabout$6billion,andtenyearsofphaseIIallowanceswouldhavebeenworth$45–63billion(Ellermanetal.2000:36).Thosesumswereinasensenotional,inthatautilitywouldneedtosurrenderallowancescorrespondingtoitsemissions,buttheydoindicatethatadesirablecommoditywasabouttobedistributed.Afavourabledistributionwouldbevaluable,andanunfavourableonecouldbeexpensive.‘Withthatsortofrentonthetable,onewouldcertainlyexpecttoseeseriousrentseeking,andWashingtondidnotdisappoint’(Ellermanetal.2000:36).Thebasicallocationrulewasthatgeneratingunitsreceivedallowancesproportionaltothecalorificvalueofthefueltheyburnedinthebaselineyears1985–7.12Therewas,however,enormouslycomplexjostlingoverdeviationsfromthebaselinerule,withthestaffofSenatorsandRepresentativeslobbyingforprovisionsthatwouldfavourminingand/orutilityinterestsintheirstates,andotherstates,suchasFlorida,receivingfavourableallocationsbecausetheywereexpectedtobefinelybalancedinupcomingelections(Ellermanetal.2000:13–76).Forsomeoftheeconomistsinvolved,itwasaneducationinthepoliticalprocess.ThusMIT’sRichardSchmalenseerecalledlaughingwhenaspecialprovisionconcerninglignite,the‘browncoal’commoninNorthDakota,wasproposedatameetingofCongressionalstaffmembers.‘HewasforcefullyremindedthatNorthDakotawasarelativelypoorstatewithbleakprospectsand,moreimportant,thatChairmanBurdick[QuentinBurdick(D-ND),1908–1992,ChairoftheSenateCommitteeonEnvironmentandPublicWorks]wasnottobetrifledwith.’Theprovisiondulybecamelaw.13 ConstructingEmissionsMarkets/147TheRatchetTheneedtoassuagepowerfulpoliticalactorsbyhavingrulesthatwouldgeneratefavourableallocationsfortheirconstituentsmadethesectionsoftheCleanAirActAmendmentsgoverningthoseallocationscomplex.TheEnvironmentalProtectionAgency(EPA)hadthreeinternalteamsplusanexternalAcidRainAdvisoryCommitteeatworkontranslatingthosesectionsintodetailedrulesandactualallocations.‘Inordertorecordanddefenditsinterpretation,EPAdocumentedtheallowance-allocationmethodsindetail’,andproducedandmadepublicwhatwere‘essentially[two]largespreadsheets’containingthecalculationsthatgeneratedtheallocationstoeachofthe3,842unitscovered.14Accommodatingspecialinterestsbygrantingfavourableallocationscouldeasilyhaveunderminedtheenvironmentalgoalsofthesulphurdioxidemar-ket.Crucially,however,thespreadsheetbywhichallowanceswereallocatedhadtohave,ineffect,afixedtotalfrom2000onwards.Thatwastheresultofaprovisionknownas‘theratchet’,whichwasaddedtotheCleanAirActAmendmentsearlyinthepoliticalinfighting.Section403(a)oftheAmend-mentssetamaximumonthetotalannualallowancesthatcouldbeissuedfrom2000onof8.9milliontons.Iftheconsequenceofdetailedrule-makingwasatotalentitlementinexcessofthat,theallocationsofeachunitwouldbereducedproratatobringthetotalbackdowntotherequisitelevel.Theratchetsurvivedthein-fightingalmostintact.Exceptinonecase,15itturnedouttobepossibletoassuagespecialinterestswhilekeepingtheratchetmechanism—‘whichwasnotcontroversial’(Ellermanetal.2000:37)—outsidetheconflict.Theratchet’seffectsseemtohavebeenunder-estimatedbythoseinvolved:ascaling-backofallowancesofnomorethan5percentwasexpected,butinfactallocationswerereducedacrosstheboardbynearly10percent(Ellermanetal.2000:37).Itisworthnotingthatattackingtheratchetwasacollective-actionproblem(seeChapter2):anybenefitfromdoingsowouldhavebeensharedbyalltheutilitiesinvolved,makingthebalanceofcostandbenefitinfightingagainsttheratchetquitedifferentfromfightingforarulethatwouldhavespecificadvantagesforone’sownstateorcompany.Theratchetmadethegameofallocationzero-sum:theextraallowanceswonbyspecialinterestswereclawedbackfromallparticipants.Theclaw-back 148/ConstructingEmissionsMarketstookplaceoverayearafterthelegislationhadbeenpassedbytheSenateandHouseofRepresentatives,bywhichtimeWashingtonpoliticalcircleshadmovedontootherthings.16SotheassuagingofspecialinterestscausednomorethanfairlyminorlooseningoftheoverallcaponSO.172FromSulphurtoCarbonThe‘ratchet’thuskeptthepoliticsofallocation,whichwascertainlyintense,largelyseparatefromtheoveralleffectsofthenewmarket,whichseemsingeneraltohaveworkedwell.Certainly,themeasurementsystem,sketchedinChapter2,whichunderpinnedthemarketappearstohavefunctionedimpeccably,givingrisetonodisputeofwhichIamawareaboutthequantitiesofsulphurdioxideactuallyemitted.Largecutsinemissionswereachieved,18atacostfarlowerthanindustrylobbyistshadpredicted:around$1billionperyear,ratherthan$10billionormore(Kerr1998).Thepriceofone-tonphaseIallowances,originallyexpectedtobeintherange$290–410,infactaveragedonlyaround$150(Ellermanetal.2000:172–3;Swift2000).Amongstthecausesofreducedcostswerethatthecapitalcostofscrubbersfellsharply,thecostofoperatingthemturnedouttobelowerthananticipated,andtheoptiontobuyallowanceswhennecessaryavoidedtheneedtoinstallsparemodulesinscrubbersystemstobeusedwhenothermoduleswereoutofaction(Burtrawetal.2005:268).Crucially,too,rail-freightderegulationreducedthecostoftransportationfromWyoming’sPowderRiverBasin,themainsourceoflow-sulphurcoalintheUnitedStates.Someoftheabovefactorswouldhavereducedcompliancecostsevenifsulphuremissionshadbeencontrolled(astheywereinEurope)bymech-anismsotherthanemissionstrading:indeed,emissionswereindeclineintheUSAinthe1980sbeforethetradingschemebegan,forexamplebecausecoalwithlowlevelsofsulphurwasalreadybecomingmorereadilyavailable(Lohmann2006:101).Theeffectsoftradingperse—whichwassometimesquitelimitedinscale,forexamplebecausestate-levelregulationsconstrainedutilities’options(Burtrawetal.2005:265)—aredifficulttodisentanglefromtheeconomicadvantagesofsimplybeingabletochoosebetweenachievingabatementbyinstallingscrubbersorbyswitchingtolow-sulphurcoal.Despite ConstructingEmissionsMarkets/149thesecomplications,however,thetwomainstudiesoftheissue(Ellermanetal.2000,andCarlson,Burtraw,Cropper,andPalmer2000)areinbroadagreementinsuggestingsavingsofabout50percentbycomparisonwithatraditionalregulatoryapproach.Sulphurdioxidetradingcanthusbejudgedatleastaqualifiedsuccess,anditwascertainlyperceivedassuccessful.19Thatsuccess—particularlythefactthatthemarketpriceofallowanceswasmuchlowerthanhadbeenpredicted—movedemissionstradingfirmlyintothepoliticalmainstreamintheUSA.ThustheeconomistWilliamNordhaus‘happenedtobesittingbetweenRichardSandor[theChicagoBoardofTrade’sformerchiefecono-mist,mentionedinChapter4]andVicePresidentGoreattheWhiteHouseClimateConferenceinOctober1997.Theywerebothpointingtothesulfurprogramasanargumentagainstthe“dismal”economistswhowereprojectingthatthecostsofthe[December1997]KyotoProtocolwouldbeveryhigh.TheywereassertingthatoncethemarketsforCO2wereopened,thecostswouldplummet’(Nordhaus2000:65–6).TheClintonadministrationenteredtheKyotonegotiationsdeterminedthat‘theagreementshouldincludeanarrayofflexible,market-basedapproachesforreducingemissions’,andtheadministrationwasconvincedthatthoseapproacheswouldkeepthenetcosts‘relativelymodest’.Detailedeconomicmodellingconductedfortheadministrationsuggestedthatun-restrictedinternationaltradingcouldreducethecoststotheUSAtoafifthofwhattheywouldbeifthenecessaryreductionshadtobeachieveddomes-ticallyalone.Theresultantcost,theadministrationclaimed,wouldbeamere$7billionayear,0.07percentofanticipated2010grossdomesticproduct(ClintonAdministration1998:21,39,53).Inpushingfortheuseofemissionstradingincurbingglobalwarming,theClintonadministrationfacedoppositionbothfromtheEuropeanUnion(whichpreferredmandatorymeasures,andwhichhadalreadysoughtaEurope-widecarbontax)andfromthird-worldcountries,manyofwhichsawtheargumentthatgreenhousegasemissionscouldmorecheaplybecurbedthereas‘carboncolonialism’.TheUSnegotiators,however,pos-sessedenergy,professionalism,thecloutoftheworld’ssolesuperpower,andstrongconvictionofthecorrectnessofwhatoneobserverofthenegotiationscalled‘economicideology...dominatedbygeneralequilibriumconceptsthatfocusuponeconomicefficiencyandimplythatflexibilityachievesthesame 150/ConstructingEmissionsMarketsenvironmentalbenefitsatlowercosts:hence,themoreflexibilitythebetter’(Grubb1999:99,112).TheUSAprevailed.AkeymomentcameinNovember1997,justbeforetheKyotomeeting.Brazilhadproposedthattheindustrializedcountriesshouldfaceafineiftheybreachedthecommitmentstheywereabouttoenterinto.Thefine‘wouldbepaidintoaCleanDevelopmentFundthatwouldbeusedtosupportappropriateprojectsindevelopingcountries’.Needlesstosay,‘[t]heideathatindustrializedcountrieswouldagree’tosuchafine‘seemedfar-fetchedtoanyseasonedpolitician’.However,aseniorBraziliannegotiatorstartedtocanvasstheideaofsettingthefine‘ataratesettofundprojectsthatwouldsaveemissionsequivalenttothedegreeofnon-compliance’(Grubb1999:101–2).USnegotiatorssuddenlyrealizedthatwithaminortweak—turningthepaymentfromafineintoacontributiontowardsmeetingone’sobligations—theBrazilianproposalwasaroutetointernationalemissionstrading.A‘USteamdasheddowntoRiotoexploretheoptions’(Grubb1999:103),andKyoto’s‘CleanDevelopmentMechanism’wasborn.ThecoreoftheKyotoProtocolwastheundertakingofits‘AnnexI’signatories(theindustrializedcountries)thatbythetimeoftheprotocol’s2008–12‘commitmentperiod’theywouldhavelimitedtheirgreenhouse-gasemissionstoagreedproportionsoftheir1990levels(93percentfortheUSA,92percentfortheEuropeanCommunityoverall,andsoon).However,undertheprotocolthosecoun-triescanfundemission-reductionprojectsin‘PartiesnotincludedinAnnexI’(developingcountries),andcountthe‘certifiedemissionreductions’fromsuchprojectsagainsttheirKyotocommitments.20TheKyotoProtocolalsoprovidesforasimilar‘JointImplementation’mechanismwherebyoneindustrializedAnnexIcountrycanpayforandreceive‘emissionreductionunits’fromaprojectinanother(forexample,intheRussianFederationorUkraine,bothofwhichareAnnexIparties).Moresweepingly,theprotocolallowstheAnnexIcountries,whichhaveundertakentolimitemissionstothe‘assignedamounts’determinedbytheirKyotocommitments,totradethoseamountsamongstthemselves,increas-ingoneparty’sallowedemissionswhilemakingtheequivalentreductioninanother’s.TheKyotoProtocolwasthebarestskeletonofamarket—itpostponedforlaterdiscussion‘therelevantprinciples,modalities,rulesandguide-lines,inparticularforverification,reportingandaccountabilityforemissions ConstructingEmissionsMarkets/151trading’21—yettheUSnegotiatorshadsucceededinforginganagreementthatlargelyreflectedtheirpreferences.Kyoto’sironyisthus—asthereaderwilldoubtlesshaveanticipated—thathavinggotmuchofwhatitwanted,theUnitedStatesthenwalkedaway.InMarch2001,thenewadministrationofGeorgeW.BushannouncedthewithdrawaloftheUSAfromtheKyotoProtocol.EuropeanCarbonTradingAlthoughitwasnosurprise,theBushadministration’sdecisionwasahugeset-backforgreenhousegasemissionstrading.However,theideadidnotdiewiththewithdrawalofthecountrythatwasitsauthor;ithadstartedtotakerootoutsidetheUSA.AmongstthemechanismsbywhichthishappenedweretheeffortsoftheEnvironmentalDefenseFund,thekeynon-governmentalproponentofthesulphurdioxidemarket,whichalsopushedforcarbontrad-ing.Itviewed‘emissionstradingasawaytoendthepolaritybetweenbusi-nessandenvironmentalgroupsovertheclimatechangeissue’(Fialka2000),andworkedhardtosecuretheinclusioninKyotooftheprotocol’smarket-basedprovisions.Inparticular,EnvironmentalDefenseforgedapartnershipwithBritishPetroleum(BP),whichin1997,underitsnewchiefexecutiveJohnBrowne,brokerankswiththeothermajoroilcompanies,announcingthatitacceptedthatglobalwarmingwasamajorthreatandthatactionwasneeded.EnvironmentalDefense’sPresident,FredKrupp,‘lobbiedJohnBrownetoadoptacapandtradesystem’(VictorandHouse2006:2102).BPsetupaninternaltradingschemethatEnvironmentalDefensedescribedas‘amicrocosmoftheglobalemissionstradingsystemenvisionedatKyoto’(EnvironmentalDefenseFund1998:7).BPundertookthatby2010itwouldcutitsemissionsbyatenthfromtheirlevelin1990(agoalroughlycomparableinsizetosalientKyototargets).Afterpilottradinginautumn1998and1999,BPallocatedemissionsceilingstoallitsbusinessunitsfortheyear2000ofroughly99percentoftheir1998emissions.Theseceilingstooktheformofallowancesrecordedina‘centraldatabase,wheretheycouldbeelectronicallymovedfromone[businessunit]toanother’.Businessunitshadeithertolimittheiremissionstotheirallocationofallowancesorto‘buy’allowancesfromunitsthatexpectedtoemitlessthantheirallocations.Purchaseswere 152/ConstructingEmissionsMarketsmadewithoutmoneychanginghands,buttheschemewasgivenforcebyincorporatingitintoBP’sinternalaccountingandperformance-measurementsystems.The‘income’and‘expenses’earnedorincurredviacarbontradingbyeachbusinessunitenteredintothecalculationofacloselymonitoredinternalfinancialmetric:thebusinessunit’s‘returnoncapitalemployed’(VictorandHouse2006:2102).BPrapidlyfoundthatitsyear2000goalofa1percentreductioninemissionswasveryeasilyachieved;thegoal’smodestyhadarisenfromanoveroptimisticviewofthegrowthofBP’sbusinessandwhatturnedouttobeexaggeratedestimatesbybusinessunitsofprojectedgrowthintheiremissions.Ineffect,thecompanyhadover-allocatedemissionsallowances.Thecapfor2001wasthereforemademuchtighter,ineffectforcingalmostalltheremainderofthe10percentreductiontobemadeinasingleyear,andwithbusinessunits’capsbeingfurtherreadjustedquarterlyinorder,inthewordsofoneparticipant,‘continuously[to]pullthebehavioralleverstore-incentivize’(JohnMogford,headofBP’sClimateSteeringGroup,quotedinVictorandHouse2006:2105).Thesereductionstookplaceagainstabackgroundofmeasurementuncer-tainty:BP’sestimateoftheuncertaintyofitsestimateofits1990emissionswas30–40percent,andevenits1998estimatehadanerrorbandof5percent(anon.2002).WhentheschemeclosedinDecember2001,theequivalentof4.5milliontonnesofcarbondioxide(aroundatwentiethofBP’soverallemissions)hadbeentradedbetweenitsbusinessunitsatanaverage‘price’of$39.63pertonne,andBPwasabletoconcludethatthegoalofa10percentreductioninemissionshadalreadybeenachieved,nineyearsearlierthananticipated.Althoughthemeasurementbaselinepriortotheschemewasacknowledgedtobesomewhatuncertain,BPwasconfidentthatalarge,genuinereductionhadbeenmade.Overall,thecompanycalculatedthattheemissionscuthadbeenatnonetcosttothecompany;indeed,thereductionshadapositivenetpresentvalueof$650million,mainlybecauseofenergysavingsandbecausethecompanycouldsellnaturalgasthatpreviouslywouldhavebeenventedorflared(VictorandHouse2006:2105).BP’smotivesinintroducinginternaltradingweremixed:‘thedesiretoheadoffastandards-basedoratax-basedpolicy’(VictorandHouse2006:2101)wasamongstthem.AswithSO2,itisunclearhowmuchthetradingmechanismitselfhadcontributedtothesuccessofBP’sexperiment.Allofthereductions‘madeeconomicsensewithoutthefinancialreturnfromemissionstrading’ ConstructingEmissionsMarkets/153(VictorandHouse2006:2105).Thebulkofthereductions,especiallyinitially,wereinemissionsofmethane(themaincomponentofnaturalgas),notCO2.MethaneisamorepotentgreenhousegasthanCO2—morethantwentytimessoovera100-yearperiod(IntergovernmentalPanelonClimateChange2007:33)—andaccordinglyatonneofavoidedemissionsofmethanewasaccountedforintheBPschemeasequivalenttotwentytonnesofCO2(VictorandHouse2006:2103).Becausereducingnaturalgasventingandflaringwasrelativelystraightforward,itwasthuspossiblecheaplytogeneratelargesavingsin‘CO2equivalents’.Theverycreationofacompany-widemetrologicalnetwork,sys-tematicallymeasuringemissions,focusedmanagementattention.‘Whatgetsmeasuredgetsmanaged’,asJohnBrowneputitinthespeechatStanfordUniversityinwhichhefirstannouncedthatBPwastakingclimatechangeseriously(Browne1997).BP’sembraceofcarbontradingwasinfluential,particularlybecauseoftheinvolvementoftheincreasinglyprominentJohnBrowne(nowLordBrowne),anditsbroadlysuccessfuloutcomewaswidelynoted.BPstaffbecamepros-elytizersforemissionsmarkets,‘relentlessinpromoting[theBPscheme]asamodelforinternationaltrading’(Engels2006:342).AsaworkingexampleofcarbontradinginoneofEurope’sleadingcompanies,theBPschemewastobecome‘akeydriverinthepolicydebates’inEurope(ChristiansenandWettestad2003:9).22In2001,DenmarklaunchedaschemeintendedtopromoteCO2tradingamongsteightlargeelectricpowergenerators.23TheUKfollowedsuitwithabroader—butvoluntary—tradingschemelaunchedin2002.24TheBlairgovernmentthathadcometopowerintheUKin1997wasaparticularenthusiastforemissionstrading:likemany‘NewLabour’policies,green-housegasmarketsseemedawaytocreateanewmarketmechanismtoachieveagoalthatunfetteredmarketscouldnotachievebythemselvesandthattraditionalgovernmentregulatorytoolswouldachieveonlyatgreaterexpense.Whiletheotherschemesdiscussedinthischapterinvolvedfreedistributionofallowancestoincumbents,usuallyonthebasisofabaselineemissionslevel,theallocationmechanismintheUKschemewasanauction.However,theUKcarbonauctionwasagainnotthesaleofpermitsenvisagedbyDales.Thethirty-fourvoluntaryparticipantsintheUKschemesubmittedbidsforthequantitiesofemissionstheywouldundertakeinreturnfordifferentlevelsofgovernmentsubsidy.The‘descendingclock’auctionbeganatasubsidy 154/ConstructingEmissionsMarketsof£100pertonneofcarbondioxide,andthepricewasreduceduntiltheproductofpricesandquantitiesbidfellwithinthescheme’s£215millionbud-get,whichoccurredataclearingpriceof£53.37/tonne(NationalAuditOffice2004).Thathighprice—andthussubstantialsubsidy—madetheschemeanexpensivewayofsecuringabatement,andaswiththeinitialexperienceoftheBPscheme,participantsintheUKschemefoundthattheycouldeasilymeetthetargetstheyhadsetthemselvesviatheauction.Allowanceswereaccordinglyinsurplus.By2004,‘thedemandforallowances[had]essentiallyvanished’(ChristiansenandArvanitakis2004:7).TheEuropeanUnionEmissionsTradingSchemeIntheUSA,the‘ChicagoClimateExchange’establishedbyRichardSandor(seeChapter4)setupavoluntary—butlegallybinding—tradingscheme,initscasenon-governmentalandnon-subsidized.Butthe‘geography’ofemis-sionsmarketshadshiftedtowardsEurope.Byfarthemostimportantcarbon-tradingscheme—farmoreimportantthantheearlyinitiativesdescribedabove—hasbeentheEuropeanUnionschemewithwhichIbeganthischap-ter,andwhichisthecentreofwhatispotentiallyaglobalcarbonmarket.AlthoughtheEmissionsTradingSchemecanbeanalysedinseveralways,25fromtheviewpointofthisbookitismostinterestingtoseeitasanexerciseinmarketconstructionand,asdiscussedbelow,asaninvivoeconomicexperiment(inthesenseofMuniesaandCallon2007).AttheheartofmarketconstructionwasasmallteamofofficialsoftheEuropeanCommission,ledbyeconomistJosDelbekeoftheCommission’sEnvironmentDirectorate-General.AsPeterZapfelandMattiVainiooftheEnvironmentDirectorate-Generalputit,atthetimeofthe1997KyotoProtocol‘theconceptofemissionstradingwasknown[inEurope]onlyinnarrowscientificcirclesencompass-ingprofessorsandstudentsofEuropeanenvironmentaleconomicsandtoaverylimitedaudienceofenvironmentalpolicymakersandadministratorsinenvironmentalagencies’(ZapfelandVainio2002:1).Thatsituationrapidlychanged.Academicsandnon-governmentalorganizationsfromtheUSA(includingEnvironmentalDefenseandtheCenterforCleanAirPolicy)weredrawnintoEuropeanpolicydiscussions,andintervieweesreportedthattheEuropeanCommissionteamstudiedtheSO2experienceintensively.European ConstructingEmissionsMarkets/155consultantsandauditingandcertificationfirms‘sawapotentialmarketaris-ing’,andsomeindustryassociationsandcorporationsbecameinterestedinemissionsmarketsasapotentially‘cheap’alternativetoexpensiveregulatorymeasures(ZapfelandVainio2002:1and7).AmongstfactorsthatturnedthisgrowinginterestinEuropeintoanactualcarbonmarketwasonethatwasspecifictothepoliticalstructureoftheEuropeanUnion.Asnoted,thepreferenceofmanyEuropeanenvironmentpolicymakerswouldhavebeenforharmonizedcarbontaxesratherthanemissionstrading.However,plansin1992–5tolevysuchtaxesacrossEuropehadmetwithfierceindustryopposition,andtherewasastructuraldisadvan-tagetotryingtorevivethem.IntheEuropeanUnion,taxmeasuresrequireunanimity:asingledissentingcountrycanblockthem.TheinitialproposalforaharmonizedEuropeancarbontaxwas,asonepolicy-makerintervieweeputit,‘killedbasicallybytheUK’,whichwasconcernedtokeepjurisdictionovertaxatthenationallevel,althoughothermemberstateswerepotentialopponentsaswell,and‘ifonememberstate,evenaverytinymemberstate,saidno’toataxmeasure,‘itwasno’.Asanotherintervieweesaid,‘taxationisalwaysfraughtwithobstacles,andifitisn’tonememberstatewhoisobjectingit’sanother.Andit[theobjection]mightrotatearoundthememberstatesbecausetheyallhavetheirelectoralcycles.’TheunanimityrequirementwastoodemandingforaEurope-widecarbontaxatanysignificantleveltobefeasiblepolitically:‘Welearnedourlesson’inthatrespect,saidanintervieweeinvolvedinthetaxproposal.EuropeanUniondecisionmakingonenvironmentalmatters,however,fallsintotheterrainnotofunanimitybutof‘qualifiedmajorityvoting’.Nosinglecountrycanstopanenvironmentalmeasure:doingsotakesacoalitionofcountriessufficientlypopulous(sincevotingweightsroughlyfollowpopulation)toforma‘block-ingminority’.TheEuropeanCommissionsoughtlegaladviceonwhetheracap-and-trademarketcouldbeclassifiedunderEuropeanUnionproceduresasanenvironmentalissue,and,asanintervieweereported,‘gottheconfirmation’thatitcould:aCO2capwasanenvironmentalmatter,andthuswouldnotrequireunanimity.WhilethepoliticalstructureoftheEuropeanUnionmeantitwasnotfeasibletogivecarbonasubstantial,harmonizedpriceviaatax,itpotentiallycouldbegivenapriceviaacap-and-tradescheme.Emissionstrading,whichhadpreviouslyseemedinEuropetobe,inaninterviewee’swords,‘anacademicideawhich...wasseenastoofar...awayfromreality’,couldinfactbemaderealmoreeasilythanataxcould. 156/ConstructingEmissionsMarketsTherewascertainlyoppositiontothecompulsoryEurope-widetradingschemethateventuallyemerged.Environmentalcampaignerswere(andoftenstillare)scepticalofemissionsmarketsthatallowpollutersthecapacitytopaytokeeppolluting:astandardanalogyisbuyingindulgencesforsins.TheEuropeanCommissionteam,however,tookcaretoengagewithleading,mainstreamenvironmentalNGOs,andthelattergenerallycametosupportthenewmarket,albeitcritically.Industryoppositionwasgenerallymuted,withsomesectorswelcomingtrading(certainlypreferringittoatax),thoughmuchofGermanindustrywasopposed,feelingithadalreadytakenononer-ouscommitmentsundervoluntaryagreementswiththeGermangovern-ment.TheUKwasagainapotentialobstacle:intervieweesreportedthatitwantedtheEuropeanmarkettobesimilarinitsdesigntotheUKschemedescribedabove.ButevenifGermanyandtheUKhadunitedinopposition,theywouldhavefallenshortofthevotesrequiredtoformablockingminority(ChristiansenandWettestad2003:13).Bytheautumnof2001,therewere‘[i]ndicationsthataqualifiedmajority’inthepremierdecision-makingbody,theCounciloftheEuropeanUnion(the‘CouncilofMinisters’,asitusedtobecalled),wouldindeedbackemissionstrading,andsupportintheEuropeanParliament,whoseagreementwasalsoneeded,wasstrong(ChristiansenandWettestad2003:7).TheemergenceoftheDanishandUKtradingschemes,andaparliamentarycommissiononemissionstradingsetupinSwedenin1999,werealsoaspurtoaction.Theyposedthethreatthatapartial‘patchworkofincompatiblenationaltradingschemes’couldemerge(ChristiansenandWettestad2003:7).SomesmallercountriessuchasAustria,Finland,andIrelandwereinclinedtosupportemissionstradingbutfelttheirdomesticmarketstobetoosmalltosupportanationalscheme(ZapfelandVainio2002:10).TheEuropeanCommission,oneofwhosekeygoalsisharmonizedEuropeanmarkets,saw‘thedevelopmentofacoordinatedEU-wideschemewithcommonrules’asawayofavoiding‘marketfragmentation’(ChristiansenandWettestad2003:7).Anintervieweereported,forexample,thatrepresentativesofDenmarkandtheUKhadmettodiscusslinkingtheirtwoschemes,andfoundthatthedifferencesbetweenthedesignofthetwomadeitinfeasible.ItalsograduallybecameclearthattheUnitedStateswasnotgoingtoplaytheleadingroleitearlierhad.TheClintonadministrationseemedunabletoturnitsKyotocommitmentsintoconcreteproposalsthatstoodachanceofsucceedinginCongress,and,asnoted,theBushadministrationthen ConstructingEmissionsMarkets/157abandonedtheprocess.TheinitialbeliefinEuropehadbeenthatasystemofinternationalemissionstrading,basedaroundtheUnitedNationsbutledbytheUSA,wasonitsway.Asitbecameclearthatthiswasnotso,itwasreplacedbyasensethat‘wehavetomakeithappenathomefirst’(ZapfelandVainio2002:8).Indeed,thewithdrawaloftheUSAseemstohavereducedoppositioninEuropetocarbontrading,perhapsbecausethattradingwasnolongersomethingtheUSAwastryingtoimposeonreluctantpartners.26America’snewunilateralism—onclimate,andonmuchelse—provokedhos-tility.ThetortuousdiplomaticprocessofaddingfleshtoKyoto’sskeletonmarketmay,paradoxically,havebeenfacilitatedbythewithdrawaloftheUSA:‘Inasinglestroke,theUnitedStatesmanagedtofocustheentireplanetona...convolutedpieceofdiplomaticliterature’(Benedick2001).FouryearsaftertheKyotoProtocolwassigned,the2001MarrakeshAccordsfinallysetoutthedetailedrulesgoverningtheinternationalschemes—theCleanDevelopmentMechanismandJointImplementation—withwhichaEuropeanmarketcouldbelinked.WithasignificantcurrentofopinionamongstEuropeanpolicymakersgalvanized,momentumbuiltupbehindtheprojectofbuildingaEurope-widetradingscheme.Bylate2002,thedesignoftheEuropeanschemewascom-pleted.ByOctober2003,thedirectiveestablishingtheschemewasagreedandpublished(EuropeanParliament,Council2003),anditwasfollowedinOctober2004bythe‘linkingdirective’(EuropeanParliament,Council2004)layingthefoundationfortheinterconnectiontotheKyotomarkets.InJanuary2005,Europe’scarbon-tradingschemebeganoperation.MarketDesignThefactorsinfluencingthedesignofthenewmarketwereheterogeneous.Measurement,forexample,iskeytoanyemissionsmarket,andthepracti-calitiesofmetrologymeantthatthemarket’sarchitectsquicklyconcludedthatitwouldhavebeenoverambitioustoincludeeithergasesotherthanCO2orland-usechangesthataffectthebiosphere’scapacitytoabsorboremitgreenhousegases.Inacontextofactualorpotentialoppositiontothenewmarket,unreliablemeasurementandmonitoringwouldhavebeenfatal,saidoneofmyinterviewees: 158/ConstructingEmissionsMarketsFromaveryearlystage,thecommunications[policypapers]of’98or’99,wehadidentifiedaCO2-onlyschemeasbeingthemostfeasible...Itwasbasicallyamonitoringissue...wefeltthatifothergasescouldn’tbemonitoredrobustlyweweren’tabletoaffirm[thescheme’sintegrity].Andthecriticismweweregettingwasthatthiswillallbesmokeandmirrorsandwecouldn’trefutethatcriticismunlesswe’vegotaproposedrobustmonitoring.Ratherthanbuildinganewsystemofmetrology,asthedesignersofthesulphurmarkethad,theEuropeanschemeasfaraspossiblemeasuresCO2emissionsbyre-employingtheexistingmetersandothermechanismsbywhichsupplierschargeforenergy.Themeasurementsofcoal,gas,andoilinputsthusobtainedareturnedintoamountsofcarbondioxidebymultiply-ingquantitiesoffuelbyemissionandoxidationfactors,sometimesspecific,forexampletoaparticulartypeofcoal,butmoreusuallystandardfactorsforthefuelinquestion.Thescopeoftheschemewas,similarly,keptdeliberatelylimited.ItcoversonlyaroundhalfofEurope’sCO2emissions(Delbeke2006:293),essentiallythosefromlargefixedsourcesofCO2.Publicpowerandheat(primarilyelectricitygeneration)isthelargestsector,accountingforroughly60percentoftheEmissionsTradingScheme,withtheothersectorsinvolvedbeingenergy-intensiveindustries,notablymetalsproduction,oilandgas,andcement,lime,andglass.Groundtransport,marinetransport,andaviationwereallomitted,aswasthedomesticsectorotherthanviatheeffectsoftheschemeonprices,especiallyofelectricity.(Aviationistobebroughtintothescheme,probablyin2012.)Despitethedecisiontokeeptheschememodestinitscoverage,awholenewapparatusofverificationhadtobecreatedtoauditCO2measurements,alongwithNationalRegistries,whicharedatabasesinwhichtheissuanceandtransferofCO2allowancesarerecorded,asistheircancellationwhenemit-terssurrenderallowancescorrespondingtotheiremissionsoverthepreviousyear.A‘CommunityIndependentTransactionLog’,registeringtransfersofallowancesacrossnationalboundaries,wasdesignedandmadeoperationalimpressivelyquickly.(ThecorrespondingsystemforKyotoProtocoltransac-tionswas,incontrast,muchdelayed.)Ofcourse,notalltheworktopreparethenewmarketwasdonebytheEuropeanCommissionanditscontractors.Europe’saccountingregulators,forexample,hadtoconsiderhowCO2allowancesshouldbetreated,whichinvolveddecidingwhatkindofaccountingitemanallowanceis.Therewas ConstructingEmissionsMarkets/159noexistingbodyofpracticethatwasdirectlyrelevant:becauseaccountingregulationintheUSAisdifferent,thewaySO2allowancesareaccountedforcouldnotsimplybeadoptedforCO2inEurope.Withanunfamiliaritemtoclassify,thefinitistissuesdiscussedinChapter6cametothefore.CO2allowanceshadanalogiestoaccountingitemswithestablishedtreatments,buttomultiplesuchitems,notjustone.Asanaccountantintervieweeputit,thecarbonmarketis:anemergingmarketinacommoditythathasaspectsofdifferenttypesofmarketplaces....Ithassomecharacteristicsofagovernmentgrant,hassomecharacteristicsofanintangibleasset,hassomecharacteristicsofinventory,hassomecharacteristicsofafinancialinstrument...anditdependshowit’sappliedacrosswhichsectors.Indeed,nofullysatisfactoryaccountingtreatmentwasfound:aproposed‘Interpretation’ofexistingaccountingstandardstoencompassthenewcom-moditywaswithdrawninJune2005inthefaceofcriticism,andatthetimeofwritingnoreplacementhasyetbeenagreed(Casamento2005;Cookforthcoming).ThePoliticsofAllocationThetrickiestissueofall,however,wastheallocationofCO2allowances.TheMarch2000EuropeanCommission‘GreenPaper’thatwasthebasisfortheconsultationexerciseontheproposedmarketlaidoutthecaseforauctioningallowancesratherthanallocatingthemfreeofcharge:Periodicauctioningistechnicallypreferable,asitwouldgiveanequalandfairchancetoallcompaniestoacquiretheallowancestheywantinatransparentmanner.Auc-tioningappliesthe‘polluterpays’principle.Therevenuesraisedbygovernmentscouldbere-cycledinavarietyofways,evenkeepingtheoverallrevenueeffectneutral,orbyusingtherevenuestopromoteenergyefficiencyinvestments,researchanddevel-opmentorpublicinvestmentinothergreenhousegasabatementefforts.Auctioningavoidstheneedtotakethedifficultandpoliticallydelicatedecisionsabouthowmuchtogiveeachcompanycoveredbythetradingscheme.Thecomplexissues...aboutstateaidandcompetitionwouldlargelydisappear.Itwouldalsoguaranteefairtermsfornewentrantstojointhesystemasthey,likeexistingsources,wouldalsohavethesameopportunitytobuytheallowancesthattheyneeded.27 160/ConstructingEmissionsMarketsUnsurprisingly,however,themarket’sarchitects‘were...lobbiedveryhardbyindustrytoallocatefree’(interviewee).Theyknewtheconstraintstheywereunder:Interviewee:...ourdecisions...weredrivenbywhatwethoughtwecouldachieve...MacKenzie:Achievepolitically?Interviewee:Indeed.Whilethereweresomeactors—membersoftheEuropeanParliamentandoneparticularmemberstate,Sweden—infavourofauctioning,thebalanceofforceswasfirmlyinfavouroffreeallocation.TheSDPinGermany,forexample,wasunenthusiasticaboutthenewmarket,andcouldbepersuadedbyitscoalitionpartner,theGreens,tosupportitonlyifallocationwasfree.‘Oncethey’dagreedthat,itwasthefoundationoftheirgovernmentandneithersidecouldgobackonit.Andtosomeextentweknewthat,sowecouldn’tforcetheissue’,reportedaninterviewee.Allthatwasfeasiblewasaprovision—inpracticenotheavilyused—thatgavememberstatestheoptiontoauctionupto5percentofallowancesinthefirstphaseofthescheme(January2005toDecember2007),and10percentinthesecondphase(January2008–December2012).InaEuropeanUnionmadeupofindependentnationstates,withtheirdis-tinctiveinterests,priorities,andapproachestoenvironmentalmatters,itwassimilarlyinfeasiblepolitically,whensettingupthenewmarket,toinsistuponstandard,centralizedrulesforfreeallocation.AttheheartoftheEmissionsTradingSchemeisadeepcompromisebetweenitsarchitects’desirefor,inaninterviewee’swords,‘anabsolutelyopenmarketwithonesinglemarketprice’andtherootednessofmanydecisionmakers,evenintheenvironmen-talsphere,intheirnationalsettings.‘Themostnationalofthings...wastheNationalAllocationPlan’,whichlaysdownbothacountry’soverallallocationofCO2allowancesandtherulesbywhichtheyaredistributedtosectorsandcompanies.Althoughthemarket’sarchitectsknewthatthisdecentralizationheldthepotentialfortrouble,theyalsoknewitwasanunavoidablecompro-mise:‘thequidproquowaswegotourmarketandthememberstatesgottheirNationalAllocationPlan.’Forthemarkettoworkatall,however,NationalAllocationPlanshadtoberegulated,becauseoftheobvioustemptationformemberstatestobe ConstructingEmissionsMarkets/161toogenerousinallocationstotheirindustriessothatthelattercouldprofitbysellingallowancestheydidnotneedtoemittersinothercountries.TheDirectiveestablishingtheEmissionsTradingSchemelaiddownthatproposedplanshadtobesubmittedtotheEuropeanCommission,gavethelattertherighttorejectthem,andsetouttherulestobeappliedbytheCommissioninjudgingthem.Thetotalquantityofallowancestobeissuedunderamemberstate’sNationalAllocationPlan:(a)mustbecompatiblewithmeetingthestate’scommitmentsundertheKyotoProtocolanditstargetunderthe1998‘BurdenSharingAgree-ment’,whichallocatestoeachofthethenfifteenmemberstatesanagreedshareoftheEuropeanUnion’s8percentcutinemissions;and(b)‘shallnotbemorethanislikelytobeneeded’(EuropeanParliament,Council2003:43).Thelattercriterionisparticularlypertinenttothosememberstatesthatareontracktomeetorover-fulfiltheirKyotoor‘BurdenSharing’commitments(as,forexample,mostoftheformerSoviet-bloccountriesthathavejoinedtheEuropeanUnionare,becauseofthecollapseofmanyoftheirheavyindustries).Thecriterionwasintendedtostopsuchcountriesissuingmoreallowancesthantheirindustrywaslikelytorequire,takingaccountofanticipatedeco-nomicgrowthandtechnologicalandotherimprovementsthatcouldreducecarbonintensity(emissionsperunitofgrossnationalproduct).ApplyingtherulesforassessingNationalAllocationPlanswas—asafinitistanalysisofthekindsketchedinChapter2wouldsuggest—farfromstraight-forward.Theimplicationsofthetwocriteria—consistencywithKyotoandnoexcessoverwhatwas‘needed’—werecontestable.WhentheNationalAllo-cationPlanssubmittedforthefirst(2005–7)phaseoftheschemewerebeingassessed,theKyoto‘commitmentperiod’(2008–12)stilllayinthefuture,andcountries’undertakingsforitcouldbemetbyabatementinthe‘non-traded’sectorsnotcoveredbytheschemeaswellasviathescheme’scaps.Thepredictionsofeconomicgrowthandtechnologicalimprovementneededtoestimate‘need’forallowancescouldlikewisebedisputed.Aboveall,though,existingpatternsofemissionswerenotknownwithanygreatcertainty.Atthelevelofnationalemissionstotals,anintervieweetoldme,‘youfindridiculousthingslikeleakagefromgaspipelines,orcompletegaspipelines,nothavingbeenconsidered’.Typicallevelsofuncertaintyinoverallnationalgreenhousegasemissionsinventorieswereestimatedtorangefrom±4percentto 162/ConstructingEmissionsMarkets±21percent(Monni,Syri,andSavolainen2004).WhilethereisreckonedtobelessuncertaintyinestimatesofCO2producedfromfuelcombustion(perhaps±2percent:seeMonni,Syri,andSavolainen2004:93),suchestimatesweretypicallyproducedfromaggregatetotalsoffuelconsumption.Keydataneededtoassesshowmanyallowanceswere‘likelytobeneeded’—inparticu-lar,existingpatternsofplant-levelemissions—hadinmanycasessimplyneverbeencollectedsystematically.Intherun-uptothelaunchofthetradingscheme,theinstallationstobecoveredwereencouragedtosupply,totheirnationalgovernments,dataontheirexistingemissions,andmostdidso,‘perhapsbecausetheallocationstoinstallationsdependedonthesedata’(EllermanandBuchner2007:70).Laterinvestigationsuggestedthat,despitetheincentivetoexaggerateexistingemissionstogetmoreallowances,mostofthisreportingwasdoneingoodfaith,butitneverthelessformedaless-than-securebackdropagainstwhichtoassessNationalAllocationPlans.ThearchitectsoftheEmissionsTradingSchemeknewperfectlywellthatthecombinationofhavingeachmemberstatedrawupitsownNationalAllo-cationPlanandincompletedataagainstwhichtoassessthoseplanswaslikelytogiverisetoanovergenerousdistributionofallowances.Asanintervieweeputit:‘Wewerefearfulofafirstphasethatwasn’tsufficientlystringent’,andforthatreasonrefusedtoallow‘banking’ofunusedallowancesforuseinthesecond(2008–12)phase.The‘firewall’thuscreatedwasdesignedsothatovergenerositydidnot‘contaminateasecondphase’.Thescheme’sarchitectsalsoknew‘wecouldn’thaveawaronallfronts’byrejectingalmostallNationalAllocationPlans.WiththeEmissionsTradingSchemenotyetlaunched,itwasalltooeasy,asthisintervieweeputit,toimaginedisgruntledmemberstatesconvening‘aspecialEuropeansummittopostponeitbysixmonths,orayear,orthreeifthegoinghadbeentootough’.Accordingly,whilethemostclearlyexcessiveNationalAllocationPlanswereindeedcutbacksubstantially,othersthatinretrospectshouldalsohavebeencutwereleftintact.ThePriceofCarbonDespitetheseandotherproblems,afunctioningmarketwascreated,andfromthestarttradinginitmeantthatcarbonfinallyhadaEuropean-wide ConstructingEmissionsMarkets/163price:becauseallowancescouldbeusedinanyoftheEuropeanUnion’smem-berstates,thepotentialforarbitragekeptthatpriceclosetouniformacrossEurope.Furthermore,withallowancestradedbothonorganizedexchangesandviabrokersthepricewaspublic,andatleasttoadegreeafact,inthesenseofreflectingperceptionsofthechangingeconomicvalueofallowances.On25February2005,forexample,theFinancialTimescarried,forwhat(asfarasIamaware)wasthefirsttime,anarticlediscussingthepriceofcarbonandadeterminantofitintermsverysimilartothoseituseswhentalkingaboutothercommodities:‘Carbonpricessoaredyesterday,sparkedbyasnapofcoldweatheracrossEurope...CarbonreachedC9.08(£6.23)atonneattheendoftrade...Brokersyesterdayreportedmorethan800,000tonnesofcarbonchanginghandsinvolatiletrade’(Harvey2005).TheEuropeancarbonpricecontinuedtoriseduringthefirsthalfof2005,reachingalmostC30/tonnebyJuly.Bothmarketparticipantsandeconometricanalysisidentifiedtwomaindeterminantsofthechangingprice,bothrelatedtothewayinwhich,asoneintervieweeputit,theEuropeancarbonmarket‘sitsontopoftheelectricitymarket’,thedominantsectorinthescheme.Thefirstwasthepatternoffuelprices:asnaturalgaspricesroserelativetothoseofcoal,therewasanincentiveforelectricitygeneratorstoshifttothelatter,increasingtheirneedforallowancesbecauseofcoal’smuchhigheremissions.Thesecondwasweather.AsontheFebruarydayreportedbytheFinancialTimes,coldwinterweatherincreasesthedemandforelectricityandhencetheneedforallowances.Dryweather,similarly,wasseenastendingtoincreasethecarbonprice,byreducingthefutureamountsofelectricitythatcanbeproducedbyhydropower(PointCarbon2006).Sothecarbonpricewasseenasreflectingfactorsthatwere‘real’.AsPointCarbon,themarket’sleadinganalysts,putitinFebruary2006:‘themarketistoalargeextenttradingonchangesinthefundamentals...Thisisyetanothersignalofthemarketworkingeffectively...themarketpriceisnotarbitrary.’However,politics(notgenerallyseenbymarketparticipantsasamongstthe‘fundamentals’)remainedprominent:asurveyofPointCarbon’ssubscribersrevealedthatintheiraggregateopinion‘politicalfactors’werethesecondmostimportantdriver(nexttofuelprices)ofthecarbonpriceintheshortterm,anditsmostimportantdriverinthelongterm.‘Itwouldclearlybeapositivedevelopmentiftheimportanceofpoliticswasreducedandreplacedbyamorepredictablefundamentalbothasariskandapricedriver’(PointCarbon2006:19,21). 164/ConstructingEmissionsMarketsThatcarbonneverthelesshadwhatwasseenasatleasttosomeextentanon-arbitrarypriceinEuropehadglobaleffects,aboveallviathelinkbetweentheEuropeanschemeandKyoto’sCleanDevelopmentMechanism.TheCDMwas,assuggestedabove,thecentrepieceofKyoto’sefforttolinktheinterestsofthedeveloped‘north’anddeveloping‘south’.AsoneintervieweewhohadbeeninvolvedinnegotiatingtheCDMputit:‘Iknowfirst-handthatwasagenuinebargain,oneofthefewrealnorth/southbargainswheretherewassomethinginitforusandsomethinginitforthemandeveryoneknewit.’But,hewenton,itwasabargainthatwouldbekept‘onlyifitworks,onlyiftheyactuallyseemoneyflowtothemandweseecreditsgoback’.TheCDMisbroaderinscopethantheEuropeancarbonmarket.Itsremitistheentiredevelopingworld(includingChinaandevenSouthKorea),andpotentiallyitencompassesallthegreenhousegasesandallthemeansofcon-trollingthem,includingforexamplereforestation.Totheresultantdauntingproblemsofmetrologywereaddedaslow,cumbersomeapprovalmechanismand,crucially,thenecessitytodemonstrate‘additionality’.ToqualifyundertheCDM,itmustbeshownthataprojectwill,inthewordsofarticle12oftheKyotoProtocol,resultin‘[r]eductionsinemissionsthatareadditionaltoanythatwouldoccurintheabsenceofthecertifiedprojectactivity’—arequirementthatgivesrisetothepotentiallytrickymethodologicalproblemofdemonstrating,withcredibility,whatwouldhappenintheabsenceoftheprojectbeingassessed(see,e.g.,Lohmann2005;2006).Unsurprisingly,therefore,thebulkofemissionsreductionsundertheCDMhavesofarcomefromdevelopingcountrieswithreasonablysophisti-catedinfrastructures.Inparticular,anumberof‘niches’havebeenfoundinwhichtheproblemsofmetrologyandthedemonstrationofadditionalityaretractable.Particularlynoteworthy—becausetheyhavebeenthelargestsinglesectoroftheCDM—havebeenlarge-scaleprojectsinKorea,India,andChinatopreventemissionsofHFC23(trifluoromethane),anespeciallydamaginggreenhousegasthatisaby-productoftheproductionoftherefrigerantandchemicalfeedstockHCFC22(chlorodifluoromethane).28Alsoprominenthavebeenprojectstoreduceemissionsofnitrousoxidefromplantsproducingadipicandnitricacids.AsoneintervieweeinvolvedintheCDMputit,‘whatwe’vediscoveredisthattherearepeopleouttherewhoarefantasticallyexpertatnitrousoxideandHFC23...andweputalittlebitofmoneywiththemandgoroundthe ConstructingEmissionsMarkets/165worldandcollectinemissionreductionsbyapplyingcatalytictechnologiestoindustrialprocesses’.However,forthat‘littlebitofmoney’tobeforthcomingfrominvestorsforprojectsofthissort—or,forexample,forthesmallerbutmorenumerousrenewableenergyprojects,suchastheMongolianwindfarmofwhichtheaboveintervieweewasproud—thecredits(‘certifiedemissionsreductions’)fromsuchprojectshavetohaveamonetaryvalue.ThelinkbetweenEuropeancarbontradingandtheCDMgivesprojectsunderthelatterthismonetaryvaluebecausethosecreditscanbeusedinEuropeinplaceofEmissionsTradingSchemecarbonallowances.Europe’semittershavethereforebeenfundingprojectsundertheCleanDevelopmentMechanisminordertoearncreditsthattheycanuse,andotherfirmsfundprojectsinordertohavecreditstosellataprofit.Theselatterfirmswereineffectconducting‘carbonarbitrage’(seeChapter5).Intheautumnof2005,forexample,withallowancestradinginEuropeatC20–5/tonne,therightstocertifiedemis-sionreductionsfromCDMprojectscouldstillbepurchasedataroundC7–8/tonne.Itwasonlyapotentialarbitrage.IfCDMcreditswerepaidforonlywhentheywerecertifiedanddelivered,fundingproblemsmightcausetheunderly-ingprojectsnottobecompleted;ifcreditswerepaidforinadvance,thentherewastherisktheywouldnotbeforthcoming,forexamplebecauseoffailuretoachievecertificationorbecauseofthebankruptcyoftheentityrunningtheproject.Nevertheless,thepotentialarbitragecastitsshadowaheadofit,infusingfactualstatusintothepriceofcertifiedemissionreductions(CERs)fromCDMprojects.Oneclimate-changenewslettercommentedinApril2005that‘theoperationoftheEUETS[EmissionsTradingScheme]seemstohavemadethevalueofCERsmore“real”toinvestors’(Latham&WatkinsLLP2005:3).Asanintervieweeputit:‘we’recreatingavalueassociatedwithreducingemissions.Andit’smonetizable.’Therewereothersourcesofmonetizability—viavoluntarycarbonoffsetschemes,andbecauseCDMcreditscanbeboughtbycountriestomeettheirKyotocommitments—butbecauseofitssizetheEuropeancarbonmarketwasthecrucialsource.Aspricesinthatmarketrose,sotoodidthepriceofcontractsthatwouldproduceCERs.Byearly2006,CERsfromfirmcontractswithnear-termdeliveryandperceivedlowriskwerecommandinguptoC14/tonne(PointCarbon2006:25).AbatementprojectsthroughouttheworldhadthusbecometightlylinkedtotheEuropeanEmis-sionsTradingScheme. 166/ConstructingEmissionsMarketsEmissionsMarketsasEconomicExperimentsEmissionsmarketshadmovedfromaproposalbyeconomiststohavingglobaleffects.TheseventhofthepreceptssketchedinChapter2is‘economicsdoesthings’,andthatisclearlythecasehere:emissionsmarketswouldnothavecomeintobeingwithouteconomicsandeconomists.AsinthecelebratedcaseoftheuseofeconomicsinthedesignoftheauctionsofthefrequencyspectrumbytheFederalCommunicationsCommissionandotheragencies,notjusteconomists’proposalsbut‘economiststhemselveshadtocirculate’(MuniesaandCallon2007:183).Inthatcase,economistsbecamedirectlyinvolvedinthedesignoftheauctionsandinadvisingfirmsontheirbiddingstrategies.Here,especiallyinthecaseofthepivotalearlymarket—inSO2emissions—akeyroleofeconomistswasaspolicyentrepreneurs.Whatpolicy-entrepreneureconomistsdidwastohelplaunchasetofwhat,followingMuniesaandCallon(2007),canbedescribedas‘economicexper-iments’.Bytheterm,theyincludemorethantheincreasinglyprominentfieldof‘experimentaleconomics’,inwhich‘markets’arecreatedinlaboratoryconditionsandsubjects’behaviourinthemisstudied.Economicexperimentsare:researchactivitiesinthesensethattheyaimatobservingandrepresentingeconomicobjects,butalso—andquiteexplicitly—inthesensethattheyseektointerveneontheseeconomicobjects:toseizethem,tomodifyandthenstabilizethem,toproducetheminsomespecificmanner.Toexperimentistoattempttosolveaproblembyorganizingtrialsthatleadtooutcomesthatareassessedandtakenasstartingpointsforfurtheractions.Experimentationisactionandreflection.(MuniesaandCallon2007:163)Ofthetwomainmarketsdiscussedinthischapter,theSO2marketandEuropeanUnionEmissionsTradingScheme,thelatterismoreexplicitlyanexperimentinthissense.Asnoted,itisbeingimplementedinphases:phase12005–7;phase22008–12;andphase3from2013onwards.Phase1waslargelydominatedbythepracticalities(bothtechnicalandpolitical)ofensuringthatthescheme—MuniesaandCallon’s‘economicobject’—wasindeedbroughtintobeing.Thebroadframeworkofrulesforphase2wasestablishedinthe2003Directiveestablishingthescheme,butimportantaspectsofwhatisbeingdoneinthatphase,especiallyinrespecttoallowanceallocation,wereinfluenced—asdiscussedbelow—bywhatwaslearnedfromphase1and,crucially,bythe ConstructingEmissionsMarkets/167€/t3530252015105001.12.0401.01.0501.02.0501.03.0501.04.0501.05.0501.06.0501.07.0501.08.0501.09.0501.10.0501.11.0501.12.0501.01.0601.02.0601.03.0601.04.0601.05.0601.06.0601.07.0601.08.0601.09.0601.10.0601.11.0601.12.0601.01.0701.02.0701.03.0701.04.0701.05.0701.06.0701.07.0701.08.0701.09.0701.10.0701.11.0701.12.07Figure7.1.Thepriceofallowancesforphase1oftheEuropeanUnionEmissionsTradingSchemeSource:CourtesyPointCarbon.wayinwhichtheverycreationofthe‘economicobject’reshapeditstechnicalandpoliticalcontext.Theprocess,underwayatthetimeofwriting,ofdesign-ingphase3(theplansforwhichwerenotlaiddownatthestartofthescheme)is,likewise,beingshapedinpartbytheaccumulatingeffectsoftheevolvingprocessofmarketconstruction.ThatthefirstphaseoftheEmissionsTradingSchemeshouldbeseenasaneconomicexperimentisimportantbecausemanyobserversconcludedonthebasisofitthattheschemewasafailure.Aftergenerallyclimbinginearly2005and2006—atonepointreachingC31/tonne(seeFigure7.1)—theEuropeancarbonpricefellbyover30percentinasingleday,26April(Morrison2006).Bymid-May,allowancesweretradingaslowasC9/tonne.Despiteaslightrecoveryinsummer2006,29pricesthencontinuedtoslide.ByDecember2007,onecouldbuyaone-tonnecarbondioxideallowanceforaslittleas4eurocents,30littlemorethanathousandthofthepeakpricesofthepreviousyear.Theinitialcauseofthepricecollapsewastheleaking-outoftheinformationthatanumberofcountries—theCzechRepublic,theNetherlands,Estonia, 168/ConstructingEmissionsMarketstheWalloonregionofBelgium,andFrance—hadreportedtotheEuropeanCommission2005emissionsofCO2bytheirindustriesthatweresubstantiallylowerthanthenumberofallowancesissuedtothoseindustries.Itgraduallythenemergedthatthiswastheoverallpattern:bythestartofJune2006,itwasbeingcalculatedthatthetotalofallowancesissuedexceeded2005’sactualemissionsby3.4percent,oraround60milliontonnes(CarbonMarketData2006).Installationsinseveralcountries—notablyItaly,Spain,and,especially,theUK—wereleftwithsizeablenetshortagesofallowances,butthoseshort-ageswerefaroutweighedbythesurpluseselsewhere.TherewerevariouspossiblereasonsforEurope’snetsurplusofallowances.Theweatherlatein2005hadbeenmild,andsomeabatementhadmostlikelybeenpromptedbythestartoftheschemeandbytherelativelyhighcarbonpricein2005andearly2006(seebelow).Byfarthemostplausibleexplanation,however,wasthattoomanyallowanceshadbeenissued.Asnotedabove,althoughthemostexaggeratedclaimsofnational‘needs’hadbeenrejected,theexigenciesofgettingtheschemeupandrunningandthepaucityofdetaileddataagainstwhichtoassessclaimshadledtomoreallowanceshavingbeencreatedthanwereactuallyneededfor‘businessasusual’.Itwasadamagingoutcome.TheEuropeancarbonprice,havingearlierseemeda‘fact’reflectingfundamentalsofthemarketsuchasweather,wasrevealedbythe2006crisistohavebeenarbitraryafterall,theresultofpoliticalprocessesofallocationtheeffectsofwhichhadinitiallynotbeenunderstoodfully.AlsoproblematicforthereputationoftheEmissionsTradingSchemewasthatanalysisofelectricitypricessuggestedthatdespitedoingrelativelylittletoreducetheiremissionsgeneratorshadearnedsignificantincomefromthescheme,ineffectpassingontoelectricityconsumersthepriceofcar-bonallowances,whenthesehadbeenallocatedtothegeneratorsfreeofcharge(see,e.g.,Sijm,Neuhoff,andChen2006;PointCarbon2007).Standardeconomicspredictedthattheywoulddothat.Evenwithfreeallocationofallowances,companieswill,otherthingsbeingequal,increasepricestotakeintoaccountthe‘opportunitycost’ofusingthoseallowancestoproduceelectricity(orothergoods),ratherthansellingthem:‘[I]nprincipleandinlinewitheconomictheory,acompanyisexpectedtoaddthecostsofCO2emissionallowancestoitsothermarginal(variable)costswhenmaking(short-term)productionortradingdecisions’(Sijm,Neuhoff,andChen2006:50). ConstructingEmissionsMarkets/169Forexample,Sijm,Neuhoff,andChenestimatethatwhentheaverageallowancepricewasaroundC20/tonnetheelectricitysectorofNetherlandswasmakingtheequivalentofaroundC300–600millionperyearprofitfromtheEmissionsTradingScheme(2006:67).TheUK’sproducersarereckonedtohaveprofitedintotalbyaround£2billionin2005–7(Crooks2007).Myintervieweesintheelectricitysectorfirmlydefendedtheappropriatenessofaddingtheopportunitycostofallowancestoelectricityprices,citingtheaboveeconomicargument,buttootherstheresultantincomewas‘windfallprof-its’.InGermany,forexample,theBundeskartellamt(FederalCartelOffice)chargedgeneratorRWEwithbehavingillegallyinpassingonpricesinthisway(RWEAG2007);thecasewassettledoutofcourt.Sowastheexperimentafailure?AgainsttheserealproblemsoftheEmis-sionsTradingSchemearethreesubstantialachievements.First,thereisten-tativeevidencethatitdidproducesomeabatementeveninitsfirst(2005–7)phase.EllermanandBuchner(2007)reportthattotal2005emissionsfromthesectorscoveredbytheschemewerearound7percentlowerthanwouldhavebeenpredictedsimplybyextrapolatingpre-2005emissionsinthelightofeconomicgrowthandthetrendincarbonintensityinEuropeaneconomies.Ofcourse,thatapparentabatementwouldbepartiallyorentirelyspuriousifpre-2005emissionshadbeenexaggeratedtogainadditionalallowances.Asnotedabove,however,thereseemstohavebeenless‘gaming’ofthissortthanmighthavebeenexpected,sosomegenuineabatementprobablytookplace.Inparticular,myintervieweesintheelectricitysectortoldmethatwhentheEuropeancarbonpricewashighitdidprovideanincentivetosupplyelectricitywherepossiblefromgas-firedplantsratherthanthemorecarbon-intensivecoal-firedplants.Second,thecomingintobeingoftheEuropeancarbonmarkethasbeenessentialtothegrowthoftheCleanDevelopmentMechanism.ThereareunquestionablyproblemswiththeCDM,rangingfromtheprotractedanddemandingapprovalprocesstodoubtsaboutthegenuine‘additionality’ofsig-nificantnumbersofprojects,questionsaboutlocalenvironmentalandsocialimpact,andconcernsthatCDMfundingoftheeliminationofHFC23isgeneratingwindfallprofitsfortheproducersofHCFC22(andevenprovidingaperverseincentivetoincreaseproductionofHCFC22).31Furthermore,thescalesofaroftheCDMhasbeensmallinrelationtowhatwouldbeneededsignificantlytorestrainglobalemissions,especiallyfromcountriessuchas 170/ConstructingEmissionsMarketsChinaandIndia.However,aninstitutionalstructurehassuccessfullybeencreatedthatdoesallowforsignificantflowsofcapitalfromthedevelopedtothedevelopingworldinordertoachieveabatement.Whiletheproblemsofensuringthatprojectswouldnothavetakenplaceinanycaseandofmoni-toringandverifyingabatementareunquestionablyreal,thoseproblemshavetobesetincontextbyconsideringthelikelyanalogousdifficultiesoftheothermainmeansofachievingsignificantcapitaltransfers:directgovernmentaid.Third,asthefirstlarge-scaleworkingcarbonmarket,theEuropeanUnionEmissionsTradingSchemeservesasavitalexampleforthosedevelopingsuchmarketsinotherplaces.InNorthAmericainparticular,momentumisgrowingbehindtheconstructionofcarbonmarkets.Onesignificantregionalcarbonmarketisalreadywellontheway:tenstatesinthenorth-easternUSA(Connecticut,Delaware,Massachusetts,Maine,NewHampshire,NewJersey,NewYork,RhodeIsland,Vermont,andMaryland)havesetuptheRegionalGreenhouseGasInitiative,acap-and-trademarketfortheirelectricitygen-erators.InAugust2006,thestateofCaliforniaapprovedabillintendedtoreturnthestate’scarbonemissionstotheir1990levelsby2020—asignificanttask,givenCalifornia’sratesofpopulationandeconomicgrowth—andlaidthefoundationforanemissionstradingschemetohelpachievethatgoal.TheCalifornianschememayexpandintoawiderregionalmarket(includingCanadianprovincesaswellasstatesintheUSA),andindeedtherehavebeenaslewofproposalsintheformofCongressionalbillsaimedatconstructingaUSA-widemarket.TheEuropeanexperiencehasbecomeavitalpartofthedebateintheUSA:OneofthekeyareasthatHillstaffersandthewesternstatecounterpartswillhome-inonishowRGGI[theRegionalGreenhouseGasInitiative]handlestheallocationofallowances.EveryoneishopingtopreventwhathappenedinthefirstphaseofEurope’sEmissionTradingScheme,whenthespotpriceofatonneofCO2droppedfrommorethanC30($43)inthespringoflastyeartobelowC0.10today[December2007]....Additionally,thefreeallocationofpermitsinEuropeledtoaccusationsofwindfallprofitsforgenerators...(Volcovici2007:28)Whilesuchproblemscouldhavebeenpredictedonothergrounds(asnotedabove,the‘windfallprofits’aretheresultofamechanismthateconomistswouldanticipate),theirappearanceinaninvivoeconomicexperimentmakesthemfarmorecompellingasaninputintothepolicydebateintheUnitedStates. ConstructingEmissionsMarkets/171TheExperimentChangesitsContextItwould,ofcourse,bequitemistakentoregardtheEuropeanUnionEmissionsTradingSchemeasanalogoustoaclassiclaboratoryexperiment,inwhichconsiderableeffortisdevotedtotightcontrolofexperimentalconditions.The‘economicobject’thathasbeenconstructedisnotastaticone,noristhereasingle‘experimenter’whoischangingtheexperimentalconditions.Asalreadynoted,thescheme’sarchitectshavehadtoworkundertightpoliticalconstraints,buttheveryexistenceoftheexperimenthasbeguntoshiftthoseconstraints.ThewaytheexperimentchangeditscontextcanbeseenmostclearlyinthecontrastbetweentheprocessofassessingNationalAllocationPlansforphase1ofthescheme(2005–7)andphase2(2008–12).Asnotedabove,theneedtogettheschemeupandrunningandtheweaknessofthedatathatcouldbedrawnupontoassessproposedplansledinphase1toover-allocationofallowancesandtoapricecrash.Inphase2,incontrast,onlyfourofthetwenty-sevenmember-stateplanssubmittedtotheCommissionwereapproved(Denmark,France,Slovenia,andtheUK).Forexample,theNationalAllocationPlanofEurope’slargestemitter,Germany,wascutfrom482milliontonnesto453.1,andswingeingcutswereimposedonsomeothermemberstates’proposals.Latvia,forexample,haditsdesiredcapof7.7milliontonnesreducedto3.43million,andLithuania’sproposed16.6millionwascutto8.8million(anon.2007:3).KeytothecutswastheapplicationofwhatparticipantscametocalltheEuropeanCommission’s‘NAPformula’(see,e.g.,Wyns2007;NAPsareNationalAllocationPlans).Settingcomplexitiesaside,thecoreofthisisthefollowingexpressionforcalculatingacountry’s‘needed’annualaverageallo-cationofallowancesfor2008–12:(2005emissions)×(GDPgrowthfactor)×(carbonintensityimprovement)TheGDPgrowthfactorreflectstheanticipatedincreaseineconomicactivitybetween2005and2010(themid-yearofthesecondphaseofthescheme).InthecaseofLatvia,forexample,whoseGDPwasexpectedtogrowby50percentbetween2005and2010,theGDPgrowthfactorusedbytheEuropeanCom-missionwasthus1.5(CommissionoftheEuropeanCommunities2006).ThecarbonintensityimprovementfactorsimilarlytakesintoaccountanticipatedchangesinemissionsperunitofGDP. 172/ConstructingEmissionsMarketsTheNAPformulaisawayofmakingconcretetherulethatcountriesshouldnotallocatemoreallowancesthanareneeded.Again,though,itshouldnotsurpriseusthatallthethreefactorsintheformulaarecontestableandcon-tested.(Atthetimeofwriting,ninememberstatesweresuingtheEuro-peanCommissionoveritsrulingscuttingtheiremissionscaps.)Forexample,should2005emissionsbeusedasthebaseline,orshouldanaverageoversev-eralyearsbeemployed?Memberstatesarguingforaveragingtypicallysug-gestedthattheir2005emissionswereanomalouslylow,forexamplebecauseofweatherconditions.FromtheCommission’sviewpoint,however,thegreatvirtueoftheuseof2005emissionswasthatthesehadbeenverifiedundertheEmissionsTradingScheme,whichmeanttheywereunlikelytobeexaggera-tions:companies’immediateeconomicinterestswouldhavebeenservedbyunderestimatingratherthanoverestimatingemissions,becauseoftheneedtosurrenderallowancescorrespondingtothoseemissions.The2005emissionsweretherefore‘harder’factsthanestimatesofemissionsinearlieryears.InthecaseofLatvia,forexample,theCommissionnotedthat‘itcannotbeexcludedthatemissionsfiguresreportedbyLatviainrespectofearlieryearsoverstateactualemissions’(CommissionoftheEuropeanCommunities2006:4).GDPgrowthfactorsandcarbonintensityimprovementswere,likewise,challengeable.Again,though,theCommissionwasabletoavoidsimplyrely-ingontheestimatesofthesethatmemberstatessubmittedinthespreadsheetaccompanyingtheirdraftNationalAllocationPlan(thespreadsheetcanbefoundinCommissionoftheEuropeanCommunities2005).TheCommis-sion’scrucialresourceinthisrespectwasaneconomicmodel,PRIMES,‘amodellingsystemthatsimulatesamarketequilibriumsolutionforenergysupplyanddemandintheEuropeanUnion(EU)memberstates.Themodeldeterminestheequilibriumbyfindingthepricesofeachenergyformsuchthatthequantityproducersfindbesttosupplymatchthequantityconsumerswishtouse’(NationalTechnicalUniversityofAthensn.d.:3).InitsdecisioninrespecttoLatvia,forexample,theCommissionspelledouttheadvantagesofPRIMESasawayofdeterminingGDPgrowthfactorandcarbonintensityimprovement:Ofallthedataatitsdisposal,includingthoseinthepublicdomain,theCommissionconsidersthedataindicatedinthePRIMESmodelasthemostaccurateandreli-ableestimationsofbothGDPgrowthandcarbonintensityimprovementrates.ThePRIMESmodelhasbeenusedforanalysisofenergyandclimatepolicyforalongtime ConstructingEmissionsMarkets/173andthebaselineassumptionsareupdatedonaregularbasistoreflectthemostlikelyfuturetrend.Furthermore,baselineassumptionsarevalidatedwiththeinvolvementofexpertsfromMemberStates....ThereisnootherdatasourceatthedisposaloftheCommissionwhichoffersacomparabledegreeofconsistencyanduniformaccuracyacrossallMemberStates,thusensuringequaltreatmentofMemberStates.(CommissionoftheEuropeanCommunities2006:5–6)Theuseofthe‘NAPformula’,the2005data,andPRIMESmeantaprocessoftheassessmentofNationalAllocationPlansthatintheeyesofcarbon-marketparticipantscametobeseenashavingthevirtueofpredictability.AftertheJanuary2007decisionsontheBelgianandDutchNationalAllocationPlans,KjerstiUlsetoftheconsultancyPointCarbonasked,‘ShouldanyonebesurprisedbytheNAPdecisions?’andansweredno:‘anyoneshouldnowbeabletopredicthowtheCommissionwillruleontheremainingcountries,oratleastcomeveryclose.TheEC’smethodologyforNAPassessmentsiscrystalclear’(Ulset2007:1).TheveryexistenceoftheEuropeanUnionEmissionsTradingScheme,eveninitsexperimentalfirstphase,hadchangedtheformsofassessmentthatwerefeasible.First,asnotedabove,the2005emissionsmeasurementsundertheschemeprovidedacrediblebaseline.Second,theimplicitthreatinthebackgroundtophase1assessments—thatdisgruntledmemberstatesmightsucceedinhavingthescheme’slaunchpostponedorconceivablyevencancelled—wasnolongerpresentwhenphase2planswerebeingassessedin2006–7.Theassessmentofphase2allocationsthustookplaceinwhatoneintervieweecalled‘acompletelydifferentcontext’:‘Feasibility,dataavailabil-ity....Factsareinplace.Oneisthattheemissionstrading[scheme]isthere,andtheideaofitevernotbeingthereisnotontheagendaandwouldshockthepoliticalprocess.’Eveninphase2oftheEmissionsTradingScheme,theabatementrequiredofparticipantsismodest.ThetotalEurope-wideannualcap,2.10billiontonnes,isonlyslightlylowerthantheequivalent2005emissions(around2.23billiontonnes),andgiventhatcreditsfromtheCleanDevelopmentMechanismcanbeused(uptosetproportions)inplaceofEuropeanallowances,evenphase2iscompatiblewithagrowthinemissionsinEurope.Nevertheless,reasonablysubstantialphase-2allowanceprices,nearlyC30/tonneatthestartofJuly2008,indicatedperhapsthatthegoalofgivingcarbonameaningfulpricewasgraduallybeingachieved.Thetermsofpolicydebatewerealsochanging. 174/ConstructingEmissionsMarketsTheEuropeanUnion’s2007decisiontocommititselftohavecutitsemissionsby2020by20percentoftheir1990levels,whateverothercountriesdid(andby30percentifothersjoinedininternationalaction),madeitpossibletocanvassfarmorestringentphase-3caps,andlarge-scaleauctioning—politicallyinfeasibleintheearlierphasesofthescheme—waslikewisebeingproposedinfluentially.TheverdictofonecarbonbrokerIinterviewed—that‘nowyou’vegotadialthatyoucanturn’,inotherwordsthatitwasnowpossibletoreduceemissionsbytighteningtheEuropeancaps—understatedtheextenttowhich‘turningthedial’downwardswasstillpoliticallycontentious,buttherewassomejusticeinhisverdictthat‘wedoseemtohavesetupaworkingsys-tem’.Ofcourse,theproblemofcombatingclimatechangeismuchtoofar-reachingandcomplextobesolvedbyasingleclassofinstrumentssuchasemissionsmarkets:directregulation,large-scalepublicinvestmentinresearchanddevelopmentandininfrastructure,internationalaid,and(incontextsinwhichtheyarepoliticallyfeasible)carbontaxesarealllikelytobepartofthenecessaryarmoury,asistheremovalofthemanysubsidiesthatexistgloballyfortheextractionanduseoffossilfuels(see,forexample,Lohmann2006,andPrinsandRayner2007).Nevertheless,theexperimentofEuropeancarbontradinghaddemonstratedthatconstructingaworkinggreenhousegasemissionsmarket,thoughenormouslydemanding,mightindeedbefeasible.ConclusionTheconstructionofemissionsmarketshasmanyaspects.Forexample,afea-tureofeconomicexperimentsinlaboratoryconditionsisthatparticipantsoftenhavetobe‘taught’tobehaveaseconomicallyrationalagents(see,forexample,MuniesaandCallon2007).SoitiswithEuropeancarbontrading:theconstructionofeconomicagentsisinasensestillincomplete.Forexample,evenacompanyintheEmissionsTradingSchemethathasbeenallocated,free,asmanyallowancesasit‘needs’forbusinessasusualshouldneverthe-lessbemotivatedtoachieveabatement,ifthatabatementcostslessthanthemarketpriceofemissionsallowances,becauseitcanthenearnprofitbysellingunneededallowances.Interviewees,however,reportedlittleclearevidenceofthishappening.Insteadoftreatingacarbonmarketasaprofitopportunity, ConstructingEmissionsMarkets/175mostenergy-intensiveindustry(theelectricitysectoraside)wastreatingthenewandunfamiliarschemeprimarilyasacompliancematter,abusinessofensuringthatithadenoughallowancestocoveritsemissions.Inphase1,evenallowancesthatatypicalindustrialemitterwasclearlyunlikelytoneedwereoftenheldbackuntilitsemissionsintheyearinquestionwereknownwithcertainty,whichmayexplainwhyallowanceswereabletemporarilytocommandhighpriceseventhoughtheywereinaggregatesurplus.Evenaftertheextentofover-allocationbecameclearinspring2006,pricesonlyverygraduallydroppedtoclosetozero(seeFigure7.1),withmanycompaniesnotsellingunneededallowancesevenwhentheycouldstillearnsubstantialincomebydoingso.Suchissuesmust,however,beleftforfuturediscussion.Instead,letmeendbyreturningtothecentralthemerunningthroughthischapter’sdiscussionoftheconstructionofemissionsmarkets:thetendencytoover-allocation.Ineachofthecap-and-trademarketsdiscussed—thesulphurdioxidemarket,theBPscheme,andthefirstphaseoftheEuropeanUnionEmissionsTradingScheme—therewerestrongforcespushinginthedirectionofover-allocation,andeventheUKscheme,thoughdifferentlystructured,manifestedanequiv-alentphenomenon.Itseemslikelythatthetendencytoover-allocationwillmanifestitselfwheneveranewemissionsmarketiscreated.Freeallocationwilltypicallybemoreattractivepoliticallythanauctioning,measurementsofexistinglevelsofemissionsareoftenfarmoreuncertainthanonemightimagine,thepotentialstraightforwardlyandcheaplytoachieveabatementisoftengrosslyunderestimated,and—crucially—thepoliticalexigenciesofgettingmarketsestablishedmaymakestringencyinallocationinfeasible.Atendencytoover-allocationis,therefore,apredictablefeatureofwhatmightbecalledthe‘technopolitics’ofconstructingemissionsmarkets,andanover-allocatedmarketispointless.32Yetthischapterhasalsoemphasizedthepossibilityoffindingtechnopoliticalmeansofcombatingthistendency.The‘ratchet’inthesulphurdioxidemarketwasacrucialsuchmeans,andtheEuropeanCommission’smobilizationoftheNAPformula,the2005emissionsdata,andthePRIMESmodeltomoveinphase2oftheEmissionsTradingSchemetowhatoneintervieweecalled‘moremechanistic’allocationmayalsoprovetohavebeeneffectiveinendingover-allocation,althoughatthetimeofwritingitisstilltooearlyinphase2tobecertain.Iusetheexpression‘technopolitics’becausethepoliticsofemissionsmar-ketsrunsfardeeperthanthequestionsonwhichmuchoftheexisting 176/ConstructingEmissionsMarketsliteraturetendstofocus,suchasthereasonsforthechoiceofamarketratherthanatax.Takethephase-2allocations,forexample.Apparently‘technical’matterssuchastheuseof2005emissionsasthebaseline(ratherthananaverageoveralongerperiod)ortheemploymentofthePRIMESmodelarenotmattersof‘meredetail’butcentraltowhetherover-allocationcanbeavoided,andthustowhetheranemissionsmarketachievesitsenvironmentalgoals.Suchmattersoftenescapenotice:forexample,Ihavebeenabletodiscoveronlyonebriefdiscussionofsulphur’s‘ratchet’intheexistingliterature,despiteitscrucialroleinthesuccessofthemuch-discussedSO2market(Ellermanetal.2000:37,38,and49).Yetsuchspecifics—thetechnopolitical‘nutsandbolts’ofthedesignofmarketsandofallocationmechanisms—cannotbeignored.Hence,Iwouldargue,theappropriatenesstounderstandingemissionsmar-ketsofanapproachofthekindproposedinthisbook.Scalesaren’tstable:‘micro’phenomena(details,technicalities)arecrucialtothesuccessorfailureofemissionsmarkets.Emissionsmarketsarepoliticallyattractive,makingpos-siblecoalitionsof‘left-wing’environmentalismand‘right-wing’pro-marketsentiment.Yettomakethemsuccessfulweneedapoliticsofmarketdesign,onethatfocusesnotjustontheoverallvirtuesanddemeritsofmarketsolu-tionsbutontechnopoliticalspecificssuchastheratchetandNAPformula.Theneedforsuchapoliticsislarge.Asnoted,carbon-marketconstructionisgatheringmomentuminNorthAmerica,anditislikelythatifasuccessortotheKyotoProtocolcanbenegotiatedaglobalcarbonmarketislikelytobeatitsheart.Inthisevolvingcontext,restrainingthetendencytoover-allocation,andinotherwaysmakingemissionsmarketseffective,aredaunting,difficultproblemsthataresimultaneouslytechnicalandpolitical.Theyare,however,problemsthatmustbefaced:theissuesatstakearefartooimportantforthemtobeevaded. 8Conclusion:OpeningtheBlackBoxesofFinanceFriday,14September2007:Britain’shighstreets.Itwasascenefromadifferenttimeoradifferentplace:fromDepression-eraAmerica,perhaps,ormodernArgentina.TheUnitedKingdom’sfirstlarge-scalebankrunsinceVictoriantimeshadbegun,afterthepreviousnight’sleaktotheBBCthatNorthernRockhadhadtoturntothelenderoflastresort,theBankofEngland,foremergencyfunding.Asonedepositorqueuingoutsidethestrickenbank’sbranchinKingston-upon-ThamestoldajournalistfromtheFinancialTimes:‘I’mheretotakethelotout,becausethey’regoingunder,aren’tthey?’(Braith-waiteandTighe2007).Forafewdays,confidenceintheUK’sbankingsystemwavered,asdeposi-tors’concernsaboutotherbanksgrew,andasinternationalfinancialinstitu-tions,alarmedbythedramaticTVpictures,grewwaryoflendingtoBritishbanks.PanicwaseasedonlybyagovernmentguaranteetoNorthernRock’sdepositors—andtothoseofanyotherUKbankinsimilardifficulties—andbytheBankofEnglandreversingitsearlierpolicyandmakingfundsmorereadilyavailabletocash-strappedbanks.TherunonNorthernRockandthetemporarylossofconfidenceinUKbankingwerepartofaglobalfinancialcrisisthatbeganinthesummerof2007.Aseriousbutapparentlylimitedproblem—risingdefaultrateson‘sub-prime’mortgagesintheUnitedStates—hadspreadthroughthefinancialsystem.Amongstthevectorsofthecreditcrisiswasacollapseofconfidenceinthemarkets’fact-generationsystems.Aslossesanddowngradeshiteven 178/Conclusionthemosthighlyrated,AAA,mortgage-backedsecurities,theratingssystemitselfwascalledintoquestion.Asonemanagerofamoney-marketfund—previouslyasubstantialinvestorintheapparentlysafe‘commercialpaper’issuedbythecreditworld’sspecialpurposevehicles—putittoBloombergMar-ketsinearlyautumn:‘Intoday’smarket,youreallycan’ttrustanyratings’(Evans2007:46).Thebuyersofcommercialpaperarethe‘photoplankton’ofthemarkets,oneintervieweetoldme.Themetaphormixedscornforthosewhobasedtheirinvestmentdecisionssolelyonratingswithacknowledgementoftheircrucialimportancetofundingthesystem.Astheystoppedbuying,thespecialpurposevehiclestheyhadpreviouslysupportedturned,wheretheycould,totheirparentbanks.Thosebanksthemselveswereinmanycasesalreadycarryinglossesthatwerebothlargeandalarminglyindeterminate.Modernaccountingregulationforcesmuchofbanks’portfoliostobe‘markedtomarket’—revaluedasmarketpriceschange—butsecuritiesofmanykinds(bondsbackedbyNorthernRock’smortgages,forexample)eithercouldnotbesoldatall,orsoldonlyatdistressedprices.Banks’preparednesstolendmoneytoeachother—exceptperhapsfortheshortestpossibleperiod,overnight—evaporated.AsnotedinChapter4LIBOR,themostsolidoffinancialfacts,lostsomeofitsauthority,criticizedforexampleasameasureofinterestratesinwhatwasinsomecasesamarketinwhichonecouldnolongeractuallyborrow.‘TheLiborratesareabitofafiction’,saidonelargebank’streasurer(Tett2007).Thecreditcrisis—farfromoveratthetimeofwriting—is,amongstotherthings,acrisisoftheinfrastructureofthefinancialworld:notofitstech-nologicalinfrastructure,whereonlylimiteddifficultiesweremanifest,butofitscognitiveinfrastructure,ofitsfact-generationmechanisms.Itremainsunclearjusthowwidetheramificationsofthecollapseoffactwillbe.ManyAmericans—someestimateasmanyastwomillion,othersevenlargernumbers—willlosetheirhomes,unabletorepayloansarrangedforthembyintermediarieswhoknewtheythemselvescarriednoriskofdefault,becausemortgagesandthecreditrisktheyinvolvecouldbesoldon.Pensionsandmanyformsofinvestmenthavealreadysuffered.Theoverallsqueezeoncreditwilldampeneconomicactivity,buthowmuchandforhowlongis,atthetimeofwriting,unclear.Confidenceinfactswillnodoubtreturn,butnotinallfacts. Conclusion/179Fromtheviewpointofthisbook,thecreditcrisisillustratestheimpor-tanceofthekindofphenomenauponwhichthesocialstudiesoffinancefocuses.Ofcourse,thecrisisreflects‘big’phenomena:theglobalimbalancescausedbyaglutofsavingsincountriessuchasChinaandmassivebor-rowingandbalance-of-paymentsdeficitsincountriessuchastheUSAandUK;thederegulationoffinancialsystems;theneoliberalideasthatpromotedderegulation;andsoon.Yetequallycrucialwereapparently‘little’things.Attheheartoftheexpansionofcredit,forexample,wasaparticularsetofinvestmentvehicles—thecollateralizeddebtobligations(CDOs)mentionedinChapter2—thataresufficientlycomplicatedthattheirpropertiescanbegraspedonlywithmathematicalmodels,inparticularthe‘Gaussiancopula’modelstouchedoninthatchapter.WithoutthecapacitytomodelCDOs,theratingagencieswouldhavebeenhardpressedtoawardratingstoproductsbasedonthem,andwithoutratingsmostinvestorswouldnothaveboughtthem.Inparticular,CDOswereacrucial—perhapsthecrucial—sourceofdemandforsecuritiesbaseduponsub-primemortgages:centraltotheattrac-tivenessofCDOstoinvestorswasthataCDOcouldbeusedtopackagehigh-riskdebtintoformsthatcouldstillachievehighratings,whileofferinggreaterratesofreturnthansimplerproductssuchasbondswiththesameratings.Thecrisisshowed,inotherwords,thatscalesaren’tstable.Attheheartofanepochalevent—intheviewofmany,themostseriousfinancialcrisissincetheSecondWorldWar—were‘technical’matterssuchasmathematicalmodelsandcreditratings.Ifthisbookhasasingledominantthemetoconvey,itisthattechnicalmattersofthiskindarenot‘meredetails’thatcansafelybesetasidebysocialscientistslookingfor‘thebigpicture’.Yes,globalimbalancesandtheinfluenceoftheneoliberalideasmatter,butsotoodoesthegamutofphenomenaofthekinddiscussedhere,frombodilycapacitiessuchasbroker’seartothealgorithmthatgeneratesLIBOR.AssuggestedinChapter1,thesocialstudiesoffinanceisamaterialsociol-ogyofmarkets,onethatemphasizestheirphysicality,theircorporeality,andtheirtechnicality.Iam,ofcourse,painfullyawareofthebook’slimitationsintheserespects.AccessdifficultiesmeanthatthedatauponwhichIdrawarepredominantlyinterviewbasedratherthandirectlyobservational:gettingpermissiontoconductethnographicobservationinfinancialmarketscanbeextraordinarilyhard.Thisinturnmeansthattheprecedingchaptershave 180/ConclusiongivenlessemphasistophysicalityandcorporealitythanIwouldhaveliked:thoseaspectsofmarketsaremoreeasilydocumentedbywitnessingobjects,technologicalsystems,andhumanbodiesininteractionthanbylisteningtopeopletalkaboutthem.Evenontheparticulartopicsdiscussedhere,thereisinconsequencemuchmoretobediscoveredintheserespects,andmanyotheraspectsofmarketsremainterraincognitafromtheviewpointofmaterialsociology.Itisalsoappropriatetobemodestabouttheacademicambitionsofthesocialstudiesoffinance.Itisalltooeasyfortheproponentsofanewapproachtooveremphasizeitsnovelty,ortoclaimthatitsupplantsallpre-viousapproaches.Neitheristhecasehere.Take,forexample,therelationshipbetweensocialstudiesoffinanceandmainstreameconomicsociology.Thereismuchworkinthelatterfieldthatisfullycompatiblewithasocialstudiesoffinanceapproach,forexampletheinsightfulanalysisbyCarruthersandStinchcombe,alreadytoucheduponinChapter2,ofhow‘liquiditypresumesassetsthatareknowablebyalargegroupofpotentialbuyersandsellers’andinvolvescreating‘generalizedimpersonalknowledgeoutofidiosyncraticper-sonalknowledge’(1999:356).Thecollapseofliquidityfollowingthelossofconfidenceinfinancialfactsinthecreditcrisisisaperfectillustrationoftheirargument.Similarly,thesocialstudiesoffinancecomplements,ratherthandisplaces,whatisperhapsthecentraltraditionofrecenteconomicsociology,whichfocusesontheconsequencesforeconomicactionoftheembeddingofactorsinnetworksofinterpersonalconnections,andisalineofresearchinspiredaboveallbytheworkofMarkGranovetter(see,forexample,Granovetter1973;1985;1992).Itisnotthathumanbeingsareembeddedinsystemsoftechnolo-gies,cognitiveframeworks,simplifyingconcepts,andcalculativemechanismsratherthaninnetworksofpersonalconnections,butthattheyareembedded,simultaneouslyandinextricably,inboth.Taketraders,forexample.TheresearchreportedhereandtheearlierworkdescribedinMacKenzie(2006)hasinvolvedsomefiftyinterviewswithtraders.Eventhoseinvolvedinthemost‘technical’formsoftrading,suchasoptionsorcreditderivatives,reportedtheneedtopayattentionnotjusttocalculativetoolssuchasmodelsbutalsotowhowasdoingwhatandwhy,whichisinformationtypicallyacquiredthroughnetworksofpersonalconnec-tions.Furthermore,modelsmatterinpartsimplybecauseothersusethem.Tradersincreditderivatives,forexample,reportedthatitwassometimes Conclusion/181productivetorequestpricequotationsforsetsofderivativesfromseveraldifferentinvestmentbanktradingdesks,workoutfrompatternsofquotedpricesthecharacteristicsofthemodelsthattradingdesksmustbeusing,andarbitragethedifferencesbetweenthemodelsusedbydifferentbanks,ineffectbuyingcreditprotectioncheaplyfromonebankandsellingitexpensivelytoanother.Theexampleofhowtradersusemodelsillustratesageneralpoint:the‘technical’andthe‘social’arenottwoseparatespheres,buttwosidesofthesamecoin,asalongtraditioninthesocialstudiesofscienceandtechnologyhasemphasized.1Anymarket—indeedanysociety(see,forexample,Latour2005)—isasociotechnicalconstruction.Acentralroleofthesocialstudiesoffinanceandsimilar‘materialsociology’approachestomarketsisthustoaddtothewell-honedsetoftoolsforanalysingthemoredirectly‘social’aspectsofmarkets—suchastheworkpioneeredbyGranovetter,thedifferentformofnetworkanalysisdevelopedbyHarrisonWhite(1981;2001),the‘political-cultural’approachofFligstein(2001),andsoon—asetoftoolsformakingsenseoftheirmore‘technical’aspects.Becausethe‘social’and‘technical’areinextricablylinkedinmarketconstruction,thetwosetsoftoolswillulti-matelyneedtobeintegratedfully:achallengingbutimportantacademictask.Theambitionsofthesocialstudiesoffinance,however,shouldnotbeacademicalone.Therehasbeeninrecentyearswithinsociologyagreatlyenhancedattentiontothetopicof‘publicsociology’(formsofsociologythatgobeyondacademia,andengagewithmultiplepublicsinmultipleways),attentionsparkedaboveallbyMichaelBurawoy’sPresidentialAddresstotheAmericanSociologicalAssociation(Burawoy2005).WhilethespecificsofwhatBurawoycallsforarecontentious—amongstmanycritiquesisHolmwood(2007),andIofferacriticalcommentofmyownbelow—hisimpulseissurelycorrect.Whendealingwithtopics,suchasmarkets,thathaveenor-mousimplicationsforpeople’slives,researchersshouldsurelyaimnotjustathigh-qualityacademicresearch,butalsoshouldseektoreachoutbeyondacademiatowiderpublics.Thisisanissueforallsocialscientists,notjustsociologists:atstakeisnotjustpublicsociology,butpublicsocialsciencemoregenerally.Whatsortofpublicsocialsciencemightthesocialstudiesoffinancefoster?Therearemanypossibilities,butoneparticularlyattractiveoneisapoten-tialbroadeningoftheformsofpoliticalengagementwithmarkets.Alltoo 182/Conclusionfrequently,existingpoliticaldebatetreats‘themarket’asasingularentitywithinherentcharacteristics,whichshouldeitherbepromoted(asthepoliticalrightnormallysuggests)oropposed(astheleftoftenprefers).Forexample,forallthestrengthsofhisadvocacyofpublicsociology,Burawoy’swritingtendstoremainlodgedwithinthisdichotomy,treating‘themarket’assomethingwithafixedessenceagainstwhich‘society’needsprotection(see,especially,Burawoy2007).Attheheartofthesocialstudiesoffinanceistheconvictionthattreating‘themarket’asasingularentityismistaken:thebest-knownbookbyMichelCallon,onethathasbeendeeplyinfluentialinthefield,istitledTheLawsoftheMarkets(Callon1998).Ifthisconvictioniscorrect,andmultipleformsofmar-ketswithdiversecharacteristicsandsubstantiallydifferentconsequencesarepossible,thenpoliticsdividedbetween‘pro-market’and‘anti-market’(bothinthesingular)iswhollyimpoverished,andevena‘thirdway’thatseekssimplytopositionitselfbetweenthetwoisinsufficient.Ofthemanymarketsthatarepossible,whichmarketswehavematters,andthatisaquestionnotsimplyoftheiroverallcharacteristicsbutofthedetailsoftheirdesign,thetechnologicalinfrastructuresthatsupportthem,andthewayeconomicagentsinthemareconstructed:thesystematicformsofknowledgethoseagentsdeploy;thephenomenatowhichtheypayattentionandtowhichtheydonot;thewaysinwhichcomplexitiesaremadesimpleenoughforeconomicagentstograsp;andsoon.Theneedforaformofpoliticalengagementwithmarketsthattacklessuchphenomenaisperhapsclearestinemissionsmarkets.AsChapter7emphasizes,suchmarketscertainlydonothavefixedcharacteristics.Awell-designed,well-functioningemissionsmarketcanbeaneffectiveandeconomicaltoolofabatement,astheUSSO2marketlargelywas.Ontheotherhand,apoorlydesignedemissionsmarketcanbeuselessorworse.Whichoutcomehappensdependsnotjustontheoveralladvantagesanddisadvantagesofmarketsaspolicytools,butonthedetailsofmarketdesign,forexampleoftheallocationmechanismsdiscussedinChapter7.AdetailedpoliticsofmarketdesignisalreadybeingpractisedbyleadingenvironmentalNGOs,whichhavegonebeyondthe‘pro-market’/‘anti-market’dichotomytoseektoinfluencethekey,detailedmattersthatmakeemissionsmarketseffectiveorotherwise.Butthatisunusual.Thereare,forexample,onlylimitedexamplesofsimilarengagementwiththedesignoffinancialmar-kets.Suchmarketsarejustas‘political’asemissionsmarkets—theyoperate Conclusion/183withinframeworksstilllargelysetbygovernmentsandregulatorybodies,andchannelhugeamountsofresourceswithmajorconsequencesforpat-ternsofeconomicgrowth,employment,wealth,andpoverty—butthecru-cialspecificsthatshapefinancialmarketsseldombecomemattersforpublicdebate.MuchofwhatgoesonintheprocessesshapingmarketsiswhatBeck(1996)calls‘subpolitics’,involvingconsequentialdecisionmakingtakingplaceout-sidetheformalpoliticalsystemandconcerningphenomenanottraditionallythoughtofaspolitical.Provisionforoldage,forexample,isatraditional,recognizablypoliticaltopic.Theaccountingrulesthatgovernthefinancialreportingofpensionprovision,incontrast,aretreatedasatechnicalmattertobedecidedbyspecialists.Yetthoserulesdeeplyaffectwhatkindofpensionprovisionprevails.Afinancialinnovationthathasastrongcaseforbeingthetwentiethcen-tury’sfinestisnotanysophisticatedderivativebutthe‘finalsalary’,defined-benefit,pensionscheme,inwhichitispredominantlyemployerswhobearthefinancialrisksofprovisionforoldage.(Intheothermaintypeofscheme,definedcontribution,itistheemployeewhobearstheriskofmarketfluctu-ations.)Inrecentyears,defined-benefitschemeshavedeclinedsharplyintheprivatesectorintheUK,andamongstthecauses,arguesAvrahampour(2007),wastheintroductionofUKFinancialReportingStandard17(andnowInterna-tionalAccountingStandard19),governingemployees’pensionbenefits.Thesestandardsmakeafirm’spensionprovisionfarmorecentraltoitsfinancialreportingthanpreviously,causingthefinancialrisksinsuchprovisiontobecomefarmoresalienttoinvestors.Inthiscontext,defined-contributionschemestendtobefarmoreattractivetofirmsthandefined-benefitschemes,preciselybecausetheformershifttheriskawayfromthefirmanditsinvestors.Evenapparentlyverydetailedtechnicalitiescanbedeeplyconsequential.Forexample,anemployer’spensionliabilitieslargelylieinthefuture,soadiscountratemustbechosentoworkouttheirpresentvalueforreportingpurposes.WhetheroneusestheyieldofanAA-ratedcorporatebondasthediscountrate(asiscurrentpractice)or(assomearenowadvocating)usestheloweryieldonagovernmentbondhasamajoreffect;thelatteroptionmakesliabilitiesseemmuchlarger.Anotherexampleoffinancialreportingmatteringintermsofwiderout-comesconcernswhethermultinationalcompaniesinsectorssuchasoilandminingcansimplyproduceaccountsfortheiraggregateglobalactivitiesor 184/Conclusionhavetoreportontheiroperationscountrybycountry.In2006agroupofeightyNGOspromptedtheInternationalAccountingStandardsBoardtobeginconsideringrequiringcountry-by-countryreporting;thecoalition’saimwastoforcecorporationstoprovidefiguresonmatterssuchastheirpaymentstoeachspecificgovernment(Jopson2006).Ofcourse,the‘finitist’issuesout-linedinChapter6wouldbearupontheeffectivenesswithwhichsuchreport-ingcouldbeturnedintoatooltoincreasetransparencyanddetercorruption,buttheinitiativeisanexcellentexampleofthe‘nutsandbolts’politicsofmarketsthatthisbookadvocates.Consider,too,theregulationofbanking.TheBritishgovernment’srescueofNorthernRockmadeclearthatithasbecomeclosetoinconceivableforadeveloped-countrygovernmenttoallowamajorbanktocollapseinsuchawaythatlarge-scalelosseswerecausedtosubstantialnumbersofthevotingpublic.Inconsequence,thereisanasymmetry(long-standing,butcurrentlyespeciallyprominent)intheconstructionofabankasaneconomicagent:therewardsoftakingrisksinbankingarecapturedprivately,whilethemostextremepotentiallossesareimplicitlybornebytaxpayers.Thisfirstasym-metryisfurtherreinforcedbyasecond:theprevailingsystemofrewardingbanks’employees,especiallytraders.Theirsalariesandbonusesincrease,oftendramatically,iftheygenerateprofits,butlossesarenotpenalizedbyneg-ativesalariesorbynegativebonuses.Becauseofthefirstasymmetry—thesecond,thoughalsoconsequential,hasuntilveryrecentlyfeaturedmuchlessofteninpolicydebate—itisalmostuniversallyacknowledgedthatbankingrequirescloseregulation.Yetthespecificsofsuchregulation—themodelsemployedtocalculatebanks’‘valueatrisk’;whethersuchmodelsandthemarkingtomarketofportfoliosare‘pro-cyclical’(thatis,encouragetoomuchrisktakinginboomsandfeeddownwardspiralsinbadtimes:seeGood-hartandPersaud2008);theroleofcreditratingsindeterminingthecapi-talreservesbanksmusthold;waystoreduceincentivestoadoptstrategiesthatpromiseenhancedreturns(andthusbonuses)mostofthetime,butruntheriskofoccasionalcatastrophiclosses;andsoon—are‘subpolitical’ratherthan‘political’matters,seldom,forexample,thetopicofdebatebylegislators.Ofcourse,matterssuchasaccountingstandardsandbankingregulationsaretechnical:thatisthemainbarriertomakingthemobjectsofwiderdebate.Animportantformofpublicsocialsciencethatthesocialstudiesoffinance Conclusion/185couldpractiseisthereforetohighlighttheimportanceofsuchmattersandtoexplaintheminwaysthatwideraudiencescangrasp.Howacademicsinthefieldwriteisthusaffected,butalsowheretheywrite.Publishinginacademicjournalswithlimitedcirculationisnotenough,andevenacademicbookshavetheirlimitations.Anarticleforamoregeneralistmagazine(see,forexample,MacKenzie2007b)canreachareadershipatleasttentimesaslarge;anews-paperarticlecanboostthepotentialaudienceahundredfold.Furthermore,thenecessarypublicsociology,asBurawoy(2005)emphasizes,goesbeyondsimplywritingforprintorInternetoutletswithwidereadership.Weacad-emicshaveatleastasmuchtolearnfromthosewhoareseekingtoalterthecharacteristicsofmarketsastheyfromus.Especiallyinspheressuchasemissionsmarketswherethespecificsofdesignarealreadythesubjectofwiderdiscussionandpoliticalaction,‘publicsociologybringssociologyintoaconversationwithpublics,understoodaspeoplewhoarethemselvesinvolvedinconversation’(Burawoy2005:7).Thematerialsociologyofmarketsexemplifiedinthesocialstudiesoffinancehasthusthepotentialnotjusttobroadentheacademicstudyofmarketsbuttohelpwiden—andinturntobeenrichedby—publicdiscussionandactioninrelationtothem.Marketsareofcoursecentraltomodernlife,andareheretostay:acomprehensivemoveawayfrommarketformsofeconomicprovisionisclosetoinconceivable.Yetmostpeople’sdirectexper-ienceofmarketsislimited.Thecreditcrisis,forexample,suddenlythrewintothespotlightentitiessuchasSIVs(structuredinvestmentvehicles),struc-turedproductssuchasCDOs(collateralizeddebtobligations),andfirmssuchasmonolines(insurersofbonds),whichfewoutsidershadpreviouslyevenheardof.Financialmarketsarelitteredwithwhatthesocialstudiesofsci-enceandtechnologytendstocall‘blackboxes’ofthiskind.AsnotedinChapter2,‘blackboxes’aredevices,practices,regulations,organizations,mod-els,andsoon,theinternalstructureofwhichcanbedisregardedorwhichareopaquetooutsiders,oftenbecausetheircontentsareregardedas‘technical’(seeMacKenzie2005).Researchthatopenstheblackboxesoffinancecanthuscontributetopublicaswellastoacademiclife.Thisbookhassought,inmodestandpreliminaryways,toopensomeblackboxes:hedgefunds(oftenregardedasparticularlyopaqueorganizations),derivativesmarkets(oftendeeplytechnicalintheiroperations),arbitrage(acrucialmarketoperationbutseldomexaminedin 186/Conclusionitssociotechnicalspecifics),accounting(seldomstudiedbynon-accountants),andemissionsmarkets(whichintheircomplicationcanbedeeplydauntingwhenfirstencountered).Ihopethatinsodoingithasshownthatblackboxescanbeopened,andthatgainingabetterunderstandingoftheircontentsisbothinterestingintellectuallyand,atleastinsomecases,potentiallyconse-quentialforreal-worldaction.Ifthebookhassucceededevenpartiallyindoingthis,thenitsauthor’shopesforitwillhavebeenfulfilled. GLOSSARYThefinancialmarkettermsinthisglossaryoftenhavearangeofmeanings,andIhavegivenonlythemeaningrelevanttothisbook.Forwidermeanings,seeMolesandTerry(1999),whichhasbeenmymainsource.1arbitrageInfinancialmarketparlance,arbitrageistradingthatseekstomakelow-riskprofitsfrompricediscrepancies,forexamplebetweenthepricesofsimilarassets.Infinancetheory,arbitrageistradingthatgeneratesrisklessprofitwithnonetcapitaloutlay.bondAtradeableformofdebt.Bondsnormallycommittheissuer(mostcom-monly,agovernmentorcorporation)torepayafixedsum(theprincipal)onagivendateandtomakeperiodicinterestpayments(coupons)offixedamountsuntilthatdate.Theownerofabondcansellittoanotherinvestor.collateralizeddebtobligation(CDO)Aninvestmentvehiclethatacquirespoolsofdebtinstrumentssuchasbonds,loans,securitiesbackedbymortgages,etc.,andsellsontoinvestors,repackagedinstructuredform,thecashflowfromtheseinstrumentsbutalsotheriskthattheirissuerswilldefault.couponSeebond.derivativeAcontractorsecurity(suchasaforward,future,option,orswap),thevalueofwhichdependsuponthepriceofanother‘underlying’asset,onthelevelofanindex,exchangerate,orinterestrate,oronotherparameterssuchastheprobabilityofabondissuerdefaulting.discountTocalculatetheamountbywhichfuturecosts,payments,orotherbene-fitsmustbereducedtogivetheirpresentvalue.forwardAcontractinwhichonepartyundertakestobuy,andtheotherpartytosell,asetquantityofanassetofaparticulartypeatasetpriceatagivenfuturetime.Ifthecontractisstandardizedandtradedonanorganizedexchange,itisreferredtoasafuture.futureAstandardizedcontracttradedonanorganizedexchangeinwhichonepartyundertakestobuy,andtheothertosell,asetquantityofanassetofaparticular 188/Glossarytypeatasetpriceatagivenpointintimeinthefuture.Thetermisalsousedforcontractsthatareeconomicallyclosetoequivalenttosuchfuturepurchases/salesbutwhicharesettledbycashpayments.hedgefundAspecialcategoryofinvestmentvehicle,oftenregisteredoffshoreand/orfallingwithinthe‘privatefunds’exemptionfromtheUSInvestmentCompanyActof1940,whichtypicallypermitsonlyverylargeinvestmentsand/orastrictlylimitednumberofinvestorsandisbannedfromadvertising,butisexemptfrommanyregulatoryrequirementsandisfreetoadoptstrategies(suchasshortsellingandusingborrowedfundstoenhancereturns)thatmanyothercategoriesofinvestorareprohibitedfromusing.impliedvolatilityThevolatilityofastockorindexconsistentwiththepriceofoptionsonthestockorindex.LIBORLondonInterbankOfferedRate,theaveragerateofinterestatwhichapanelofmajorbanksreportotherbanksasbeingpreparedtolendthemfundsinaparticularcurrencyforaparticularperiod.SeeChapter4forhowtheaverageiscalculated.longpositionAportfolioofanassetand/orderivativeofthatassetthatwillriseinvalueifthepriceoftheassetrises.Cf.shortposition.openoutcryTradingbyvoiceand/orhandsignalsthattakesplacewithinafixedarenaandisaudible/visibletothoseinthearena.optionAcontract,thepurchaserofwhichgainstheright,butisnotobliged,tobuy(‘call’)ortosell(‘put’)anassetatagivenprice(thestrikepriceorexerciseprice)on,orupto,agivenfuturedate(theexpiration).Theseller(or‘writer’)oftheoptionisobligedtofulfilhisorherpartofthecontractifsodemanded.pitThephysicallocation(oftensteppedaroundthesides)ofopenoutcrytrading.shareSeestock.shortpositionAportfolioofanassetand/orderivativeofthatassetthatwillriseinvalueifthepriceoftheassetfalls.Ashortpositioncan,forexample,beconstructedbyshortsellinganasset.Cf.longposition.shortsellingAprocessinwhichatradersellsasecurityheorshedoesnotyetown,orownsonlytemporarily.Shortsellingisoftenaccomplishedbyfindinganownerofthesecuritywhoisprepared,forafee,to‘lend’ittothetrader:inotherwords,totransferownershipofittothetrader,whointurnundertakestoreplaceit.Thetraderwhoshortsellsmay,forexample,expectthatthepriceofthesecuritywillhavefallenbythetimeheorshehastoreplaceit,soheorshecankeepthedifferenceinprice(minusthefee). Glossary/189stock(US)orshare(UK)Asecuritythatconferspart-ownershipofacorporation.swapAcontracttoexchangetwoincomestreams,forexamplefixed-rateandfloating-rateinterestonthesamenotionalprincipalsum.volatilityTheextentofthefluctuationsofthepriceofanasset,conventionallymeasuredbytheannualizedstandarddeviationofcontinuouslycompoundedreturnsontheasset.yieldTheyieldofabondistherateofreturnitoffersoveritslifetimeatitscurrentmarketprice,normallymeasuredbyfindingtherateofinterestatwhichabond’scouponsandprincipalhavetobediscountedsothattheirtotalpresentvalueisthebond’scurrentprice. NOTESChapter11.Unfortunately,precisionisimpossiblebecausetherelevantdatasource,theBankforInternationalSettlements(,accessed13June2007),doesnotspecifytheinterestrateunderlyinginterest-ratederivatives.Myroughestimateof$170trillionisbaseduponthetotalnotionalamountsofinterest-rateswapsoutstanding($169,106billion)attheendofDecember2005.(Inaninterest-rateswap,thenotionalamountisthesumonwhichonepartyagreestopayafixedrateofinterestwhilereceivingavariablerate.Thelatterismostusually,butnotalways,aspecifiedLIBOR.)Myestimatethusassumesthatinterest-rateswapsbaseduponratesotherthanLIBORareatleastequalledbyotherkindsofinterest-ratederivatives(suchas‘forwardrateagreements’,andfuturesandoptionsoninterestrates)thatarebaseduponLIBOR.(Thenotionalamountsofinterest-ratederivativesotherthanswapstotalled$95trillion.)Thatassumptionisplausible,butIhavemademyestimateconservativebyignoringtheriseinthetotalnotionalamountsofinterest-rateswapsbetweenendDecemberandthetimeofmyobservationsinFebruary2006(bytheendofJune,thetotalhadrisento$207trillion).2.See,forexample,theclassicworkofMaxWeber,recentlytranslatedasWeber(2000a;2000b),alsoRose(1951;1966),Smith(1981),AdlerandAdler(1984),andBaker(1984a;1984b).3.Foracollectionofarticlescoveringsocialstudiesoffinanceinboththebroadandthenarrowersensediscussedinthetext,seeKnorrCetinaandPreda(2005);anotherusefulcollectionisKalthoff,Rottenburg,andWagener(2000).Asidefromthechaptersinthosevolumes,othernoteworthyrecentcontributionstothesocialstudiesoffinanceincludeAbolafia(1996;1998),Arnoldi(2004),BeunzaandStark(2003;2004;2005),Caliskan(2005),Clark(2000),Godechot(2000;2001;2004),deGoede(2005),Hassoun(2000),Hertz(1998),HolzerandMillo(2005),Izquierdo(1998;2001),KnorrCetinaandBruegger(2000;2002a;2002b),Lépinay(2004;2007),Levin(2001),LiPumaandLee(2004;2005),McDowell(1997),Maurer(2001;2002;2005),Millo(2003),Millo,Muniesa,Panourgias,andScott(2005),Miyazaki(2003;2005),Muniesa(2003;2005),Podolny(1993;2001);Preda(2001a;2001b;2004a;2004b;2006),PrykeandAllen(2000),Riles(2004),Thrift(1994), Notes/191Tickell(1998;2000),Uzzi(1999),Widick(2003),Zaloom(2003;2004;2006),Zorn(2004),andZuckerman(1999;2004).4.Amongstthefield’sleadingscholarlybodiesaretheSocietyforSocialStudiesofScience()andtheEuropeanAssociationfortheStudyofScienceandTechnology().AmongstitsmostprominentjournalsareSocialStudiesofScienceandScience,Technology,andHumanValues.5.Onmaterialitymoregenerally,seee.g.Miller(2005).6.Bakerdidnotnametheexchangehestudied,butwhenIbegantoresearchoptionstrading(MacKenzieandMillo2003;MacKenzie2006),itquicklybecameclearthathisfieldworksitemustbetheChicagoBoardOptionsExchange.7.Foranothersetofpreceptsforsocialstudiesoffinance,withadifferentemphasis,seePreda(2001c).Chapter21.ShorterOxfordEnglishDictionary,5thedition.2.Onthehistoricallyspecificnatureofthemodern‘experimentallife’,seeShapinandSchaffer(1985).3.AccordingtoKnorrCetina(2005:45),therewereover300,000Reutersterminalsand150,000Bloombergterminalsworldwidein2001,andthenumbers(espe-ciallyofBloombergterminals)willhaverisenconsiderablysincethen.4.ChicagoMercantileExchangetraderspointedtothecloseregulationofthemarketbytheCommodityFuturesTradingCommission,andsuggestedthat‘complaintsarearisingnowlargelybecausemilkpricesarerelativelylow’(Grant2006).5.SeeMacKenzie(2006),and,foracontraryview,MooreandJuh(2006)andMixon(2006).6.Amarketthatisofparticularinterestfromtheviewpointofchartismisforeignexchange,whereverylargenumbersoftradersarechartists,andeconometricanalysissuggeststhatchartisttechniquesdoindeedsuccessfullypredictshort-termratemovements(seetheliteraturecitedbyOsler2003:1791).Thustwoprominentchartistbeliefsarethat‘trendstendtoreversecourseatpredictablesupportandresistancelevels’andthat‘trendstendtobeunusuallyrapidafterratescrosssuchlevels’(Osler2003:1791).Oslersuggeststhattheclusteringof‘take-profit’ordersatround-numberrates,andof‘stop-loss’ordersimmedi-atelybeyondthoserates,mayexplaintheapparentempiricalvalidityofthesetwobeliefs.DonaldsonandKim(1993)similarlydemonstrateanomalousbehav-iouroftheDowJonesIndustrialAverageinthevicinityofround-numberindexlevels. 192/Notes7.‘Basecorrelation’iswellexplainedbyitsdevelopersinMcGinty,Beinstein,Ahluwalia,andWatts(2004).8.TheworkofHutchinsbelongsprimarilywithincognitivescienceratherthanscienceandtechnologystudies,butitsrelationtotheworkofBrunoLatour(especiallyLatour1986)isclearandmadeexplicitbyHutchins(1995a:132).9.Theso-called‘largepool’approximationdoesleadtoanalyticalsolutions,buttheapproximationseemstoberegardedastooradical,andthemodelsthatareusedinpracticeareatbest‘semi-analytical’(thatis,theyinvolvetechniquessuchasnumericalintegrationandfastFouriertransforms).10.,accessed4September2006.11.Onmetrology,see,forexample,Latour(1987:247–57),Schaffer(1992),andAlder(1997).Thesignificanceofmetrologyformarketsisnoted,forinstance,byCallon(1998:23)andLevinandEspeland(2002).12.Theconcentrationofnitrogenoxidesisalsomeasured,asis‘diluentgas’(oxygenorcarbondioxide)andopacity(‘thepercentageoflightthatcanbeseenthroughthefluegas’):,accessed4September2006.13.See,accessed4September2006.14.See,forexample,EPA(1997).15.See,forexample,Greimas(1987).16.Thefewsocial-sciencestudiesofthecontentofwhatbook-keepersdothatIhavebeenabletotracecomeexclusivelyfromethnomethodologicalandsimilarethnographicresearchonworkandonitsautomation:see,espe-cially,Anderson,Hughes,andSharrock(1989:123–37),andalso,forexample,Suchman(1983)andButtonandHarper(1993).17.Forexamples,seeAbolafia(1996).18.See,forexample,Ledyard(1995:121),Baker(1984a;1984b),Pirrong(1996),andMacKenzieandMillo(2003).19.SeeLevy(1993).Unfortunately,theextenttowhichbehaviourwasactuallychangedby‘tote-board’diplomacyisunclear,sincetherelevantprotocolsmay‘merely[have]codifiedwhatmostofthepartieswereplanningtodoanyway’(Barrett2003:10).20.There’sasenseinwhichthefailurewasindeedonlyapparent.TheUnitedKingdomrefusedtoadoptthe1985HelsinkiProtocol,whichrequiredacutof30%insulphuremissions,butthenUKemissionsactuallyfellbytherequisiteamount(Barrett2003:9).Thiswaslargelyaby-productoftheshiftinelectric-itygenerationtowardsnaturalgas,butLevy(1993:124)suggeststhatpoliticalpressuremayhavebeeninvolvedtoo. Notes/19321.EthnomethodologyasdevelopedbyGarfinkel(especiallyinGarfinkel1967)isanothermajorsourceoffinitism,andLatour(2005:54)describesactor-networktheoryas‘halfGarfinkelandhalfGreimas’.Ithinkhehasgottheproportionswrong,atleastasfarasGarfinkel’sfinitismisconcerned:itdoesnotseemastronginfluenceonactor-networktheory.22.Aleasecanbeineffectawayofborrowingmoneytobuyanasset,andregulatorshavebeenconcernedthatsuchleases—‘financeleases’—shouldappearonacorporation’sbalancesheetandsoenterintocalculationsoftheextentofacor-poration’sborrowingandthelevelofreturnonitsassets.SeeIAS[InternationalAccountingStandard]17,‘Leases’(IASB2005:887–914).23.Onmicroworlds,see,forexample,Collins(1990).24.Inthesummerof2005thepawn-promotionrulewaschangedwiththeappar-entaimofblockingtheinterpretationdiscussedinthetext.Theword‘new’wasadded,soitnowreads‘exchanged...foranewqueen,rook...’(FIDE2005b:rule3.7.e,emphasisadded).However,thesolutiontothepuzzlemightstillundersomecircumstancesbearguedtobeallowable.Imaginethegameisbeingplayedwithanoldwoodenset,butthatsomepieceshavebeenlostandthewhiterookcurrentlyontheboardisamodernplasticreplacement.Isthatnot‘anew...rook’?25.SeeMacKenzie(2006)andMacKenzie,Muniesa,andSiu(2007).26.ThusMirowskiandNik-Khah(2007)useNik-Khah’srichcasestudyoftheroleofeconomistsintheFederalCommunicationCommission’sspectrumauctionsintheUSAasarefutationoftheideaoftheperformativityofeconomics,whileIseeitas(strong)evidenceagainstlinearviewsofperformativity.27.FordissentfromEdgerton’sclaimthatthelinearmodelissimplya‘strawman’,seeHounshell(2004).PaulFormanhasfiercelyattackedcriticsofalinearview,especiallywithinthehistoryoftechnology,forexamplesuggestingthatthey(includingthisauthor)areguiltyof‘animositytowardscience’(2007:62).Hedoesnot,however,putforwardevidencethatalinearviewiscorrect,andindeeddoesnotappeartobelievethatitis.28.Forotherpossibleinnocentexplanations,seeIngebretsen(2002:136–7).29.Theissuewasthe‘insidespread’—thedifferencebetweenhighestbroker-dealerbidandlowestbroker-dealerofferintheentiremarket—notthespreadquotedbyanyindividualdealer(DepartmentofJustice1996:9–10).Chapter31.Indeed,thischapterandthearticle(HardieandMacKenzie2007)onwhichitbuildscontainwhatis,toourknowledge,thefirststudyofahedgefundthatincludesdirectobservation(albeitbriefobservation)ofitsoperations(fora 194/Notesprevioussociological—butretrospective—studyofahedgefund,seeMacKenzie2003).Therehave,however,beenanumberofsociologicaloranthropologicalobservationalstudiesofotherkindsofactorandactioninfinancialmarketsthatareparticularlypertinenttoouranalysis:inparticular,KnorrCetinaandBrueg-ger(2002a)andBeunzaandStark(2004);thelatterisdrawnoninChapter5.Seealso,forexample,Heath,Jirotka,Luff,andHindmarsh(1993),Abolafia(1996),andZaloom(2003;2006).2.InvestmentCompanyAct,section3(especiallyparagraphc.1)andsection12,paragrapha.Thetextoftheactisavailableat,accessed11May2005.3.DatafromInternationalFinancialServicesLondon,,accessed23May2005.4.Data(forend2003)from,accessed23May2005.5.Incertainmarkets—notablythatforsmall-capitalizationstocks—behaviourinDecemberandJanuaryhasunusualaspectstodowithmatterssuchastheendoftheUStaxyear(Reinganum1983),butasfaraswecouldtelltherewasnosucheffectonwhatwewereobserving,thoughitispossiblethataperiodofobservationrightatthestartofthecalendaryearandafterthemarketswereclosedforaholidaymayhavebeenbusierthannormal.6.Fordiscussionsontherelativeinfluenceofendogenousandexogenousfactorsonemergingmarketbondspreads,seeforexampleManzocchi(2001)andEichengreenandMody(2000).7.Onthecontroversy,seeforexampleCollinsandYearley(1992).8.WeowethiswayofframingthepointtoFabianMuniesa.Chapter41.See,forexample,Tufano(2003)andBlack(1986).2.See,especially,Tickell(1998;2000),PrykeandAllen(2000),Maurer(2001;2002),LiPumaandLee(2004;2005),andArnoldi(2004).3.Mostcomputersystemsemploybothfast‘mainmemory’(thecontentsofwhichprogramscanaccessandmodify),whichintheearlyyearsofcomputingwasexpensiveandlimitedinitscapacity,and‘secondarystorage’,whichisslower,notdirectlyaccessible,butlargercapacity.Inthelate1950sandearly1960s,computerscientistslearnedhowtodesignoperatingsystemsthatauto-maticallytransferdatabetweenthetwoinsuchawayastofreeprogramsfromthelimitedphysicalcapacityofmainmemorybygivingthemaccesstoan‘addressspace’(‘virtualmemory’)thatismuchlarger. Notes/1954.TherecentgrowthofpropertyderivativesisthesubjectofongoingresearchbySusanSmithofDurhamUniversity.5.See,e.g.,SandorandSosin(1983:260–7).6.USCourtofAppeals,FederalCircuit,149F.3d1368.7.See,accessed4December2006.8.ItisalsoworthnotingthattheextenttowhichChicago’scompetitiveethostranslatedintotheactualityoffiercecompetitionwasinfactvariable,asbeauti-fullydemonstratedbyBaker(1984a;1984b).9.Onweatherderivatives,seePryke(forthcoming).Thepotentialdemandforlongevityderivatives—stilllargelyintheplanningstage—arisesfromthedesireofpensionfundstohedgetheriskthattheirmembersmaylivelongerthananticipated.10.LeoMelamed,electronicmailmessagetoauthor,13January2006.11.See,e.g.,.Chapter51.Apuristshouldreplacetheword‘arbitrage’inwhatfollowswith‘relative-valuetrading’.2.Threeofthefifty-onearerepeatinterviewsoftradersinterviewedinMacKenzie’soriginalstudy.3.IowethispointtoJuanPabloPardo-Guerra.Chapter61.Onthehistoryofinterchangeableparts,seeAlder(1997).2.Iowetheideaofthislist,andmanyitemsonit,toacommentbyMikePoweronanearlierversionofthischapter.3.Ananalogousissuearisesinthemeasurementofliabilities,whichforreasonsofspaceIignore.4.See,forexample,Hines(1988:253).5.Onthehistoryofratioanalysis,seeMillerandPower(1995).6.Again,seeHines(1988).7.ThelatterisofcourseMerton’sfamousexampleofself-fulfillingprophecy:seeMerton(1948).8.IAS39canbefoundinInternationalAccountingStandardsBoard(2005:1657–946).9.‘[A]man’sincome[is]themaximumvaluewhichhecanconsumeduringaweek,andstillexpecttobeaswelloffattheendoftheweekashewasat 196/Notesthebeginning’(Hicks1946:172).Thedifficultyofthisdefinitionliesinmakingprecisewhatismeantby‘aswelloff’.AsHickspointedout,thatleadsintoissuessuchasfutureinterestrates,futureprices,anddepreciation.Ontheepisode,seeHopwoodandBromwich(1984).10.,accessed5July2008.Forexample,Standard133(‘AccountingforDerivativeInstrumentsandHedgingActivities’)stretchesover165pages.11.Intheoriginal,thissentenceisinbold,whichishowtheInternationalAccount-ingStandardsBoardsignalsaprecept’sstatusasaprinciple.12.OC[OpticalCarrier]192isastandarddatarate(9,953,280kbit/s)forhigh-speed,high-capacitysynchronizeddatatransmissionviaopticalfibre.13.SECStaffAccountingBulletin,No.101:‘RevenueRecognitioninFinancialStatements’,availableat,accessed19February2008.14.FinancialAccountingStandardsBoard,StatementNo.91:‘AccountingforNon-refundableFeesandCostsAssociatedwithOriginatingorAcquiringLoansandInitialDirectCostsofLeases’,availableat,accessed19February2008.15.Tsalavoutas’ssamplesuggeststhatthecostsofacquiringcustomerrelationshipsviamergersandacquisitionsareincreasinglybeingrecognizedasadistinctasset,ratherthansimplypartof‘goodwill’.16.IowethispointtoDavidLeung.17.SeeSuchman(1983),whichisoneofonlyalimitednumberofempiricalstudiesofthecontentofbook-keepingwork,andalso,e.g.,ButtonandHarper(1993).18.ThisparagraphdrawsuponaconversationwithChristineGrimm,anexperi-encedSAPimplementer.19.Theethnomethodologicallyinfluencedcorpusofworkplaceethnographiesofferstheclosestapproachtothestudyofbook-keepers’classifications.Againsee,forexample,Suchman(1983)andButtonandHarper(1993).Chapter71.Alsorelevant,thoughlessdetailedthanDales’sproposal,wastheemissionstradingproposalputforwardbyThomasD.Crocker,thenoftheUniversityofWisconsinatMilwaukee(Crocker1966).Montgomery(1972)isanearlyformalanalysis.2.ForvonHayek,seeCaldwell(2004).3.MarketswerealsosetupintheUSAinemissionsofnitrogenoxides,startingwiththeLosAngelesRegionalCleanAirIncentivesMarket(RECLAIM)in1994.Theirhistoryischequered(seeBurtrawetal.2005),andtheyweremuchless Notes/197influentialthanthesulphurdioxidemarketasexemplarsdrawnoninshapingpolicyforcarbondioxide.4.TitleIVofthe1990CleanAirActAmendmentsisavailableat,accessed2September2006.The‘tons’inquestionareUSor‘short’tons(2,000lb;around907kg).5.SeeStavins(1998:78)andSorrell(1994:76).6.,accessed2September2006.7.Sorrell(1994:76);seealso,accessed11September2006.8.OntheissueofthelegalstatusofSO2allowances,seeDennis(1993).9.MarrakeshAccords,Decision15/CP.7.ThetextoftheMarrakeshAccordsisavailableat,accessed19September2006.10.Foraneloquentargumentthatdespitetheseassertionsemissionallowancesareneverthelesspropertyrights,seeLohmann(2006:73–87).11.Ellermanetal.(2000:38).Inviewofthesalienceofauctiondesigninthedebateovertheperformativityofeconomics(see,especially,MirowskiandNik-Khah2007),itisworthnotingthattheformofauctionthattheEnvironmentalProtectionAgencybelievesismandatedbysection416oftheCleanAirActAmendmentsisnotwhatmanyeconomistswouldhaverecommended.Insteadoftheauctionbeingusedtoidentifyasinglemarket-clearingpricewhichallbuyerspayandallsellersreceive,bidsandasksarematchedinsuchawaythatthesellerwiththelowestaskingpricereceivesthehighestpricethatisbid,andsoon.ThisformofauctionwassharplycriticizedbyexperimentaleconomistsTimothyCasonandCharlesPlott,whoconcludedonthebasisoflaboratorystudiesthatit‘mayprovidepoorpricesignals’andinparticular‘biasesmarket-clearingpricesdownward’(CasonandPlott1996:133–4).Inpractice,however,theroleoftheauctionintheemissionsmarkethasbeentoosmallforthemechanismtomattermuch.ItalsoseemsthattheformofbehaviourpredictedbyCasonandPlott’sexperimentsdidnotoccurinpractice:‘infactprivatesellershavetendedtosettheirreservationpricestoohightoproducesales’(Ellermanetal.2000:299).12.Theratewas2.5lb(1.13kg)ofsulphurdioxideforeachmillionBritishthermalunitsinput(Ellermanetal.2000:7).13.Ellermanetal.(2000:46).Theligniteprovisionissection405(b)(3)oftheCleanAirActAmendmentsof1990.14.Ellermanetal.(2000:38).Thecurrentversionofthemainspreadsheet,theNationalAllowanceDataBase,isavailableat,accessed18September2006. 198/Notes15.Theexception,containedinsection404(a)(3)oftheAmendments,wasanaddi-tionalallocationtoIllinois,Indiana,andOhio(withtheexceptionof‘unitsatKygerCreek,CliftyCreek,andJoppaStream’).Allthreestateshadwell-placedRepresentativesintheHouse.ThethreeexcludedplantsmainlysoldelectricitytoDepartmentofEnergyfacilitiesthatprocesseduranium,whichinvolved‘costplus’contractsinwhichoperatorscouldclaimbackthecostsofallowances(Ellermanetal.2000:40).16.Thecomplexityofthecalculationsinvolvedmeantthattheactualsizeoftheratchetdidnotbecomeknownuntilearlyin1992(Ellermanetal.2000:37).17.Theconcession,overandabovetheratchet,toIllinois,Indiana,andOhioseemstohavehadtheeffectofincreasingthecaponemissionsby0.24milliontonsperyear(calculatedfromthefiguresinEllermanetal.2000:37n.10).18.See,e.g.,KinnerandBirnbaum(2004).19.Aparticularworryhadbeenthatevenifemissionswerereducedoverall,thetradingofallowancesmighthaveledto‘hotspots’:localizedincreasesinSO2emissionstolevelsthatweredamagingtohumanhealthortheenvironment.Onestudyreportedthatof617facilitiesthatwereinoperationin1990,282hadactuallyincreasedemissionsby2000–2(KinnerandBirnbaum2004).Thetotalofsuchincreases(1.2milliontons)was,however,muchlessthanthetotalreductions(6.3milliontons)bytheremainder.Thoughconcernsremain,‘geographiceffects’suchashotspotsseem‘smallcomparedtotheaggregatebenefits’(Burtrawetal.2005:262).20.ThetextoftheKyotoProtocoltotheUnitedNationsFrameworkConventiononClimateChangeisavailableat,accessed24March2006.21.Thequotationisfromarticle17oftheProtocol:seepreviousnote.22.RoyalDutch/ShellandtheUSchemicalsgiantDuPontalsosetupinternaltradingschemessimilartoBP’s(Fialka2000).23.Infact,littleornotradingseemstohavetakenplace.24.TwootherUKschemesinvolvinganelementoftradingaretheRenewablesObligation,underwhichelectricitysuppliersmustbuyincreasingproportionsoftheirelectricityfromcertifiedrenewablesources,andtheEnergyEfficiencyCommitment,underwhichgasandelectricitysuppliershaveobligationstodeliverenergysavings,withaparticularfocusonlow-incomecustomers.Fortheschemes,seeSorrell(2003).25.See,e.g.,ChristiansenandWettestad(2003),DamroandMéndez(2003),Wettestad(2005),Cass(2005),andEngels(2006).26.AlthoughIknowofnodirectevidenceonthepoint,anotherfactorinreducingoppositionmayhavebeenthatthewithdrawaloftheUSA,potentiallyahuge Notes/199purchaserofcreditssuchasCleanDevelopmentMechanism-certifiedemissionreductions,reducedthecoststoothersofmeetingKyotocommitments.27.TheGreenPaperisreproducedinDelbeke(2006:279–313,quoteonp.299).28.HCFC22isanozonedepleter,albeitnotamongstthemostdamagingsuchagents.ItsproductionwillbephasedoutundertheMontrealProtocol,butwillceaseonlyin2030.HCFC22’smainfeedstockuseisintheproductionofpolytetrafluoroethylene(PTFE,orTeflon®),theusesofwhichrangefromaerospacetonon-stickfryingpans.29.AfactorinthepartialrecoverywasthattheGermanfederalgovernmentinsertedintoitsplanaprovisionfor‘ex-postadjustments’,givingitselftherighttoclawbackallowancesiftherewas‘malpractice’or‘undesirableover-allocations’(FreshfieldsBruckhausDeringer2005:3).AlthoughthelegalityofGermany’sproposedex-postadjustmentswaschallengedbytheEuropeanCommission,on15May2006Germanyannouncedthatitwouldemploytheprovisiontoclawbackallowancesequivalenttoaroundtenmilliontonnesofcarbondioxide.SinceGermanyhadmadewhatappearedtobeinabsolutetermsthelargestsingleover-allocation,theannouncementwassignificantnotjustbecauseitwouldreducetheoverallsurplusofallowances:itwasreadas‘newsthatGermanyhadacknowledgedbeingtoogenerousinits2005allocation’(MilnerandGow2006).AlthoughtheEuropeancarbonpricedidnotreturntoitspre-crisislevel,itroseby80%onthedayoftheannouncement.30.Spotofferprice,asreportedon,accessed3December2007.31.See,e.g.,Lohmann(2006)andWara(2007).32.Ontechnopolitics,seee.g.Mitchell(2002).Chapter81.Particularlycentralherehasbeentheactor-networktheoryofLatour(e.g.1987)andCallon(e.g.1986),discussedinChapter2.Glossary1.MostofthedefinitionsinthisglossaryaretakenfromD.MacKenzie,AnEngine,notaCamera:HowFinancialModelsShapeMarkets(Cambridge,Mass.:MITPress,2006).IamgratefultoMITPressforpermissiontoreusethemhere. REFERENCESA,MY.1996.MakingMarkets:OpportunismandRestraintonWallStreet.Cambridge,Mass.:HarvardUniversityPress.1998.‘MarketsasCultures:AnEthnographicApproach’.Pp.69–85inTheLawsoftheMarkets,editedbyMichelCallon.Oxford:Blackwell.A,D,andB,MD.2002.‘SynchronizationRiskandDelayedArbitrage’,JournalofFinancialEconomics,66:341–60.A,PA.,andA,P(eds.).1984.TheSocialDynamicsofFinancialMarkets.Greenwich,Conn.:JAIPress.A,K.1997.EngineeringtheRevolution:ArmsandEnlightenmentinFrance,1763–1815.Princeton:PrincetonUniversityPress.A,F,andG,D.1994.FinancialInnovationandRiskSharing.Cambridge,Mass.:MITPress.A,R.J.,H,J.A.,andS,W.W.1989.WorkingforProfit:TheSocialOrganisationofCalculationinanEntrepreneurialFirm.Aldershot:Avebury.A-G,F,G,C,andR,K.1998.MakingUpAccountants:TheOrganizationalandProfessionalSocializationofTraineeCharteredAccountants.Aldershot:Ashgate.A.2002.‘BP’sCredibilityGapoverCarbonEmissions’.ENDSReportno.326:3–4.2005a.‘TheVisionThing[LexColumn]’,FinancialTimes,20April:18.2005b.‘CaseforaCloserLookatHedgeFunds’,FinancialTimes,12May:18.2007.‘EUETS2008–2012AllocationOverview’,CarbonMarketEurope,6/43(2November):3.A,J.2004.‘Derivatives:VirtualValuesandRealRisks’,Theory,CultureandSociety,21/6:23–4.A,W.B.1984.‘CompetingTechnologiesandEconomicPrediction’,Options,April:10–13.A,M,M,AS.,andR,ME.2005.‘ArbitragingArbitrageurs’,JournalofFinance,60:2471–511.A,J.L.1962.HowToDoThingsWithWords.Oxford:Clarendon.A,Y.2007.‘Agency,NetworksandProfessionalRivalry:TheValua-tionandInvestmentofUKPensionFunds(1948–2006)’.Ph.D.thesis:UniversityofEssex. References/201B,WE.1981.‘MarketsasNetworks:AMultimethodStudyofTradingNet-worksinaSecuritiesMarket’.Ph.D.thesis:NorthwesternUniversity.1984a.‘TheSocialStructureofaNationalSecuritiesMarket’,AmericanJournalofSociology,89:775–811.1984b.‘FloorTradingandCrowdDynamics’.Pp.107–28inTheSocialDynamicsofFinancialMarkets,editedbyPatriciaA.AdlerandPeterAdler.Greenwich,Conn.:JAIPress.B,B.1982.T.S.KuhnandSocialScience.London:Macmillan.1983.‘SocialLifeasBootstrappedInduction’,Sociology,17:524–45.1988.TheNatureofPower.Cambridge:Polity.1995.TheElementsofSocialTheory.London:UCLPress.B,D,andH,J.1996.ScientificKnowledge:ASociologicalAnalysis.London:Athlone.andE,D(eds.).1982.ScienceinContext:ReadingsintheSociologyofScience.MiltonKeynes:OpenUniversityPress.B,S.2003.EnvironmentandStatecraft:TheStrategyofEnvironmentalTreaty-Making.Oxford:OxfordUniversityPress.B,W,andB,H-G.2003.‘MultinationalCompaniesandInternationalCapitalMarkets’.Pp.385–404inInternationalAccounting,editedbyPeterWalton,AxelHaller,andBernardRaffournier.London:Thomson.B,R,andT,G.2007.‘HedgeFundsRivalBanksforShareofUSTreasuryMarket’,FinancialTimes,9March:1.B,K,H,Z,E,P,andG,A.2006.‘TacklingLatency:TheAlgorithmicArmsRace’.London:IBMUK.B,V,F,S,andB,R.2001.BehindClosedDoors:WhatCompanyAuditisReallyAbout.Basingstoke:Palgrave.B,U.1996.‘WorldRiskSocietyasCosmopolitanSociety?EcologicalQuestionsinaFrameworkofManufacturedUncertainties’,Theory,CultureandSociety,13/4:1–32.B,RE.1991.OzoneDiplomacy:NewDirectionsinSafeguardingthePlanet.Cambridge,Mass.:HarvardUniversityPress.2001.‘StrikingaNewDealonClimateChange’.Availableat,accessed26May2006.B,DR.,K,NB.,andR,C.B.,J.2003.‘ReportofInvestigationbytheSpecialInvestigativeCommitteeoftheBoardofDirectorsofWorldCom,Inc’.Availableat,accessed17March2006.B,D,andG,R.2004.‘SecurityAnalystsasFrame-Makers’.Paperpresentedtoworkshop‘ThePerformativitiesofEconomics’,Paris,29–30August.H,I,andMK,D.2006.‘APriceisaSocialThing:TowardsaMaterialSociologyofArbitrage’,OrganizationStudies,27:721–45. 202/ReferencesB,andM,F.2005.‘ListeningtotheSpreadPlot’.Pp.628–33inMakingThingsPublic:AtmospheresofDemocracy,editedbyBrunoLatourandPeterWeibel.Cambridge,Mass.:MITPress.andS,D.2003.‘TheOrganizationofResponsiveness:InnovationandRecoveryintheTradingRoomsofLowerManhattan’,Socio-EconomicReview,1:135–64.2004.‘ToolsoftheTrade:TheSocio-TechnologyofArbitrageinaWallStreetTradingRoom’,IndustrialandCorporateChange,13:369–400.2005.‘ResolvingIdentities:SuccessiveCrisesinaTradingRoomafter9/11’.Pp.293–320inWoundedCity:TheSocialImpactof9/11,editedbyNancyFoner.NewYork:RussellSageFoundationPress.B,B.2006.HedgeHogging.Hoboken,NJ:Wiley.B,DG.1986.‘SuccessandFailureofFuturesContracts:TheoryandEmpir-icalEvidence’.NewYorkUniversity:SalomonBrothersCenterfortheStudyofFinancialInstitutions,Monograph1986–1.B,F,andS,M.1973.‘ThePricingofOptionsandCorporateLiabilities’,JournalofPoliticalEconomy,81:637–54.B,J.2002.ADictionaryofEconomics.Oxford:OxfordUniversityPress.B,D.1997.Wittgenstein,RulesandInstitutions.London:Routledge.B,C,andR,R.2006.‘WhitePaperonSpreadBetting’.London:CassBusinessSchool.B,T,andT,C.2007.‘I’mheretotakethelotout;they’regoingunder,aren’tthey?’FinancialTimes,15–16September:1.B,A,andH,J.B.2002.‘CompetingTheoriesofFinancialAnomalies’,ReviewofFinancialStudies,15:575–606.B,J.1997.‘AddressingClimateChange’.SpeechatStanfordUniversity.Availableat,accessed19September2006.B,M.2005.‘ForPublicSociology’,AmericanSociologicalReview,70:4–28.2007.‘PublicSociologyvs.theMarket’,Socio-EconomicReview,5:356–67.B,D,andD,I.1997.‘EarningsManagementtoAvoidEarningsDecreasesandLosses’,JournalofAccountingandEconomics,24:99–126.B,RS.2005.BrokerageandClosure:AnIntroductiontoSocialCapital.Oxford:OxfordUniversityPress.B,D,E,DA.,K,A,P,K,andT,R.2005.‘EconomicsofPollutionTradingforSO2andNOx’,AnnualReviewofEnvironmentalResources,30:253–89.B,G,andH,R.H.R.1993.‘TakingtheOrganisationintoAccounts’.Pp.98–107inTechnologyinWorkingOrder:StudiesofWork,InteractionandTechnology,editedbyGrahamButton.London:Routledge. References/203C,B.2004.Hayek’sChallenge:AnIntellectualBiographyofF.A.Hayek.Chicago:UniversityofChicagoPress.C,K.2005.‘MakingaGlobalCommodity:TheProductionofMarketsandCottoninEgypt,Turkey,andtheUnitedStates’.Ph.D.thesis:NewYorkUniversity.C,M.1986.‘SomeElementsofaSociologyofTranslation:DomesticationoftheScallopsandtheFishermenofStBrieucBay’.Pp.196–233inPower,ActionandBelief:ANewSociologyofKnowledge?,editedbyJohnLaw.London:RoutledgeandKeganPaul.(ed.)1998.TheLawsoftheMarkets.Oxford:Blackwell.2005.‘WhyVirtualismPavestheWaytoPoliticalImpotence:AReplytoDanielMiller’sCritiqueofTheLawsoftheMarkets’,EconomicSociology:EuropeanElectronicNewsletter,6/2(February):3–20.2007.‘WhatdoesitmeantosaythatEconomicsisPerformative?’Pp.311–57inDoEconomistsMakeMarkets?OnthePerformativityofEconomics,editedbyDonaldMacKenzie,FabianMuniesa,andLuciaSiu.Princeton:PrincetonUniversityPress.andC,K.2005.‘NewandOldDirectionsintheAnthropologyofMarkets’.PaperpresentedtoWenner-GrenFoundationforAnthropologicalResearch,NewYork,9April.andL,B.1981.‘UnscrewingtheBigLeviathan:HowActorsMacro-StructureRealityandHowSociologistsHelpThemToDoSo’.Pp.277–303inAdvancesinSocialTheoryandMethodology:TowardanIntegrationofMicro-andMacro-Sociologies,editedbyKarinKnorrCetinaandA.V.Cicourel.Boston:RoutledgeandKeganPaul.andL,J.2005.‘OnQualculation,AgencyandOtherness’,EnvironmentandPlanningD:SocietyandSpace,23:717–33.C,GA.,andM,E.G.2000.‘Capital-MarketsCrisesandEconomicCollapseinEmergingMarkets:AnInformational-FrictionsApproach’,AmericanEco-nomicReview,90:59–64.CMD.2006.‘CarbonMarketDataPublishesKeyFiguresontheEuro-peanEmissionsTradingScheme’.Availableat,accessed26June2006.C,C,B,D,C,M,andP,K.2000.‘SO2ControlbyElectricUtilities:WhatAretheGainsfromTrade?’JournalofPoliticalEconomy,108:1292–326.C,BG.,andS,AL.1999.‘TheSocialStructureofLiquidity:Flexibility,Markets,andStates’,TheoryandSociety,28:353–82.C,R.2005.‘AccountingforandTaxationofEmissionAllowancesandCredits’.Pp.55–70inLegalAspectsofImplementingtheKyotoProtocolMechanisms:Mak-ingKyotoWork,editedbyDavidFreestoneandCharlotteStreck.Oxford:OxfordUniversityPress. 204/ReferencesC,TN.,andP,CR.1996.‘EPA’sNewEmissionsTradingMechanism:ALaboratoryEvaluation’,JournalofEnvironmentalEconomicsandManagement,30:133–60.C,L.2005.‘NormEntrapmentandPreferenceChange:TheEvolutionoftheEuropeanUnionPositiononInternationalEmissionsTrading’,GlobalEnvironmentalPolitics,5/2:38–60.C,AC,andA,A.2004.‘WhatDeterminesthePriceofCarbonintheEuropeanUnion?’N.p.:PointCarbonandChicagoClimateExchange.andW,J.2003.‘TheEUasaFrontrunneronGreenhouseGasEmissionsTrading:HowDidItHappenandWilltheEUSucceed?’ClimatePolicy,3:3–18.C,WG.,andS,PH.1994.‘WhyDoNASDAQMarketMakersAvoidOdd-EighthQuotes?’JournalofFinance,49:1813–40.C,G.2000.PensionFundCapitalism.Oxford:OxfordUniversityPress.CA.1998.‘TheKyotoProtocolandthePresident’sPoliciestoAddressClimateChange:AdministrationEconomicAnalysis’.Availableat,accessed24April2006.C,R.H.1960.‘TheProblemofSocialCost’,JournalofLawandEconomics,3:1–44.C,P,K,H,L,M,andL,R(eds.).1982.Harrap’sShorterFrench–EnglishDictionary.London:Harrap.C,HM.1974.‘TheTEASet:TacitKnowledgeandScientificNetworks’,ScienceStudies,4:165–86.1990.ArtificialExperts:SocialKnowledgeandIntelligentMachines.Cambridge,Mass.:MITPress.andY,S.1992.‘EpistemologicalChicken’.Pp.301–26inScienceasPracticeandCulture,editedbyAndrewPickering.Chicago:ChicagoUniversityPress.CEC.2005.‘FurtherGuidanceonAllocationPlansforthe2008to2012TradingPeriodoftheEUEmissionTradingScheme’.Brus-sels:CommissionoftheEuropeanCommunities.Availableat,accessed16March2007.2006.‘CommissionDecisionof29November2006ConcerningtheNationalAllocationPlanfortheAllocationofGreenhouseGasEmissionAllowancesNotifiedbyLatviainAccordancewithDirective2003/87/ECoftheEuropeanParliamentandoftheCouncil’.Brussels:CommissionoftheEuropeanCommunities.Availableat,accessed6January2008. References/205C,A.Forthcoming.‘EmissionRights:FromCostlessActivitytoMarketOper-ations’,Accounting,OrganizationsandSociety.C,DJ.,andR,K.2006.‘Accounting,ProfessionsandRegula-tion:LocatingtheSitesofProfessionalization’,Accounting,OrganizationsandSociety,31:415–44.C,JJ.2002.ConfessionsofaStreetAddict.NewYork:SimonandSchuster.C,TD.1966.‘TheStructuringofAtmosphericPollutionControlSys-tems’.Pp.61–86inTheEconomicsofAirPollution,editedbyHaroldWolozin.NewYork:Norton.C,W.1991.Nature’sMetropolis:ChicagoandtheGreatWest.NewYork:Norton.C,E.2007.‘ElectricityGeneratorsGainfromEmissionsTrading’,FinancialTimes,18June:4.D,J.H.1968a.‘Land,Water,andOwnership’,CanadianJournalofEconomics,1:791–804.1968b.Pollution,PropertyandPrices:AnEssayinPolicy-MakingandEconomics.Toronto:UniversityofTorontoPress.D,AR.1995.Descartes’sError:Emotion,ReasonandtheHumanBrain.London:Picador.D,C,andM,PL.2003.‘EmissionsTradingatKyoto:FromEUResistancetoUnionInnovation’,EnvironmentalPolitics,12/2:71–94.D,PA.1992.‘Heroes,HerdsandHysteresisinTechnologicalHistory:ThomasEdisonand“TheBattleoftheSystems”Reconsidered’,IndustrialandCorporateChange,1:129–80.D,PM.,R,SA.,andT,I.2003.‘WhyAreEarn-ingsKinky?AnExaminationoftheEarningsManagementExplanation’,ReviewofAccountingStudies,8:355–84.D,J(ed.).2006.EUEnvironmentalLaw:TheEUGreenhouseGasEmissionsTradingScheme.Leuven:Claeys&Castells.D,G,andG,F.2004.AThousandPlateaus:CapitalismandSchizophrenia.London:Continuum.D,JM.1993.‘SmokeforSale:ParadoxesandProblemsoftheEmissionsTradingProgramoftheCleanAirActAmendmentsof1990’,UCLALawReview,40/4:1101–25.DJ.1996.‘UnitedStatesofAmerica,Plaintiffv.Alex.Brown&SonsInc.[andothers]:CompetitiveImpactStatement’.Availableat,accessed25September2006.D,R.G,andK,HY.1993.‘PriceBarriersintheDowJonesIndus-trialAverage’,JournalofFinancialandQuantitativeAnalysis,28:313–30.E,D.2004.‘“TheLinearModel”DidNotExist:ReflectionsontheHistoryandHistoriographyofScienceandResearchinIndustryintheTwentiethCentury’. 206/ReferencesPp.31–57inTheScience-IndustryNexus:History,Policy,Implications,editedbyKarlGrandin,NinaWormbs,andSvenWidmalm.SagamoreBeach,Mass.:ScienceHistoryPubli-cations.E,B,andM,A.2000.‘WhatExplainsSpreadsonEmerging-MarketDebt:FundamentalsorMarketSentiment?’NationalBureauofEconomicResearchWorkingPaper6408.E,A.D,andB,BK.2007.‘TheEuropeanUnionEmissionsTradingScheme:Origins,Allocation,andEarlyResults’,ReviewofEnvironmentalEco-nomicsandPolicy,1:66–87.S,R,B,EM.,J,PL.,andM,J-P.2000.MarketsforCleanAir:TheU.S.AcidRainProgram.Cambridge:Cam-bridgeUniversityPress.E,A.2006.‘MarketCreationandTransnationalRule-Making:TheCaseofCO2EmissionsTrading’.Pp.329–48inTransnationalGovernance:InternationalDynamicsofRegulation,editedbyMarie-LaureDjelicandKerstinSahlin-Andersson.Cambridge:CambridgeUniversityPress.EDF.1998.‘1998AnnualReport’.NewYork:Environmen-talDefenseFund,availableat,accessed2September2006.EPA(EPA).1997.‘EPATraceabilityProtocolforAssayandCertificationofGaseousCalibrationStandards’.ResearchTrianglePark,NC:EPANationalExposureResearchLaboratory,EPA-600/R-97/121.Availableat,accessed7September2006.EP,C.2003.‘Directive2003/87/ECoftheEuropeanParlia-mentandoftheCouncil’,OfficialJournaloftheEuropeanUnion,275(25October):32–46.2004.‘Directive2004/101/ECoftheEuropeanParliamentandoftheCouncil’,OfficialJournaloftheEuropeanUnion,338(13November):18–23.E,D.2007.‘UnsafeHavens’,BloombergMarkets,October:36–54.F,WD.1998.MarketMaker:ASesquicentennialLookattheChicagoBoardofTrade.Chicago:ChicagoBoardofTrade.FASB(FASB).1985.‘StatementofFinancialAccountingConceptsNo.6:ElementsofFinancialStatements’.Norwalk,Conn.:FASB.F-O’C,M,N,N,S,E,andW,P.2005.Traders:Risks,Decisions,andManagementinFinancialMarkets.Oxford:OxfordUniversityPress.F,JJ.2000.‘GreenLeaderLobbiesHardtoGainCorporateSupport’,WallStreetJournalEurope,22November:2.FIDE(FIÉ).2005a.‘FIDELawsofChess’.Availableat,accessed16May2005. References/2072005b.‘FIDELawsofChess’.Availableat,accessed11October2005.FSA.2005.‘FinalNoticetoCitigroupGlobalMarketsLim-ited,June28’.Availableat,accessed30June2005.F,J.1994.‘LearningbyTrying:TheImplementationofConfigurationalTech-nology’,ResearchPolicy,23:637–52.F,N.2001.TheArchitectureofMarkets.Princeton:PrincetonUniversityPress.F,P.2007.‘ThePrimacyofScienceinModernity,orTechnologyinPost-modernity,andofIdeologyintheHistoryofTechnology’,HistoryandTechnology,23:1–160.FBD.2005.‘EUETSTradingCommences’.Availableat,accessed26September2006.G,A.2007.‘ChicagoTrioShakeUpMarkets’,FinancialTimes,20March:39.G,H.1967.StudiesinEthnomethodology.EnglewoodCliffs,NJ:Prentice-Hall.G,G,andD,P.1986.‘AnExaminationoftheEfficiencyoftheLondonTradedOptionsMarket’,AppliedEconomics,18:995–1010.G,O.2000.‘LeBazardelarationalité:Versunesociologiedesformesconcrètesderaisonnement’,Politix,13/52:17–56.2001.LesTraders:Essaidesociologiedesmarchésfinanciers.Paris:LaDécouverte.2004.‘L’Appropriationduprofit:Politiquesdesbonusdansl’industriefinancière’.Ph.D.thesis:ConservatoireNationaldesArtsetMétiers.2007.Workingrich:Salaires,bonusetappropriationduprofitdansl’industriefinancière.Paris:LaDécouverte.G,M.2005.Virtue,Fortune,andFaith:AGenealogyofFinance.Minneapolis:UniversityofMinnesotaPress.G,C,andP,A.2008.‘AProposalforHowtoAvoidtheNextCrash’,FinancialTimes,31January:13.G,M.1973.‘TheStrengthofWeakTies’,AmericanJournalofSociology,78:1360–80.1985.‘EconomicActionandSocialStructure:TheProblemofEmbeddedness’,AmericanJournalofSociology,91:485–510.1992.‘EconomicInstitutionsasSocialConstructions:AFrameworkforAnalysis’,ActaSociologica,35:3–11.G,J.2006.‘CMEinCheesePrice-FixInvestigation’,FinancialTimes,18August:19.G,AJ.1987.OnMeaning:SelectedWritingsinSemioticTheory.London:Pinter.andC,J.1982.SemioticsandLanguage:AnAnalyticalDictionary.Blooming-ton,Ind.:IndianaUniversityPress. 208/ReferencesG,M.1999.TheKyotoProtocol:AGuideandAssessment.London:RoyalInstituteofInternationalAffairs.H,I.1983.RepresentingandIntervening:IntroductoryTopicsinthePhilosophyofNaturalScience.Cambridge:CambridgeUniversityPress.H,RW.1989.‘EconomicPrescriptionsforEnvironmentalProblems:HowthePatientFollowedtheDoctor’sOrders’,JournalofEconomicPerspectives,3/2:95–114.H,PA.,andS,D(eds.).2001.VarietiesofCapitalism:TheInstitutionalFoundationsofComparativeAdvantage.NewYork:OxfordUniversityPress.H,I.2004.‘“TheSociologyofArbitrage”:ACommentonMacKenzie’,EconomyandSociety,33:239–54.2007.‘TradingtheRisk:Financialisation,LoyaltyandEmergingMarketGovern-mentPolicyAutonomy’.Ph.D.thesis:UniversityofEdinburgh.andMK,D.2007.‘AssemblinganEconomicActor:TheAgencementofaHedgeFund’,SociologicalReview,55:57–80.H,T.2006.‘GreenTaxesandPosturingPoliticians’,FinancialTimes,21July:13.H,L.1991.‘StockPriceClusteringandDiscreteness’,ReviewofFinancialStudies,4:389–415.H,F.2005.‘ColdSnapPushesUpCarbonPrice’,FinancialTimes,25February:3.H,J-P.2000.‘TroisinteractionshétérodoxessurlesmarchésàlacriéeduMATIF’,Politix,13/52:99–119.H,D,L,D,andMK,D.Forthcoming.‘TheFinitistAccountant:Classifications,RulesandtheConstructionofProfits’.InLivinginaMaterialWorld:OntheMutualConstitutionofTechnology,Economy,andSociety,editedbyTrevorPinchandRichardSwedberg.Cambridge,Mass.:MITPress,inpress.H,G,andV,A.2007.‘ABriefHistoryofYieldApproximations’.Pp.19–30inPioneersofFinancialEconomics,ii:Twentieth-CenturyContributions,editedbyGeoffreyPoitrasandFranckJovanovic.Cheltenham:Elgar.H,PM.,andW,JM.1999.‘AReviewoftheEarningsManagementLiteratureanditsImplicationsforStandardSetting’,AccountingHorizons,13:365–83.H,C,J,M,L,P,andH,J.1993.‘UnpackingCollaboration:TheInteractionalOrganisationofTradinginaCityDealingRoom’.ProceedingsoftheThirdEuropeanConferenceonComputer-SupportedCollabo-rativeWork,13–17September,Milan.H,T,andM,M.2005.‘IndexFuturesArbitragebeforeandaftertheIntroductionofSixteenthsontheNYSE’,JournalofEmpiricalFinance,12:353–73.H,D.2004.‘FuzzyNumbers’,BusinessWeek,4October:50–4.H,E.1998.TheTradingCrowd:AnEthnographyoftheShanghaiStockMarket.Cambridge:CambridgeUniversityPress. References/209H,M.1974.TheStructureofScientificInference.London:Macmillan.H,JR.1946.ValueandCapital:AnInquiryintoSomeFundamentalPrinciplesofEconomicTheory.Oxford:Clarendon.H,RD.1988.‘FinancialAccounting:InCommunicatingReality,WeConstructReality’,Accounting,OrganizationsandSociety,13:251–61.H,J.2007.‘SociologyasPublicDiscourseandProfessionalPractice:ACritiqueofMichaelBurawoy’,SociologicalTheory,25/1:46–66.H,B,andM,Y.2005.‘FromRiskstoSecond-OrderDangersinFinan-cialMarkets:UnintendedConsequencesofRiskManagementSystems’,NewPoliticalEconomy,10:223–45.H,AG.,andB,M.1984.‘AccountingResearchintheUnitedKingdom’.Pp.133–61inEuropeanContributionstoAccountingResearch:TheAchievementsoftheLastDecade,editedbyAnthonyG.HopwoodandHeinSchreuder.Amsterdam:FreeUniversityPress.H,DA.2004.‘IndustrialResearch:Commentary’.Pp.59–65inTheScience-IndustryNexus:History,Policy,Implications,editedbyKarlGrandin,NinaWormbs,andSvenWidmalm.SagamoreBeach,Mass.:ScienceHistoryPublications.H,E.1995a.CognitionintheWild.Cambridge,Mass.:MITPress.1995b.‘HowaCockpitRemembersitsSpeeds’,CognitiveScience,19:265–88.IASB(IASB).2005.InternationalFinancialReportingStandards(IFRSs)2005.London:IASB.I,M.2002.NASDAQ:AHistoryoftheMarketthatChangedtheWorld.Roseville,Calif.:Forum.IPCC.2007.ClimateChange2007:ThePhysicalScienceBasis.Cambridge:CambridgeUniversityPress.IM,A.J.1998.‘Eldeclivedelosgrandesnúmeros:BenoitMandel-brotylaestadísticasocial’,Empiria:Revistademetodologíadecienciassociales,1:51–84.2001.‘ReliabilityatRisk:TheSupervisionofFinancialModelsasaCaseStudyforReflexiveEconomicSociology’,EuropeanSocieties,3:69–90.J,AB.,N,RG.,andS,RN.2005.‘ATaleofTwoMarketFailures:TechnologyandEnvironmentalPolicy’,EcologicalEconomics,54:154–74.J,P,andM,R.2005.‘HedgeFundsHoldaQuarterofGermany’sBlue-Chips’,FinancialTimes,2September:17.J,LW.2003.Disconnected:DeceitandBetrayalatWorldCom.Hoboken,NJ:Wiley.J,S,andS,I.2006.‘Iceland’sCollapseHasGlobalImpact’,Finan-cialTimes,23February:42.J,JJ.1991.‘EarningsManagementduringImportReliefInvestigations’,JournalofAccountingResearch,29:193–228. 210/ReferencesJ,B.2006.‘ActivistsTriggerAccountingReview’,FinancialTimes,22Novem-ber:6.K,D,andT,A.1979.‘ProspectTheory:AnAnalysisofDeci-sionunderRisk’,Econometrica,47:263–91.K,H,R,R,andW,H-J.2000.ÖkonomieundGesellschaft,Jahrbuch16.FactsandFigures:EconomicRepresentationsandPractices.Marburg:Metropolis.K,J.2004.‘IgnoretheWisdomofAccountingatYourOwnRisk’,FinancialTimes,7September:21.K,RA.1998.‘AcidRainControl:SuccessontheCheap’,Science,282/5391(6November):1024–6.K,A,andB,R.2004.‘TheAcidRainExperience:ShouldWeBeConcernedaboutSO2EmissionsHotspots?’.Availableat,accessed2September2006.K,LM.,andL,A.1993.‘GenderandtheConstructionofthePro-fessionalAccountant’,Accounting,OrganizationsandSociety,18:507–58.KC,K.2005.‘HowAreGlobalMarketsGlobal?TheArchitectureofaFlowWorld’.Pp.38–61inTheSociologyofFinancialMarkets,editedbyKarinKnorrCetinaandAlexPreda.Oxford:OxfordUniversityPress.andB,U.2000.‘TheMarketasanObjectofAttachment:ExploringPostsocialRelationsinFinancialMarkets’,CanadianJournalofSociology,25:141–68.2002a.‘GlobalMicrostructures:TheVirtualSocietiesofFinancialMarkets’,AmericanJournalofSociology,107:905–51.2002b.‘InhabitingTechnology:TheGlobalLifeformofFinancialMarkets’,CurrentSociology,50:389–405.andP,A(eds.).2005.TheSociologyofFinancialMarkets.Oxford:OxfordUniversityPress.K,TS.1970.TheStructureofScientificRevolutions.Chicago:ChicagoUniversityPress,2ndedition.K,D.1997.LIFFE:AMarketanditsMakers.Cambridge:Granta.L,I.1976.ProofsandRefutations:TheLogicofMathematicalDiscovery.Cambridge:CambridgeUniversityPress.L&WLLP.2005.‘GlobalClimateChangeNews:April’.Availableat,accessed26September2006.L,B.1986.‘VisualizationandCognition:ThinkingwithEyesandHands’.Pp.1–40inKnowledgeandSociety:StudiesintheSociologyofCulturePastandPresent,vol.vi,editedbyHenrikaKuklickandElizabethLong.London:JAI.1987.ScienceinAction.Cambridge,Mass.:HarvardUniversityPress. References/2112005.ReassemblingtheSocial:AnIntroductiontoActor-NetworkTheory.Oxford:OxfordUniversityPress.andW,S.1986.LaboratoryLife:TheConstructionofScientificFacts.Princeton:PrincetonUniversityPress.L,J.1999.‘AfterANT:Complexity,NamingandTopology’.Pp.1–14inActorNetworkTheoryandAfter,editedbyJohnLawandJohnHassard.Oxford:Blackwell.L,JO.1995.‘PublicGoods:ASurveyofExperimentalResearch’.Pp.111–94inTheHandbookofExperimentalEconomics,editedbyJohnH.KagelandAlvinE.Roth.Princeton:PrincetonUniversityPress.L,V-A.2004.‘TheConcreteAbstractionofCapital:HilferdingMeetsCalloninaTradingRoom’.Paperpresentedtoworkshop‘ThePerforma-tivitiesofEconomics’,Paris,29–30August.2007.‘DecodingFinance:ArticulationandLiquidityaroundaTradingRoom’.Pp.87–127inDoEconomistsMakeMarkets?OnthePerformativityofEconomics,editedbyDonaldMacKenzie,FabianMuniesa,andLuciaSiu.Princeton:PrincetonUniversityPress.L,J.2002.‘WhereDoesStateStreetLead?AFirstLookatFinancialPatents,1971to2000’,JournalofFinance,57:901–30.L,J,andW,G.1992.‘FuturesandOptions’.Pp.85–110inMarketsandDealers:TheEconomicsoftheLondonFinancialMarkets,editedbyDavidCobham.Har-low:Longman.L,P.2001.‘GenderingtheMarket:Temporality,Work,andGenderonaNationalFuturesExchange’,WorkandOccupations,28:112–30.andE,WN.2002.‘PollutionFutures:Commensuration,Com-modification,andtheMarketforAir’.Pp.119–47inOrganizations,Policy,andtheNaturalEnvironment:InstitutionalandStrategicPerspectives,editedbyAndrewJ.HoffmanandMarcJ.Ventresca.Stanford,Calif.:StanfordUniversityPress.L,A.1998.‘The“NumbersGame”’.Availableat,accessed7April2004.L,MA.1993.‘EuropeanAcidRain:ThePowerofTote-BoardDiplomacy’.Pp.75–132inInstitutionsfortheEarth:SourcesofEffectiveEnvironmentProtection,editedbyPeterM.Haas,RobertO.Keohane,andMarcA.Levy.Cambridge,Mass.:MITPress.LP,E,andL,B.2004.FinancialDerivativesandtheGlobalizationofRisk.Durham,NC:DukeUniversityPress.2005.‘FinancialDerivativesandtheRiseofCirculation’,EconomyandSociety,34:404–27.L,L.2005.‘MarketingandMakingCarbonDumps:Commodification,CalculationandCounterfactualsinClimateChangeMitigation’,ScienceasCulture,14:203–35. 212/ReferencesL,L.2006.‘CarbonTrading:ACriticalConversationonClimateChange,PrivatisationandPower’,DevelopmentDialogue,48(September):4–359.L,CJ.1966.‘TheJonesNobodyKeepsUpWith’,Fortune,April:237–47.L,C.2004.LegtheSpread:AWoman’sAdventuresinsidetheTrillion-DollarBoys’ClubofCommoditiesTrading.NewYork:Broadway.MD,L.1997.CapitalCulture:GenderatWorkintheCity.Oxford:Blackwell.MG,L,B,E,A,R,andW,M.2004.‘CreditCorrelation:AGuide’.London:JPMorgan.Availableat,accessed24June2006.MK,D.1990.InventingAccuracy:AHistoricalSociologyofNuclearMissileGuidance.Cambridge,Mass.:MITPress.2003.‘Long-TermCapitalManagementandtheSociologyofArbitrage’,EconomyandSociety,32:349–80.2004.‘TheBig,BadWolfandtheRationalMarket:PortfolioInsurance,the1987CrashandthePerformativityofEconomics’,EconomyandSociety,33:303–34.2005.‘OpeningtheBlackBoxesofGlobalFinance’,ReviewofInternationalPoliticalEconomy,12:555–76.2006.AnEngine,notaCamera:HowFinancialModelsShapeMarkets.Cambridge,Mass.:MITPress.2007a.‘TheMaterialProductionofVirtuality:Innovation,CulturalGeographyandFacticityinDerivativesMarkets’,EconomyandSociety,36:355–76.2007b.‘ThePoliticalEconomyofCarbonTrading’,LondonReviewofBooks,29/7(5April):29–31.2008.‘ProducingAccounts:Finitism,TechnologyandRuleFollowing’.Pp.99–117inKnowledgeasSocialOrder:RethinkingtheSociologyofBarryBarnes,editedbyMassimoMazzotti.Aldershot:Ashgate.andM,Y.2003.‘ConstructingaMarket,PerformingTheory:TheHistoricalSociologyofaFinancialDerivativesExchange’,AmericanJournalofSociology,109:107–45.M,F,andS,L(eds.).2007.DoEconomistsMakeMarkets?OnthePerformativityofEconomics.Princeton:PrincetonUniversityPress.andWJ(eds.).1999.TheSocialShapingofTechnology.Buckingham:OpenUniversityPress.M,J.2006.‘HedgeFundsAheadofEstimates’,FinancialTimes,29November:30.MN,MF.2000.‘ResearchDesignIssuesinEarningsManagementStudies’,JournalofAccountingandPublicPolicy,19:313–45. References/213M,S.2001.‘CapitalFlowstoDevelopingEconomiesthroughouttheTwentiethCentury’.Pp.51–73inFinancialGlobalizationandDemocracyinEmergingMarkets,editedbyLeslieElliottArmijo.Basingstoke:Palgrave.M,W.1999.‘RateSettinginLondon’.BritishBankers’Association,typescript.M,B.2001.‘EngineeringanIslamicFuture:SpeculationsonIslamicFinancialAlternatives’,AnthropologyToday,17/1:8–11.2002.‘RepressedFutures:FinancialDerivatives’TheologicalUnconscious’,EconomyandSociety,31:15–36.2005.MutualLife,Limited:IslamicBanking,AlternativeCurrencies,LateralReason.Princeton:PrincetonUniversityPress.M,L,andT,B.1996.LeoMelamed:EscapetotheFutures.NewYork:Wiley.M,RP.2000.‘OneHundredYearsofSolicitude:IntellectualPropertyLaw1900–2000’,CaliforniaLawReview,88:2187–240.M,RC.1973.‘TheoryofRationalOptionPricing’,BellJournalofEconomicsandManagementScience,4:141–83.andB,Z.2005.‘DesignofFinancialSystems:TowardsaSynthesisofFunctionandStructure’,JournalofInvestmentManagement,3:1–23.M,RK.1948.‘TheSelf-FulfillingProphecy’,AntiochReview,8:193–210.M,D(ed.).2005.Materiality.Durham,NC:DukeUniversityPress.M,P,andP,M.1995.‘CalculatingCorporateFailure’.Pp.51–76inProfessionalCompetitionandProfessionalPower:Lawyers,AccountantsandtheSocialConstructionofMarkets,editedbyYvesDezalayandDavidSugarman.London:Routledge.M,Y.2003.‘WhereDoFinancialMarketsComeFrom?HistoricalSociologyofFinancialDerivativesMarkets’.Ph.D.thesis:UniversityofEdinburgh.M,F,P,NS.,andS,SV.2005.‘OrganizedDetachment:ClearinghouseMechanismsinFinancialMarkets’,InformationandOrganization,15:229–46.M,M,andG,D.2006.‘InTheory,OnlytheVirtuousAreRewarded’,TheGuardian,16May:4.M,P.2002.MachineDreams:EconomicsBecomesaCyborgScience.Cambridge:CambridgeUniversityPress.andN-K,E.2007.‘MarketsMadeFlesh:Performativity,andaProbleminScienceStudies,AugmentedwithConsiderationoftheFCCAuctions’.Pp.190–224inDoEconomistsMakeMarkets?OnthePerformativityofEconomics,editedbyDonaldMacKenzie,FabianMuniesa,andLuciaSiu.Princeton:PrincetonUniversityPress. 214/ReferencesM,T.2002.RuleofExperts:Egypt,Techno-Politics,Modernity.BerkeleyandLosAngeles:UniversityofCaliforniaPress.M,S.2006.‘OptionMarketsandImpliedVolatility:PastversusPresent’.Typescript.M,H.2003.‘TheTemporalitiesoftheMarket’,AmericanAnthropologist,105/2:255–65.2005.‘TheMaterialityofFinanceTheory’.Pp.165–81inMateriality,editedbyDanielMiller.Durham,NC:DukeUniversityPress.M,P,andT,N.1999.TheHandbookofInternationalFinancialTerms.Oxford:OxfordUniversityPress.M,S,S,S,andS,I.2004.‘UncertaintiesintheFinnishGreenhouseGasEmissionInventory’,EnvironmentalScienceandPolicy,7:87–98.M,W.D.1972.‘MarketsinLicensesandEfficientPollutionControlPrograms’,JournalofEconomicTheory,5:395–418.M,L,andJ,S.2006.‘WarrantPricing60YearsbeforeBlackScholes:EvidencefromtheJohannesburgStockExchange’,JournalofFinance,62:3069–98.M,K.2006.‘LowerPollutioninEUSeesCO2PermitsFall30%’,FinancialTimes,27April:19.M,L.2003.GlobalCapitalandNationalGovernments.Cambridge:CambridgeUniversityPress.M,F.2003.‘Desmarchéscommealgorithmes:SociologiedelacotationélectroniqueàlaBoursedeParis’.Ph.D.thesis:ÉcoleNationaleSupérieuredesMines.2005.‘Contenirlemarché:LatransitiondelacriéeàlacotationélectroniqueàlaBoursedeParis’,Sociologiedutravail,47:485–501.andC,M.2007.‘EconomicExperimentsandtheConstructionofMarkets’.Pp.163–89inDoEconomistsMakeMarkets?OnthePerformativityofEconomics,editedbyDonaldMacKenzie,FabianMuniesa,andLuciaSiu.Princeton:PrincetonUniversityPress.NAO.2004.TheUKEmissionsTradingScheme:ANewWaytoCombatClimateChange.London:StationeryOffice.NTUA.n.d.‘ThePrimesEnergySystemModel:SummaryDescription’.Athens:NationalTechnicalUniversity,availableat,accessed6January2008.N,WD.2000.‘FromPorcopolistoCarbopolis:TheEvolutionfromPorkBelliestoEmissionsTrading’.Pp.61–73inEmissionsTrading:EnvironmentalPolicy’sNewApproach,editedbyRichardF.Kosobud.NewYork:Wiley.O’B,R.1992.GlobalFinancialIntegration:TheEndofGeography.London:Pinter. References/215O’M,P.2003.‘MoralUncertainties:ContractLawandDistinctionsbetweenSpeculation,Gambling,andInsurance’.Pp.231–57inRiskandMorality,editedbyRichardV.EricsonandAaronDoyle.Toronto:UniversityofTorontoPress.O,CL.2003.‘CurrencyOrdersandExchangeRateDynamics:AnExplanationforthePredictiveSuccessofTechnicalAnalysis’,JournalofFinance,58:1791–819.O,N,andP,T(eds.).2003.HowUsersMatter:TheCo-ConstructionofUsersandTechnology.Cambridge,Mass.:MITPress.P,K.1994.‘LessonsfromtheFOXResidentialPropertyFuturesandMortgageInterestRateFuturesMarket’,HousingPolicyDebate,5:343–60.P,R.2004.FinancialAccountingforNon-Specialists.Maidenhead:McGraw-Hill.P,AC.1920.TheEconomicsofWelfare.London:Macmillan.P,C.1996.‘MarketLiquidityandDepthonComputerizedandOpenOut-cryTradingSystems:AComparisonofDTBandLIFFEBundContracts’,JournalofFuturesMarkets,16:519–43.P,JM.1993.‘AStatus-BasedModelofMarketCompetition’,AmericanJournalofSociology,98:829–72.2001.‘NetworksasthePipesandPrismsoftheMarket’,AmericanJournalofSociology,107:33–60.PC.2006.‘Carbon2006:TowardsaTrulyGlobalMarket’.Oslo:PointCarbon.2007.‘Carbon2007:ANewClimateforCarbonTrading’.Oslo:PointCarbon.P,TM.1995.TrustinNumbers:ThePursuitofObjectivityinScienceandPublicLife.Princeton:PrincetonUniversityPress.P,MK.1991.‘EducatingAccountants:TowardsaCriticalEthnography’,Accounting,OrganizationsandSociety,16:333–53.1992.‘ThePoliticsofBrandAccountingintheUnitedKingdom’,EuropeanAccountingReview,1:39–68.P,A.2001a.‘TheRiseofthePopularInvestor:FinancialKnowledgeandInvestinginEnglandandFrance,1840–1880’,SociologicalQuarterly,42:205–32.2001b.‘IntheEnchantedGrove:FinancialConversationsandtheMarketplaceinEnglandandFranceinthe18thCentury’,JournalofHistoricalSociology,14:276–307.2001c.‘SenseandSensibility:Or,HowShouldSocialStudiesofFinanceBehave:AManifesto’,EconomicSociology:EuropeanElectronicNewsletter,2/2:15–18.2004a.‘InformativePrices,RationalInvestors:TheEmergenceoftheRandomWalkHypothesisandtheNineteenth-Century“ScienceofFinancialInvestments”’,HistoryofPoliticalEconomy,36:351–86.2004b.‘EpistemicPerformativity:TheCaseofFinancialChartism’.Paperpre-sentedtoworkshop,‘ThePerformativitiesofEconomics’,Paris,29–30August. 216/ReferencesP,A.2006.‘Socio-TechnicalAgencyinFinancialMarkets:TheCaseoftheStockTicker’,SocialStudiesofScience,36:753–82.P,G,andR,S.2007.‘TimetoDitchKyoto’,Nature,449(25October):973–5.P,MD.Forthcoming.‘Geomoney:AnOptiononFrost,GoingLongonClouds’,Geoforum.andA,J.2000.‘MonetizedTime-Space:Derivatives—Money’s“NewImaginary”’?EconomyandSociety,29:264–84.P,S.2003.‘AccountantBalked,thenCaved’,WallStreetJournalEurope,24June:M6–M7.andS,D.2002.‘How3UnlikelySleuthsUncookedWorldCom’sBooks’,WallStreetJournalEurope,31October:A8.Q,P,andH,T.2006.‘WhatisIT?SAP,Accounting,andVisibilityinaMultinationalOrganisation’,InformationandOrganization,16:212–50.R,MR.1983.‘TheAnomalousStockMarketBehaviorofSmallFirmsinJanuary:EmpiricalTestsforTax-LossSellingEffects’,JournalofFinancialEconomics,12:89–104.R,A.2004.‘RealTime:UnwindingTechnocraticandAnthropologicalKnowledge’AmericanEthnologist,31:392–405.R,A.2002.‘TheNightDetective’.Availableat,accessed2February2007.R,P.n.d.‘Arbitrage/ShortSelling:APoliticalEconomyApproach’.Typescript.R,AM.1951.‘RumorintheStockMarket’,PublicOpinionQuarterly,15:461–86.1966.‘ASocialPsychologicalApproachtotheStudyoftheStockMarket’,Kyklos,19:267–87.R,RE.,andW,J.2003.InanUncertainWorld:ToughChoicesfromWallStreettoWashington.NewYork:RandomHouse.RWEAG.2007.‘RWEerzieltEinigungmitBundeskartellamtinCO2-Verfahren’.Essen:RWEAG,27September.Availableat,accessed28September2007.S,L.1998.‘U.S.UniversitiesGetBurnedbyRashofRiskyStrategies’,WallStreetJournalEurope,13October:5B.S,RL.,andS,HB.1983.‘InventiveActivityinFuturesMarkets:ACaseStudyoftheDevelopmentoftheFirstInterestRateFuturesMarket’.Pp.255–72inFuturesMarkets:Modelling,ManagingandMonitoringFuturesTrading,editedbyManfredE.Streit.Oxford:Blackwell.S,AG.,R,JK.,A,JA.,N,LE.,andC,JD.2003.‘TheNeuralBasisofEconomicDecision-MakingintheUltimatumGame’,Science,300(13June):1755–8. References/217S,S.1992.‘LateVictorianMetrologyanditsInstrumentation:AManufactoryofOhms’.Pp.23–56inInvisibleConnections:Instruments,Institutions,andScience,editedbyRobertBudandSusanCozzens.Bellingham,Wash.:SPIEOpticalEngineeringPress.S,K.1989.‘CommentaryonEarningsManagement’,AccountingHorizons,3:91–102.S,C.n.d.‘TheNitty-GrittyofCBOTRDJIASMFuturesIndexArbitrage’.Availableat,accessed30July2005.S,S.1994.ASocialHistoryofTruth:CivilityandScienceinSeventeenth-CenturyEngland.Chicago:UniversityofChicagoPress.andS,S.1985.LeviathanandtheAir-Pump:Hobbes,Boyle,andtheExperimentalLife.Princeton:PrincetonUniversityPress.S,A,andV,RW.1997.‘TheLimitsofArbitrage’,JournalofFinance,52:35–55.S,J,N,K,andC,Y,2006.‘CO2CostPass-ThroughandWindfallProfitsinthePowerSector’,ClimatePolicy,6:49–72.S,WL.1981.‘Innovations,Competition,andNewContractDesigninFuturesMarkets’,JournalofFuturesMarkets,1:123–55.S,H.1955.‘ABehavioralModelofRationalChoice’,QuarterlyJournalofEconomics,69:99–118.S,A,andM,P.2005.‘DealersGettheSqueeze’,FinancialTimes,2February:41.S,SW.1981.‘TheScottishContribution’.Pp.12–26inBritishAccountingStandards:TheFirst10Years,editedbySirRonaldLeachandEdwardStamp.Cambridge:Woodhead-Faulkner.S,CW.1981.TheMindoftheMarket:AStudyofStockMarketPhilosophies,TheirUses,andImplications.Totowa,NJ:Rowman&Littlefield.S,KH.,andW,R(eds.).2002.ShapingTechnology,GuidingPolicy:Concepts,SpacesandTools.Cheltenham:Elgar.S,S.1994.‘PollutionontheMarket:TheUSExperiencewithEmissionsTradingfortheControlofAirPollution’.Brighton:UniversityofSussex,SciencePolicyResearchUnit.2003.‘WhoOwnstheCarbon?InteractionsbetweentheEUEmissionsTradingSchemeandtheUKRenewablesObligationsandEnergyEfficiencyCommitment’,EnergyandEnvironment,14:677–703.S,RN.1988.‘Project88:HarnessingMarketForcestoProtecttheEnvironment’.Availableat,accessed11September2006.1998.‘WhatCanWeLearnfromtheGrandPolicyExperiment?LessonsfromSO2AllowanceTrading’,JournalofEconomicPerspectives,12/3:69–88. 218/ReferencesS,D.1982.‘HowTradedOptionsStarted’,MoneyObserver,27April.S,N.2007.TheEconomicsofClimateChange:TheSternReview.Cambridge:CambridgeUniversityPress.S,LA.1983.‘OfficeProcedureasPracticalAction:ModelsofWorkandSystemDesign’,ACMTransactionsonOfficeInformationSystems,1:320–8.S,S.2002.‘“WhitePaper”SubmittedtoBoardofWorldCom,Inc’.Availableat,accessed8September2006.S,EJ.2000.BuildingtheGlobalMarket:A4000YearHistoryofDerivatives.London:Kluwer.S,B.2000.‘AllowanceTradingandPotentialHotSpots:GoodNewsfromtheAcidRainProgram’,EnvironmentReporter,31(12May):954–9.T,B.1993.TheMerc:TheEmergenceofaGlobalFinancialPowerhouse.NewYork:HarperCollins.T,SH,W,T.J.,andR,GR.1998.‘AreAccrualsduringInitialPublicOfferingsOpportunistic?’ReviewofAccountingStudies,3:175–208.T,G.2007.‘Libor’sValueisCalledintoQuestion’,FinancialTimes,26September:43.T,D.2007.‘HedgeFundAssetsSoartoalmost$2,500bn’,FinancialTimes,10October:23.T,P.1997.‘ThePyrrhicVictoryofGentlemanlyCapitalism:TheFinancialEliteoftheCityofLondon,1945–90’,JournalofContemporaryHistory,32:283–304and427–40.T,N.1994.‘OntheSocialandCulturalDeterminantsofFinancialCentres:TheCaseoftheCityofLondon’.Pp.327–55inMoney,Power,andSpace,editedbyStuartCorbridgeandRonMartin.Oxford:Blackwell.2000.‘Pandora’sBox?CulturalGeographiesofEconomies’.Pp.689–704inOxfordHandbookofEconomicGeography,editedbyGordonL.Clark,MericS.Gertler,andMaryannP.Feldman.Oxford:OxfordUniversityPress.T,A.1998.‘UnstableFutures:ControllingandCreatingRisksinInternationalMoney’,SocialistRegister:248–77.2000.‘DangerousDerivatives:ControllingandCreatingRisksinInternationalMoney’,Geoforum,31:87–99.T,H,andB,T.2004.‘ASBTellsUKCompaniestoIgnoreEURulingonAccountingStandards’,FinancialTimes,12October:1.andH,D.2004.‘AccountingWatchdogSeesTrouble’,FinancialTimes,10November:19.T,P.1989.‘FinancialInnovationandFirst-MoverAdvantages’,JournalofFinancialEconomics,25:213–40. References/2192003.‘FinancialInnovation’.Pp.307–35inHandbookoftheEconomicsofFinance,ia:CorporateFinance,editedbyGeorgeM.Constantinides,MiltonHarris,andRenéM.Stulz.Amsterdam:Elsevier.U,K.2007.‘ShouldAnyoneBeSurprisedbytheNAPDecisions?’CarbonMarketEurope,6/13(19January):1.U,B.1999.‘EmbeddednessintheMakingofFinancialCapital:HowSocialRelationsandNetworksBenefitFirmsSeekingFinancing’,AmericanSociologicalReview,64:481–504.D,A,andM,P.2004.‘HowCitigroupShookEurope’sBondMarketswithTwoMinutesofTrading’,FinancialTimes,10September:17.V,DG.,andH,JC.2006.‘BP’sEmissionsTradingScheme’,EnergyPolicy,34:2100–12.V,C.2002.HowtoWinatFinancialSpreadBetting.London:Pearson.V,SK.1996.FreerMarkets,MoreRules:RegulatoryReforminAdvancedIndustrialCountries.Ithaca,NY:CornellUniversityPress.V,V.2007.‘TheRGGIExperiment’,TradingCarbon,1/1(December):28–9.W,P,H,A,andR,B(eds.).2003.InternationalAccounting.London:Thomson.W,M.2007.‘IstheGlobalCarbonMarketWorking?’Nature,445(8February):595–6.W,M.1970.‘Class,Status,Party’.Pp.180–95inFromMaxWeber:EssaysinSociology,editedbyH.H.GerthandC.WrightMills.London:RoutledgeandKeganPaul.2000a.‘StockandCommodityExchanges[DieBörse(1894)]’,TheoryandSociety,29:305–38.2000b.‘CommerceontheStockandCommodityExchanges[DieBörsenverkehr]’,TheoryandSociety,29:339–71.W,H.2005.‘EcuadorPavesWayfor$750mBondIssue’,FinancialTimes,5December:8.W,J.2005.‘TheMakingofthe2003EUEmissionsDirective:AnUltra-QuickProcessduetoEntrepreneurialProficiency?’GlobalEnvironmentalPolitics,5:1–23.W,HC.1981.‘WhereDoMarketsComeFrom?’AmericanJournalofSociology,87:517–47.2001.MarketsfromNetworks.Princeton:PrincetonUniversityPress.W,R.2003.‘FleshandtheFreeMarket(OnTakingBourdieutotheOptionsExchange)’,TheoryandSociety,32:677–723.W,J.2007.‘TheCityvWallStreet:TheSmartMoneyison(andin)London’,FinancialTimes,27–8January:11.W,L.1980.‘DoArtifactsHavePolitics?’Daedalus,109/1:121–36. 220/ReferencesW,J.M.2005.‘Assemblage’.Pp.77–87inGillesDeleuze:KeyConcepts,editedbyCharlesJ.Stivale.Chesham:Acumen.W,L.1967.PhilosophicalInvestigations.Oxford:Blackwell.W,T.2007.‘NAPIICourtCases:ADouble-EdgedSword’,CarbonMarketEurope,6/21(1June):1.Y,F(ed.).2005.ClimateChangeandCarbonMarkets:AHandbookofEmissionsReductionMechanisms.London:Earthscan.Y,GU.1927.‘OnReadingaScale’,JournaloftheRoyalStatisticalSociety,90:570–87.Z,C.2003.‘AmbiguousNumbers:TradingTechnologiesandInterpretationinFinancialMarkets’,AmericanEthnologist,30:258–72.2004.‘TheProductiveLifeofRisk’,CulturalAnthropology,19:365–91.2006.OutofthePits:TradingandTechnologyfromChicagotoLondon.Chicago:ChicagoUniversityPress.Z,P,andV,M.2002.‘PathwaystoEuropeanGreenhouseGasEmissionsTradingHistoryandMisconceptions’.Availableat,accessed19September2006.Z,SA.1984.‘SomeJuncturesintheEvolutionoftheProcessofEstablishingAccountingPrinciplesintheU.S.A.:1917–1972’,AccountingReview,59:447–68.Z,DM.2004.‘HereaChief,ThereaChief:TheRiseoftheCFOintheAmericanFirm’,AmericanSociologicalReview,69:345–64.Z,EW.1999.‘TheCategoricalImperative:SecuritiesAnalystsandtheIllegitimacyDiscount’,AmericanJournalofSociology,104:1398–438.2004.‘StructuralIncoherenceandStockMarketActivity’,AmericanSociologicalReview,69:405–32. INDEXA.W.Jones&Co.39audit130–1,133Abolafia,Mitchel73,106Austin,J.L.30Abreu,Dilip107aviation158accounting133–4,186Avrahampour,Yally183accruals-based111–12,114,124–5classifications109–10,112–16,127–9,132,Baker,Wayne3133,134,136,158–9banking177,178,184gender22BaringsBank100localcultures117–18Barnes,Barry26,27,28,32Scotland119basecorrelation15standards117–18,119–20BearStearns82AccountingStandardsBoard[UK]118Beattie,Vivien130,131accruals-basedaccounting111–12,114,Beck,Ulrich183124–5behaviouralfinance2,23,59acidrain143,144,147betting,seegambling;spreadbettingactor-networktheory19–20,21,22,34,57Beunza,Daniel88–9,93,94,97–8,99agencements4,19–25,38,56–7,59,60,61,62BigBangderegulation73agency21,61–2Biggs,Barton53agriculturalcommodityexchanges67,73,Black,Fischer3,8775,76,78Black-Scholes-Mertonoptionpricing87,allowances144–5,174–599asaccountingitems158–9Black-Scholesmodel3,4allocation139,146,147–8,159–62,170,blackboxes34,185–6175Blair,Tony153auctioning146,153–4,159,160Bloombergterminals191n3ch2BP151–2Bloor,David29,130,131,133ETSsurplus167–8Bobbitt,Max128over-allocation175bodies,human10–13,95–7Andersen,Arthur127,128,129bonds78arb96defined44,187arbitrage5,13,85–108,185–6bookkeeping22,131–2,133–4carbon165BP(BritishPetroleum)138,151–3defined85–6,187brains,human3,13,61purism’slimitations88Brandt,Richard130,131Arditti,Fred69,81Brav,Alon107Aron,J.andCompany93Brazil:ArthurAndersen127,128,129emissionstrading150assets109,114,121–2,128governmentbonds44,55,90–2,98,99,Attari,Mukkaram107107 222/Indexbreaks45ChicagoBoardOptionsExchange3,66,70,BrettonWoods6974BritishBankers’Association1,81,82ChicagoClimateExchange154BritishPetroleum138,151–3ChicagoMercantileExchange9–10,66,67,brokers,voice11–12,8070,83broker’sear12Eurodollarcontract79,81Browne,John151,153Chile56Bruegger,Urs13,46,51,194n1ch3chlorodifluoromethane164,169Brunnermeier,Markus107Christie,William34,35Buchner,Barbara169Citadel40bundfutures104CitigroupGlobalMarketsLtd.104,105–6,Burawoy,Michael181,182,185108Burdick,Quentin146CityIndex67Burgstahler,David125–6classifications,accounting109–10,112–16,Bush,GeorgeH.W.143,144127–9,132,133,134,136,158–9Bush,GeorgeW.110,151,156–7CleanAirActAmendments1990[US]138,143,145,147California170CleanDevelopmentMechanism,seeunderCaliskan,Koray21KyotoCallon,Michel:clearinghouses84,99–100actor-networktheory34Clinton,Bill149,156agencements4,19,20,21CO2,seecarboneconomicexperiments166coal,brown146onmarkets174,182Coase,Ronald140performativity31,32,68,89cognition,distributed16–17,46–51,59calls14,96collateralizeddebtobligations(CDOs)CantorIndex6715–16,17,179,185cap-and-tradeschemes142,155defined187BP151,152,155collectiveaction24,34–5carbon:Collins,Harry21colonialism149CommodityFuturesTradingCommissionprices162–4191n4ch2taxes149,155CommunityIndependentTransactiontrading138,151–76seealsoEmissionsLog158TradingSchemeCompaq99Carruthers,Bruce10,180computers:carrytrade60–1heatissues17cashsettlement78–9location,importanceof94Cason,Timothy197n11memory65CaymanIslands41,57contractdesign69–70CDOs,seecollateralizeddebtobligationsconversion96certifiedemissionreductions(CERs)Cooper,Cynthia127–8,129,134,135165copulamodel,Gaussian15,179chartism13,14,90–1corporeality10–13,95–7cheeseauction,Chicago9–10costs114,115Chen,Yihsu168,169coupons,defined44,187chessexample28–9,29–30creditcrisis2,178–9,185Chicago,cultureof73–4creditdefaultswaps44ChicagoBoardofTrade63,66,67,78Crocker,ThomasD.196n1 Index/223cuesindealingrooms41electricityprices168,169currencyfutures66,69,71electronicmail51–3customerrelationships122Ellerman,A.Denny146,149,169embodiment10–13,95–7Daimler-BenzAG117emissionsallowances,seeallowancesDales,J.H.140–2,145–6emissionsmarkets18–19,137–76,182,186daytrading36design33Delbecke,Jos154emergence31,32Deleuze,Gilles19emissionsrights120,144–5DeloitteToucheTohmatsu111EmissionsTradingScheme137,138–9,Denmark153154–65,166–7depreciation117NationalAllocationPlans160–2,171–2,derivatives4–5,10,63–84,185173amountstraded63–4NationalRegistries158cashsettlements78–9emotionalsupport51,59defined187EnergyEfficiencyCommitment[UK]housing67,79198n24interests70Enron110,115,129,132Islamic65EnterpriseResourcePlanning(ERP)longevity79systems132pricing64EnvironmentalDefenseFund143–4,151property67,79EnvironmentalProtectionAgency(EPA)standards118[US]18,19,25,142,144,147taxtreatment71,77EONIA83virtuality64–5,83,84Ernst&Young111weather79ethnomethodology193n21Dichev,Ilia125–6EuropeanAssociationfortheStudyofdiscount,defined187ScienceandTechnology191n4ch1distributedcognition16,46–51,59EuropeanClimateExchange138diversification102EuropeanUnion:Donaldson,R.Glen191n6ch2carbontax149DuPont198n22EmissionsTradingSchemeseeEmissionsTradingSchemeearningsmanagement5,123–4,125–6,131environmentalvoting155Ebbers,Bernard123,127taxrules155economicreality116–19Eurozonegovernmentbonds104,105,106economicsociology180exchangerates69economics6,65,139–42,143,145extensionalsemantics26,27,28andemissionsmarkets139–42,166–7,externalities139–40174–5performativity30–1,32,68,89F9monkeys21Ecuador56facticity8–9,10Edge,David32andderivatives66,78–83EdinburghUniversity:Fearnley,Stella130,131carbonpermits137–8FinancialAccountingStandardsBoard[US]chartofaccounts112–13120efficientmarkethypothesis88financialeconomics14,30–1eighthsofadollar34–5,95financialreporting111,130,135–6,183–4Eire57,58technologicalbases133 224/IndexFinancialReportingStandard17[UK]183GreatLakes,pollutionof140–1,142FinancialServicesAct1986[UK]76–7,78greenhousegases139,151,161–2seealsoFinancialServicesAuthority[UK]40,106carbon;HCFC22;HFC23;nitrousfinitism5–6,26–30,133,134,184oxidedefined26Greimas,Algirdas20financialreporting129–30,135–6Grimm,Christine196n18Fleck,James32,68Fligstein,Neil181Hahn,Robert143foreignexchangemarkets191n6ch2Hardie,Iain89,90Forman,Paul193n27HarvardUniversityendowmentfund86,forwards,defined18799,101framings52–3HCFC22(chlorodifluoromethane)164,FTSE-100futures76,77169fuelprices163Heaton,J.B.107fundsoffunds40,43,89,100–1hedgefunds37–62,185futures63advertisingprohibited39bund104assets39,58carbondioxide138backofficesupport45–6,57,58currency66,69,71complexityof38defined44,66,187–8defined38–9,188FTSE-10076,77fees40grain63,75investorslimited39Islamic103gambling65,69,75–7seealsospreadlocations37,40–1bettingnumbersof39,57GamingAct1845[UK]75,76observedexample37–8,41–56Garfinkel,Harold193n21organizationalsize37Gaussiancopulamodel15,179primaryregistrations41gender57tradingvolumes40gentlemanlycapitalism24,73hedgelets84Germany40,160hedging69allowances199n29HelsinkiProtocol1985192n20electricityprices169HewlettPackard99ETS,oppositionto156HFC23(trifluoromethane)164,169governmentbonds86,104Hicks,John118–19NationalAllocationPlan171Holmwood,John181gifteconomies23,24hotspots198n19globalwarming25,137,149,151,153housingderivatives67,79globalization55Hutchins,Edwin16,47,48Godechot,Olivier61Goede,Mariekede65IAS2:135Goffman,Joe144IAS19:183goldarbitrage86,98IAS38:113Goodhart,Charles184IAS39:118,124,135Gore,Al149Iceland60–1Grady,Robert144IG(InvestorsGold)Index67,72,77–8grainfutures63,75imitation60,72–3Granovetter,Mark180,181impliedvolatility14–15GreatCrash1929119defined188 Index/225income118–19LaboratoryLife9indexarbitrage93Latour,Bruno9,19,20,34,181initialpublicofferings(IPOs)125Latvia171,172innofusion32,68Law,John19,20innovation:LawsoftheMarkets,The182financial65,68–9,71,72lawyers,roleof144–5technological31–3,65,68,70–1,72leadcredits142InstituteofAccountants[US]119,120Leeson,Nick100intellectualpropertylaw72Lépinay,Vincent-Antonin45intenttest75–6Leung,David196n16internalcombustionengine71leverage39,40InternationalAccountingStandards,Levy,Marc192n20seeIASLIBOR3,9,10,11,12,66,76,79–80,178InternationalAccountingStandardsback-uparrangements1–2Board120constructionof67,81–2,83InternationalCommercialExchange66,67,defined80,18869inputfreeofcharge23InternationalFinancialReportingprestige24InterpretationsCommitteeLIFFE67,68,71,72–3,74,76,83(IFRIC)120ShortSterlinginterest-ratecontract79InternationalMonetaryMarket66lignite146interviewsources6–7,67–8,89linearmodelsofinnovation31–3,68inventories120–1liquidity10,19,180InvestmentCompanyAct1940[US]39London,cultureof24,73Ireland,Republicof57,58LondonFuturesandOptionsExchange67,Islamicderivatives6579Islamichedgefunds103LondonInterbankOfferedRate,seeLIBORItaly,governmentbonds86LondonInternationalFinancialFuturesExchange,seeLIFFEJ.Aron&Company93LondonStockExchange74jobbers74motto23,24Jones,A.W.39LondonTradedOptionsMarket71,73,74Jones,Jennifer124–5longpositions70,103JWMPartners101defined188Long-TermCapitalManagement34,39,87,Kahneman,Daniel23,5989,99,100,101,107Kerviel,Jérôme100diversification102–3Kilcollin,Rick69longevityderivatives79Kim,Harold191n6ch2LTCM,seeLong-Termkinks125LTOM(LondonTradedOptionsMarket)KnorrCetina,Karin46,51,194n1ch371,73,74KPMG111,128,129Krupp,Fred151MacKenzie,Donald89Kynaston,David76McNichols,Maureen125KyotoProtocol1997138,145,149,150–1,Malone,Ferrell128157mark-to-market58,100,118CleanDevelopmentMechanism150,marketmaking69157,164,165,169–70,173MarrakeshAccords2001145,157transactionslog158marriage59 226/Indexmaterialsociology3,20,88Nordhaus,William149materialisttheoryofexchange75NordPool138materialityofmarkets:NorthernRock82,177,184arbitrage92–8defined2–4O’Malley,Pat75Maurer,Bill65open-outcry11,24,73,84,96,97measurement:defined188carbonmarket152,153,157–8options3–4,14–15,74,87metrology18–19carbondioxide138profit109–36defined3,66–7,188sulphurdioxidemarket144,148Osler,Carol191n6ch2Melamed,Leo81,83ostension133–4Mello,Antonio107Meriwether,JohnW.101paradigms,scientific133Merton,RobertC.3,87Pardo-Guerra,Juan-Pablo195n3ch5Merton,RobertK.195n7ch6patenting72methane153pensionprovision183,195n9metrology,seeundermeasurementperformativity,economics30–1,32,68,89Mexico,governmentbonds103Persaud,Avinash184Mirowski,Philip193n26Pigou,Arthur140Miyazaki,Hirokazu100,101pits11,24,73,96,97models180,181defined188modemspeedlimitation94Plott,Charles197n11Monitorsystem93PointCarbon163monolines185politics,importanceof55–6,163MontrealProtocol25subpolitics183,184Morse,Gene127,128,132,133Power,Mike195n2ch6Mosley,Layna56Preda,Alex13MTSbond-tradingsystem86,105,106,prices,materiality92–5108PricewaterhouseCoopers111Muniesa,Fabian166,174,194n8primebrokers45–6Myers,David128,129PRIMESmodel172–3,175Project88,144NASDAQ34,35propertyderivatives67NationalInstituteofStandardsandprospecttheory59Technology[US]18publicsocialscience6navigationanalogy16,48publicsociology181–2,185netrealisablevalue121puts14,96Neuhoff,Karsten168,169neuroeconomics61QuantumFund39NewLabour153QWERTYkeyboard70–1NewYorkInternationalCommercialExchange66,67,69ratchet147–8,175,176NewYorkStockExchange73ratings,system178Nik-Khah,Edward193n26rationality239/11,2001:94reconciliation451987crash77–8,87,93reflexivity126nitrogenoxidesmarkets196n3RegionalCleanAirIncentivesMarketnitrousoxide164(RECLAIM)LosAngeles196n3 Index/227RegionalGreenhouseGasInitiative[US]SO2,seesulphurdioxide170socialnetworks56regulation84socialstudiesoffinance4,16,23,30,179–80,RenewablesObligation[UK]198n24181Reutersterminals93,191n3ch2defined2revenue114SociétéGénérale100Robertson,Dennis119SocietyforSocialStudiesofRobotti,Paola103Science191n4ch1roublebonds103sociology:RoyalDutch99,198n22economic180Ruckes,Martin107material3,20,88rules27–8,29,135public181–2,185accounting119,120,134Soros,George39Russia103speculating69speed,transmission92–4SalomonBrothers101spreadbetting66,67,77,83–4seealsosandbagging53gamblingSandor,Richard19,69,149,154tax71,77Sarbanes-OxleyBill[US]110spreads74scales,instability33–6,109standardization78Schipper,Katherine123–4Stark,David88–9,93Schmalensee,Richard143,146StateStreetBank&TrustCo.v.SignatureFinancialScholes,Myron3,87Group72Schultz,Paul34,35statisticalarbitrage97–8Scotland,accounting119Stavins,Robert143–4,145scrubbers143,148Steen,David73–4SecuritiesandExchangeCommissionSternreview25[US]35–6,95Stinchcombe,Arthur10,180securitymasters45stocks,defined189securityprices,arithmetic34–6structuredinvestmentvehicles(SIVs)185selfishness23,24Stupka,John127semiotics20sub-primemortgages[US]177September11,2001:94Suchman,Lucy196n17Shapin,Steven101Sullivan,Scott121–3,127,128,129,132,134shares,defined188sulphur-dioxidetrading138,143–9,166Sharia103swaps79–80sheetmonkeys21,22creditdefault44Shell99,198n22defined189Shleifer,Andrei88,107Sweden156,160shoes,platform97shortpositions,defined188tax71,77shortselling39,70,103–4carbon149,155defined188year[US]194n5Signature72technicalmatters179Sijm,Jos168,169technicalities33–4Simon,Herbert47technologicalinnovation31–3,65,68,70–1,Siu,Lucia2272Smith,Glyn127technopolitics175–6Smith,Susan195n4ch4Telerate81 228/Indextickers13voicebrokers11–12,80tizzysnatching73–4volatility,defined189tote-boarddiplomacy24vonHayek,Friedrich141tradeblotters45vonMises,Ludwig141tradepress67traders46–51,89,180–1,184wagers,seegamblingtradingscreens13WallStreetJournal82–3trifluoromethane164,169weather:trust23,24,101derivatives79Tsalavoutas,Yannis122effectsof163,168Tversky,Amos23,59Weber,Max73Tweedie,SirDavid118,131Wheeler,Stuart67White,Harrison181Ulset,Kjersti173Wittgenstein,Ludwig27,135UnitedKingdom56women11,22,57carbontrading153,155,156Woolgar,Steve9dirtymanofEurope24–5WorldCom109–10,112,121–2,123,127–9,UnitedStates56130,132KyotoProtocol149–51,156–7Wyoming148ozonedepletion25universityendowmentfunds86Yearley,Steven21uptickrule33,103yields15,54,55calculators47–8Vainio,Matti154defined189virtualderivatives64–5,83,84Vishny,Robert107Zaloom,Caitlin78,84,95–6Vogel,Steven27Zapfel,Peter154