资源描述:
《burgess phd thesis a computational methodology for modelling the dynamics of statistical arbitrage(1)外语英文电子书》由会员上传分享,免费在线阅读,更多相关内容在教育资源-天天文库。
1、AComputationalMethodologyforModellingtheDynamicsofStatisticalArbitrageAndrewNeilBurgessDecisionTechnologyCentreDepartmentofDecisionSciencesAthesissubmittedtotheUniversityofLondonforthedegreeofDoctorofPhilosophyUNIVERSITYOFLONDONLONDONBUSINESSSCHOOL1O
2、ctober1999Tomyparents,ArnoldandCarol.©A.N.Burgess,199923AcknowledgementsThankstomysupervisor,PaulRefenes,forbringingmetoLBS,keepingmeinbreadandwater,helpingmewithideasandevenreadingthebloodything!ThankstopastandpresentcolleaguesattheNeuroforecastingU
3、nit/DecisionTechnologyCentre.EspeciallyYvesBentz,PeterBolland,JerryConnor,StefaniaPandelidaki,NevilleTowersandPeterSchreiner;fordiscussions,hardwork(notallthetime),companyandsixgoodyears.Alsoalltheex-CRLerswhoshowedthatevenarealjobcanstillbefun.Thank
4、stothevisitorstotheLBSgroup,FernandoandPaul,forgoodtimesandhardworkinLondon,HelsinkiandMelbourne.Thanksforthepeoplewhohelpedkeepitreal,especiallytoPratapSondhiforideasandsupportwhenhewasatCitibankatthebeginning;theothersponsorsoftheNeuroforecastingCl
5、ubandthentheDecisionTechnologyCentre;BothaandDerickdowninS.A.fortrustingmetobuildatradingsystem;andDavidandAndrewforalltheireffortsonbehalfofNewSciences.AlsotowhoeverfirstdecidedtoholdNIPSworkshopsatskiresorts;andtotheregularsatNNCM/CFconferences:Joh
6、nMoody,AndreasWeigend,HalWhite,YaserAbu-Mostafa,AndrewLoandBlakeLeBaroninparticularfortheirenthusiasmandinspiration.Finally,myloveandthankstoDeborah,whohadtoputupwithmewhilstIwaswritingup-andwhosephotographsofchromosomeshadtocompeteforcomputertimewit
7、hmyequitycurves.4AbstractRecentyearshaveseentheemergenceofamulti-disciplinaryresearchareaknownas“ComputationalFinance”.Inmanycasesthedatageneratingprocessesoffinancialandothereconomictime-seriesareatbestimperfectlyunderstood.Byallowingrestrictiveassu
8、mptionsaboutpricedynamicstoberelaxed,recentadvancesincomputationalmodellingtechniquesofferthepossibilitytodiscovernew“patterns”inmarketactivity.Thisthesisdescribesanintegrated“statisticalarbitrage”frameworkforidentifying,modellingandexploitingsmallbu