CFA二级公式表

《CFA二级公式表》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、CFA公式表-LevelIIQuantitativeMethods1.SimpleLinearRegressionCovxyCorrelation:rxy()()ssxyrn2t-testforr(df=n-2):t21rCovxyEstimatedslopecoefficient:2xEstimatedintercept:bybˆˆx01^ConfidenceintervalforpredictedY-value:ytSEofforecastc2.MultipleRegressionYbbX()()bX()bXii01122i33ii

2、Teststatisticalsignificanceofb;Hb:00ˆtdb,1fnksbRejecthif

3、

4、tcriticaltorp-value<0ConfidenceInterval:btsˆ()jcbjSST=RSS+SSEMSR=RSS/kMSE=SSE/(n-k-1)Teststatisticalsignificanceofregression:F=MSR/MSEwithdfkandn-k-1(1-tail)Standarderrorofestimate(SEEMSE),SmallerSEEmeansbetterfit.2

5、Coefficientofdetermination(RRSSSST/).%ofvariabilityofY2explainedbyX’s;HigherRmeansbetterfit.3.RegressionAnalysis-ProblemsHeteroskedasticity.Non-constanterrorvariance.DetectwithBreusch-Pagantest.CorrectwithWhite-correctedstandarderrors.Autocorrelation.Correlationamongerrorterms.Detectwit

6、hDurbin-Watsontest;positiveautocorrelationifDW

7、ncorrectlypoolingdata.Usinglaggeddependentvariableasindependentvariable.Forecastingthepast.Measuringindependentvariableswitherror.5.EffectsofMisspecificationRegressioncoefficientsarebiasedandinconsistent,lackofconfidenceinhypothesistestsofthecoefficientsorinthemodelpredictions.6.LinearT

8、rendModel:ybbttt017.Log-LinearTrendModels:ln()ybbttt018.CovarianceStationary:Meanandvar.don’tchangeovertime.Todetermineifatimeseriesiscovariancestationary,(1)plotdata,(2)runanARmodelandtestcorrelations,and/or(3)performDickeyFullertest.9.UnitRoot:Coefficientonlaggeddep.vb1=1.seriesw

9、ithunitrootisnotcovariancestationary.Firstdifferencingwillofteneliminatetheunitroot.10.Autoregressive(AR)Models:Specifiedcorrectlyifautocorrelationofresidualsnotsignificant:xbbxbx......bxtt01122tptptb011.MeanRevertingLevelForAR(1):(1b)112.RMSE:squar