7113

6640

كارشناسي ارشد

آمار اقتصادي و اجتماعي

اصفهان: دانشگاه صنعتي اصفهان، دانشكده علوم رياضي

1390

[هشت]، 115ص.: مصور، جدول

ص.ع. به فارسي و انگليسي

سروش عليمرادي، علي رجالي

امير مظفر اميني

29/7/91

ايرج كاظمي، علي زينل همداني

1396/09/20

كتابنامه

علوم رياضي

رياضي

ID6640

به فارسي و انگليسي: قابل رويت در نسخه ديجيتالي

Abstract Current research on predictive regressions studies the case where for each t Ytis regressed on a lagged predictor variable Xt 1 and predictive series Xt is offirst order autoregressive The problem with these predictive models is that theOLS estimated slope coefficient is biased in the finite sample case when theerrors of the autoregressive model for Xt are correlated with the errors in thepredictive regression model However there are predictor series havingautoregressive structure of order greater than 1 Predictive regression withautoregressive predictors that are not necessarily of the AR 1 structure arequite common in finance and economics So a method that could reduce bias ofestimators for coefficients in predictive regressions with order p autoregressivepredictors can be useful In this thesis we consider predictive model where for each t Yt is regressed onXt 1 Xt 2 Xt p and Xt being AR p with p 1 Using generalizedaugmented regression method for this case where the predictor variable isAR p a method has been proposed for having bias reduced point estimationfor the predictive coefficients and a corresponding hypothesis testing procedure This method has been generalized to the case of multiple AR p predictors Wecompare OLS and augmented regression methods in terms of the bias inestimating the predictive coefficients and in terms of the size of the statisticaltests on hypothesis tests for the coefficients using simulation and empiricalanalysis For New York Stock Exchange data our method applied to a model in whichquarterly stock returns are predicted by dividend yields shows that the predictorseries is AR 2 For these data dividend yield is a significant predictor of stockreturn not only based on OLS but also based on the standard bias correctionmethod that assumes that the predictor series is AR 1 However the predictorseries is found to be AR 2 The result of applying this method was that theestimated predictor coefficients are insignificantly different from zero Also inthis thesis Iran Stock Exchange data were evaluated based on two models Weshowed that the predictability of stock return using dividend yield does notexist

سروش عليمرادي، علي رجالي

امير مظفر اميني

ايرج كاظمي، علي زينل همداني