MISS
Professor: G. Phillips
Course: Empirical Finance
Code: BST264
Empirical Finance
Name: Ce Jiang
Student Number: C1240653
Required words（excluding the tables）:2000- 2500
Total words（excluding the tables）：2335
Date Submitted: 29/04/2013
Critical evaluation of Heteroskedasticity in Stock Returns
Motivation
A large numbers of researchers pointed that variance of aggregate stock returns changes over time. They figuring out that the standard deviation of aggregate monthly returns are different between two periods. Prior researches took heteroskedasticity as a purely statistical problem, just as a potentially confounding factor in estimating the market model. However, many authors opposed this view since the market return is a poor estimate of market volatility and evidences have been found to against homoscedasticity while relate error variances to squared market returns. The authors in this paper want to illustrate how a simple weight least squares heteroskedasticity correction could effect by testing the capital asset pricing model. And the authors in this paper and others believe that many types of tests should take heteroskedasticity into account though there are predictable movements in stock volatility.
Objective
This paper trying to find out whether Heteroskedasticity in stock returns is a pervasive phenomenon. The authors also need to measure aggregate monthly stock volatility within five portfolios of stocks sorted by firm’s size, this could judge the correlation between the volatility of monthly returns to the size portfolios and autoregressive predictions of this market volatility factor. The paper has four separate parts. First, it will analyze the single index model of heteroskedasticity. From this model it will also discuss the “market model” regression equation. In this part they will also show how CAPM are affected by the use of a weighted least squares estimation procedure that accounts for heteroskedasticity.
Course: Empirical Finance
Code: BST264
Empirical Finance
Name: Ce Jiang
Student Number: C1240653
Required words（excluding the tables）:2000- 2500
Total words（excluding the tables）：2335
Date Submitted: 29/04/2013
Critical evaluation of Heteroskedasticity in Stock Returns
Motivation
A large numbers of researchers pointed that variance of aggregate stock returns changes over time. They figuring out that the standard deviation of aggregate monthly returns are different between two periods. Prior researches took heteroskedasticity as a purely statistical problem, just as a potentially confounding factor in estimating the market model. However, many authors opposed this view since the market return is a poor estimate of market volatility and evidences have been found to against homoscedasticity while relate error variances to squared market returns. The authors in this paper want to illustrate how a simple weight least squares heteroskedasticity correction could effect by testing the capital asset pricing model. And the authors in this paper and others believe that many types of tests should take heteroskedasticity into account though there are predictable movements in stock volatility.
Objective
This paper trying to find out whether Heteroskedasticity in stock returns is a pervasive phenomenon. The authors also need to measure aggregate monthly stock volatility within five portfolios of stocks sorted by firm’s size, this could judge the correlation between the volatility of monthly returns to the size portfolios and autoregressive predictions of this market volatility factor. The paper has four separate parts. First, it will analyze the single index model of heteroskedasticity. From this model it will also discuss the “market model” regression equation. In this part they will also show how CAPM are affected by the use of a weighted least squares estimation procedure that accounts for heteroskedasticity.