Output details

Heteroscedasticity and interval effects in estimating beta: UK evidence

This paper compares beta estimates from Ordinary Least Squares (OLS) regression with estimates corrected for heteroscedasticity of the error term using Autoregressive Conditional Heteroscedasticity (ARCH) models for the stocks from the London Stock Exchange. The main finding is that OLS tends to overestimate the beta coefficients compared with ARCH models. The bias in betas is also associated with the volatility of the share’s daily returns. The importance of this paper is that this is the first study to directly investigate both the intervalling effects and ARCH effects (heteroscedasticity) in beta estimation which has direct practical relevance for stock market investors.