@Article{mfj:766,
title={U.K. Stock Market Inefficiencies and the Risk Premium},
author={Antonis Demos and George Vasillelis},
journal={Multinational Finance Journal},
volume={11},
number={1/2},
pages={97--122},
year=2007,
publisher={Multinational Finance Society; Global Business Publications},
url={http://www.mfsociety.org/../modules/modDashboard/uploadFiles/journals/MJ~745~p16tq6pf6t1mm3qaoe23por1a0t4.pdf}
keywords={conditional heteroskedastic latent factor; LM test; stock returns},
abstract={The stock market predictability has been a favorite topic of scholars and practitioners alike. It seems that some small predictability is present in all major stock markets worldwide. This predictability can be attributed to the risk premium structure and/or to inefficiencies present in the markets. This paper investigates the predictability of returns of some major shares listed in the London Stock exchange, using economic as well as accounting variables. We first measure the predictability of these variables by regressing individual stock returns on their corresponding accounting variables and the economic ones. Second, we estimate for the returns a seasonal latent factor model with time varying volatility. Provided that our measure of risk is an adequate one, the residuals of this estimation are free of the predictability of risk premium, and consequently one expects that any accounting and factor economic variables would have no predictive power. An LM-type test is developed and employed to indicate that indeed the U.K. stock market predictability is due to the risk premium structure, and the explanatory power of the variables considered here is due to them being an approximation of risk. However, when we perform the test jointly for all assets, we reject the zero predictability hypothesis at 5% but not at 1%..},
}