The Value Premium - Is the Failure To Explain it as a Compensation for Risk a Consequense of Mis-specified Models?

Abstract: Many studies have tried to explain the stock market value premium identified by Fama and French and Rosenberg, Reid and Lanstein. To the proponents of conventional asset pricing theory the value premium, measured by HmL (high book-to-market minus low book-to-market), is a bit of a dilemma. This is due to that the value of growth stocks depend more on business cycles than on value stocks, whose values are less dependent on economic circumstances. Accordingly growth stocks are expected to have higher returns and betas but empirical evidence confirms the contrary. The purpose of this study is to examine the value premium on the Swedish stock market by applying the Sharpe Lintner CAPM as well as two additional models, LCAPM and HCPAM. The models are empirically tested in an unconditional and conditional manner where the latter uses changes in industrial production and the implied volatility as conditional variables. These two variables could be seen as proxies for the state of business cycle as well as the investors’ uncertainty regarding the future. The empirical results show that all the unconditional versions of CAPM show a significant valuepremium, which is consistent with other studies. This is also true for all the conditional versions of HCPAM, but not for LCAPM, where the intercept no longer significant. This implies that both the states of business cycle as well as the future uncertainty have explanatory power when it comes to the investors’ preferences regarding the relationship between risk and return. These results show that the failure of explaining the value premium as a compensation for risk is in fact the result of mis-specified models.