Stock Market Return Predictability Before and After the Dodd-Frank Act

Abstract

We analyze the stock market return predictability for three different periods. We evaluate the conditional variance (CV) and the variance risk premium (VRP) as predictors of stock market returns for which we are using well-established versions of the heterogeneous auto-regressive (HAR) model and propose two new semi-parametric extensions. Results show that the CV and VRP are predictors of future stock market returns in the period before the global financial crisis (GFC). However, these variables lose predictive power after the Dodd-Frank Act (DFA) and change sign, indicating that investors are willing to pay a risk premium for "good uncertainty".

Keywords: Dodd-Frank Act; Non-Parametric Methods; Predictability; Realized Variance; Variance Risk Premium

JEL Classification: C22; C51; C52; C53; G1; G32