That the volatility of the equity risk premium and the risk-free rate are larger than most models predict has long been known and discussed (for example the seminal Mehra and Prescott 1985 paper). Bansal and Yahon try their hands at explaining why this puzzle exists (their paper will appear in an upcoming Journal of Finance (JF)). They model the "consumption and dividend growth rates as containing (i) a small long-run predictable component and (ii) fluctuating economic uncertainty." In other words, market participant's risk aversion varies with conditions and hence drives the market to be more volatile than the underlying economic conditions would suggest. Interesting paper! [FinanceProfessor.com Annotation]