SOLVING ASSET RETURN PUZZLES BY FORCE OF PARAMETER UNCERTAINTY: A MIXTURE-OF-DISTRIBUTIONS APPROACH

SOLVING ASSET RETURN PUZZLES BY FORCE OF PARAMETER UNCERTAINTY: A MIXTURE-OF-DISTRIBUTIONS APPROACH

I provide a potential solution to asset return puzzles as been noted in macrofinance literature since the seminal work of Mehra & Prescott (1985). It is well known that the standard consumption-based asset pricing model (Lucas 1978) leads to asset return puzzles such as high equity premium (6-8%) and low rikfree rate (1-2%) that are hard to explain with a reasonable degree of risk aversion in the CRRA utility function, which are commonly employed in macroeconomic literature. Hence, these puzzles raise critical questions for the common macroeconomic modeling exercise. I address these puzzles by showing that a non-Gaussian error distribution, perceived by the consumerinvestor due to the uncertain nature of parameters underlying the dividend process, can explain the historically high equity premium and low riskfree rate; thus, providing a potential solution to those puzzles. Mehra & Prescott (1985) find that about 0.35% is the maximum risk premium the standard C-CAPM can explain with a risk-free rate between 0 and 4% when the mean and standard deviation of consumption growth is assumed 1.8% and 3.6% respectively. I also take 1.8% mean and 3.6% standard deviation for consumption growth but explain 6-7% equity risk premium, 0.5-2% riskfree rate and 5-7% dividend yield with the risk aversion coefficient below 10 solely relying on negative skewness and leptokurtosis that can be derived from the stochastic nature of the uncertain (variance) parameter underlying the economy’s endowment process.