Which Mechanism, Illiquidity or Volatility Risk, Better Explains the Equity Premium Puzzle?

Abstract

We show in a fairly general setting of a buyer and seller with the same preferences trading two related assets so as to share volatility risk that illiquidity and virtually all impediments to trade cannot be priced. This is because the buying and selling counterparties must both be optimizing. The buyer values the asset at the low Bid-price and seller at the high Ask-price. When market clears, the buyer and seller order-flows cancel out leaving the midpoint price independent of the tax wedge. Using this admittedly “stylized” model and very modest CARA and CRRA risk preferences, we generate Mehra and Prescott’s (1985) equity premium of about 8% and negligible yield on T-bills, together with the observed turnovers of both T-Bills and equity securities, to show that market volatility is systematically priced. Moreover, only in the presence of binding short-selling restrictions that effectively eliminate the supply responses to adverse demand shocks, can illiquidity be priced and a number of other anomalies explained. We reconcile our findings with Amihud and Mendelson’s (1986) treatment of the buyer by adding in the essential seller’s perspective which only then provides the necessary theoretical underpinnings to explain the asymmetric empirical findings of Brennan, Chordia, Subrahmanyam, and Tong (2012) and Brennan, Huh, and Subrahmanyam (2015).