Can Illiquidity Be Priced in an Active Secondary Market? Theory and Evidence

Commencing with a Lucas (1978)-type representative investor but with differing endowments, we develop a new theoretical model of counterparty trading inclusive of frictions to show that symmetric liquidity costs, which could arise either from exogenous costs or from order-flow asymmetric information, are not priced. This is because seller costs cancel out the buyer costs correctly identified in Amihud and Mendelson's (1986a) seminal theoretical model. We test our generalization of the Lucas model utilizing NYSE (US) equity market microstructure data to show that we cannot reject our main hypothesis concerning the absence of liquidity pricing effect on stock returns. We split transaction costs into their buy (upside) and sell (downside) components to find they are priced with similar magnitudes in contemporaneous returns. Based on our NYSE sample, the balanced effect of buy and sell lambda price impact does not generate a downside lambda premium in future stock returns. We further report a positive pricing effect of the bid-ask spread on future returns on the extreme quintile of lambda asymmetry