Extrapolative Expectations: Implications for Volatility and Liquidity

This paper presents a model of the joint behavior of liquidity and volatility. In the model, investors extrapolate recent price movements to forecast the volatility of a risky asset. When the perceived volatility is high, the risk premium is high, the current return on the risky asset is low, the risk-free rate is low, and the market is illiquid. Illiquidity amplifies supply shocks thus increasing realized volatility of prices, which in turn, feeds into volatility forecasts for the next period. As a result, clustering of volatility and liquidity arises endogenously even though fundamentals are homoskedastic. The model helps to unify several known facts about liquidity and volatility, and I find support for its new predictions using a measure of misperception of risk based on the implied volatility of options.