Volatility Tests and Efficient Markets: A Review Essay

Abstract

This essay examines what volatility tests tell us about the data and what implications we should derive from them. It argues that volatility tests do not tell us that "prices are too volatile", implying that "markets are inefficient", but rather that "(discounted) returns are forecastable", implying that "current discount rate models leave a residual". It also argues that the discount rate residuals documented by volatility tests (and equivalent return forecasting regressions or Euler equation tests) are suggestive of rational, business cycle-induced discount rate movements, rather than "fads" or other inefficiencies.