Volatility Noise

This study shows that fitting errors of equity-option-implied volatility surfaces are informative about intermediary frictions. For each stock and day, we quantify the goodness of fit between the observed implied volatilities of all available options and the corresponding estimates from OptionMetrics' smoothed volatility surface using root-mean-square errors. In the cross-section of stocks, this error metric increases in idiosyncratic stock volatility and several measures of option and stock illiquidity. Based on these insights, we propose an overarching measure for intermediary frictions given by the value-weighted average of the stock-specific fitting errors. This measure of volatility noise peaks during episodes of market distress and exhibits sensible correlations to standard economic state variables like the market return, the TED spread, or the VIX. Moreover, we uncover a close link between volatility noise and the constraints on intermediary equity and debt. Finally, our asset pricing results indicate that volatility noise is informative for the cross-sectional variation in expected returns beyond the equity option market. In particular, we find volatility noise to constitute a priced risk factor in returns of stocks as well as equity and bond mutual funds