Volatility Forecasting Using Financial Statement Information

Abstract

This paper examines whether financial statement information can predict future realized volatility incremental to the volatility implied by option market prices. Prior research establishes that option-implied volatility is a biased estimator of future realized volatility. I use an analytical framework to identify accounting-based drivers of equity returns volatility. My main hypothesis is accounting-based drivers can be used to forecast future volatility incremental to either past volatility or the market’s expectation of future volatility quantified as option-implied volatility. I confirm this empirically and show that my findings are robust to the measurement of option-implied volatility using either a model-free approach or the Black-Scholes model. I also document abnormal returns to a option-based trading strategy that takes a long (short) position in firms with financial statement information indicative of high (low) future volatility. Additionally, I provide evidence that contradicts a risk-based explanation for the incremental predictive ability of accounting-based variables. Taken together, my results indicate that the market’s failure to fully process accounting-based fundamental information explains some of the previously documented bias in implied volatility.

Keywords: volatility, fundamental analysis, option returns

JEL Classification: M41, G12, G13, G14