Investigating the Economic Utility of Implementing Implied Volatility and Select Derived Metrics in Investment and Trading Strategies

Abstract

This thesis explores whether option-implied volatility contains useful information for trading strategies. It evaluates the performance of volatility-managed portfolio strategies based on implied volatility and the predictive power of implied volatility-derived metrics around corporate earnings announcements. The study begins with a literature review covering capital asset pricing, options markets, portfolio strategies, volatility modelling, and corporate earnings announcements. This informs a theoretical framework that combines mean-variance optimization, factor models, and event study methodology.Two applications of implied volatility are explored. First, the volatility management strategy by MOREIRA and MUIR (2017) is extended using implied volatility as a forward-looking volatility proxy and testing whether it produces enhanced performance. Second,the thesis examines the predictive power of implied volatility-derived metrics, specifically the call-put spread and the curve coefficient of the volatility smile, on abnormal earnings announcement returns. This analysis builds on and modifies the methodology presented by Lei et al. (2020) by incorporating the smile coefficient and the second moment of the volatility spread. Using S&P 500 large-cap stocks from 1996 to 2022 and Option Metrics data, the thesis finds that volatility-managed strategies, whether based on realized or implied volatility,yield statistically significant positive alphas over the sample period. While neither yields significant positive alphas, the realized volatility management strategy yields higher average returns and Sharpe ratio. For the event study, the thesis finds that neither the volatility call-put spread or skew of the volatility smile are able to significantly predict abnormal post-earnings announcement returns. Although the thesis finds that including the second moment of the volatility spread in the event study regression increases the statistical significance and coefficient of the spread, it remains insignificant.