Option-based variables reflect investors’ assessment of future risk and therefore contain information about expected stock returns. Early studies show that information flows from the options market to the equity market. Empirical evidence suggest that portfolios created using option-based variables have returns that cannot be fully explained by traditional asset pricing variables. Following Bali, Chabi-Yo and Murray (2022), this thesis examines the predictive power of option-based variables, such as the difference between call and put implied volatilities, the difference between realized volatility of the underlying stock and option implied volatility, and the change of the open interest in options. The options on stocks traded in the US stock exchanges in the period between 1996 and 2015 are analyzed. The study also investigates whether the predictive power of the option-based variables changes during periods of economic recession. The findings show that option-based variables increase the predictive power of the models when used with the traditional asset pricing variables. Option-based variables are found to be useful predictors of stock returns during recessions as well. The estimation model which includes option-based variables and stock characteristics outperforms CAPM and Fama-French three-factor model during both recession and expansion periods but the accuracy of the model is significantly lower during recessions. The model fails to estimate the future returns of high beta stocks as accurately as low beta stocks. Portfolios formed based on quintile values of the option-based variables create economically large but statistically insignificant abnormal returns.