Abstract
This thesis studies the content of the implied volatility (IV) of SPX and SPY options, and explores the IV for the prediction of equity premiums and variance risk premiums, and compares these two option premiums. My thesis is organized in three chapters as follows,
In Chapter 2, we provide a comprehensive study of the IV smirk in the SPDR S&P 500 exchange-traded fund (SPY ETF) option market. The IV curves are downward sloping with little curvature, exhibiting an almost straight line. However, the shape of the IV curves becomes more curved during the global financial crisis (GFC) period, showing that the commonly accepted IV smirk shape is driven by the GFC. In addition, based on in-sample and out-of-sample tests and asset allocation analysis, we show that the first difference of the slope factor can predict the next month's SPY excess returns.
In Chapter 3, we study the pattern of S&P 500 index (SPX) options' IV curves and their predictive ability for the variance risk premium (VRP). We find that the commonly-accepted SPX IV smirk shape is driven by the global financial crisis (GFC) period. We find that the daily and monthly VRP can be significantly predicted by the IV factor term spreads. The level factor (at-the-money IV) term spread shows significant predictive power for all our monthly VRP proxies, namely the investable variance swap and S&P 500 straddle returns in both in-sample and out-of-sample tests. At a daily frequency variance swap returns can be predicted in-sample, but only the straddle returns can be predicted out-of-sample. This disparity is likely due to monthly variance returns aligning more with the expectation horizons of the quantified IV factors.
In Chapter 4, we show that SPY ETF options (American) are more expensive than SPX index options (European) on average. However, there are times when SPX options are more expensive than SPY options throughout the sample. This result is more pronounced for call options. Dividend payments seem to play an important role in explaining this phenomenon as the dates explain the difference between SPY and SPX premiums over time. The dividend has a negative effect on the SPY and SPX call option price difference in both time-series and cross-sectional tests. By forming long-short option portfolios around dividend dates, we can earn significant abnormal portfolio returns.