Research

Dozens of firm characteristics predict the returns on delta-hedged equity options, with different explanations for different characteristics. I show that most of them are the same anomaly. Their long-short portfolios lose money because they hold options on volatile stocks, and options on volatile stocks are repriced every month when the standard contracts expire and positions roll into the next maturity. Implied volatility alone accounts for four fifths of the average long-short return, and for nearly all of it on the two rollover days; among stocks of similar volatility, the idiosyncratic-volatility and maximum-return anomalies point the other way. What the expiration weekend costs the holder of a contract that does not expire, beyond what any weekend costs, is not time decay or realized variance but a missing recovery in implied volatility, the rise that every other weekend shows and this one does not; it grows with volatility and with the open interest that rolls. The anomalies are not separate puzzles but different exposures to one recurring repricing, and each portfolio's exposure is known a month ahead, from which stocks are volatile and when the contracts expire.

Using option-implied volatility to measure firms' exposure to the FOMC statement and to the Chair's press conference separately, I find that only press-conference exposure is priced in the cross-section of announcement-day returns. Firms most exposed to the press conference earn 108 basis points per event more than the least exposed since Chair Powell's arrival in 2018, while statement exposure is not priced beside it. When the statement stands alone, exposure to it is priced. The premium is mostly earned in the hours before the Chair speaks, and on both hawkish and dovish news. The firms earning it are the fastest-growing. How the Federal Reserve explains a decision, and not only what it decides, shapes which firms bear priced monetary-policy risk.

The cross-strike geometry of the implied-volatility smile predicts 30-minute equity option returns. Smiles that are more sharply curved or show larger cross-strike deviations are followed by higher returns; at-the-money options priced rich relative to the fitted smile mean-revert. Yet the same signals carry no information about the underlying stock over the same horizon, and a return decomposition rules out a dominant underlying-move channel. The spreads survive controls for past option returns and standard liquidity proxies. They concentrate in the final half-hour of the session and are largest in stocks where open positions cluster at a small number of strikes, consistent with rebalancing of concentrated inventory toward the close. The cross-strike dimension of the option market has its own short-horizon dynamics, distinct from the behavior of the underlying stock.

Dispersion strategies capture the difference in variance dynamics between a basket and its components. Even though smart-beta indices intend to load heavily on a particular factor, factor dispersions based on such baskets are exposed to risks of other factors and idiosyncratic variances. Analyzing factor dispersions through a linear factor model and equicorrelation representations, we recover driving forces behind dispersion dynamics and work out an attribution of a dispersion risk premium. As a balanced combination of systematic and idiosyncratic variance components, dispersion and its risk premium provide signals about future changes in systematic and alpha-based investment opportunities.