Finance Seminar by Pascal François (HEC Montréal)
December 2018, Monday 3 (2:30 pm) - N1 – Room 1701
Invoking the scale invariance property used by most option pricing models, Bates (2005) presents a model-free method to compute the delta and the gamma from the shape of the volatility smile. That promising approach yields hedging strategies which solely rely on market information and whose performance had, to the best of our knowledge, yet to be tested on a large-scale sample. That is what this paper does using roughly 6 million daily hedges of S&P500 calls and puts from January 1996 to April 2016.
Our first major finding is that smile-implied delta-neutral and delta-gamma-neutral hedging strategies cannot outperform their Black-Scholes counterparts neither in a mean-variance framework, nor with a tail risk metric.
In light of these disappointing results, we turn our attention to managing volatility-related Greeks. We extend the relation that holds in the stochastic implied volatility model to build smile-implied vegas, vannas and volgas from smile-implied gammas. We find that the smile-implied delta-gamma-vega neutral hedging strategy strongly outperforms its Black-Scholes counterpart, especially for put options. Adding the vanna or the volga into the hedging strategy does not improve the quality of replication. All in all, our findings suggest that option market information is particularly worthwhile for managing volatility risk, not underlying price risk.
P. François (HEC Montréal) & L. Stentoft (Western University)