Simulated testing of nonparametric measure changes for hedging European options

We test the accuracy and hedging performance of the deltas given by a range of nonparametric measure changes. The nonparametric models accurately estimate deltas across a number of asset price dynamics. The optimal nonparametric measure change displays superior estimation bias, which depends on how the models capture the stylised features of the dynamics, moneyness, and time-to-expiry. Differences in estimation error appear negligible. The optimal measure change produces superior static hedging outcomes compared to the Black–Scholes model. Differences in dynamic hedging outcomes are negligible.