As the title suggests, I am currently trying to implement a dual regime-switching options pricing model. In its simplest form, I am fitting a risk-neutral GARCH(1,1) to a crash and normal regime. However, because the volatility in the crash regime is higher, I am finding that the options actually have higher prices in the crash regime. I am wondering how to reconcile this, or introduce a term that provides a negative relationship between returns and vol. But I don't know how to do this, as risk neutral pricing implies the discounted expected value of an option must be the risk-free rate. Thanks!