AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
How to Use Permutation Tests in Strategy Development Processes
There are statistical techniques to get around this problem of the out of sample trades not being a real legitimate predictor of what you're going to get in the future, but rather being a random sample. So that's one way. But another thing that we often want to do is come up with probabilities for drawdowns. That will give you reasonably valid probability estimates for extreme drawdowns. If you have to do a thousand subsamples of a thousand samples, right there you're talking about a million replications. And now you're getting into some serious computer time. On the other hand, if that's what it takes to get reliable estimates of the probability of catastrophic drawdowns, then it might be