To view this content you must be an active member of the CMT Association.
Not a member? Join the CMT Association and unlock access to hundreds of hours of written and video technical analysis content, including the Journal of Technical Analysis and the Video Archives. Learn more about Membership here.

Can you quantify the likelihood your trading system will continue to perform well in the future? Do you think Monte Carlo simulation, walk-forward testing, or out of sample testing can provide the answer? In this session, Mr. Walton discusses why common system validation methods often provide dangerously misleading results and then introduces an alternative called stochastic modeling, a process commonly used by insurance companies to price insurance policies and other financial products. When properly applied to trading systems, stochastic modeling can overcome common development challenges including data mining bias, stateful luck, buying power limitations, and position sizing / system dependencies.