Statistical testing of proprietary strategies & improvement consulting


It is a common practice for boutique & medium sized trading desks to manually trade on custom proprietary strategies.  Manual trading can be biased or inefficient for the following reasons.

  1. Manually traded strategies are not uniquely identified. Instead, they are based on broad and intuitive market perceptions of the trader in charge. This means that either these strategies cannot be codified and thus cannot be statistically tested or if codified they result in bundling together fundamentally different market states and thus in negative expected returns.
  2. In the highly unlike case that a strategy is well formed, then manually trading is probably inefficient end prone to human error. For a trader cannot be 24/7 alert and mentally fit in order to exploit all the favorable market conditions whenever they arise.

With our experience and infrastructure we offer solutions to both the aforementioned cases.  We statistically test manually traded strategies and also suggest possible improvements. These improvements could be market based, and/or machine learning based.

Leave a Reply

Your email address will not be published.