QuanterLab is a quantitative research platform built for rule-based methodology tests. We deliver complete validation and test packages, with audited results and the records behind them.
For asset managers, index builders and ETF sponsors
Method
A firm that sends a strategy out for review usually expects the reviewer to re-run the backtest. The backtest reproduces. What fails is the part of the rule that was fitted to the sample it was tested on, and that shows only when the rule meets data it has not seen. So the work starts at walk-forward construction: the rule is registered at an anchor date, sealed, scored on the window that follows, and moved forward one window at a time, every step recorded before it is scored. Twenty windows of that is the standard shape of a study here.
Deliverable
A report in the form the published studies take: the question, the sealed design, every window's result, the exclusions printed, and the frozen artifact behind it: parameters, windows, data vintage, code path. A validation report that cannot be re-run is an opinion. This one can be re-run on the platform by anyone you give access to. The report is private to you; it appears on this site only if you ask for that.
Scoping
Scoping comes first and looks at the data before anything else. Some markets have point-in-time constituent lists and some do not, and that one question usually decides whether a study takes two weeks or cannot be done at all. You hear which it is, with a fixed quote, before you commit to anything.
Contact
What the rule is, what it runs on, and what you need to know about it. Replies come from Serhat Girgin, who writes the studies. Or write to contact@quanterlab.com.
Serhat Girgin, QuanterLab