I will validate your trading strategy and detect overfitting
Python Quant Developer I Backtesting and Strategy Validation
About this Gig
Most backtests look profitable for the wrong reasons the strategy was tuned on the same data it was tested on, or the code quietly peeks into the future. I find out which.
I validate trading strategies with the methods institutional desks rely on (López de Prado): Combinatorial Purged Cross-Validation, Probabilistic & Deflated Sharpe Ratio, and a static look-ahead/leakage code audit.
Send me your strategy (rules or code) and trade log. You get back a clear report answering one question: is this edge real, or will it die live?
What you receive:
A PASS/FAIL verdict with the probability your edge is real (PSR/DSR)
Out-of-sample results across multiple independent CPCV paths not one lucky history
Monte-Carlo stress and drawdown analysis
(Standard/Premium) code audit flagging look-ahead bias and leakage
A clean report: executive summary + technical appendix
My open-source validation toolkit (CPCV, Deflated Sharpe, look-ahead audit) is on GitHub with a full test suite see exactly how I work before you order.
Not sure which package fits? Message me a one-line description of your strategy and I'll advise honestly.
Programming language:
Python
Technology:
Excel
•
Jupyter Notebook
Analysis Type:
Statistical Analysis
•
Predictive Analysis
Tools:
Other
Other Data Analytics Services I Offer
FAQ
Do you need my secret strategy code?
No. For the statistical packages a trade log is enough. The code audit (Standard/Premium) needs the source — happy to sign an NDA.
What if my strategy fails the validation?
Then you've saved real money. A clean FAIL with the reasons (overfitting / leakage) is a valid, valuable result — that's what you're paying me to find.
Can you build or fix the strategy too?
My focus is validation, not signal-selling. If fixes are needed I can scope that separately, but the audit itself stays unbiased.

