I will backtest your trading strategy in python with an honest report
About this Gig
Most backtests look great because they ignore costs, peek at future data, or are tuned on the same data they're tested on. I'll test your strategy properly and tell you what the numbers really say, good or bad.
What's included:
- Realistic commission and slippage, plus a harsher stress test
- Train / validation / untouched holdout split, so you see results on data it was never tuned on
- Profit, drawdown, win rate, profit factor and worst day, by year
- A check of whether profits depend on a few lucky days
- Premium: robustness check (do nearby settings still work?) and prop-firm rule simulation
- A plain-English verdict, plus all the Python code
See my sample report in the gallery: a strategy I tested and told the buyer not to trade.
What I need from you:
- Your strategy rules or Pine Script
- Historical data as a CSV (from TradingView, your broker or a data vendor)
Message me before ordering and I'll confirm your data is suitable.
Please note: I report what the data shows. I don't give trading advice or guarantee future results.
Platform:
TradingView
•
Prop firm
•
Custom
Development technology:
Python
FAQ
Does EdgeProof provide the historical data?
No. Please supply your own CSV export from TradingView, your broker or a data provider. I can tell you what format and length work best.
Will you optimise my strategy to make it profitable?
Not by curve-fitting. Premium tests nearby settings to check whether results hold up. Tuning until the past looks perfect usually fails live.
What if the backtest shows my strategy loses money
Then I’ll tell you clearly, with the reasons, before you risk real money. That’s the point of the test.
Can you check it against prop-firm rules?
Yes. Premium simulates trailing drawdown and daily loss limits, so you can see how often it would pass or fail an evaluation.
How much data do I need?
Ideally several years, enough for a few hundred trades. Less data gives a weaker answer, and I’ll tell you how reliable the result is.

