Seven checks, run in order, on whatever record you have. You get a written answer and all the working behind it. The first system I pointed this at was my own, and it didn't survive.
Not deliberately. You tune a parameter, the curve gets nicer, and you have quietly fitted the strategy to the last three years. The hard question is whether you are looking at an edge or at a period of luck that happened to suit you, and it is much cheaper to answer that before you fund it.
I don't sell signals and I don't sell courses.
I tell you whether the thing works.
Any one of them can end it. If your strategy dies at check two I'll tell you at check two, rather than running the other five so the report looks thicker.
Against the boring alternative, which is usually just holding the thing. Beating zero is not the bar.
Results split by market condition. Plenty of strategies are really just a bet that last year repeats.
Where trades actually closed against where your rules said they would. These disagree more often than people expect.
Whether your test universe quietly drops everything that went to zero. This one flatters almost every backtest it touches.
What is left after fees, spread and fills you'd actually get. Thin edges usually die right here.
Whether you have enough trades to tell skill from luck. Forty trades is a story, not evidence.
What I found, which check decided it, and the working attached so you can argue with me.
The rules in enough detail that I can reproduce them, plus whatever record you have: backtest output, live fills, or both. If the rules only live in your head, we'll write them down first. That step alone finds problems surprisingly often.
No. The first system I pointed this at was mine, and it failed. It's still the top row on the ledger above. An audit you can only pass is worth nothing to you.
No. Your rules stay confidential, I don't trade systems I've audited, and I don't name clients without asking. That independence is the whole product.
You get the answer and the check that produced it, which is often more useful than a pass. It tells you whether the idea is wrong or only the implementation. Failing on costs or sample size is usually fixable.
Depends on the job. One strategy with clean data is a different piece of work from a full book and fifteen years of messy records. Tell me what you've got and I'll come back with a number.
The strategy, the market it trades, and what record you have of it. I'll tell you honestly whether it's worth auditing. Sometimes the answer is that you already know enough.
I read every message myself. If it isn't a fit, I'll say so.