Is your backtest real, or luck?

Upload the results of a trading strategy, yours or one you are about to buy, and get a verdict in seconds: it holds up, it doesn't, or the data can't tell. Same checks I used on my own AI trading bot. Your file never leaves your browser.

Check a backtest Free. No sign-up.

Check a backtest

Everything runs in this page. The file is not uploaded or stored anywhere.

Four ways a backtest fools you

Hindsight

I rerun the strategy on random sets of stocks or periods. If the result depends on what was picked, it shows.

Real costs

Minimum commissions, spreads and slippage, at the capital you will actually trade. Small accounts often lose here.

Unseen years

How much of the result survives outside the period the strategy was built on.

Hidden attempts

How many variants were tried before the one you see, and how much that inflates the numbers.

What the check gives you

  • One clear verdict: holds up, doesn't, or can't tell.
  • The numbers behind it, with costs spelled out.
  • The reasons behind it, in plain words.

What I don't do

  • Tell you what to buy or sell.
  • Manage money or connect to your broker.
  • Promise returns. I judge the method, not the investment.

The full check, done by me

  1. Send the filesThe strategy rules, the backtest and the data or code you have.
  2. Get a quoteAfter a first look I tell you what I can check and the fixed price.
  3. Read the verdictAll four checks, including the hindsight test, with the numbers and the code.

Most backtests I would expect to fail at least one of these checks. Better to find out before real money does.

This is a review of research method, not investment advice. Past results, including mine, do not predict future returns.