We ran the exact opposite of our strategy — the account zeroed out

Most strategy sellers only show you the nice equity curve. We do the opposite: we took the exact inverse of our strategy and ran it on the same data. The logic is simple — if inverting it makes money, then our real signal points the wrong way, or there's a leak in our test rig.

Result

Out-of-sample (data the model never saw):

• Strategy: Sharpe +1.31
• Its inverse: Sharpe −6.42 — the account effectively zeroed out

This proves two things at once: our signal is real and pointing the right way, and our test rig is clean (a hidden lookahead would have made the inverse look like a "winner" too). We also tried random entry on the same data — it lost as well, at −1.19 Sharpe. So the strategy beats random by a wide margin.

That's why our promise is "proof, not a promise": instead of polishing a number, we stress-test our own claim.