Find out if your edge is real.
Most backtesters sell you a beautiful equity curve. This one tells you the truth. Describe a strategy in plain English, run it through a look-ahead-safe simulation, and get an honest verdict: held-out testing, walk-forward, Monte-Carlo and a Deflated Sharpe that penalizes every variation you tried. The quant rigor, shown simply.
The Lab is in early access. Browse the example freely; the live builder and runner unlock with an access code.
A bar-by-bar simulator that can only see the past. Costs, slippage and risk limits are modeled, so the curve is what the rules would actually have done, not hindsight.
Every run holds out data it never trains on, walks the strategy forward through time, and reports a Deflated Sharpe penalized for the number of variations tried.
A Monte-Carlo fan shows the luck around the result, breakdowns split it by regime, session and symbol, and a plain-English verdict says whether the edge holds up.
From an idea to an honest verdict
No code required. Compose from a library of researched modules, or just write what you mean. The same engine that prizes apart edge from noise runs underneath either way.
- 01Describe it in plain English
Write your idea as a sentence, or assemble it from the module library with no-code blocks: universe, session, location, setup, risk and exits.
- 02Run the honesty suite
One run gives you the simulation, the held-out test, walk-forward, Monte-Carlo and the overfitting math: no peeking, no survivorship, no cherry-picked window.
- 03Read the verdict
Approved, tuning, or did not pass, with the three reasons why, the headline metrics, and the candle chart with your entries and exits marked.
A verdict, not just a curve
SAMPLE data- Deflated Sharpe
- 0.31
- Sharpe (naive)
- 1.74
- Prob. of overfit
- 78%
- Trials run
- 240
- CAGR
- +12.4%
- Max drawdown
- -18.6%
Deflated Sharpe and probability-of-overfit penalize the 240 variations tried: the math that turns a pretty curve into an honest verdict. A high naive Sharpe next to a low deflated Sharpe means the edge is most likely noise.
Hypothetical performance disclaimer (CFTC 4.41)
HYPOTHETICAL PERFORMANCE RESULTS HAVE MANY INHERENT LIMITATIONS, SOME OF WHICH ARE DESCRIBED BELOW. No representation is being made that any account will or is likely to achieve profits or losses similar to those shown. In fact, there are frequently sharp differences between hypothetical performance results and the actual results subsequently achieved by any particular trading program. … Simulated trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profits or losses similar to those shown.
SAMPLE data shown for illustration. No live engine is connected on this preview.
This is the honesty panel that accompanies every result: the verdict, the deflated metrics, and the hypothetical-performance disclaimer. See the full example result (equity curve included), or unlock the builder to run your own.