A document built to be disputed.

1. 01VerdictOne result, its reason, and what happens next.

   The first page answers the question that was asked, in one line, in plain language. It states which of the three results applies, why, and what the next step is. Nothing later in the document contradicts it.

2. 02Validation Scope, confirmedThe exact claim, asset, timeframe, costs and benchmark you agreed to.

   The scope you confirmed before anything ran, restated verbatim: the strategy version, the asset, the timeframe, the period, the cost assumptions and the benchmark. It sits directly under the verdict because everything below is only valid inside it.

3. 03Executive summaryWhat was found, in the length of a page.

   What the strategy does, what was measured, what held and what did not — written for someone deciding rather than someone auditing. The detail is below; this is the part that survives being forwarded.

4. 04Key metricsRecomputed from the frozen inputs, reconciled against what you declared.

   Return, volatility, Sharpe, maximum drawdown and turnover, recomputed from the frozen material rather than copied from your report. Where a recomputed figure differs from the declared one, both are shown and the difference is explained.

5. 05Benchmark comparisonLike for like, against the comparison agreed at scoping.

   The strategy against the benchmark you confirmed, over the same period, with the same costs and the same exposure basis. Without a fair comparison an improvement cannot be attributed to the strategy rather than to the market it was in.

6. 06Classic Monte CarloHow the same edge could have played out along other plausible paths.

   Alternative paths generated while preserving part of the observed time dependence, so drawdown, time under water and losing streaks can be read as ranges rather than as the one number that happened.

7. 07Permuted Monte CarloThe same returns, reordered — how much of the risk was sequence luck.

   The observed returns kept exactly, their order changed. If the risk profile falls apart under reordering, the historical experience depended on a sequence rather than on an edge.

8. 08Drawdowns and time under waterNot only how deep, but how long.

   Every drawdown with its depth and its duration. Depth is what gets quoted; duration is what decides whether a strategy is actually held long enough to recover.

9. 09FindingsWhat was found, located precisely enough to act on.

   Each finding names what was checked, what was observed and why it matters, with the place in the material where it applies. A finding that cannot be located is an opinion.

10. 10LimitationsWhat this validation could not establish.

    Gaps in the data, assumptions that had to be made, parts of the claim outside the agreed scope. A limitation is only acceptable if it is specific and written down, which is why this is a section rather than a sentence of general caution.

11. 11RecommendationsWhat would have to change, in priority order.

    Concrete, ordered by how much each one would move the result, and written so you can judge whether meeting them is worth the effort before committing to it.

12. 12Next stepWhat to do with this, stated explicitly.

    Proceed to unfunded shadowing, revalidate after specific changes, or supply the missing material. The report never ends on an ambiguity about what happens next.

Sample report

I cannot show you a client's report. So I applied the process to a strategy of my own, published openly, and the verdict was that it does not work. Every figure below comes from that validation, and anyone can clone the repository and check it.

The full report

This summary is a view of the page. The document a client receives is longer: it explains what the strategy does, what it is compared against and why, and what falls outside the scope. Download it and judge it yourself.

Download the report (PDF)3 pages · PDF

View the repository on GitHub

And an evidence annex

Alongside the report comes an annex holding the series, metrics and figures behind every number: equity curve, drawdowns with their duration, rolling metrics, return distribution, and integrity fingerprints for the input data. The report is what you read to decide; the annex is what holds up when someone disputes a figure.