Platform · Deploy

Strategy Health

Edges decay. A strategy exported six months ago is quietly rotting on someone's terminal, and nothing tells them. This does.

What you get
Live measured against its own backtest

The comparison is to the result that sold you the strategy — never a fresh re-run, which would move the goalposts every time we add history.

Warned once, when it crosses

A drift alert fires on the transition, not on every check. A warning repeated daily stops being a warning.

It won't cry wolf

Under a meaningful number of forward trades it says so rather than grading. A handful of paper trades cannot tell decay from variance.

Better variants are shown to you

When the evolution loop seals an improvement it runs as a paper challenger on the desk, and you see it in the app under Strategy Performance → Evolved. You promote it from the desk. Nothing changes a strategy you are running without you.

Whatever you build here is tested against real market conditions and graded out-of-sample, so you always know whether it's working — on your own strategy, not a marketing figure. How the proof engine grades.

What we don't publish

We don't publish the health score, its threshold, or how the drift comparison is constructed.

We publish what a module does and what it's for. We don't publish how it works. The method is the product — and anyone holding the method holds the product.

Evaluating this for a desk? Talk to sales and we'll go as deep as an NDA allows.

Worth knowing

Health is a signal to look, not a verdict. Markets change, edges decay — that is the job, not a failure.

How it works
  1. 01
    Compare live against the tested baseline

    A deployed strategy's realised behaviour is measured against what the backtest said it should do — not against whether it made money last week, which is noise at small sample sizes.

  2. 02
    Drift is a signal, not an alarm bell

    Divergence is reported with its sample size and its size relative to normal variation, so a bad fortnight inside expectations does not read as a broken system.

  3. 03
    It tells you; it does not act

    Health reporting never disables, resizes or retires anything. The decision stays with you, because the cost of a wrong automatic retirement is paid by you.

Worked example
Why a bad fortnight is not a verdict

The evidence bar a strategy must clear before its live record is treated as saying anything.

Trades before a promotion is judged
25 minimum
What a 6-trade drawdown proves
inside normal variationnothing yet
What the radar does at trade 6
reports, with the sample size

Small samples produce dramatic-looking numbers in both directions. Stating divergence alongside the evidence behind it is what separates an early warning from a false alarm.

What it refuses to do

Every one of these is a thing we could ship and choose not to. They are here because the limits are the part of a research tool you actually have to trust.

It won't turn your strategy off

Drift is reported. Acting on it is your call.

It won't cry wolf on a small sample

Divergence is stated with the evidence behind it.

Where strategy health sits
Describe → Build → Prove → Train → Paper → You arm liveevery improvement re-proves before it counts1DescribePlain English, a pastedscript, or a chart photo2BuildA runnable strategy —one shape everywhere3ProveBacktested and gradedout-of-sample4TrainThe Conscious works it,condition by condition5PaperA real forward record,no money at risk6You arm liveOff by default. Only youcan turn it on

One pipeline, one strategy shape end to end. What you backtest is byte-for-byte what papers and what exports — there is no re-implementation step where drift can hide.

Questions people actually ask
What is decay, in plain terms?

A strategy behaving differently live than it did in testing. Markets change, and an edge that was real can stop being real — the radar's job is to tell you early and with enough context to judge, rather than after a drawdown has made it obvious.

Does the platform stop a strategy that is decaying?

No. It reports the divergence and the evidence behind it. Nothing here switches your trading off or on for you.

How is drift different from a normal losing streak?

A losing streak sits inside the variation the backtest already predicted; drift is behaviour outside it, sustained, on enough trades to mean something. The radar reports both the divergence and the sample behind it so you can tell which one you are looking at.

How soon will I know?

As soon as there is enough evidence to say anything — which is deliberately not the same as immediately. An alert fired on three trades would be noise, and noise is how real warnings get ignored.

Also in deploy

Bring a strategy you already trade.

Test it yourself — or bring your desk's questions to us.

No card required · sales is for desks and teams