🚦 Edge Health Monitor
Everything else in AniQuant looks backwards (backtest, Survival, Walk-Forward…). This module looks forwards: once you deploy a strategy, it watches whether it is still alive or whether its edge is fading. It is the link between the lab and real trading.
You get in through 💰 Capital & Risk → Edge Health Monitor. It works on the strategies of a Portfolio, all at 1 contract (it measures the edge, not the leverage).
Deploying = freezing the footprint (starting the clock)
You set a deployment date and the module freezes the strategy's reference footprint (its statistics as of that date). From then on, the trades that close afterwards are the forward — the real out-of-sample acid test. The numbers are frozen like a photograph: in futures the history shifts with every roll-over, so they are not recomputed.
Expectancy per trade and its deviation, profit factor, historical drawdown, the 95th percentile of the drawdown (Monte Carlo), the worst losing streak and the AQ Confidence.
Set a deployment date in the past: the later trades in your history already count as forward and the traffic light fills up instantly.
The traffic light: is it still alive?
When you check health, the current series is re-tested, the forward trades are isolated and compared against the footprint. The verdict combines three signals:
- ·z-score of the mean — how many deviations the forward mean per trade sits away from the expected one. Very negative = performing below expectation (the main deterioration signal).
- ·DD ÷ Profit (pain ratio) — the forward drawdown relative to the profit. It is scale-fair: it does not punish a very profitable run for having large swings in dollars. (The absolute «DD/p95» is kept purely as information.)
- ·Losing streak — the worst forward streak against the historical one.
In line with (or above) expectation.
Mild deterioration: keep a close eye on it.
Outside expectation: review it or retire it.
No later data yet (or too little).
Forward equity vs the expected band
Select a strategy and you see its real forward curve (green) against what was expected (expectancy × trades, dotted gray) and the lower 95% band (red). If the green sinks below the red, the strategy is performing worse than chance alone would suggest — the visual signal that the edge is fading. The grid is sortable, with the selection in orange and keyboard navigation, like everywhere else in AniQuant.
No edge lasts forever. The backtest tells you it existed; this monitor tells you whether it still exists. Deploying a strategy is not the end of the analysis: it is the beginning of watching it — so you can retire it in time, before it eats what you made.
AniQuant can be tried free for 30 days, with every module and no card.