Home › Documentation › Portfolio, capital & risk
Money management

⚖️ Confidence-based Allocation

You have several good strategies and a limited risk budget. How much do you give each one? This module splits your risk in proportion to each strategy's confidence: more to the robust ones, less to the doubtful ones, and nothing to those that do not clear the bar.

You get in through 💰 Capital & Risk → Confidence-based Allocation. You choose a portfolio, tick the strategies, and the module re-tests them, estimates their AQ Confidence (from their trades alone) and works out the split.

🎯

The AQ Confidence (0..100)

A robustness grade computed from the trades alone — a quick, honest proxy. It combines five ingredients:

Sample

Number of trades: the more there are, the more reliable the evidence.

Regularity

The share of positive months: that it wins often, not off one big score.

Consistency

That both halves of the history win by similar amounts (not everything at the start).

Risk

Ret/DD: profit for every unit of drawdown.

Edge

Profit Factor: how much it makes for every dollar it loses.

It is summarized as an A / B / C / D grade. It does not replace the Survival Test or the Walk-Forward — for the final word on one specific strategy, put it through those.

💰

How it splits the risk

You set a total risk budget (heat) — a % of capital — and it is split among the strategies that clear your minimum confidence, in proportion to their confidence. Optionally it penalizes correlation: the ones that move together weigh less, so the portfolio does not end up depending on a single engine.

Capital and heat

Account capital and the total % to put at risk (6%, say). Total risk = capital × heat, and that is what gets split.

Minimum confidence

Below that grade, the strategy is excluded — we do not put capital where the evidence is thin.

Penalize correlation

It lowers the weight of highly correlated strategies (Pearson on their monthly P&L).

🏁

What you see

  • ·Allocation per strategy — a table with Confidence (grade), Trades, Ret/DD, Corr., Weight, Risk %, Risk $ and Contracts. It sorts by any header, the selected row is highlighted in orange and you move with the ↑/↓ arrows. Every header carries its explanation on hover.
  • ·Risk allocation — bars with the % and the contracts of each included strategy.
  • ·Summary — total risk, strategies included, average confidence, average correlation and total contracts.
  • ·A reading in plain language — the verdict and the warnings (what was excluded and why, or which strategy does not even cover 1 contract).
About the Contracts: Contracts = the $ risk allocated ÷ the risk per contract (rounded down). If it comes out as 0, that strategy's budget does not cover even 1 contract: raise the capital or the heat, or include fewer strategies. On strategies with no stop, the "risk per contract" is estimated from their worst historical loss (a conservative criterion), which tends to be large — so 0 is easy to hit there.
💡
The idea behind it

Do not risk the same on something you have watched work 500 times as on something you have barely tested. Confidence-based allocation puts your money where the evidence is, and leaves out whatever has not earned it yet.

Try it yourself

AniQuant can be tried free for 30 days, with every module and no card.

← Previous
Position sizing
Next →
Edge health monitor
More in Portfolio, capital & risk