An adaptive sizing rule changes risk according to recent results. The method below is one example, with clear rules and clear trade-offs, that you can test against plain percentage risk on your own trades.
The rules
Start from a base risk percentage. Measure momentum as the change in equity over the last 10 trades, then scale the base risk up or down. Finally, clamp the result between 0.5% and 5% of equity.
momentum = (equity now − equity 10 trades ago) / equity 10 trades ago
if momentum > 0: risk = base × (1 + momentum × 0.5)
otherwise: risk = base × (1 + momentum × 0.3)
risk is then limited to 0.5% – 5% of equityThe two factors are asymmetric by design: gains raise risk by half the momentum, while losses reduce it by only 30% of the momentum. The constants are choices of this method, not laws of the market.
Worked example
Risk after a good and a bad stretch
- Base risk 2%, equity 10 trades ago €10,000.
- Equity now €11,000: momentum = 1,000 / 10,000 = +0.10, so risk = 2% × (1 + 0.10 × 0.5) = 2.1%, which is €231.
- Equity now €9,000 instead: momentum = −0.10, so risk = 2% × (1 − 0.10 × 0.3) = 1.94%, which is €174.60.
- Both values are within the 0.5% to 5% range, so no clamping applies.
A ±10% equity move changes the risk by +5% and −3% of its size, so 2.1% and 1.94% instead of a flat 2%.
Trade-offs
Strengths
- Risk responds to recent performance without manual decisions.
- The clamp stops risk from becoming extremely small or large.
- Reducing risk after losses can slow a drawdown, as percentage sizing also does.
Weaknesses
- Raising risk after a winning stretch means more is risked after a streak that may be luck.
- The reduction after losses is gentle, so risk falls only slightly in a drawdown.
- Extra parameters (window, factors, limits) can be tuned to past data, which risks overfitting.
- It adds complexity over a plain fixed percentage.
Key takeaways
- The method scales a base risk by 10-trade equity momentum, then limits it to 0.5% – 5%.
- Gains raise risk by momentum × 0.5, while losses cut it by momentum × 0.3.
- It is one example of an adaptive rule, with trade-offs rather than guaranteed benefits.
- Compare it with plain percentage risk on your own trades before relying on it.