Risk management is the part of a forex process that determines whether you can survive the gap between an idea and reality. The goal is not to avoid every loss. It is to prevent a normal sequence of losses, slippage or changing volatility from becoming catastrophic.
Position sizing from the stop—not from conviction
Define the price at which the trade idea is invalid before calculating size. Then use:
Risk amount = account equity × chosen risk percentage
Position size = risk amount ÷ (stop distance × value per price unit)
Worked example
Account equity: €10,000. Planned risk: 0.5%, or €50. Stop distance: 40 pips. If the instrument’s pip value is €1 per micro-lot-equivalent unit, the maximum size is €50 ÷ (40 × €1) = 1.25 units of that size. The final platform quantity must be rounded down and checked against the broker’s contract specification.
This is an illustration, not a recommended risk level. Include commission and an allowance for slippage when the risk limit must be strict.
Why a fixed lot size is unstable
A fixed lot size produces different monetary risk when volatility or stop distance changes. Volatility-scaled sizing—such as using Average True Range to help define the stop—can make risk more comparable across trades. It does not make the underlying signal profitable.
Drawdown mathematics
Recovery is asymmetric. A 10% loss requires an 11.1% gain to return to the starting balance; a 50% loss requires a 100% gain.
| Drawdown | Gain required to recover |
|---|---|
| 5% | 5.3% |
| 10% | 11.1% |
| 20% | 25% |
| 30% | 42.9% |
| 50% | 100% |
Small per-trade risk does not eliminate drawdown because losses can cluster. If several positions depend on the same US-dollar move, they may be one large exposure disguised as separate trades.
R-multiples and expectancy
Let 1R equal the amount planned to be lost if the initial stop is executed. Results can then be compared across different position sizes: a +2R trade earned twice the planned risk; a −1R trade lost the planned risk.
Expectancy = (win rate × average win in R) − (loss rate × average loss in R).
A system can have positive expectancy with more losing trades than winners if the average win is sufficiently larger. Conversely, a high win rate can hide rare, very large losses. Estimate expectancy after fees and with enough independent observations to show uncertainty.
Do not optimise away uncomfortable history
Backtests often look strongest after many parameter choices. That raises the probability that the rules fit historical noise. Keep the model simple, reserve untouched out-of-sample data and stress assumptions for wider spreads, slippage and delayed fills.
Practical limits
- Per-trade limit: cap the planned loss before entry.
- Aggregate currency limit: add exposures that depend on the same currency or macro event.
- Daily/weekly stop: stop opening trades after a predefined loss or process breach.
- Event rule: decide whether positions may remain open through central-bank decisions, inflation data or employment reports.
- Leverage limit: use a personal limit below the broker maximum.
- Operational reserve: keep enough free margin for adverse movement without relying on additional deposits.
European product-intervention rules require retail CFD protections such as leverage limits, margin close-out and negative-balance protection. They are a safety boundary, not a trading plan. ESMA’s 2026 risk review reiterates that leverage amplifies market risk and costs and increases the likelihood of forced exits.
Pre-trade checklist
- Is the setup allowed by the written rules?
- Where is the idea objectively invalid?
- What is the monetary loss at the intended size, including costs?
- How much correlated exposure is already open?
- Is a scheduled high-impact event inside the holding period?
- Are spread, liquidity and financing conditions normal?
- What is the exit rule for profit, loss and time?
- Will the trade be logged regardless of outcome?