Definition and purpose of risk of ruin in forex
Risk of ruin describes the likelihood that a trading account will reach a predefined “ruin” level (for example, near-zero equity or a stop level where the strategy cannot continue). It is a probability concept, not a measure of profitability. In a forex context, you model a sequence of trades and compute how often the account could drift downward to that ruin boundary.
A key limitation is that forex market outcomes are not guaranteed to follow the simplified assumptions used by the calculation. Therefore, risk of ruin should be treated as a model-based estimate under specified conditions, not as a promise about future results.
Core mechanics: inputs you need
To calculate risk of ruin, you typically need a model of trade outcomes and how they affect account balance. The most transparent inputs are:
- Ruin boundary and starting point
- Define the current account size (or starting equity).
- Define the “ruin” threshold (e.g., a fraction of starting equity).
- Trade-level payoff distribution (simplified) Many formulas use a discrete approximation: each trade either wins or loses by fixed or effectively fixed amounts.
- Let p be the probability of a winning trade.
- Let q = 1 − p be the probability of a losing trade.
- Let b represent the average win size relative to average loss size (often called a payoff ratio in simplified models).
- Step size / bet sizing rule A model must specify how trade results translate into equity changes. Common simplified setups assume a constant fraction risk per trade or a constant amount per trade. If the bet size changes with equity, the model has to incorporate that rule.
A common simplified model (constant win/loss steps)
If you assume:
- each trade moves equity by a constant “up” amount when winning,
- and by a constant “down” amount when losing, then risk of ruin can be approximated using standard probability-of-boundary-hit calculations from random-walk style models. The exact formula depends on how you map p, payoff ratio (b), and the number of steps implied by the starting equity vs. ruin boundary.
If you want a practical calculation approach without hidden complexity, the workflow is:
- Convert the ruin boundary and starting equity into an equivalent number of losing steps (or net steps) under your bet-sizing assumption.
- Use the win/loss probabilities and payoff ratio to estimate the probability that the cumulative sequence ever reaches that negative boundary.
Example calculation approach and independent checks
Because detailed formulas vary by assumption, a reliable way to proceed is to keep the calculation consistent with your own model choices and then test sensitivity.
Example setup (conceptual)
- Choose a ruin threshold (for instance, an equity level you consider unable to continue).
- Choose a bet-sizing rule (e.g., fixed fraction of equity or fixed per-trade risk).
- Estimate p from historical data or from a stated model. Use the same definition of a “win” as the one used to estimate p.
- Determine a payoff ratio b from the simplified win/loss magnitudes.
Then compute a ruin probability using the boundary-hit method appropriate to that simplified random-walk model.
Checks that do not require future data
- Consistency check: If you increase the ruin buffer (move the ruin threshold farther away), the ruin estimate should generally decrease under the same model assumptions.
- Direction check: If p increases (holding payoff ratio and step sizing constant), the ruin probability should typically decrease.
- Sensitivity check: Recalculate using slightly different plausible values for p and payoff ratio. If small input changes cause large swings in ruin probability, the estimate is fragile.