Losing Streak Probability Calculator
Explore the exact probability of at least one entered-length losing run within a finite trade horizon, conditional on one assumed unchanged win rate and independent outcomes. This is assumption arithmetic, not a forecast, risk grade or position-sizing tool.
Enter finite-horizon assumptions
The model does not estimate or verify the win rate.
A breakeven must be treated consistently as either a loss or a non-loss before entering this binary assumption.
Conditional run probability
Finite-Horizon Losing-Run Probability 1.0.0.
Out of date — recalculate after changing an input.
How losing-run probability is calculated
The model keeps one state for each possible current loss-run length below the target. A non-loss returns probability mass to state zero, a loss advances it, and the transition that reaches the target enters an absorbing hit total.
State j next = Loss probability × previous state j-1
Run probability = accumulated transitions into the target run
This is exact finite-state arithmetic up to floating-point precision, not Monte Carlo simulation. Hit probability is accumulated directly and checked against the remaining no-run state total.
Worked example from the audited fixture
With a 60% assumed independent win rate, the entered loss probability is 40%. Across a finite horizon of 10 trades, the exact recurrence gives a 31.412224% probability of at least one run of three consecutive losses and a 68.587776% probability of no such run.
How to interpret the result
The probability that one particular three-trade block is all losses is 0.4³ = 6.4%, but the 10-trade horizon contains multiple overlapping places where the run can occur; simply multiplying 6.4% by the number of blocks would double-count overlaps. The 31.412224% result is conditional on the unchanged 40% loss probability and independence assumptions—it does not say when a run will occur or that losses are due.
Assumptions and limits
- Trials are independent and use one unchanged binary loss probability.
- The page does not verify the entered probability, dependence, regime changes or clustering.
- No account balance, payoff, position size, drawdown, recovery or ruin floor is modeled.
- The output is not the probability that the next trade loses and does not mean a loss is due.
- No result is a real-world forecast, safety threshold, expected maximum streak or recommendation.
Frequently asked questions
- It calculates at least one run of the entered number of consecutive losses anywhere within the entered finite trade horizon.
- No. It uses a deterministic finite-state recurrence and returns the same result for identical inputs.
- The model is binary. Define breakevens consistently before entering the win rate; a non-loss resets the loss run.
- The exact probability is zero because that many consecutive losses cannot fit inside the entered number of trials.
- No. Risk of ruin needs a monetary path, payoff, position-risk and account-floor assumptions. This route models only binary run occurrence.
- No. The model contains no timing claim and does not change the assumed probability of the next independent trial.
- No. The win rate is a user-entered assumption and is not derived from Trade Journal or broker records.
- Not in version 1.0.0. Dependence, changing probabilities and regime shifts are outside the model.
Sources and methodology
- NIST/SEMATECH — Binomial Distribution — constant-probability independent-trial assumptions.
- NIST/SEMATECH — Runs Test — definition and relevance of runs in ordered binary data.
- CFTC — Trading System Claims Advisory — hypothetical-performance limitations.
Continue the sequence-risk workflow
Verify the records behind an assumption
If an assumption comes from account history, first confirm complete records, applicable charges and account terms.
Risk and affiliate disclosure: Leveraged forex and CFD trading can result in substantial losses. These are affiliate links, so ForexMT4Indicators.com may receive compensation if you register or trade through them, at no additional cost to you. Availability and terms vary by jurisdiction and broker entity.