Risk of Ruin: How Traders Can Survive Losing Streaks
Last saved:
Executive Summary
- ✦Ruin is the point where the trading plan can no longer operate, not necessarily a zero balance.
- ✦A positive expectancy does not prevent long or clustered losing streaks.
- ✦Risk per trade determines how severely the same streak damages equity.
- ✦Size from a logical stop and stress the whole portfolio before risking capital.
Risk of Ruin: How Traders Can Survive Losing Streaks
A strategy can have positive expectancy and still suffer a long losing streak. Risk of ruin asks a more useful question than “How much can this trade make?”: Can the account remain tradable long enough for the edge to appear?
বাংলায় পড়ুন: Risk of Ruin: Losing Streak থেকে Trading Account কীভাবে বাঁচাবেন
Learning objective: define a practical ruin threshold, estimate the effect of losing streaks, choose position risk from the account’s survival constraint, and know when a model is too optimistic.
1. Ruin does not have to mean zero
For a leveraged trader, “ruin” often arrives before the balance reaches zero. Define it as the point where the plan can no longer operate. Examples include:
- a 25% or 30% maximum tolerable drawdown;
- falling below a broker’s margin requirement;
- losing enough capital that the minimum position is too large;
- reaching a pre-agreed stop for a funded account;
- suffering a drawdown that makes the trader abandon the rules.
Write the threshold in money and percentage terms. An undefined threshold cannot control position size.
2. Expectancy is only the starting point
Express each outcome in R, where 1R is the amount planned to be lost if the stop is reached. A simple pre-cost expectancy is:
Expectancy = (win rate × average win in R) − (loss rate × average loss in R)
Suppose a strategy wins 45% of trades, its average winner is 1.5R and its average loser is 1R:
(0.45 × 1.5R) − (0.55 × 1R) = +0.125R per trade
That is a positive estimate, but 55% of trades still lose. The sequence can contain several losses together. Fees and slippage also reduce the estimated edge.
3. A losing streak is not proof the edge disappeared
If losses were independent and the estimated loss probability were 55%, the probability that one specified block of eight trades all loses would be:
0.55⁸ ≈ 0.84%
Across 100 trades there are many possible overlapping eight-trade blocks, so the probability of seeing such a streak somewhere is much higher than 0.84%. The exact answer depends on the model. A Monte Carlo test can shuffle or simulate thousands of trade sequences and report the distribution of the worst streak and maximum drawdown.
Do not assume independence blindly. Trend strategies may lose repeatedly in choppy markets; short-volatility strategies may cluster losses during a shock. Regime dependence and changing volatility can make historical simulations too comfortable.
4. Worked example: 1% risk versus 5% risk
Consider a £10,000 account and ten consecutive full-R losses. Position risk is recalculated from current equity after every trade.
| Risk per trade | Equity after 10 losses | Drawdown | Gain needed to recover |
|---|---|---|---|
| 1% | £9,043.82 | 9.56% | 10.57% |
| 2% | £8,170.73 | 18.29% | 22.39% |
| 5% | £5,987.37 | 40.13% | 67.02% |
The arithmetic is starting equity × (1 − risk fraction)¹⁰. The strategy and streak are identical; position sizing changes whether the trader is bruised or close to failure.
This is a simplified illustration. Gaps, slippage, fees and correlated positions can make the realised loss larger than the planned R.
5. Recovery becomes harder as drawdown grows
Loss and recovery percentages are asymmetric:
| Drawdown | Return required to recover |
|---|---|
| 10% | 11.1% |
| 20% | 25.0% |
| 30% | 42.9% |
| 50% | 100.0% |
After a 50% loss, a 50% gain does not restore the account; it leaves it 25% below the starting value. Survival protects both capital and the time needed to recover.
6. Size from the stop, not from conviction
CME Group’s position-sizing lesson starts with two inputs: the logical stop distance and the account amount or percentage the trader is willing to risk.
Use:
Position size = cash risk budget ÷ loss per unit at the stop
If equity is £10,000, risk is 1% (£100), entry is £50 and the invalidation stop is £48, the price risk is £2 per share. Before fees and slippage, the maximum size is £100 ÷ £2 = 50 shares.
Do not move the stop closer merely to buy more units. The stop should mark where the trade thesis is invalidated; size adapts to that distance.
7. Measure portfolio heat
Five positions risking 1% each do not necessarily create only 1% risk. If all are long technology shares, equity indices or other correlated assets, one market shock can hit every stop together.
Add the planned loss across open positions and group exposures by common driver. Stress a gap beyond the stop. Portfolio heat should respect the same survival threshold as a single trade.
The CFTC warns that leverage amplifies gains and losses and that traders can lose all margin and, in some arrangements, more than the initial deposit. Its guidance also says speculative trading should use risk capital, not money needed for living expenses or emergencies.
8. Build a survival test
Before trading a strategy live:
- Estimate win rate, average win, average loss and costs from representative data.
- Define the capital floor that counts as ruin.
- Simulate many sequences, including worse win rates, larger losses and clustered losses.
- Record the 95th or 99th percentile maximum drawdown, not only the average.
- Reduce risk until the stressed drawdown fits below the capital floor with a safety margin.
- Forward-test at small size and compare live slippage with the model.
- Pause and investigate if live results leave the pre-agreed confidence range.
A model does not prove safety. It makes assumptions visible and lets the trader reject fragile sizing before capital does it for them.
Common mistakes
- Treating zero balance as the only definition of ruin.
- Sizing from confidence instead of stop distance and equity.
- Using the backtest win rate as a guaranteed future probability.
- Assuming trades are independent when losses cluster by regime.
- Ignoring gaps, slippage, fees, financing and minimum contract size.
- Calculating risk trade by trade while ignoring correlated portfolio heat.
- Increasing size after losses to recover faster.
- Using emergency savings or borrowed money as trading capital.
Practical checklist
- Ruin threshold written in cash and percentage terms?
- Win rate and payoff estimated after costs?
- Losing streaks and maximum drawdown simulated?
- Stress test includes clustered losses and gaps?
- Risk per trade recalculated from current equity?
- Stop placed at thesis invalidation before size is calculated?
- Correlated open risk included in portfolio heat?
- Minimum tradable size still fits the risk budget?
- Live results compared with the model?
- Pause rule defined before the drawdown?
The Money Man View
Risk of ruin is less about predicting the next trade and more about controlling the sequence you cannot predict. A modest positive edge needs time and many observations. Position size determines whether an ordinary losing streak remains data for the journal or becomes the end of the account.
Educational risk disclosure
This tutorial uses simplified hypothetical examples and was prepared with AI-assisted research and writing under The Money Man editorial workflow. It is educational information, not personalised investment advice. Probability estimates depend on assumptions that may fail, stops do not guarantee execution prices, and leveraged trading can cause substantial losses.