Risk of Ruin: Losing Streaks, Drawdowns and How Much to Risk

Risk of ruin is the probability that a trading strategy will lose so much of the account that the trader has to stop, either because the money is gone or because the drawdown is too deep to recover. It depends on your win rate, your average win and loss, and above all on how much you risk per trade. Even a strategy with a genuine edge can be ruined by normal losing streaks if each trade risks too much, which is why professionals usually risk 0.5% to 1% per trade.

The maths of drawdowns

Losses and gains are not symmetric. After a loss, you have less capital to earn it back with:

Gain needed to recover = 1 ÷ (1 − drawdown) − 1

Drawdown	Gain needed to recover
10%	11.1%
20%	25.0%
30%	42.9%
50%	100%
75%	300%
90%	900%

Beyond about 30% to 40%, recovery becomes very slow for any realistic strategy. That is why "ruin" in practice usually means a drawdown of around 50%, not zero.

Losing streaks are normal

If each trade is independent, the chance of losing n in a row is the loss rate to the power n. More useful is the longest streak you should expect over many trades. We simulated 20,000 runs of trades with random outcomes:

Loss rate	Over 100 trades: typical longest losing streak	Over 500 trades: typical longest losing streak	Over 500 trades: chance of 8 or more in a row
50%	about 6	about 8	about 63%
60%	about 8	about 11	about 97%
65%	about 9	about 13	almost certain

A trend-following system that loses 60% to 65% of its trades will almost certainly have a streak of 10 or more losses at some point. That is not a sign that it is broken; it is arithmetic.

What a streak costs at different risk levels

Ten losses in a row, with risk recalculated on current equity each time:

Risk per trade	Account after 10 losses	Drawdown
1%	90.4%	−9.6%
2%	81.7%	−18.3%
5%	59.9%	−40.1%
10%	34.9%	−65.1%

At 1% a ten-trade streak is painful but recoverable. At 10% it leaves you needing almost a 190% gain to get back.

Simulating risk of ruin

There is no single formula that captures every real strategy, so traders use Monte Carlo simulation: generate thousands of random trade sequences with your win rate and payoff, and measure how often the account suffers a given drawdown.

These simulations assume each trade is independent and that your win rate and payoff are known exactly. Real trading is worse: edges shrink, streaks cluster when markets change, and gaps produce losses bigger than 1R. Treat simulated drawdowns as a minimum, not a maximum.

The Kelly criterion, and why to use less

The Kelly criterion gives the fraction of capital to risk that maximises long-run growth, if you know the true edge:

Kelly fraction = W − (1 − W) ÷ R, where W is the win rate and R the average win divided by the average loss.

Look at the simulation table: even 5%, less than two-thirds of Kelly here, produced a typical worst drawdown above 60%. Full Kelly is far too aggressive for real traders, because:

• Your true win rate and payoff are uncertain. If you overestimate the edge, Kelly sizing overbets and can turn a winning strategy into a losing one.
• Kelly-sized drawdowns are psychologically unbearable, and people abandon the plan at the bottom.

Many professionals use a quarter of Kelly or less, which usually lands near the familiar 0.5% to 2%.

Reducing your risk of ruin

1. Risk 0.5% to 1% per trade until you have hundreds of live trades of evidence. See Position Sizing.
2. Cap total open risk and correlated exposure (see Portfolio Risk and Correlation).
3. Cut risk in drawdowns. For example, halve the risk per trade after a 10% drawdown and restore it at a new equity high.
4. Set a maximum drawdown at which you stop and review, decided in advance.
5. Use your backtest's longest losing streak and worst drawdown, then assume live trading will be worse (see How to Backtest a Forex Strategy).

Next, put everything together in the capstone: Build and Test Your Forex Trading Plan.