Streaks are normal, not a sign of failure
Ten losses in a row feels like proof that something is broken. For many strategies it is simply what randomness produces. If you do not know how often streaks happen, you will abandon strategies at the wrong moment and size positions too large to survive them.
The figures below are illustrations computed from probability alone, assuming each trade is independent with a fixed chance of losing. Real trades are not perfectly independent, so treat these as a guide, not a forecast.
The same randomness works in both directions. A strategy with no edge at all will also produce winning streaks, and a run of eight or ten wins feels just as convincing as a run of losses feels alarming. In our research, a trend-day rule that looked excellent on five and a half years of data matched coin flips on the next two. Streaks, good or bad, are what you should expect to see; they are not evidence on their own.
How likely is a long losing streak?
| Win rate | Trades | Chance of at least one losing run of | Probability |
|---|---|---|---|
| 50% | 100 | 5 or more | 81.0% |
| 50% | 100 | 6 or more | 54.6% |
| 50% | 300 | 8 or more | 44.2% |
| 40% | 100 | 8 or more | 49.0% |
| 40% | 100 | 10 or more | 20.5% |
| 40% | 200 | 10 or more | 38.0% |
| 40% | 300 | 10 or more | 51.6% |
A strategy that wins 40% of the time, typical of trend and breakout rules with targets larger than their stops, has about a coin-flip chance of a run of ten or more losses somewhere in 300 trades.
What a streak does to your account
With a fixed percentage risked per trade, each loss is taken from a slightly smaller balance. The drawdown after n losses in a row is 1 - (1 - risk)^n, and the gain needed to get back is 1 / (1 - drawdown) - 1.
| Risk per trade | After 10 losses | Gain needed to recover | After 20 losses |
|---|---|---|---|
| 0.5% | -4.9% | +5.1% | -9.5% |
| 1% | -9.6% | +10.6% | -18.2% |
| 2% | -18.3% | +22.4% | -33.2% |
| 5% | -40.1% | +67.0% | -64.2% |
Remaining balance = $10,000 x 0.98^10 = about $8,171
Drawdown = 1 - 0.817 = about 18.3%
Gain needed to recover = 1 / 0.817 - 1 = about 22.4%
At 5% risk: 0.95^10 = 0.599, a 40.1% drawdown needing a 67% gain
Two lessons. The damage grows much faster than the risk: five times the risk turns a painful drawdown into a disaster. And losses and recoveries are not symmetric: the recovery must come from the same strategy that just lost.
Plan for streaks before they happen
- Estimate your strategy's win rate from at least a hundred recorded trades.
- Look up (or compute) how long a losing run you should expect over the next few hundred trades.
- Choose a risk per trade that keeps that streak's drawdown at a level you could live with and recover from.
- Write down in advance how many trades the strategy gets before you judge it, and what result would make you stop.
- Do not switch strategies in the middle of a streak you had already planned for. Each switch is another try, chosen by luck, and our research counts tries for a reason.
How many trades before you judge?
Streaks also explain why short records mislead. If your results vary by about 1R from trade to trade (typical for trades that hit either a stop or a target), the average needed to reach a t of 2, the usual minimum for 'probably not luck', is roughly:
| Trades | Average needed for t of about 2 |
|---|---|
| 50 | about 0.28R per trade |
| 200 | about 0.14R per trade |
| 1,000 | about 0.063R per trade |
To detect an edge half the size you need about four times as many trades. A strategy can therefore spend a long time in drawdown while still being exactly what it was, and a strategy can spend months above water while having no edge at all. Decide the test length first, using numbers like these, and judge only at the end.
Limits
Gaps over weekends and through news can also make a single loss larger than 1R, and correlated positions can lose together. Regulators report that 74% to 89% of retail CFD accounts lost money in the EU studies cited by ESMA in 2018, and 82% in the UK FCA's 2016 sample. Plan for the bad run first.
Key takeaways
- With a 40% win rate, a run of 10+ losses in 300 trades is about a coin flip (51.6%).
- Ten losses at 2% risk cost about 18.3% and need a 22.4% gain to recover; at 5% risk, 40.1% and 67%.
- Choose risk per trade for the streak you will meet, not the average you hope for.
- Decide in advance how many trades a strategy gets before you judge it.
Sources
Education only, not investment advice. Results quoted from TradeLab research are historical simulations; trading can lose more than you expect.
