Losing Streaks
Why consecutive losses occur even with a positive expectancy, and how risk sizing affects survival through them.
Losses cluster
Even a strategy with a positive historical win rate can produce several consecutive losses. Random sequences do not alternate neatly between wins and losses, so a trader needs enough capital and emotional tolerance to survive ordinary variance.
Win rate does not tell you the next outcome
A 60% historical win rate does not mean every block of ten trades will contain exactly six winners. Short samples can look dramatically better or worse than the long-run average.
Risk size determines streak damage
Ten consecutive losses at a small fraction of equity are very different from ten losses taken with aggressive exposure. Compounding also means percentage-based risk naturally reduces dollar exposure as equity falls.
Responding to a streak
- Check whether trades followed the written setup.
- Compare current market conditions with the strategy's tested environment.
- Review execution and slippage.
- Avoid increasing size solely to recover losses.
- Use predefined drawdown or pause rules rather than improvising.
A practical protocol
- Reduce or pause after a predefined number or amount of losses.
- Check whether losses followed the plan.
- Separate execution errors from strategy losses.
- Review market regime and liquidity.
- Resume normal size only under defined conditions.
The psychological danger
After multiple losses, traders may hesitate on valid setups, take profits too early, double size to recover or abandon the process. A prewritten losing-streak protocol can reduce those discretionary reactions.
Win rate does not tell you the likely worst streak
The number of trades matters. Over hundreds of trades, longer losing runs become more likely than over ten trades. Historical backtests and simulations can help estimate the kinds of streaks a strategy has experienced, though future sequences can always be worse.
Streaks are inevitable in probabilistic systems
Even a strategy with a positive win rate can produce several losses in a row. Humans tend to interpret streaks as evidence that something has “changed,” but some streaks are simply normal randomness. Risk sizing must be able to survive them.