After a losing streak: cut size first, or change the strategy?

Separate the two kinds of loss first

Consecutive losses have two fundamentally different sources. The first is a normal drawdown: a strategy's win rate and payoff ratio mean a losing run can appear in any sample, and as long as risk per trade stayed inside the plan, the drawdown is not evidence that the strategy has failed. The second is a broken strategy: the market regime has changed structurally and the original assumption no longer holds.

Confusing the two leads to two opposite errors — misreading a normal drawdown as failure and switching strategies constantly, or misreading failure as a drawdown and adding to positions.

Three tests

First, the nature of the losses. If every loss occurred at the planned stop, it is a normal drawdown. If several losses show obvious slippage or exceed the planned stop, there is a problem in execution.

Second, whether the market regime has changed. The current environment is characterised by higher volatility: Brent above 100 dollars, spot gold swinging by hundreds of dollars in a session. If your strategy was designed in a low-volatility regime, the change in volatility is itself a factor requiring reassessment.

Third, whether the sample is large enough. For a strategy with a 40% win rate and a 2:1 payoff, a run of five consecutive losses inside a twenty-trade sample is within normal bounds. With a small sample, any conclusion is unstable.

Sequence: cut size, then diagnose

The recommended order is: first, immediately halve the risk per trade — this step needs no conclusion, its purpose is to buy time for judgement. Second, run the three tests above. Third, based on the diagnosis, decide whether to restore the original size, adjust parameters, or retire the strategy.

What not to do

First, do not reverse direction immediately after a losing streak; that usually means emotion has entered the decision. Second, do not increase size to "win it back" — that pushes risk per trade beyond the planned limit. Third, do not overhaul parameters on a small sample: the more parameters you change, the harder it becomes to tell which change helped.

What to record

At the end of each losing streak, record three things: the number of losing trades, the ratio of maximum drawdown to account equity, and whether every loss stayed within the planned stop. Accumulating those three data points is the only way to answer whether the current drawdown is still inside the strategy's normal range.