How Many Trades Do You Need to Evaluate a Strategy?

Every trader eventually asks whether a strategy is working or whether they have just been lucky. The honest answer depends on how many trades there are, and it is usually more than people would like. A handful of results, however good, says very little.

Here is why. A win rate measured from a limited number of trades is an estimate of the true win rate, and the estimate has a margin of error. For a strategy whose true win rate is 50%, the usual 95% margin of error on the measured rate is about plus or minus 18 percentage points with 30 trades. After 100 trades it is about 10 points. After 400 trades it is about 5 points, and after 1,000 it is about 3.

So a measured 60% win rate from 30 trades is entirely consistent with a true rate of 45%, or of 75%. You do not know which. The same strategy, at the same true rate, could show 38% in one batch of 30 trades and 62% in another by chance alone. This is why a good month, or a bad one, tells you little about whether the approach works.

Averages of win size and loss size are noisier still, because a few large trades can dominate them. That makes expectancy, which depends on all three numbers, harder to pin down than win rate. The practical consequence is humility about short records, and patience with a strategy that is performing near its expected range.

None of this means you can do nothing until you have thousands of trades. Practicing in a simulator can add decisions at no cost, and the first purpose of an analysis of your own trades is to spot large effects, like losses much bigger than wins, that stand out even in a modest sample. The analyzer states how many trades each finding rests on, and labels results as limited, preliminary, meaningful or strong by sample size.