How many trades to prove a per-trade edge is real?

Question

A trading tactic earns an average of 0.10% per trade, with a standard deviation of 0.50% across trades.

Roughly how many trades do you need before the average edge is statistically significant at the usual 5% level (t of about 2)? Assume trades are independent.

Accepted Answer

Set up the test. For a run of N independent trades, the t-statistic for "the true average edge is zero" is the per-trade Sharpe times the square root of the number of trades: t = \frac{\bar{r}}{s}\sqrt{N} = s{	ext{trade}}\sqrt{N}, where s{	ext{trade}} = \bar{r}/s is the average edge divided by the trade-to-trade standard deviation. In plain terms, each trade adds a little more evidence, and the evidence piles up with the square root of the trade count. Solve for the number of trades. We want t \geq 2, so s{	ext{trade}}\sqrt{N} \geq 2 \quad\Longrightarrow\quad N \geq \left(\frac{2}{s{	ext{…

How many trades to prove a per-trade edge is real?

Your answer

More Statistics questions