Case Study, Ines: Judging a Swing Strategy After Eight Trades

[Fairman]

Case Study, Ines: Judging a Swing Strategy After Eight Trades

Ines is a composite character. She trades a daily-chart swing strategy and takes about four trades a month. After two months, she has eight trades: five losses of minus 1R and three wins totaling plus 4.5R, a net of minus 0.5R. She is thinking of dropping the strategy.

What does the data say? Average result: minus 0.5 divided by 8, minus 0.06R per trade. The standard deviation of the eight results is about 1.3R, so the standard error is 1.3 divided by the square root of 8, which is 0.46R. The average of minus 0.06 is about 0.14 standard errors below zero. It is indistinguishable from zero, and also from plus 0.4R, which would be a good strategy.

How many trades would she need? To distinguish an expectancy of 0.3R from zero at about two standard errors, with a standard deviation of 1.2, she needs (2 times 1.3 divided by 0.3) squared, which is about 75 trades. At four trades a month, that is 19 months.

Options. Backtest on years of history to get 100 trades quickly. Apply the setup to several pairs. Combine a historical test with live trading as evidence. Define a dropping rule before continuing: a drawdown beyond the tested maximum, or poor execution.

What not to do: abandon it because of eight trades, or scale up after eight good ones.

Lesson: slow strategies need a verification plan, and patience is part of it.

Exercise: compute how many months your strategy needs to reach 75 trades.