Lessons › Advanced track › Forward test, review and improve

World 10 · Advanced track · lesson 15 · level 5

Forward test, review and improve

Trade your strategy in practice on new data, measure it honestly, and change only what the evidence tells you to change.

In one line

Your backtest looked great. What will 20 brand-new trades say?

Explained simply

A forward test is like a dress rehearsal before opening night. You perform the show exactly as written, on a stage you haven't practised on, while a friend keeps score. Afterwards you decide: keep the show, fix one scene, or rewrite it.

The lesson

A forward test runs the frozen strategy on new data as it arrives, here in the app's simulation, for a set number of trades. Review it with expectancy, win rate, average win and loss, maximum drawdown (the biggest drop from a peak) and profit factor, which is gross profit divided by gross loss, and compare these with the backtest. Decide in advance which results mean continue, adjust one rule or stop. Simulated results never guarantee real ones, because real trading adds real emotions, fills and costs.

A worked example

Illustrative example: a forward test of 20 trades has 8 wins totalling 18R, an average of 18 / 8 = 2.25R, and 12 losses of 1R each. Gross profit is 18R and gross loss is 12 x 1 = 12R, so the profit factor is 18 / 12 = 1.5 and the net result is 18 - 12 = 6R. That is an expectancy of 6 / 20 = 0.3R per trade and a win rate of 8 x 100 / 20 = 40%. The deepest drop ran from a peak of +6R after trade 7 to +2R after trade 11, a maximum drawdown of 6 - 2 = 4R. The backtest showed +0.36R, so the forward test is in line, and the rule written in advance says continue.

The same idea at four levels

1. Beginner. A forward test trades your strategy on new data without changing it.
2. Foundation. Track the number of trades, win rate, average win and loss, expectancy and the biggest drop from a peak.
3. Intermediate. Profit factor is gross profit divided by gross loss, and above 1 means more was won than lost.
4. Advanced. Compare with the backtest while allowing for luck in small samples, and write your continue, adjust or stop thresholds before the test starts.
5. Expert. Change at most one rule per review cycle and restart the count, because simulated success is evidence, never proof that real money would do the same.

Mistakes to avoid

• Changing the strategy halfway through the forward test.
• Treating a good simulated result as proof that real money would do the same.
• Deciding the continue, adjust or stop thresholds after seeing the results.

Goal of this lesson: Forward-test a strategy in simulation, review its numbers, and decide whether to continue, adjust or stop.