What a zero-trade backtest means.

A backtest that produces no trades usually means the rules never became eligible to enter. Start by finding where the path closed.

Do not loosen the rules first

Changing parameters until trades appear can hide a setup error and turn one question into another. Freeze the current specification, save the result, and inspect each gate in order.

First confirm that the backtest ran to completion and used the intended market, dates, and timeframe. Then review the six parts of a testable rule.

Six places to look

Read the event counts

If the system exposes signal or condition counts, follow them from broad to narrow. A market filter may pass many bars while a later entry condition passes none. That tells you where the path ended without changing the hypothesis.

If no intermediate counts are available, test each condition alone in a copy of the strategy. Keep the original version unchanged so the diagnostic work has a clean reference.

Timeframe is part of the hypothesis

A daily rule tested on minute bars is not automatically equivalent to a daily test. Indicator values, signal timing, and execution assumptions can all change. Review why timeframe configuration matters before you compare results.

Warm-up can remove the whole sample

Long lookback periods need enough earlier data to become valid. If the test window is short, every eligible bar may occur before the indicator is ready. Extend the history only when that change matches the intended research design.

Record one of three conclusions

1. Configuration problem. Correct the documented setup error and rerun the same specification.
2. Rules are too restrictive for this sample. Keep the result and design a new hypothesis rather than editing the old record.
3. No valid opportunity occurred. The strategy may be working as written; test other reserved samples only if that was part of the plan.

A zero-trade result becomes useful when it narrows the next question.