Check an entry before you design exits
An entry check measures one thing: what the market did in the bars after your conditions fired. No stop, no target, no exit rule. It is the first test to run on a new strategy, because it answers the question every other test assumes you already know the answer to.
Why the order matters
A strategy can fail for two completely different reasons. Either the entries move no differently from any random moment, or they do move and the exits give it back. A backtest reports the combined result and cannot separate the two. Designing exits around an entry that was never worth anything is weeks of work on a foundation that was not there — so find out about the foundation first.
Step by step
- Pick the strategy and a market. The check uses the strategy's own direction, so a short strategy is measured the way it would trade: a fall reads as a positive move.
- Choose how far ahead to look. The horizon, in bars of the strategy's fastest timeframe. Long enough to cover how long you would normally hold.
- Run it. Every historical moment the conditions fired is found — using the same transition rule the live scan uses, so the moments measured are the ones you would have been alerted to — and the move after each one is recorded bar by bar.
Reading the result
The firing count comes first, on purpose. It decides whether the rest of the page is worth reading. A handful of firings tells you almost nothing; the shape of a path drawn from nine events is mostly noise.
Then the chart: a path, not a bar per horizon. The solid line is the median move after your entries, bar by bar. The dashed line is the baseline — what the same market did after a random moment over the same history. The band around the path is the spread of outcomes. The shape is the answer: an entry that pops at once and fades is a different strategy from one that drifts upward for two days, and that difference is what decides how long you would want to hold.
Under the chart, a table of holding periods — the same numbers at each horizon, with the baseline beside them — for when you want the figure rather than the picture.
What the two outcomes look like
- The path sits on the baseline. Your entries moved no differently from any other moment on this market over this history. Whatever a backtest shows afterwards came from the exits and the costs, not the entry. Change the conditions before you spend time on anything else.
- The path separates from the baseline. The entries did move differently, over a certain number of bars. That tells you roughly how long a hold the exit should allow for, which is the thing to carry into the backtest. It does not tell you the strategy will do this next month.
What it does not do
It does not simulate a trade, so it has no return, no drawdown and no win rate to report — those need exits, and exits are exactly what it leaves out. It measures one market over one window, and a strong reading on one pair is a reason to check the next one, not a conclusion. And like every test here, it describes the past.