When your conditions fire, does the market actually move?
A strategy can fail for two completely different reasons: the entry has no edge, or the entry has an edge and the exits give it back. A backtest tells you the strategy lost money. It does not tell you which of those happened.
Measure one thing at a time
An entry check takes your conditions, finds every historical moment they fired, and measures what price did over the following bars. No stop, no target, no exit rule — because the question is whether the entry is worth anything at all, and every exit you add is another variable standing between you and that answer.
Get this answer first. Designing exits around an entry that has no directional edge is weeks of work on a foundation that was never there.
How it runs
- Pick a strategy and a window. How many bars after the signal you want measured.
- Every historical firing is found. The same transition rule the live scan uses, so the signals measured are the signals you would have been alerted to.
- The move after each one is recorded, bar by bar, in the direction the strategy trades — mirrored for a short, so a setup that profits from a fall reads as a positive number.
- You get the distribution, plus the signal count, which is what decides whether any of the rest means anything.
The chart it produces
A path, not a bar per horizon. The shape is the answer: an entry that pops immediately and fades is a different strategy from one that grinds upward for two days, and the holding period worth choosing is wherever the line stops climbing.
The baseline is the part that matters. It is the same measurement taken from arbitrary moments in the same window — what the instrument did anyway, signal or no signal. In a market that spent the period drifting upward, every entry produces a rising line, and a chart without a baseline would make all of them look like an edge. Only the gap between the two lines belongs to your strategy. Here that gap is about 0.73% over twenty bars; the signal line's own 0.90% is not the number to read.
The shaded band is the interquartile spread of individual trades — a quarter did worse than its lower edge, a quarter better than its upper. It widens with time because dispersion does, and it is the band an exit has to live inside.
Illustrative figures, to show what the chart reports and how to read it.
A finished check
- Strategy
- Mean Reversion · LONG
- Signals found
- 318
- Measured over
- up to 20 bars
- Moved in favour
- 61%
A result like this says the entry is worth building on. A result at 50% says the conditions are picking moments no different from any other moment — and no exit strategy will rescue that.
It tells you what it assumed
If a strategy was edited after a check ran, the result says so — the conditions measured are not the ones it uses now. If optional conditions were measured as required, it says that too, because that finds fewer signals than the strategy generates live. Both of those look exactly like an ordinary result otherwise, and both would make the product look wrong when it was simply answering a stricter question.
What this does not do
It is not a backtest and it does not report profit. There are no exits in it, so there is no trade to have a return — only movement after a signal. It also cannot tell you a strategy is good: a strong entry with badly chosen exits still loses money, which is exactly why the backtest exists as a separate step.
Questions
What does an entry check measure?
What the market did in the bars after your conditions fired, and nothing else. No stop, no target, no exit rule — just the movement that followed, measured over a window you choose. It answers whether your entry has any directional edge before you spend time designing exits around it.
How is it different from a backtest?
A backtest measures a complete strategy: entries, exits, costs, and the trades that result. An entry check measures only the entry. Separating them matters because a strategy can fail for two very different reasons — the entry has no edge, or the entry has an edge and the exits give it away — and a backtest alone cannot tell you which.
Does it work for short strategies?
Yes. The measurement is mirrored for a short, so a strategy that profits from a fall reads as a positive move. A short setup that works does not appear as a negative number.
How many signals do I need for the result to mean anything?
More than you think. A handful of signals tells you almost nothing, which is why the number of signals is shown first on every result — it is what decides whether the rest of the card is worth reading at all.