One gold candle touched both exits. Which came first?

Two price paths fit the same bar

Consider a hypothetical long position in one ounce of gold. The entry was $2,000 before the bar began, and two exit orders were already active. The stop is $1,995, the target $2,010. The first exit cancels the other. For this exercise assume continuous prices, immediate fills at the stated levels and zero costs. The next 15-minute candle opens at $2,000, reaches a high of $2,012 and a low of $1,993, and closes at $2,007.

Suppose the price visits $2,012 first. It crosses the target on the way up, producing a $10 gain under those assumptions. Suppose it visits $1,993 first. It crosses the stop on the way down, producing a $5 loss. The same open, high, low and close accommodate both journeys. Even the strong close supplies no evidence about which exit came first. All prices here are invented solely for this calculation.

What the simulator decides for you

TradingView documents how its broker emulator chooses a path through a historical bar when it has only bar prices. In this example the open sits closer to the low, so the default path visits the low first. A tester who records the resulting stop as a fact about the historical tape would be importing the emulator’s assumption into the evidence. Another tester who chooses the target because the bar finished higher makes the opposite mistake.

Source notes: [1] TradingView — Pine Script strategy documentation

Ask the smaller bars one specific question

Look up the same instrument on the same feed and price basis. Confirm that the position and both exits were active before either level could qualify. Read the available one-minute bars in timestamp order. If an earlier minute crosses only the stop and a later minute crosses the target, the quote sequence resolves the reference-level test. If both levels first appear inside the same minute, the ambiguity survives. A gap in the data before the first visible crossing can also leave the result open.

TradingView can use lower-timeframe bars to improve historical detail where it has coverage. Record that coverage for the disputed interval. A sequenced quote can settle which reference level appeared first. Verifying an actual trade also requires the relevant trigger side, order terms and broker execution record.

Source notes: [1] TradingView — Pine Script strategy documentation

Show how much the missing sequence matters

In a hypothetical file of 20 eligible cases, suppose eight reach the target first, nine reach the stop first and three have unresolved ordering. Under the two-exit assumptions, the target-first share could be as low as eight out of 20 or as high as 11 out of 20. Printing 40% to 55% makes the price of the missing history visible. This is a range of possible classifications for this file. It makes no statistical claim about the next trade.

My research record would keep the three disputed bars, both possible labels and the source needed to settle each one. If better intrabar history becomes available later, another reader can reproduce the change without silently altering the rules.