Touch or through: when a backtest fills a limit order
Posted: Fri Oct 30, 2026 2:56 pm
Touch or through: when a backtest fills a limit order
A backtest with limit orders must decide when they fill. If price touches the limit level, is that enough? Or must it trade through the level, by at least a tick?
The "touch" rule is generous. It assumes the order was at the front of the queue and filled at the first contact. The "through" rule is stricter, requiring price to move beyond the level, which implies the order was filled by the flow that passed it.
In live trading, the truth depends on the broker. Many fill on touch of the bid or ask. Some require a trade through. And a touch on the chart, which shows bids, may not touch the ask that a buy limit relies on.
The difference in a backtest can be large. A strategy that uses limit entries at precise levels can show twice the fill rate under the touch rule compared with the through rule, and the extra fills are disproportionately the ones that reverse.
The conservative choice is the through rule, plus a small buffer. If the strategy survives it, fills in live trading are likely to be at least as good.
A middle path used in some testing setups is to require the price to trade through the level by a small buffer, such as 0.2 pips, and to assume a fill at the level itself. The buffer approximates the idea that a resting order competes with others and may need the market to pass through to execute.
Another refinement is to apply the rule differently to entries and targets. Limit entries are more likely to be subject to adverse selection, so a stricter rule is justified. Limit targets are often placed in front of everyone else's, so a slightly looser rule is defensible. Whatever choice is made, writing it down alongside the results prevents later confusion about why live fills disagree.
Is the through rule used in the backtests, or the touch rule?
A backtest with limit orders must decide when they fill. If price touches the limit level, is that enough? Or must it trade through the level, by at least a tick?
The "touch" rule is generous. It assumes the order was at the front of the queue and filled at the first contact. The "through" rule is stricter, requiring price to move beyond the level, which implies the order was filled by the flow that passed it.
In live trading, the truth depends on the broker. Many fill on touch of the bid or ask. Some require a trade through. And a touch on the chart, which shows bids, may not touch the ask that a buy limit relies on.
The difference in a backtest can be large. A strategy that uses limit entries at precise levels can show twice the fill rate under the touch rule compared with the through rule, and the extra fills are disproportionately the ones that reverse.
The conservative choice is the through rule, plus a small buffer. If the strategy survives it, fills in live trading are likely to be at least as good.
A middle path used in some testing setups is to require the price to trade through the level by a small buffer, such as 0.2 pips, and to assume a fill at the level itself. The buffer approximates the idea that a resting order competes with others and may need the market to pass through to execute.
Another refinement is to apply the rule differently to entries and targets. Limit entries are more likely to be subject to adverse selection, so a stricter rule is justified. Limit targets are often placed in front of everyone else's, so a slightly looser rule is defensible. Whatever choice is made, writing it down alongside the results prevents later confusion about why live fills disagree.
Is the through rule used in the backtests, or the touch rule?