Limit targets fill at your price, stops fill at the next one: what that costs per trade

[Fairman]

Limit targets fill at your price, stops fill at the next one: what that costs per trade

A take profit is a limit order and a stop loss is a stop order, and the two don't get the same treatment when the market is moving. A limit fills at its price or better. A stop becomes a market order once touched, so it fills at whatever is available next.

Here is the effect with made-up numbers: a 10 pip stop, a 15 pip target, a 45% win rate and an average stop slippage of 0.4 pips. The targets fill at exactly 15 pips because there's no price improvement in my sample. The losing trades, which are 55% of the total, average 10.4 pips.

Expectancy with perfect fills: 0.45 x 15 minus 0.55 x 10 = 6.75 minus 5.50 = 1.25 pips per trade.
With the slipping stops: 6.75 minus 0.55 x 10.4 = 6.75 minus 5.72 = 1.03 pips per trade.

The gap is 0.22 pips a trade, or $2.20 on one lot. Over 200 trades that's 44 pips and $440 per lot, from a sum that never shows up as a line in the statement. It simply sits inside the closing prices.

Two consequences. First, a backtest that fills stops exactly at the stop price is more generous than live trading by this amount, and it grows with the share of losing trades. Second, you can measure your own value: take the last 30 stop-outs, subtract the stop price from the fill price, and average it. That number, not my 0.4, is the one to use.

[Fairman]

Measuring it takes a few minutes. In the platform history, each closed trade has the stop price you set and the price it closed at. For longs, slippage is the stop minus the close price, and for shorts it's the close minus the stop, so a positive number means a worse fill. Add them up for 30 or more stop-outs, then divide by the count. If a few of the 30 are large because of a news spike, report the average with and without them. The version that includes them is the one to use in a backtest, because they are part of what stops cost.