forex-scalping.com

Sizing Across Multiple Targets

Many trades have more than one exit: a first target to bank profit, and a second to let the remainder run. Sizing such a trade needs care, because the risk calculation does not change.

Risk is determined by the stop and the total position size, regardless of how many targets you use. If you risk 1 percent in total, split the position, for instance half at target one and half at target two, both share the same stop.

The reward side is where the planning happens. Suppose the stop is 20 pips, target one is 20 pips and target two is 60 pips. With half the position at each, the combined reward is 10 pips equivalent on the first half plus 30 on the second, an average of 40 pips against 20 risked, so 2R if both are hit, but only 0.5R if only the first target is hit and the rest is stopped.

A trade where the first target is hit and the second stops at breakeven yields 0.5R.

So the probability of hitting each target matters. Work out the expected value: probability of each outcome multiplied by its R.

Many traders over-focus on the best case, assuming both targets will hit. Test how often your trades reach each level.

Practical step: from your journal, calculate the share of trades reaching the first and second targets, and recompute expectancy.

A worked example. Account 5,000 dollars, risk 1 percent, so 50 dollars. Stop is 25 pips. Total size is 0.2 lots on EURUSD. Split it into 0.1 for the first target and 0.1 for the runner. The risk is still 50 dollars in total. People sometimes size each half for 1 percent and end up risking 2 percent. Another detail: if you move the stop on the runner to breakeven after the first target hits, the open risk on the trade drops to zero. That's a good moment to check whether you have room in your total open risk limit for a new trade. It's one of the cleaner ways to free up risk without changing your rules.

Another detail: if you move the runner's stop after the first target hits, the total risk goes down. But don't count that in your initial sizing. Size as if both parts could hit the original stop. Watch the lot rounding when you split, too. If the correct total is 0.03 lots, you can't split it evenly into two halves at 0.01 increments. You'd have 0.02 and 0.01, which changes the profit profile. On small accounts, it's sometimes better to use a single target than to round the split position up and risk more than planned.

The worked example makes the point clearly: risk stays tied to the stop and the total size, whatever the number of targets. One extension: once target one is hit and the stop moves to break-even, the trade's remaining risk is zero, but the expected value of the runner depends on how often price reaches target two from there.

Tracking that conditional hit rate (how often T2 is reached after T1) over 50+ trades tells you whether the split adds or subtracts value compared with a single target.

One small correction to the numbers in the opening post. If target one is hit and the runner goes back to the original stop, the trade nets zero, since plus 0.5R on the first half cancels minus 0.5R on the second. The 0.5R figure only holds once the runner's stop sits at breakeven, which is the case the next line describes.

FXS Prop Desk's conditional hit rate then gives a clean break even number for this setup. With the runner at breakeven, the split pays 0.5R plus 1.5R times the chance of reaching target two, while closing everything at target one pays 1R.

So the split only beats a single 20 pip target when more than one runner in three goes on to reach 60. Below that, the runner is costing money even though it feels like free upside.

My EURUSD figure last quarter was close to that line, around 38 percent over roughly 70 trades that reached the first target. That is too close to call once the extra close is costed, so I am still collecting.

The case that skews this most is a breakeven stop tagged by noise just before the runner would have gone on. Do you count those as runner losses in the conditional rate, or leave them out?