Here is an uncomfortable pair of numbers. An all-out exit at 2R with a 40 percent win rate and $100 risk earns $2,000 per hundred trades. A common partial scheme on the same setup, half off at 1R, stop to entry, rest at 2R, can earn exactly the same $2,000. Scaling out often adds nothing to expectancy. What it changes, and why it is still worth doing for most humans, is everything else about the experience of trading the system.
The math, honestly
Take the $10,000 account, $100 risk, EURUSD with a 25 pip stop, 0.40 lots, targets at 1R (25 pips) and 2R (50 pips). Journal data: with the original stop left alone, 40 of 100 trades reach 2R and 60 stop out, which is the all-out result above. Under the partial scheme, 60 trades reach plus 1R, where you bank $50 on half; you move the stop to entry, and 30 of those runners carry to 2R for $100 more while 30 fade back and scratch; 40 trades lose $100 outright. Total: 60 × $50 plus 30 × $100 minus 40 × $100 = plus $2,000. Identical expectancy, and notice the honest detail: the break even move on the runner donated some full-target wins (30 versus 40), and the early banking paid for it. Real numbers vary by setup; run yours.
What did change is the shape of the results. All-out, 40 percent of your trades end green. With partials, 60 percent end with money banked. Variance drops, the equity curve smooths, and the psychological tax per trade falls. Traders rarely quit systems because expectancy is negative; they quit because the experience of trading them is unbearable. Partials buy bearability, and bearability keeps you executing through the losing stretch that every strategy serves eventually.
A simple two-target framework
Half off at 1R or at the first meaningful opposing level, whichever arrives first. If the chart shows an obvious shelf at plus 20 pips, take the partial there rather than praying it stretches to 25. After the partial is banked, and only then, the stop on the remainder goes to entry. The runner exits at 2R, at a structure trail if the move keeps printing higher lows, or on a time stop if it stalls for an hour. Every number is set before entry. The framework's job is to leave you zero decisions to make while money is moving.
The two mistakes that gut the method
First, the micro-partial. Banking half at plus 0.3R takes $15 off a $100-risk trade and caps your best case at 1.15R while your worst case stays near 1R. Run a hundred trades of that and the expectancy that looked robust at 1R partials goes negative: you have built a system that collects coins and pays out bills. Partials before 1R need an exceptional reason. Second, sizing up because 'half comes off anyway'. Doubling the entry to 0.80 lots so the runner stays big means the stop now costs $200. That is not a partial strategy, it is a 2 percent position wearing a disguise, and the sizing discipline behind the 1% rule outranks any exit scheme.
When all-out is the better call
Mean reversion trades with a defined target usually deserve a single exit: the trade's premise ends at the average, so there is nothing for a runner to run toward. And when computed size is tiny, say the formula gives you 0.02 lots on a volatile metal, splitting it into 0.01 and 0.01 adds bookkeeping without meaning. Partials are a tool for trades with genuine continuation potential and enough size to divide. Everywhere else, one clean exit at a planned level beats ceremony.