Partial Exits & Exit Planning
Execution Precision
8 min read
Plan and execute partial exits without regret by using structured rules for locking in gains progressively.
Loading
8 min read
Plan and execute partial exits without regret by using structured rules for locking in gains progressively.
A partial exit is the act of closing a fraction of an open position at a predetermined R-multiple while letting the remainder run toward a further target or trailing stop. It trades upside variance for booked P&L and only improves expectancy on left-skewed strategies. The way you exit determines whether you keep your edge — or kill it.
This lesson assumes you have read Scaling Like a Pro, which covers fractional position sizing — partial exits are scaling in reverse.
Exits create more regret than entries — but partial exits do not eliminate regret, they redistribute it. You will still feel pain when the runner runs without you, or when the partial booked at 1R after the full position would have stopped out. The goal is not zero regret. It is regret that is consistent with a written rule.
“I exited too early.” “I gave it all back.” “I should’ve held.” “I should’ve taken profit.”
Here’s the truth:
The best traders don’t try to predict the top — they plan partial exits based on structure, stats, and consistency.
This post gives you the tools to exit with confidence, not confusion.
The shape of your strategy's return distribution decides your exit style. Right-skewed strategies (a few 5R+ outliers carry the P&L) need full runners — partials starve the tail. Left-skewed strategies (frequent small wins, occasional give-back) need full exits or aggressive partials. Most retail traders wrongly apply right-skew exits to left-skew systems and call it "taking profit".
Poor exit logic leads to:
Without an exit plan, your best setups become mid-tier outcomes — and your stats lie to you.
A partial exit is a volatility tax. You shrink the variance of your P&L distribution at the cost of trimming the right tail. If your strategy has positive skew (a few 5R+ winners pay for many 1R losers), aggressive partials destroy expectancy. If your strategy is mean-reverting (high win-rate, capped upside), partials are nearly free.
Lock in some gains, let the rest ride to full target or trail.
Best for:
Example:
Weighted-R math: if the runner closes at 5R, you booked 0.5(2) + 0.5(5) = 3.5R, vs 5R for a straight runner — partial exit costs you 1.5R on the upside in exchange for guaranteeing at least 0.5(2) + 0.5(0) = 1R if the runner stops out at break-even.
R_weighted = w_TP * R_TP + (1 - w_TP) * R_runner
Emotionally easier and reduces variance — but it does cap, not "keep open", the upside.
Predetermined take-profit based on POI, liquidity, or range — see Exiting at POIs for the dedicated treatment.
Best for:
Downside:
Exit if any of: (1) candle closes back inside the level after tagging, (2) CVD diverges against position for >2 bars at the level, (3) absorption pattern (large resting size eats your continuation) prints on the tape.
Best for:
Risk:
Bail if price shows rejection plus absorption near 4H resistance.
Honest costs of this plan: if BTC runs to 8R, your weighted result is 0.3(2) + 0.3(3.5) + 0.4(8) = 4.85R, vs 8R for a straight runner — a 39% haircut on the right tail. If your edge is right-skewed (a handful of 8R+ trades carry the year), this haircut compounds into materially lower CAGR. Run the numbers on your own MFE distribution before committing to a partial schedule.
No guesswork. You’re reacting with structure, not emotion.
Use MFE (Maximum Favorable Excursion) — the methodology originated in John Sweeney's Maximum Adverse Excursion (Wiley, 1996) and is now the standard way to audit whether your exit prices match what your edge is producing — to validate your exit levels.
If your trades consistently reach 2.8R but you exit at 1.2R…
What your edge is producing — the typical favorable excursion before reversal.
What you are actually banking — the realized R at the moment you close.
You’re killing your edge.
Exit planning should reflect what your edge is actually producing, not just what “feels safe.”
After trade:
Tag outcomes:
This builds repeatability — the only real antidote to regret.
No. Taking 50% at 1R and then stopping out the rest at -1R yields 0R, not a win. Partial exits redistribute the shape of your P&L distribution — they shrink upside variance and lift the floor on small winners — but they do not move expectancy upward. Whether they help or hurt depends on your strategy's skew.
Break-even first; partials second. A break-even stop converts a loser into a scratch with zero opportunity cost on the remaining size; a partial exit always costs upside R on the booked portion. If you can move to break-even before the partial level prints, you have removed account risk for free, and only then are you choosing whether to also trim the right tail. See Break-Even vs Staggered Scale-Outs for the side-by-side.
Match the percentage to your strategy's hit-rate at that R level. If 60% of your trades reach 2R, taking 50% there is reasonable; if only 25% do, you are over-trimming. The rule of thumb: never take a percentage so large that the booked portion alone exceeds the expectancy of letting the full size run.
Pull the Maximum Favorable Excursion of your last 30+ trades and look at the distribution, not the average. If MFE peaks at 2.8R but you exit at 1.2R, you are leaving a measurable amount of edge on the table. If MFE has a long right tail (a few 5R+ excursions), avoid front-loaded partial schedules — they will systematically clip those outliers.
You do not need to predict the top. You need a written rule whose weighted-R you have computed and whose skew matches your edge. Let part of the trade go. Let part run. Let none of it be random — and let none of it be untracked.
Next: Advanced Trailing Stops — how to mechanize the runner portion of every recipe in this lesson.