Most traders set stops based on a round number below support or a fixed ATR multiple — then never check whether those stops are actually working. Reviewing your journal’s MAE data turns stop placement from a guess into a calibration exercise, and the difference is measurable in dollars.
What MAE/MFE Measures and Why It Matters
MAE/MFE analysis originated in John Sweeney’s 1996 book Campaign Trading and has since become the gold standard for quantitative stop auditing. Every closed trade has two excursion values: Maximum Adverse Excursion (MAE) is the worst intraday drawdown before the trade closed, and Maximum Favorable Excursion (MFE) is the peak unrealized gain before closing.
A trading journal that records these values lets you answer one question systematically: did price need to move against you by more than your stop allowed before ultimately going in your favor? If yes, your stop killed a trade that was working correctly — it just needed more room.
Plotting 50-100 trades on an MAE chart reveals the distribution of adverse moves your strategy tolerates. Winning trades cluster at lower MAE values. Losing trades show a bimodal pattern: some have low MAE (price moved against you and stayed there — a correct stop) and some have high MAE (price blew through your stop and kept going — a sizing or placement error). The group you don’t want to see is winners with MAE close to your stop distance. That cluster represents money left on the table.
Classifying Your Stopped-Out Losers
Not all stopped trades are equivalent. The most useful audit you can run is a post-exit review: after each stopped trade, track what price did in the next one to two sessions. This produces three categories:
Premature stop: Price recovered past your entry after the stop triggered. You were right about direction, wrong about timing or placement. This is the most fixable failure mode.
Correct stop: Price continued lower and never returned. The stop did its job — it capped a losing thesis.
Stop too wide: Price reversed briefly, stopped you out, then collapsed further. A tighter stop would have saved you the same or more.
The 2019 analysis of 10,000 retail forex trades found that 43% of losing trades hit their stop within the first 15 minutes — before the setup’s intended timeframe had any chance to play out. If your journal shows a similar pattern, time-based stops (exiting if the trade doesn’t move in your favor within a defined window) may work better than price-based stops alone.
If more than 30% of your stopped-out losers fall into the “premature stop” category, your stops are too tight relative to the noise your strategy operates in. That’s the threshold worth acting on.
The 80th Percentile MAE Rule
The most actionable insight from MAE analysis is computing the 80th percentile MAE across your winning trades. Here’s why that number is the right anchor for stop placement:
Your winners survived some adverse move before becoming profitable. The 80th percentile MAE represents the maximum noise your strategy routinely tolerates on trades that ultimately worked. A stop placed just beyond this level cuts actual losers — trades where price moved further against you than any of your winners did — while giving legitimate setups enough room to breathe.
Setting stops at an arbitrary 1 ATR or 0.5% ignores this completely. A breakout entry in a high-volatility environment might have an 80th percentile MAE of $1.40, while a pullback entry in the same stock might show $0.60. One stop distance doesn’t fit both.
To apply this: export your last 60-90 days of winning trades, calculate the MAE for each, sort ascending, and find the value at the 80th rank. That’s your calibrated stop distance for that setup type. Run the same calculation separately for each major setup in your playbook — breakouts, pullbacks, fades — because they’ll produce different numbers.
The Case Study: 36% Premature Stops Dropped to 14%
Here’s what this looks like in practice. A swing trader took 60 SPY trades over 90 days and exported her journal. Of 22 stopped-out losers, 8 (36%) had SPY recover past her entry within two sessions. Her stop was set at 0.18% below entry — roughly $0.85 on SPY at $472.
Pulling the MAE data on her 31 winning trades, she calculated the 80th percentile MAE at $1.10. She adjusted her stop to $1.15 below entry and recalculated position size: previously 350 shares at $300 max risk, now 260 shares to keep the same dollar exposure.
She ran this new rule for 30 days, logging each trade result and the stop rule applied in her journal notes. Premature stop-outs fell from 36% to 14%. Net P&L improved by $840 over the test period, despite taking slightly larger per-share losses on trades that were correctly stopped. The dollar risk per trade never changed — only the distribution of outcomes did.
This is the arithmetic of stop optimization: fewer premature exits at the same dollar risk per trade means more winners completing, which compounds into meaningful P&L improvement without any change in setup selection or win rate.
Segmenting by Setup Type and Calculating Stop Efficiency
Aggregate MAE data is a starting point. Segmenting by setup type — breakout entries versus pullback entries, for example — often reveals that you need different stop widths depending on how you’re entering. Risk managers and active day traders deal with this constantly: a breakout into new highs has a different noise profile than a mean-reversion entry at a key level.
One useful metric for quantifying stop quality is the stop efficiency ratio: divide your average actual loss on stopped trades by your average winner. A ratio above 0.6 is a warning sign. It means your losers are cutting almost as deep as your winners run, which compresses profit factor regardless of win rate. Either your stops are too wide (you’re letting losers run), or your setups need filtering to eliminate low-quality entries that don’t deserve the full risk allocation.
Position sizing is where the math gets concrete. On a $30,000 account risking 1% ($300) per trade, a $0.30 stop on a $50 stock allows 1,000 shares. Widening to $0.45 requires cutting to 667 shares — same $300 max loss, but the trade now has 50% more price room before stopping out. This adjustment costs nothing in dollar risk; it only requires the discipline to take fewer shares.
The position sizing framework and stop placement are inseparable. Optimize one without adjusting the other and you’ll change your risk exposure unintentionally.
Running a 30-Day Test Cycle
Once you’ve identified a new stop rule from your MAE data, don’t immediately apply it to full-size positions. Run a structured 30-day test:
Write the new rule explicitly in your journal: “Testing stop at 80th percentile MAE ($1.15 on SPY entries) vs. previous $0.85 rule. Hypothesis: premature stop-out rate drops below 20%.” Track each trade result against this hypothesis in your notes field.
After 20-30 trades, calculate the premature stop rate under the new rule and compare it to your baseline. If the improvement holds and the stop efficiency ratio has improved, adopt the rule. If results are mixed, examine whether the new rule underperformed on a specific setup type or market condition and segment further.
This is the same discipline that winning traders apply to every edge: hypothesis, test, measure, iterate. The difference is that stop optimization has a clean before/after metric — premature stop rate and net P&L over the same number of trades.
Swing traders have a natural advantage here: fewer trades mean each data point carries more weight, and a 30-day test window typically yields 15-25 trades rather than hundreds, making the review manageable without statistical noise overwhelming the signal.
Key Takeaways
- Filter your stopped-out losers and check how many recovered past entry within two sessions. Above 30% means your stops are consistently too tight for your strategy’s noise profile.
- Calculate the 80th percentile MAE across your winning trades to find the data-driven stop distance — not an ATR multiple or round number.
- Segment MAE data by setup type. Breakout entries and pullback entries in the same instrument often need different stop widths.
- Widen stops by reducing share size to keep dollar risk constant. A $0.45 stop with 667 shares carries the same $300 risk as a $0.30 stop with 1,000 shares.
- Run every stop rule change as a 30-day hypothesis test with a documented baseline, tracked in journal notes. Adopt it only after 20-30 trades confirm the improvement.
JournalPlus automatically captures MAE and MFE for every trade, lets you segment by setup type, and tracks custom metrics like stop efficiency ratio across rolling periods. If you’re ready to audit your stops with actual data instead of intuition, it’s a one-time $159 and the analysis pays for itself quickly.
People Also Ask
What is MAE/MFE analysis in trading?
MAE (Maximum Adverse Excursion) measures how far price moved against you before a trade closed. MFE (Maximum Favorable Excursion) measures how far it moved in your favor. Together, they let you audit whether your stops are placed too tight, too wide, or correctly — using your actual historical trade data.
How do I know if my stop losses are too tight?
Export your stopped-out losers from the last 90 days and check how many had price recover past your entry within the same session or next two sessions. If more than 30% recovered, your stops are likely too tight relative to normal market noise.
What is the 80th percentile MAE rule?
Look at your winning trades and find the 80th percentile MAE — the adverse move that 80% of winners stayed within before turning profitable. Setting your stop just beyond this level cuts real losers while surviving the normal volatility your winners experience.
Does widening my stop increase my risk?
Not if you adjust position size accordingly. A wider stop with proportionally fewer shares keeps your dollar risk per trade constant. For example, at $300 max risk, a $0.30 stop allows 1,000 shares while a $0.45 stop requires 667 shares — same dollar exposure, more breathing room.
How long should I test a new stop rule before relying on it?
Run a 30-day test cycle with the new stop rule before committing full capital. Track each trade under the new rule in your journal notes, compare premature stop-out rates to your baseline, and only adopt the rule permanently once you have at least 20-30 data points confirming improvement.
