Cutting Winners Short: How to Put a Dollar Figure on the Trades That Got Away

The trade closed green. It still cost you money.

You took the entry, the move went your way, you held through the first wobble, and you took the exit at a respectable number. The trade is logged as a winner. Your win rate ticks up. The P&L line goes up.

Then you look at the chart five minutes later and see that the trade went another fifty points before it stopped out at break-even. The exit was thirty points off the high. The math says you took a winner; the chart says you took a fraction of a winner.

That gap has a name and a dollar figure. The name is cutting winners short, and the figure is the difference between where the trade went and where you sat.

The metric that exposes it: Maximum Favorable Excursion

A log that stores only the exit price cannot measure this gap. To compute the cost of cutting winners short, you need a second number per trade: the most favorable price the trade reached before it closed. In trade-analytics literature this is called Maximum Favorable Excursion, or MFE.¹

Per trade, MFE is straightforward: scan the price action between entry and exit, find the highest favorable price the position touched, record it. The "money left on the table" for that trade is the dollar gap between MFE and the realized exit, scaled by position size and instrument multiplier.

It only becomes interesting in aggregate. Across a quarter of trades, the distribution of that gap tells you something a single number cannot: whether your exit method is consistently early on the winners that mattered, or whether it is just doing its job and capturing a reasonable slice.

How to compute the dollar cost across a window

Run this calculation against your last 90 days of trades:

1. For each winning trade, record the realized exit price and the MFE price.
2. Compute the per-trade gap in dollars: (MFE price − exit price) × position size × instrument multiplier. Use the contract size in your trade log; do not approximate.
3. Sum the gaps across all winning trades in the window. This is your gross "left on the table" number for the period.
4. Bucket the gaps: how many trades left less than 0.25R on the table, how many left 0.5R to 1R, how many left more than 1R. The shape of the distribution is more informative than the total.

The dollar total tells you the size of the problem. The bucketing tells you whether the problem is uniform (every winner closes a little early) or concentrated (most winners are fine, a small number of trades end at MFE/4 or worse).

Uniform exits are usually a method issue — the stop or target is sitting in the wrong place relative to the move's typical extent. Concentrated exits are usually a behavioral issue — fear of giving back, fatigue, or session-time effects pulling specific trades in early.

Why this is not "every trade should run to the high"

Hindsight bias is a real risk here. No exit method captures every move's full extent. A trade that closes thirty points off the high is not necessarily a mistake — it might be the rule working correctly to protect against the moves that reverse hard.

The point of the MFE-vs-exit gap is not to feel bad about every trade that did not close at the absolute peak. The point is to compute the cost so the comparison between your current method and any proposed change is in dollars, not vibes.

If you are considering a trailing stop instead of a fixed target, the MFE distribution tells you the realistic ceiling on what a trailing method could capture from the same set of trades. If the dollar gap is small, the proposed change is not worth the additional management complexity. If the dollar gap is large and concentrated, the change has room to pay for itself.

What you usually find when you run this

The pattern shows up consistently when traders first compute their MFE distribution against the last 90 days:

• The total dollar gap is larger than expected. Most traders underestimate it because they remember the trades that closed near the top and forget the ones that did not.
• A small number of trades drive most of the gap. The "money left on the table" total is rarely uniform; it is usually 70 to 80 percent driven by 10 to 20 percent of the trades.
• Those trades cluster around specific times of session. The afternoon and the late-session hours are common offenders, especially for traders who scale down attention as the session goes on.
• The gap correlates with smaller-than-normal position size. Smaller positions get exited earlier because the dollar urgency is lower; the trader takes the green and moves on. The size that would have been worth holding gets treated as the size that wasn't worth managing.

None of these are universal rules. They are observation patterns. The actual numbers belong to your own data.

What to do with the number

Once the dollar figure is on the page, the next move is structural, not motivational.

• Compare two methods on the same trades. Take the realized exits as Method A. Take a defined trailing rule (e.g., trail at 1.5× ATR from entry, never tighten) as Method B. Recompute period P&L on the same trade set under Method B's exit logic. If Method B closes ahead by a meaningful margin and the bucketing supports it, the trailing rule is paying for itself on the data you already have.
• Tag the late-session offenders. If the concentration is at a specific time, set a session-state rule — for example, switch to a defined trailing exit only during the last hour, leaving the rest of the session on the regular target.
• Recompute monthly. The number is most useful as a trend, not a snapshot. If the dollar gap is shrinking month over month, the intervention is working. If it is not, the intervention was not the right one.

The discipline here is the same as everywhere else in trading: any change to method should be evaluated against the data you already have, not against the version of yourself that you would like to be.

Bringing it back to the journal

Cutting winners short is invisible until two things sit next to each other on the page: the realized exit and the most favorable price the trade reached. When the log stores only the exit, the favorable excursion stays off the page. The exit lands in the win column, the trade looks fine, the equity curve climbs at a slower slope than it could, and the gap stays a feeling rather than a figure.

Putting the figure on the page does not by itself fix the behavior. It moves the behavior out of the "I should probably hold longer" bucket and into the "this method has cost me $1,840 across the last 90 days, drawn from these 23 trades" bucket. The first bucket is durable. The second bucket is something you can act on.

This is the kind of pattern Imperial Analytics is built to surface — MFE alongside realized exit on every trade, the dollar gap broken out per trade and aggregated by window, and the sample-size disclosure called out before any aggregate claim is made.

Sources

Footnotes

1. Sweeney, John. Maximum Adverse Excursion: Analyzing Price Fluctuations for Trading Management. Wiley Trader's Advantage Series, 1996. The MFE/MAE framework introduced in this work remains the standard reference for excursion-based trade analytics. ↩