What Is Maximum Adverse Excursion in Futures Trading?
Maximum adverse excursion is the deepest unrealized loss a futures trade carried before it closed. Here is how to measure it and what it reveals about your stops.
By Imperial Analytics
Maximum adverse excursion is the deepest unrealized loss a futures trade carried at any point between entry and exit. Read across many trades, it answers a question the realized result hides: how much heat a position took before it became a winner or a loser. This post defines the term, shows how to measure it step by step, and explains how the distribution of those numbers informs a stop without inventing a rule from a sample too small to support it.
By Imperial Analytics
What maximum adverse excursion actually is
Maximum adverse excursion is the largest unrealized loss a trade carried at any single moment between the entry fill and the exit fill. It is measured per trade, in dollars, against the entry price. Its companion, maximum favorable excursion, is the highest unrealized profit the same trade carried in the same window. Together the two numbers describe the price journey of the trade, not just where it ended. John Sweeney formalized the measure for trade management.1
A trade is usually filed by its result: a winner or a loser of a certain size. That number is the endpoint, and the endpoint throws away the path. Maximum adverse excursion, or MAE, recovers part of the path by recording the worst point the position ever reached while it was open. A trade can close green and still have spent most of its life deep in the red, and a trade can close at a small loss after never moving far against the entry at all. The realized result cannot tell those two apart. MAE can.
The measure is most useful as a companion to maximum favorable excursion. MFE is the best the trade ever looked; MAE is the worst it ever looked. One trade carries both numbers, and the pair turns a single realized figure into a short description of how the price actually traveled between the two fills.
How MAE differs from the realized loss and from the stop
The realized loss is where the trade closed. The stop is where the trade was planned to close if it went wrong. Maximum adverse excursion is the worst point the trade actually reached in between, whether or not it closed there. A winning trade can carry a large MAE, and a trade stopped out has an MAE at least as deep as its stop.
These three numbers are easy to blur and worth keeping apart. The realized loss is an outcome. The stop is an intention set before entry, the same planned exit covered in average true range and how to size a stop. MAE is neither a plan nor an outcome; it is a fact about the path. A trade that closed at breakeven might have an MAE of several hundred dollars, meaning it went well against the entry before recovering. The breakeven result records none of that heat, but the MAE does.
The distinction is what makes MAE useful. Reviewing only realized results tells a trader which trades won and lost. Reviewing MAE tells a trader how close the winners came to being losers and how far the losers ran before they were cut. That is information the outcome column does not carry, and it is the input a stop decision actually needs.
How to measure it, step by step
For a long trade, maximum adverse excursion is the entry price minus the lowest price reached while the position was open, converted to dollars by the contract's point value and size. For a short, it is the highest price reached minus the entry. The result is always recorded as a non-negative number, the depth of the worst moment.
The calculation is mechanical once the lowest traded price during the holding period is known. Take a long position in MES, entered at 5,000.00, with a point value of 5 dollars per contract, held for one contract. Suppose the lowest print between entry and exit was 4,994.00. The adverse excursion in points is 5,000.00 minus 4,994.00, or 6.00 points. At 5 dollars per point that is -$30 of maximum adverse excursion for that trade, regardless of where the trade finally closed.
Data note
The numbers above are illustrative, chosen to show the arithmetic rather than drawn from any account or sample. Imperial Analytics surfaces a stop or sizing pattern from a trader's own data only when it meets the sample minimum in the AI Operating Charter, which is twenty trades in the matching condition.
Two details decide whether the number is trustworthy. The first is the price source: MAE depends on the lowest price actually reached, so it should be measured from the same data feed the trade was executed on, not a coarser bar that misses the wick. The second is the window: MAE is bounded by the entry and exit timestamps, so the holding period has to be recorded exactly. Get those two right and the figure is just subtraction on a known low.
What the distribution of MAE reveals
The single most useful view is the MAE of winning trades. It shows how much heat the trades that worked typically had to absorb first. A stop set tighter than the heat your winners routinely take will cut those winners before they turn, which the realized record alone would never explain.
One MAE is an anecdote. The distribution of MAE across many trades is a signal. Sort the trades that ended in profit and look at how deep each one went against the entry before it recovered. If most winners reached, say, four to six points of adverse excursion before working, a stop placed at three points is sitting inside the ordinary breathing room of a winning trade. It will turn a share of future winners into losers, and the trade log will record those as losses with no obvious cause. MAE is the cause.
The losing side carries the mirror lesson. If losing trades routinely ran far past the point where the original reason was gone, the stop was too loose or was being moved, and the adverse excursion on those trades will show it as a cluster of deep numbers. The pairing with maximum favorable excursion sharpens both reads: a trade with small MAE and large MFE was a clean entry, while a trade with large MAE and small MFE was a fight from the first tick.
↳ Note
A stop is a bet about how much a winning trade is allowed to hurt before it works. Maximum adverse excursion is the only number that tells you what that bet has actually cost.
How MAE shows up in the trade log
In the log, maximum adverse excursion is a per-trade column next to the realized result. Two patterns matter most: winners with deep adverse excursion, which warn that the stop is near the edge of the heat winners need, and losers with shallow adverse excursion, which suggest exits that fired before the trade was given room to be wrong.
The column makes the pattern countable instead of remembered. A winner with deep MAE is a row where the trade nearly hit the stop before paying out; a handful of those is normal, a wall of them means the stop is set inside the noise. A loser with shallow MAE is a row where the trade was cut almost immediately; a few protect capital, a wall of them can mean the entries are early or the hand is twitchy. Neither read requires a guess about intent. Both are arithmetic on the entry, the low or high reached, and the exit, the same kind of measurable signature described in what makes a behavioral pattern claim trustworthy.
How to use MAE without overfitting
MAE describes; it does not prescribe. Move a stop on the strength of a distribution, not a single ugly trade, and only once the sample is large enough to mean something. One deep adverse excursion is noise, and a stop rebuilt around it is a rule fitted to a coincidence.
The danger with any path statistic is treating one vivid trade as evidence. A trader who takes a single winner that went eight points against the entry and widens every stop to eight points has fitted a rule to one observation. The discipline is the same one that governs win rate: wait for enough trades in the matching condition before the number is allowed to change a setting, the standard set out in sample size for strategy win rate. MAE earns a stop change only when the distribution is stable across a real sample, not when the last trade stung.
Used that way, MAE keeps the stop honest in both directions. It guards against a stop set so tight it sits inside the heat winners normally take, and against one set so wide it lets losers run long past their reason. The number does not decide where the stop goes. It replaces the story a trader tells after a stop-out with a count of what the winners actually endured, and leaves the decision where it belongs, with the trader.
Frequently asked questions
- q: Is maximum adverse excursion the same as the realized loss on a trade? a: No. The realized loss is where the trade closed. Maximum adverse excursion is the deepest unrealized loss the trade reached at any point while it was open, whether or not it closed there. A winning trade can carry a large MAE, and a trade that closed at a small loss may have an MAE no deeper than its exit.
- q: How is MAE different from maximum favorable excursion? a: Maximum adverse excursion is the worst the trade ever looked, the deepest unrealized loss between entry and exit. Maximum favorable excursion is the best it ever looked, the highest unrealized profit in the same window. Each trade carries both, and together they describe the price path rather than just the endpoint.
- q: Can I use MAE to set my stop? a: It is one of the better inputs, because the distribution of MAE on winning trades shows how much heat your winners typically absorb before they work. A stop set inside that range will cut winners early. The caution is sample size: change a stop on the strength of a stable distribution, not a single deep trade.
- q: What data do I need to measure MAE accurately? a: The lowest price reached for a long, or the highest for a short, during the exact holding period, taken from the feed the trade executed on. A coarse bar can miss the wick and understate the excursion, and an imprecise holding window can include prices from before entry or after exit.
Sources
Footnotes
-
John Sweeney, Maximum Adverse Excursion: Analyzing Price Fluctuations for Trading Management (New York: John Wiley & Sons, 1996). ↩