MAE and MFE Explained

A position spends its whole life moving, yet the history keeps only the endpoints. Between open and close the price may have gone deep against the trade, far beyond the eventual exit, or both. Maximum adverse excursion and maximum favourable excursion record those two extremes, turning the part of the trade the ticket forgets into something measurable.

The realised result says what was kept; the excursions say what the trade lived through. Two trades can both close at +0.5R while one coasted there and the other spent hours a hair above its stop — identical in the statement, completely different in the path data.

Both values come from scanning every price the market printed between the open time and the close time and keeping the running extremes. That requires the intratrade price path: ideally tick data, or at minimum the highs and lows of bars covering the holding period. The ticket alone is not enough — open and close prices say nothing about what happened in between.

Raw pip values are hard to compare across symbols, so excursions are usually normalised by the planned risk: dividing by the entry-to-stop distance restates them in R, the framing covered in the R-multiple guide. An MAE of 0.7R means the trade used 70% of its allotted risk before turning; that reads the same whether the instrument was a currency pair or an index.

Consider a long GBP/USD position opened at 1.2710 with a stop-loss at 1.2670 — a 40-pip risk, so 1R = 40 pips. After entry the price first slides to 1.2682, recovers, runs up to 1.2772, then fades, and the trade is closed at 1.2746.

The statement shows a tidy +0.90R winner. The excursions add the texture: the trade used 70% of its risk budget before working, ran to more than one and a half times the risk, and surrendered 26 pips between its peak and the exit. None of that is visible from the open and close prices alone.

A single trade’s excursion is an anecdote. The descriptive value appears when MAE is plotted across fifty or a hundred trades of the same strategy, because the shape of the distribution characterises how that strategy’s entries behave. These are patterns traders examine in their own data — observations, not instructions:

None of these shapes dictates a change by itself; each one is a question to put to the sample. How the stop distance was chosen in the first place is its own topic, covered in the stop-loss and take-profit guide.

The MFE distribution answers a different question: how much profit was open at some point, and how much of it the exits actually converted. The gap between the two is the giveback, and comparing realised results with MFE across a sample shows whether it was occasional or systematic.

• Winners closing far below their MFE — say a typical peak of 1.8R kept as 0.6R — describe exits that repeatedly left profit on the table after a larger move had already happened.
• Losers with substantial MFE, such as trades that were +0.8R ahead before closing at −1R, describe positions that reversed all the way through the entry to the stop.
• Realised results sitting close to MFE describe exits that captured most of what the path offered — the favourable extreme and the close were near each other.

Excursion columns sit naturally beside win rate and average R in a monthly trading review, because they explain why the headline numbers look the way they do: the same average result can come from clean entries with modest targets or from volatile paths where much of the open profit evaporated before the close.

MAE and MFE are reconstructions, and the reconstruction has moving parts. Three of them change the measured values enough to matter:

• Data resolution. Tick data captures the true extremes; one-minute bars approximate them well for trades held hours or days, but a spike inside a bar can be missed, and for very short-lived trades bar-based excursions are rough estimates.
• Bid/ask side. A long position is valued — and stopped — on the bid, a short on the ask. Excursions computed from a single price series ignore the spread, which understates MAE on longs measured from bids and matters most where the spread is large relative to the stop distance, such as during news.
• Partial closes. A position scaled out in slices has one underlying path but several closed tickets with different holding windows. Per-ticket excursions and per-position excursions diverge, so it matters which convention the analysis uses — and that it uses the same one throughout.