Trailing drawdown does not measure your losses. It measures your give-back.

A static drawdown is measured from where you began. A trailing drawdown is measured from the best you have ever done. That single substitution changes everything downstream of it, and almost every misunderstanding of the rule is a failure to carry the substitution through.

Write it precisely. Let B be the starting balance, A the drawdown allowance in dollars, and E your equity now. Let M be the high-water mark: the greatest value E has ever taken. The floor is then

Substitute and the breach condition for a trailing floor becomes M − E ≥ A. Read that carefully, because it is the whole article: the starting balance is not in it. The rule does not know what you paid, what you started with, or whether you are up or down. It watches one number — the decline from your peak — and ends the account when that number reaches the allowance.

Your platform reports your position relative to a fixed origin. The rule reports your position relative to a moving one that only ever moves against you. Both are correct; they are simply not the same measurement, and only one of them is authorised to close your account.

Because M is a running maximum, it is monotonically non-decreasing: it rises when you make a new high and is otherwise unchanged. It never falls. The floor inherits that property exactly. This is why the mechanism is better described as a ratchet than as a trailing line — a trailing line implies something that follows you both ways, and it does not follow you down.

One consequence is worth stating immediately, because it is the opposite of how the rule is usually described. If an account never makes a new high, M stays at B and the trailing floor sits at B − A, which is exactly the static floor. A trailing drawdown is inert until you become profitable. It is not a penalty on losing. It is a penalty that switches on with success and scales with how much success you reach but do not hold.

The figure above asserts something that sounds like an error. It is not, and it takes four lines of arithmetic to confirm. Take the account in the figure: it starts at $50,000 with a $2,500 trailing allowance, and at its best moment it is worth $53,600.

The account is eliminated $1,100 above the balance it was funded with. Nothing has gone wrong with the arithmetic and nothing unusual has happened in the trading: the peak-to-current decline reached $2,500, which is the only event the rule is watching.

Once your peak exceeds your starting balance by more than the allowance, the floor is above your starting balance — and from that moment “profitable” and “alive” are independent states.

That threshold is worth naming because it is computable and it is the moment the account’s character changes. It arrives when M > B + A — in this case, the first time the account trades above $52,500. Before that point, being in profit and being alive coincide. After it, they can disagree, and the size of the possible disagreement is exactly M − B − A: here, $1,100.

This is also the precise answer to a question traders ask in the wrong units. “How much can I lose?” has no single answer under a trailing floor. You can lose A from your peak, whatever your peak happens to be. If you want the answer in the units your platform shows you — dollars from your starting balance — it is A − (M − B), which shrinks every time you set a new high and goes negative once you clear B + A. A negative answer means the floor is above water: you would be stopped out while still up.

Rearrange the breach condition once more. Your remaining room is

Room depends on M. M is a property of your history, not of your current position. Two accounts holding identical equity, on identical instruments, with identical open risk, can therefore have materially different amounts of room — and the difference is decided entirely by paths already walked and closed.

Take two traders who both end the month at $53,000, both up exactly $3,000. The first went up in something close to a straight line. The second reached $54,000 in the second week, gave most of it back, and recovered.

A trailing drawdown prices the volatility of the path, not the quality of the outcome. Two identical results are not the same position, and the one that oscillated is closer to elimination.

This inverts a piece of received wisdom. Traders are taught to judge a month by its result and to treat the intra-month wobble as noise. Under a trailing floor the wobble is not noise; it is the charge. Every dollar of give-back is billed once, in full, and never refunded — while the same dollar, when it is recovered, restores equity without lowering the floor it raised.

The asymmetry is the point. Reaching a high costs you nothing at the moment you reach it, because equity and the high-water mark move together. It costs you later, and only if you fail to hold it.

Everything above assumed we knew what M is a maximum of. That definition is not standardised, it is usually one line deep in a rules page, and it changes the answer more than any other term in the contract.

The two common definitions are the maximum of your closed balance, and the maximum of your equity including open positions. Under the second, a trade that moves into profit raises the high-water mark while it is still open — before you have decided anything, and whether or not you ever realise it.

Consider a single trade taken with the account at $51,000. It runs $1,500 in your favour, retraces, and you close it for $300.

The trade was profitable. It was closed green. Under a closed-balance high-water mark it cost nothing at all — the floor rose $300 and equity rose $300 with it. Under an equity high-water mark the same trade permanently surrendered $1,200 of the account’s survival budget, because the peak was recorded at a value the trader never banked.

Where the high-water mark includes open profit, you are charged for the best moment of every trade and credited only with the exit. The difference is the give-back, and it is billed on winners and losers alike.

This deserves care, because it is easy to over-read into advice, and it is not advice. It does not follow that taking profit earlier is correct: the peak is only identifiable afterwards, and a rule that shortens every winner carries costs of its own that this article does not attempt to price. What follows is narrower and firmer: on an equity-based high-water mark, unrealised excursions are not free. They are a real, quantifiable charge against the constraint that ends accounts, and any assessment of a strategy that ignores them is measuring the wrong account.

Two independent switches produce the regimes actually sold. The first is when the peak is sampled: continuously, including open positions, or once a day on the closing balance. The second is whether the floor stops: many futures programmes freeze the floor once it reaches the starting balance, after which the account behaves as though it had a static floor all along.

These are not stylistic variations. They can be ordered, and the ordering is a proof rather than an opinion. Daily closing balances are a subset of all the values equity takes during those days, so the maximum over the closes can never exceed the maximum over everything — which means, for the same allowance and the same trading:

A higher floor is strictly less room. So the regime alone establishes a difficulty ordering that holds regardless of strategy, instrument, win rate or account size — before a single trade is simulated. Equality is possible: an account that never closes above its starting balance has an end-of-day floor identical to a static one, which is the clause-01 observation again.

The rule regime is not a detail of the offer. It is a difficulty axis that can be ranked with certainty, and it is largely independent of the headline numbers traders compare.

The figure below runs the identical thirteen days from clause 01 against all four floors. The trading does not change. The outcome does — twice.

Two accounts advertising “$2,500 maximum drawdown” can differ by more than the allowance itself, depending on how these five terms resolve. They are the terms to find before the headline number means anything.

A practical consequence follows from the arithmetic in clause 03. Under any trailing regime, your remaining room is not visible on your platform: it is A − (M − E), and M is a number from your own history that no standard trading screen displays. Balance and open profit and loss are both the wrong coordinate. If a trailing floor governs an account, the floor is the only balance that decides anything, and it has to be tracked deliberately — because the day it matters is the day it is already too late to reconstruct.

Everything in this article is arithmetic, and arithmetic has a narrow warrant. It is worth being explicit about where the warrant ends.

• Every figure and every worked number here is illustrative, constructed to demonstrate a mechanism. None of them is a measurement of any firm’s programme, and none should be read as one.
• The ordering in clause 05 holds for a fixed allowance. Firms do not offer a fixed allowance across regimes — a static-floor programme and a trailing one typically differ in allowance, target and price at the same time. The regimes are strictly ordered in difficulty; the products are not, and this article makes no claim about which is preferable.
• Nothing here estimates how often accounts fail, or what share of failures a trailing floor causes. Those are empirical questions that arithmetic cannot answer and that we do not have defensible first-party data for.
• Nothing here is a recommendation to trade in any particular way. Clause 04 in particular is a statement about a cost, not an instruction about exits.

What the arithmetic does establish, and establishes firmly, is the mechanism: the floor is a ratchet driven by a maximum, the breach condition never references your starting balance, profit and survival separate once the peak clears B + A, give-back is charged permanently, and the four regimes admit a strict difficulty ordering. Those are not estimates. They follow from the definition, and they hold for every account the definition describes.