Maximum Drawdown Explained: Which Number You Are Reading

Two people can look at the same trading account, both say maximum drawdown, and mean different numbers. One is reading a figure measured from the starting deposit. The other is reading the largest fall from any high the account reached.

MetaTrader does not help by picking one. Its own report prints six separate drawdown figures for a single account, defined in the platform’s documentation, and they do not agree with each other by design.

What follows is about which number you are actually reading. It deliberately leaves out the arithmetic of recovering from a loss, which our recovery calculator already covers in depth.

What a Drawdown Measures

A drawdown is the decline in an account from a high point to a subsequent low point. It is a statement about the path the account took, not about where it finished.

That distinction is the whole reason the figure exists. Two accounts can end a year at exactly the same value while one travelled there smoothly and the other fell by half on the way, and only a drawdown figure separates them.

The word maximum then narrows it to the worst such decline over the period examined. Everything contentious about the number comes from three choices buried inside that sentence.

Those choices are which high point you measure from, what you measure, and how often you look. Change any one and the same account produces a different maximum drawdown, without a single trade being different.

There is also a second dimension the single figure discards entirely. Depth says how far the account fell; it says nothing about how long the account stayed below its previous high, and those two properties vary independently.

An account can fall a long way and recover the high within days, or fall modestly and sit below it for a year. A maximum drawdown figure reports both as a single percentage, so duration has to be read separately or not at all.

The Six Figures MetaTrader Reports for One Account

Most explanations describe a single formula. The platform most retail traders actually use reports six values, and its documentation defines each of them separately.

They come in two families, balance and equity, and each family has three members: absolute, maximal and relative. That is where the six comes from, and none of the six is redundant.

Each of those three exists twice, once against balance and once against equity, which is what produces the six.

The definitions come from the platform’s own testing report documentation rather than from general finance usage, and the two do not always coincide. That is the source of most of the confusion.

Balance Drawdown and Equity Drawdown Are Not the Same Test

Balance changes only when a position is closed. Equity changes continuously, because it carries the unrealised profit and loss of everything still open.

A drawdown measured on balance therefore describes only the trades you decided to finish. A drawdown measured on equity describes what the account was actually worth at each moment, including the positions you were still holding.

The gap between them is diagnostic. Where balance drawdown is small and equity drawdown is large, the account was carrying open losses that had not been realised, and the balance figure was hiding them rather than reporting them.

This is what makes any track record quoted only on closed trades incomplete. A strategy that never closes a loser can show a balance drawdown near zero while equity fell to a fraction of the account.

Grid and averaging approaches make the divergence structural rather than occasional. Where the method adds to a losing position and closes only once the basket returns to profit, closed trades are overwhelmingly winners by construction, so the balance curve climbs in small steps while the equity curve carries the entire open loss.

The balance figure in that case is not wrong. It is answering a question about closed trades that nobody asked, and the number a reader assumed they were getting is the equity one.

None of that requires bad faith to happen. It is the default behaviour of the two measurements, which is why the useful habit is to read them as a pair rather than picking one.

Absolute, Maximal and Relative: Three Different Questions

Absolute drawdown is the one that departs most from the common definition. The platform defines it as the difference between the initial deposit and the lowest level reached below that initial deposit, which is not a peak-to-trough measure at all.

Work the consequence through. An account that starts at a given deposit, doubles, then falls most of the way back has a large maximal drawdown and an absolute drawdown of zero, because it never went below where it started.

Maximal drawdown is the peak-to-trough figure people usually intend. It is defined as the difference between the highest local value and the next lowest value, expressed in the deposit currency with the percentage alongside.

Relative drawdown answers the same question as maximal, in the other unit. It is the difference in percentage terms between the highest local value and the next lowest, with the money figure alongside.

That last definition is worth stating precisely, because it is frequently written up differently. Some widely read explanations define relative drawdown as the absolute drawdown divided by the initial balance, which is a different calculation and produces a different number from the one the platform prints.

The practical rule is that a drawdown percentage means nothing until you know its denominator. A percentage of the starting deposit and a percentage of the highest value the account reached are not comparable quantities.

Why the Sampling Interval Changes the Answer

A drawdown is computed from a series of account values, and that series has to be sampled at some frequency. The frequency is a choice, and it changes the result.

A maximum drawdown computed from month-end values can only see month-end lows. If the account fell sharply mid-month and recovered before the month closed, the figure never records it.

The same applies at every resolution. Daily closes miss intraday lows, and tick-level equity captures excursions that a daily series smooths away entirely.

The direction of the error is consistent: coarser sampling reports a smaller maximum drawdown for the identical account. It never reports a larger one.

The reason is straightforward. A coarser series is a subset of the finer one, so every low it contains is also present in the finer series, while the finer series contains lows the coarser one skipped over. Taking the worst value from a smaller set cannot produce a worse result.

That makes two drawdown figures from different sources non-comparable unless you know how each was sampled. Comparing a tick-level equity drawdown from a strategy test against a monthly figure from a factsheet is not a comparison at all, a point that also applies to reading a backtest result.

Static and Trailing Loss Limits on Funded Accounts

Evaluation and funded accounts turn drawdown from a descriptive statistic into a rule that ends the account. How the rule is anchored matters more than its headline percentage.

A static limit is fixed against the starting balance. It sits at one level for the life of the account and does not move when you make money.

A trailing limit is anchored to the highest value the account has reached. As equity makes a new high, the floor rises with it and stays there.

The consequence is the part worth working through. Under a trailing limit, an unrealised gain permanently raises the level you must stay above, so reaching a peak and giving it back can breach the limit while the account is still above the balance it started with.

Whether the limit trails on equity or only on closed balance is a further variation, and it decides whether an open profit counts towards raising the floor. Both variants exist, and the difference is not visible in a headline percentage. Our page on funded account rules covers what else these programmes measure.

Reading a Drawdown Figure in Someone Else’s Track Record

A published drawdown number is a claim, and it is checkable only to the extent that its definition is stated.

Four questions settle most of it. Is it measured on balance or equity, from the initial deposit or from a peak, over what sampling interval, and over what period.

The period matters because a maximum is a record. A longer history has had more opportunity to produce a bad outcome, so a small maximum drawdown over three months carries far less information than the same figure over three years.

A low drawdown is also not automatically a good sign. It can mean modest position sizing, or it can mean the record is short, or that losing positions were held open rather than closed. Those are very different situations behind the same number. A drawdown figure also says nothing about what the account earned for enduring that variability, which is the question a risk-adjusted return is built to answer.

Where the figure is used to compare copy-trading or signal providers, the same four questions apply before the comparison means anything, which is the caution set out under assessing a provider.

Who This Page Is Not For

This page is not for anyone looking for an acceptable drawdown threshold. Published rules of thumb circulate widely and none traced to a source that could be verified, so no number is offered here.

It does not cover the arithmetic of recovering from a loss. That is a separate subject and it is worked through with examples in our drawdown recovery calculator.

It also does not tell you how much to risk. That decision belongs with the account-level limits described under account-level risk limits and the probabilities set out in the risk of ruin calculator.

What this page is for is knowing which of six possible numbers you are looking at, and what would have to be true for it to mean what you think it means.

Frequently Asked Questions