Drawdown Explained

Return describes where a strategy ended up. Drawdown describes what it did on the way, and that is the number that determines whether you are still following it at the end.

How it is measured

Take the highest account value reached — the peak. Follow the account down to its lowest point before a new peak is established — the trough. The decline between them, usually expressed as a percentage, is a drawdown.

Maximum drawdown is the largest one in the period examined. It is a single number describing the worst stretch, and it says nothing about how often smaller ones occurred.

Two dimensions are routinely omitted and matter as much as depth.

Duration is how long the decline lasted. A twenty percent drawdown recovering in two weeks is a different experience from one grinding on for eight months.

Time to recovery is how long from the trough back to the previous peak. A record can show a modest maximum drawdown that took two years to recover, which is a considerably worse property than the depth alone suggests.

The recovery asymmetry

The arithmetic that makes drawdown the dominant risk measure.

Losses and gains are not symmetric in their effect on an account, because a loss reduces the base from which the recovery must be earned.

A ten percent decline needs roughly eleven percent to recover. A twenty percent decline needs twenty-five percent. A fifty percent decline needs a hundred percent. A seventy-five percent decline needs three hundred percent.

The curve steepens rather than rising linearly, which is why a large drawdown is not merely a worse version of a small one. Past a certain point, recovery stops being a matter of continuing the strategy and starts requiring returns the strategy has never produced.

This is the entire reason position sizing works the way it does. Capital divided by twenty as the ceiling on any single position, under the divide-by-20 rule, is not about maximising return. It is about keeping the account in the region where ordinary returns can still recover ordinary losses.

Why headline return misleads without it

Two strategies showing the same annual return can be entirely different propositions.

One reaches it through a steady progression with a small worst stretch. The other reaches it after a deep decline and a strong recovery. The end points match; the experience does not, and the second requires you to have held through a period where stopping would have been the natural response.

Which is why a return figure presented without its drawdown describes half the trade. When evaluating any track record, find the maximum drawdown first and treat the return as context for it rather than the other way round.

The useful test is to assume the worst drawdown in the record begins the day you start. If that outcome would make you stop, the strategy is not appropriate at that size — and finding out now is considerably cheaper than finding out during.

Backtested drawdown understates the real one

Three reasons, all systematic rather than occasional.

Fills are optimistic. A backtest fills at a modelled price. A real order crosses a spread and consumes liquidity, and the gap is worst in exactly the volatile conditions that produce drawdowns.

The sample is short. The worst drawdown in your data is the worst that happened in that period, not the worst possible. A strategy tested over three years has not seen what four years would have shown.

Optimisation flatters it. Parameters fitted to historical data produce a curve smoother than the strategy will deliver forward. The more parameters tuned, the more this applies.

A reasonable working assumption is that live drawdowns will be deeper and longer than the record suggests. Planning for the recorded figure is planning for the best case.

Drawdown as a performance measure versus an account rule

Worth separating, because the same word describes two different things.

Everything above treats drawdown as a statistic — something you measure after the fact to understand a strategy's behaviour.

In proprietary trading firm accounts, drawdown is instead a hard rule: a line that terminates the account the moment it is crossed, regardless of what the trade would have done next. That version often trails your peak upward, which makes it a different mechanic entirely and one that deserves separate treatment.

The distinction matters because a strategy with an acceptable statistical drawdown can still fail an account rule, particularly where the rule measures unrealised equity rather than closed balances.

What it means for automation

Two practical points.

A daily loss limit is a drawdown control operating at session scope, and it is worth sizing against your strategy's normal variance rather than a round number — a limit inside the range of an ordinary bad day halts you constantly on days where nothing is wrong.

And your software should track drawdown from a persisted peak rather than recomputing from current state. A system that restarts and takes its current balance as the peak has silently reset its own risk measurement, which is the same class of bug as tier state or activation state failing to survive a restart. On a self-hosted deployment that persistence is yours to implement, and keeping the calculation off the order path — the worker thread pattern — keeps it from delaying the orders it is measuring.

The honest limits

Maximum drawdown is a single observation from a limited sample. It is not a bound, and the next one can be worse.

It says nothing about frequency. A record with one deep drawdown and one with several of similar depth report the same maximum and describe different strategies.

And measuring drawdown does not reduce it. Position sizing does, by determining how much of the account any single adverse sequence can reach.

Frequently asked questions

What is drawdown? The decline from a peak in account value to the lowest point before a new peak, usually expressed as a percentage.

What is maximum drawdown? The largest such decline in the period examined — a single number describing the worst stretch in a record.

Why does a 50 percent loss need a 100 percent gain? Because the loss reduces the base the recovery is earned from. The required recovery steepens rather than rising linearly with the loss.

Is backtested drawdown reliable? It systematically understates the real figure, because fills are optimistic, the sample is short, and optimisation smooths the curve.

How should I use maximum drawdown when evaluating a strategy? Assume it begins the day you start. If that would make you stop, the strategy is not appropriate at that size.

Disclaimer: This article is educational content about trading mechanics and software. It is not investment advice, financial advice, tax advice, legal advice, or a recommendation to buy or sell any security or futures contract, nor a recommendation of any strategy, platform, broker, or firm. Any contracts, specifications, figures, or firm rules named are described for illustration and are subject to change without notice. Futures and options trading involve substantial risk of loss and are not suitable for all investors; futures are leveraged and losses can exceed the amount deposited. Please read Characteristics and Risks of Standardized Options before trading options. Automated trading carries additional risks including software defects, connectivity failures, missed or duplicated signals, broker API changes, and outages that may prevent orders from being placed, modified, or cancelled. Past performance does not indicate future results, and no configuration, position-sizing rule, or risk setting can guarantee a profit or prevent a loss.

StaxInvesting LLC sells self-hosted trading software. It is not a broker-dealer, futures commission merchant, investment adviser, proprietary trading firm, or financial institution, and it does not manage accounts, hold member funds, place trades on behalf of members, or access member brokerage accounts. Members run the software in their own cloud environment, connect their own brokerage accounts under their own credentials, and are solely responsible for their configuration, their credential security, their compliance with any third-party terms or firm rules, and every trade executed in their account. Exchange specifications, firm rules, and platform details described here reflect publicly available information as of publication and change frequently; always verify against current official sources. Consult a qualified financial adviser, tax professional, and attorney regarding your individual circumstances.