Understanding Drawdown in Automated Trading

What Drawdown Measures

Anyone reviewing a trading track record will encounter drawdown figures early, usually alongside return figures, because the two together give a far more complete picture of a strategy's behaviour than either one alone. Understanding exactly what the term means — and what it doesn't — is a foundational piece of reading any performance history sensibly.

Drawdown measures the decline from a peak in account equity to a subsequent low, before a new peak is reached. It's typically expressed as a percentage — a drop from a $10,000 peak to an $8,500 low is a 15% drawdown — and it's one of the more commonly cited ways to describe how much an account's value can fluctuate during losing periods.

Drawdown is distinct from a single losing trade. It accumulates across a losing streak or a rough patch, and it isn't officially over, technically, until a new equity peak is reached — an account can still be "in drawdown" even while recovering, right up until it exceeds its previous high.

Types of Drawdown Worth Distinguishing

The word "drawdown" is sometimes used loosely to mean several related but distinct things, and being precise about which one is being discussed avoids a common source of confusion when comparing strategies or reading a performance report.

• Maximum drawdown — the largest peak-to-trough decline observed over a given period, often used as a headline risk statistic.
• Current drawdown — how far the account currently sits below its most recent peak, which may be smaller or larger than the historical maximum.
• Duration of drawdown — how long an account spends below a prior peak, which is a different question from how deep the drawdown got.

A statistic like maximum drawdown answers "how bad has it been so far," not "how bad can it get." Those are different questions, and conflating them is a common source of misplaced confidence.

Why Historical Drawdown Is Not a Ceiling

This distinction is one of the more important ones in this entire subject, and it's worth stating plainly before going further: a historical figure describes the past, full stop. Nothing about a historical maximum guarantees it can't be exceeded once genuinely new conditions arise.

A historical maximum drawdown figure describes what has actually occurred during the period observed — nothing more. It does not describe the largest decline a strategy is capable of producing under conditions that haven't yet occurred in its track record. Markets can, and periodically do, produce conditions outside anything seen in a prior sample.

This isn't a reason to dismiss historical drawdown data — it's genuinely useful context for understanding a strategy's typical risk profile. It's a reason to treat it as a floor for what's possible, not a ceiling.

Thinking About Recovery

Recovery from drawdown is mathematically asymmetric: a 20% loss requires a 25% gain to recover, and a 50% loss requires a 100% gain. That asymmetry is one of the clearest arguments for controlling the depth of a drawdown in the first place, rather than relying solely on the expectation of eventual recovery.

• A 10% drawdown requires roughly an 11% gain to recover — a relatively small asymmetry.
• A 25% drawdown requires roughly a 33% gain to recover — the gap starts widening noticeably.
• A 50% drawdown requires a full 100% gain to recover — the account needs to double just to return to its starting point.
• A 75% drawdown requires a 300% gain to recover — a scale of recovery that becomes genuinely difficult to achieve within a realistic timeframe.

That curve is the mathematical reason deep drawdowns are treated so seriously in risk management, independent of any judgement about whether a strategy is otherwise sound. It isn't that a 50% decline makes a strategy "bad" — it's that the return required to recover from it is disproportionately larger than the loss itself, which changes the practical stakes of allowing a drawdown to run deep in the first place.

Recovery time also isn't something a track record alone can promise. A prior recovery period reflects the conditions that applied at the time; a future drawdown of similar depth is not guaranteed to resolve on a similar timeline.

Drawdown Specifically in an Automated Strategy

Automated strategies experience drawdown in the same statistical sense as any other trading approach — the automation changes how consistently the underlying rules are executed, not whether losing periods occur. An EA will continue to apply its logic through a drawdown exactly as it was designed to, which is a genuine strength in terms of execution discipline, but it also means an automated strategy won't independently recognise that a drawdown has become unusually deep relative to its own history unless that judgement was explicitly built into its rules.

This is part of why monitoring an automated strategy's drawdown against its own historical range remains a meaningful oversight task even after a system is fully automated — the automation handles execution, but comparing current conditions to the strategy's known historical envelope is a judgement call that typically still sits with the person running it.

Drawdown's Relationship to Position Sizing

Drawdown depth isn't purely a property of a strategy's entry and exit logic — it's also shaped directly by position sizing, covered in more depth in How Position Sizing Affects Trading Risk. The same sequence of winning and losing trades, run at different sizing levels, produces meaningfully different drawdown depths. This means two people running an identical strategy can experience very different drawdowns purely because of how much risk each chose to take per trade, which is worth keeping in mind before attributing an unusually deep drawdown to the strategy alone.

Key Points to Remember

• Drawdown is the peak-to-trough decline in account equity, typically expressed as a percentage.
• Maximum drawdown, current drawdown and drawdown duration are related but distinct statistics.
• Historical maximum drawdown describes what has happened, not the largest decline a strategy could produce in future conditions.
• Recovery from a loss requires a proportionally larger gain — a 50% drawdown needs a 100% gain to fully recover.