Drawdown

Context & Mechanics

Definition and measurement

The drawdown measures the distance between a peak of the equity curve and the subsequent trough: (peak − trough) ÷ peak. As long as the old high is not reclaimed, the account is “underwater”. Three quantities describe the phase completely: depth (maximum distance), duration (time from peak to peak) and frequency (how often drawdowns of a size class occur). The deepest decline ever measured is the max drawdown – the drawdown term itself means any ongoing or completed losing phase.

The asymmetric mathematics of recovery

Losses and gains do not act symmetrically: a 10% drawdown requires an 11.1% gain to recover, 20% already 25%, 50% a full 100%. This asymmetry is the strongest argument for consistent position sizing: keeping drawdowns small keeps the recovery requirements small. Since losing streaks are statistically normal – at a 55% hit rate, 5–6 consecutive losses are to be expected over a hundred trades – the expected drawdown depth can be estimated from hit rate and risk per trade.

Drawdown in prop trading

Prop firms operationalise the drawdown as hard rules: daily loss limit for the daily loss, overall loss limit for the total loss, frequently as a trailing drawdown that follows the account high. Knowing one's own strategy drawdown statistics allows checking whether it fits a program's limits at all – a strategy with historical 12% drawdowns is structurally incompatible with a 10% limit, regardless of its profitability.

Why it matters for traders

Drawdowns decide whether a profitable strategy is sustained: abandoning during the losing phase – shortly before recovery – is among the most common expensive behavioural errors. In the GlanWick simulator, a strategy's drawdown depth and duration can be measured over historical data by way of example – GlanWick is a training and simulation tool and not a prop firm itself.

Execution Example

An account grows from $100,000 to a peak of $112,000. A losing streak follows: the balance falls over 18 trades to $103,600 before the strategy works again and the account later reaches a new high.

Depth: (112,000 − 103,600) ÷ 112,000 = 7.5% drawdown or $8,400.
Recovery requirement: 8,400 ÷ 103,600 = 8.1% gain needed to reach the old high – more than the 7.5% loss, because of the smaller base.
Duration: the underwater phase spans 18 trades plus the recovery stretch – often more taxing for endurance than the depth itself.
Interpretation: at 1% risk per trade the drawdown corresponds to a good 8 loss units (8,400 ÷ 1,000) – within the normal losing streaks of a 55% strategy. Knowing this in advance helps persevere instead of switching systems at the low.

Execution Risk & Errors

Interpreting drawdowns as a sign of a broken strategy although they are statistically normal

Ignoring the asymmetric recovery mathematics and increasing risk after losses

Considering only depth and underestimating the duration of the underwater phase

Trading strategies whose historical drawdowns exceed the prop program's limits

Switching systems at the drawdown low and starting the next cycle from scratch

Frequently Asked

What is the difference between drawdown and max drawdown?

A drawdown is any losing phase from a high to the following low; the max drawdown is the deepest of these phases in the entire observation period – a strategy's central risk metric.

Why does a 10% drawdown need more than 10% gain to recover?

Because the gain is earned on the smaller base: after a 10% loss only 90% of capital remains, and 90 × 1.111 ≈ 100. The deeper the drawdown, the more extreme the asymmetry.

What drawdown depth is normal?

That depends on hit rate, risk per trade and strategy. As a guide: losing streaks of 5–6 trades are statistically expected at a 55% hit rate over 100 trades – drawdown depth scales directly with risk per trade.

How do drawdown and prop-firm limits relate?

Daily and overall loss limits are predefined drawdown boundaries. One's own strategy's historical drawdown statistics must sit below these limits with a buffer, otherwise a rule breach is only a matter of time.