Maximum Drawdown and Recovery Factor

Returns make the headlines. Drawdown decides whether you survive to read them.

What maximum drawdown is

Maximum drawdown (MDD) is the largest percentage decline from a peak in the equity curve to the subsequent trough, before a new peak is reached. It measures the worst pain the system has ever inflicted on its operator.

MDD = (Trough Value − Peak Value) / Peak Value × 100%

If your account peaks at $100, falls to $70, then recovers, your MDD is 30%.

Why MDD matters more than return

A system returning 40% with a 50% drawdown is mathematically inferior to one returning 25% with a 10% drawdown — and far less likely to be traded through to recovery.

Two reasons:

Asymmetry of recovery: A 50% drawdown requires a 100% gain to recover. A 20% drawdown requires 25%. Deeper drawdowns are exponentially harder to climb out of.
Psychological abandonment: Most traders abandon a system at 25–30% drawdown — exactly when recovery is most likely. They book the loss and miss the rebound.

Drawdown table

Drawdown	Gain needed to recover
10%	11%
20%	25%
30%	43%
40%	67%
50%	100%
60%	150%
70%	233%
80%	400%

A 50% drawdown requires doubling the remaining account. A 70% drawdown requires more than tripling. Both are extremely difficult from a depleted base.

The recovery factor

Recovery factor puts drawdown into context relative to return:

Recovery Factor = Net Profit / Maximum Drawdown

Recovery Factor	Interpretation
< 1	The drawdown is larger than the profit — poor
1–3	Acceptable
3–5	Good
5–10	Excellent
> 10	Suspicious (overfit or tiny sample)

A recovery factor of 5 means the system makes five times more than its worst drawdown. That's the kind of cushion that lets you trade through the inevitable bad stretches.

How to measure drawdown realistically

Backtest drawdowns understate live drawdowns. Reasons:

Backtests miss slippage and spread widening
Live execution is slower than simulated
Psychological slippage: traders hesitate, skip signals, exit early
Different broker behavior than assumed in backtest

Rule of thumb: Multiply backtested MDD by 1.5–2× to estimate live MDD. If backtested MDD is 15%, plan for 25–30% live.

Drawdown in R-multiples

Express drawdown not just in % but in R multiples (multiples of your average risk per trade):

A 15R drawdown on a system with 0.5% risk per trade = 7.5% equity drawdown
A 15R drawdown on a system with 2% risk per trade = 30% equity drawdown

Same R drawdown, very different equity impact. This is why position sizing is the dominant lever for drawdown control.

Position sizing to cap drawdown

If you want to limit worst-case drawdown to 25% of equity, and the 95th-percentile Monte Carlo drawdown is 20R:

Max risk per trade = 25% / 20 = 1.25%

Position sizing to the Monte Carlo 95th-percentile drawdown is the most defensible method of capping MDD.

The underwater curve

Plot the equity curve's drawdown over time — the area below the peak. A good system spends most of its time near zero drawdown with brief excursions. A bad system spends long periods deeply underwater.

Mature systems often show:

Short underwater periods (months, not years)
Quick recovery to new highs
Drawdown depth shrinking over time as edge compounds

Common drawdown pitfalls

Quitting at the bottom: traders abandon systems at maximum drawdown, then watch them recover without participating
Increasing size to "make it back": doubling down during drawdowns deepens them
Comparing MDD across systems with different risk per trade: always normalize to R multiples
Reporting MDD only in %: also report duration in months
Ignoring drawdown duration: a 20% drawdown over 6 months is far worse than over 2 weeks

How long drawdowns last

System type	Typical max underwater duration
High-frequency	Days to weeks
Intraday	Weeks to a few months
Swing (daily)	3–9 months
Trend-following	6–24 months

Long-duration drawdowns are the hardest to trade through. Plan for the duration, not just the depth.

Practical MDD planning

Backtest MDD in % and in R
Run Monte Carlo — use the 95th-percentile drawdown as your planning number
Multiply by 1.5–2× for live execution haircut
Set position size so the resulting worst-case drawdown is below your psychological limit (typically 20–25%)
Define an abandonment rule in advance: "I stop trading this system if MDD reaches X% and lasts more than Y months"
Re-evaluate quarterly — if live MDD diverges from plan, investigate before adding funds

The survival principle

A system with 50% annual return and 60% max drawdown will likely blow up its operator before it pays off. A system with 15% annual return and 8% max drawdown compounds for decades.

Next: how to recognize when a system's edge is decaying — before drawdown reveals it.