Drawdown

Drawdown measures the decline from a peak to a trough in your portfolio’s value before a new peak is reached. It answers one question: “How much did I lose from my highest point before recovering?”

Types of Drawdown

  • Maximum Drawdown (MDD): The largest peak-to-trough decline during the entire backtest. If your portfolio went from $100,000 to $70,000 before recovering, your max drawdown is 30%.
  • Drawdown Duration: How long from peak to trough and back to a new peak. A 20% drawdown lasting 6 months is psychologically very different from one lasting 3 years.
  • Average Drawdown: The mean of all drawdown periods. Shows typical pain level.
  • Calmar Ratio: Annual return divided by maximum drawdown. Measures return per unit of drawdown risk.

The formula:

Drawdown = (Peak Value - Trough Value) / Peak Value x 100%

Why Drawdown Matters More Than Return

Drawdown is arguably more important than total return. Anyone can show big returns by taking big risks — drawdown tells you the price you paid.

  • Survival metric: A 50% drawdown requires a 100% gain just to break even. A 75% drawdown requires 300%. Deep drawdowns end trading careers and trigger forced liquidation.
  • Psychological tolerance: Most traders quit during drawdowns, not during profits. If your strategy has a 40% max drawdown, ask honestly: can you watch 40% of your money disappear and keep following the system?
  • Position sizing foundation: Maximum drawdown directly determines leverage limits. If max drawdown is 30% with 1x leverage, using 3x leverage means a potential 90% loss.
  • Strategy comparison: Two strategies returning 15% annually are very different if one has 10% max drawdown (Calmar = 1.5) and the other has 40% (Calmar = 0.375).

The Recovery Problem

This table shows why deep drawdowns are so devastating:

Max DrawdownGain Needed to Recover
10%11.1%
20%25.0%
30%42.9%
40%66.7%
50%100.0%
60%150.0%
75%300.0%
90%900.0%

The relationship is non-linear. Once you lose more than 50%, recovery becomes nearly impossible without taking even more risk.

Concrete Examples

Equity Curve Drawdown

Starting capital: $100,000

MonthPortfolio ValuePeakDrawdown
Jan$105,000$105,0000%
Feb$98,000$105,000-6.7%
Mar$92,000$105,000-12.4%
Apr$101,000$105,000-3.8%
May$108,000$108,0000% (new peak)
Jun$95,000$108,000-12.0%

Max drawdown so far: 12.4% (January peak to March trough). The March-to-May recovery took 2 months.

Duration Matters as Much as Depth

Strategy A: 25% max drawdown, recovered in 2 months. Strategy B: 25% max drawdown, took 14 months to recover.

Same drawdown percentage, vastly different experiences. Strategy B tests your patience for over a year. Most traders would abandon it before recovery, locking in losses.

Historical Context

The S&P 500 buy-and-hold:

  • 2000-2002 dot-com crash: ~49% drawdown, ~7 years to recover
  • 2007-2009 financial crisis: ~57% drawdown, ~5.5 years to recover
  • 2020 COVID crash: ~34% drawdown, ~5 months to recover

If your backtest doesn’t include at least one of these periods, you’re underestimating drawdown potential.

Leverage Amplification

A strategy with 15% max drawdown looks manageable. A trader applies 4x leverage to boost returns. The leveraged max drawdown becomes approximately 60%. A margin call forces liquidation at the worst possible time, turning a temporary drawdown into a permanent loss.

This is why drawdown analysis must come before leverage decisions, not after.

Managing Drawdown in Strategy Development

1. Set a Maximum Before Backtesting

Decide your pain threshold first (e.g., 20%). If the backtest exceeds it, reduce position sizes or add risk management rules. Don’t find the drawdown and then try to tolerate it — decide your limit upfront.

2. Use Position Sizing to Control Drawdown

The most direct lever. Half-size positions roughly halve drawdown. Methods include Kelly Criterion, fixed fractional sizing, and volatility-targeting. Read more about how slippage and costs interact with position sizing.

3. Include Circuit Breakers

Consider drawdown-based stop rules. For example: “If strategy drawdown exceeds 15%, reduce position sizes by 50% until a new equity high is reached.” This reduces the risk of catastrophic drawdowns.

4. Test Across Market Regimes

Your max drawdown is likely worse than what your backtest shows. The worst drawdown is always in the future. Use Monte Carlo simulation to estimate the true drawdown distribution.

5. Measure Drawdown Multiple Ways

  • Maximum drawdown (worst case)
  • Average drawdown (typical experience)
  • Drawdown duration (time underwater)
  • Ulcer Index (depth and duration combined)

What to Report

When evaluating a backtest, always include:

  • Max drawdown percentage
  • Max drawdown duration (peak to recovery)
  • Number of drawdowns exceeding 10%
  • Calmar ratio (annual return / max drawdown)
  • Time spent in drawdown (percentage of total time)

A strategy can have acceptable max drawdown but spend 80% of its time underwater. That’s miserable to trade.

Common Mistakes

  • Backtesting without crashes: If your data doesn’t include 2008, 2020, or similar events, your drawdown estimates are dangerously optimistic.
  • Assuming past max drawdown is the worst possible: It’s not. Use Monte Carlo or simply assume the worst drawdown is 1.5-2x the backtest maximum.
  • Confusing peak drawdown with account risk: A 30% drawdown from a peak of $150,000 means losing $45,000 — not $30,000 from original capital.
  • Ignoring drawdown duration: Time underwater matters as much as depth.

Resources