What max drawdown actually measures

Max drawdown (MDD) is the largest percentage drop from a strategy’s equity peak to its next trough over a given period. In plain English: it tells you the worst loss you would have experienced if you had bought at the top of the equity curve and held through the decline until the bottom.

The formula is simple:

Drawdown at time t = (Peak Equity - Current Equity) / Peak Equity

Max drawdown = maximum drawdown observed over the full sample

Example: if your account rises from $10,000 to $14,000, then falls to $11,200 before recovering, the drawdown from the peak is:

[(14,000 - 11,200) / 14,000 = 0.20 = 20%]

If that is the worst decline in the test, your max drawdown is 20%.

Why traders should care more than raw returns

A strategy that makes 30% per year sounds great until you discover it regularly suffers 35% drawdowns. Many traders can tolerate slower growth, but not the emotional and financial stress of deep losses.

Max drawdown matters because it affects:

  • Capital preservation: A 50% loss requires a 100% gain to recover.
  • Behavioral survival: Large drawdowns often cause traders to abandon otherwise sound systems.
  • Position sizing: The same strategy can be tradable or untradable depending on how much capital it can lose peak-to-trough.
  • Leverage tolerance: High leverage can make a profitable strategy impossible to hold through normal volatility.

A useful mental model is that returns tell you how much a strategy can make, while drawdown tells you how much pain it can impose on the way there.

How max drawdown differs from volatility

Volatility measures the typical spread of returns around the average, usually using standard deviation:

σ = sqrt(average squared deviation from mean return)

Drawdown is different. It is path-dependent and asymmetric. Two strategies can have the same volatility but very different drawdowns.

For example:

  • Strategy A has many small up-and-down moves.
  • Strategy B has steady gains followed by one sharp crash.

Both may show similar standard deviation, but Strategy B likely has a much worse max drawdown. That’s why drawdown is often more intuitive for investors than volatility.

What is a “good” max drawdown?

There is no universal threshold, but rough expectations help:

  • 5%–10%: Very mild; often seen in conservative or low-volatility systems.
  • 10%–20%: Common for diversified systematic strategies.
  • 20%–35%: Material; may be acceptable if returns are strong and stable.
  • 35%+: Aggressive; usually requires high conviction, strong diversification, or very high return potential.

The key is not the number alone, but whether the drawdown is acceptable relative to the strategy’s return, frequency, and your ability to stick with it.

Pair max drawdown with return-based metrics

Max drawdown should never be viewed in isolation. A strategy with 8% annual return and 4% max drawdown may be more attractive than one with 25% annual return and 40% max drawdown.

Useful companion metrics include:

  • CAGR: Compound annual growth rate.
  • Return / max drawdown: A simple efficiency ratio.
  • Sharpe ratio: Return per unit of volatility.
  • Calmar ratio: CAGR divided by max drawdown.

A higher Calmar ratio often indicates a better balance between growth and risk. For example:

Calmar = CAGR / Max Drawdown

If a strategy earns 18% CAGR with a 9% max drawdown, Calmar = 2.0. If another earns 24% CAGR with a 24% drawdown, Calmar = 1.0. The first may be preferable even with lower absolute return.

Common pitfalls when evaluating drawdown

1) Ignoring sample length

A strategy tested over one year may show a tiny drawdown simply because it hasn’t lived through a bad market regime. Longer samples usually reveal deeper, more realistic drawdowns.

2) Overfitting to reduce drawdown

It’s easy to tune parameters until the backtest looks smooth. But a curve-fit strategy often fails out of sample. If you optimize too aggressively for shallow drawdowns, you may be fitting noise.

3) Using only closed-trade equity

Some traders calculate drawdown from trade results only, ignoring unrealized losses. That understates risk. Equity-based drawdown should include open positions.

4) Forgetting gaps and slippage

A stop-loss may look protective in theory, but real markets gap through stops. Backtests should include realistic slippage and commissions, especially for short-term systems.

5) Comparing strategies with different time horizons

A swing strategy and a market-neutral intraday strategy can have very different drawdown profiles. Compare systems only when the holding periods and instruments are similar.

How to reduce drawdown without killing returns

You usually cannot eliminate drawdown, but you can manage it:

  • Lower position size: Risking 0.25%–1% of equity per trade is common for retail systematic traders.
  • Diversify signals: Combining uncorrelated strategies often smooths the equity curve.
  • Use regime filters: Some strategies work only in trending or mean-reverting markets.
  • Add exits carefully: Stops, time exits, and volatility-based exits can reduce tail losses.
  • Avoid excessive leverage: Leverage amplifies both returns and drawdowns.

A practical rule is to design for the drawdown you can psychologically and financially tolerate, then see what return remains.

How to test drawdown properly

When backtesting a strategy, look at more than the single worst number:

  • Max drawdown depth: How bad was the worst decline?
  • Drawdown duration: How long did recovery take?
  • Number of drawdowns: Are losses frequent or rare?
  • Rolling drawdown: Does risk change across different periods?
  • Out-of-sample performance: Does drawdown stay reasonable on unseen data?

Tools like Algovex can help you visualize the equity curve, inspect drawdown periods, and run walk-forward analysis so you can see whether the risk profile is robust rather than accidental.

Bottom line

Max drawdown is one of the most important metrics in systematic trading because it captures the real cost of being wrong. Returns tell you what a strategy can earn; drawdown tells you whether you can survive long enough to realize those returns.

If two strategies have similar profits, the one with the smaller, more recoverable drawdown is usually the better choice. In practice, the best strategy is not the one with the highest backtest return — it’s the one with a return profile you can actually hold through the worst losing streak.

Disclaimer: This content is for educational purposes only and is not financial advice.