Maximum drawdown

Definition

Maximum drawdown (MDD) is the biggest loss an investor would have suffered by buying at the worst possible high and selling at the subsequent low within a given window. It is calculated on a total-return NAV series: for each date, compare the value with the highest value seen so far, record the percentage shortfall, and take the worst case. A series that rose to 140, fell to 98 and later recovered has a maximum drawdown of 30%.

Why it matters

Volatility describes the typical size of moves; drawdown describes the worst experience. Two funds with the same volatility can have very different drawdowns if one falls in a sustained trend. Drawdown, together with time to recovery, is how many investors intuitively perceive risk, so it is common on factsheets, in risk questionnaires and in quantitative screening.

In the API

The headline figure is metrics.maxDrawdown. The underlying series is available in indexedPerformance.points, so you can recompute drawdown over any window your product needs. Example values:

json
{
  "metrics": { "maxDrawdown": -33.8 },
  "indexedPerformance": {
    "points": [{ "date": "2020-02-19", "value": 141.2 }, { "date": "2020-03-23", "value": 93.5 }]
  }
}

typescript
function maxDrawdown(points: { value: number }[]) {
  let peak = -Infinity, worst = 0;
  for (const p of points) {
    peak = Math.max(peak, p.value);
    worst = Math.min(worst, (p.value / peak - 1) * 100);
  }
  return worst;
}

Common pitfalls

The value depends heavily on the window (three years, five years, since inception) and on sampling frequency: daily data finds deeper troughs than monthly data. Fund houses do not all use the same window, so compare like with like. Sign conventions vary; the API returns a negative number.