Maximum drawdown

Before you rely on it

What to check

Read it with recovery time beside it.

The common error

Where people go wrong with maximum drawdown

Looking at depth alone. A 30% fall that recovers in a year and a 30% fall that takes six are entirely different tests of patience, and only one of them is survivable for most people.

Worked through

Maximum drawdown in plain numbers

An account shows month-end values of $100, $120, $90, $110, $95 and $130. Walk through the peaks. From the $120 peak the low is $90, a fall of 25%. From the later $110 peak the low is $95, a fall of 13.6%, and the maximum drawdown for the period is the worst of these, 25%. Note that the account ended at a new high, and the maximum drawdown still records the pain along the way; a second account that finishes at the same $130 but never falls more than 8% delivers the same result with a gentler ride. Compare them and you see risk that the ending value hides.

Easily confused

Maximum drawdown and its neighbours

Maximum drawdown is a single figure taken from the drawdown series: the largest peak-to-trough decline in the period. Ordinary drawdown is the decline at any point from the latest peak. Maximum drawdown is different from volatility, which measures typical variation, and from value at risk, which estimates a loss threshold at a probability; it is also not a loss on your holding period: the peak may be before you bought; the Calmar ratio divides return by maximum drawdown, which links reward to worst-case discomfort.

In real decisions

Where maximum drawdown turns up

Investors review it when judging whether they can stomach a strategy. Fund analysts compare funds with similar returns by their maximum drawdowns. Traders set stop rules based on it; retirees test portfolios against historical worst cases to see how withdrawals would have fared; advisors show clients the maximum drawdown of a proposed allocation to prepare them. Longer histories capture more crises.

Know the edges

Where maximum drawdown stops being reliable

It depends on the period and the data frequency, and it reflects only the worst episode in the data; a future decline can exceed any in the past. It says nothing about how long it took to recover. Backtests can overfit. Use it as one input, with recovery time, volatility and your own tolerance for losses.

Quick answers

What people ask about maximum drawdown

How is maximum drawdown calculated? Track the running peak of the account. At each point, compute the decline from that peak, and take the largest. Software does this automatically from a series of values.

Is a lower maximum drawdown always better? Usually, for a given return. But lower drawdown may come with lower long-run growth. Weigh it against how much loss you can tolerate.

Does a longer history change it? Yes. More years include more crises, so the maximum can only stay the same or grow as history lengthens.

Do bonds have maximum drawdowns? Yes. They are usually smaller than stock drawdowns, but rising rates can produce declines of ten percent or more in long bond funds.

Can I compare drawdown across strategies with different lengths? Only with care. A longer history has more chances of a big decline. Compare the same period.