Maximum Drawdown Explained: The Risk Number That Actually Hurts

Maximum drawdown is the largest peak-to-trough decline an investment or strategy suffered over a period, expressed as a percentage of the peak. It answers the bluntest risk question there is: at the worst possible moment, how much of your money was gone? Volatility describes the daily wobble; maximum drawdown describes the wound. Here is the formula, the real historical numbers, and the recovery math that makes deep drawdowns so much worse than they first appear.

The maximum drawdown formula

For any point in a price or equity series, the drawdown is how far the value sits below the highest value seen so far:

Drawdown = (trough value - peak value) / peak value

The maximum drawdown is the most negative value that expression ever takes over the whole period. Concretely: a portfolio that grew to $150,000, fell to $90,000, and later recovered had a maximum drawdown of (90,000 - 150,000) / 150,000 = -40 percent, no matter how good the ending looked.

Two companion numbers complete the picture, and both matter as much as the depth:

• Drawdown duration: how long the decline took from peak to trough.
• Recovery time: how long the value stayed below the old peak before making a new high. This underwater period is where investors actually live with the loss.

Real maximum drawdowns: 2008 and 2020

Abstract percentages understate what drawdowns feel like, so anchor on two real ones from the S&P 500's history:

• The 2007-2009 financial crisis. The S&P 500 closed at 1,565.15 on October 9, 2007 and at 676.53 on March 9, 2009: a decline of about 57 percent in price terms (roughly -55 percent on a total-return basis with dividends reinvested), spread across 17 months of grinding losses. The price index did not close at a new high until March 2013, roughly five and a half years underwater from the 2007 peak.
• The 2020 COVID crash. The index fell from a closing high of 3,386.15 on February 19, 2020 to 2,237.40 on March 23, 2020: about -34 percent in 23 trading days, one of the fastest bear markets on record. Recovery was equally unusual: a new closing high arrived by August 18, 2020, under six months later.

Same index, both real, and completely different experiences: 2008 was deep and long, 2020 was violent and brief. That is why a maximum drawdown number should never travel without its dates and its recovery time. All three come straight out of clean, adjusted historical stock data; unadjusted prices distort both the depth and the recovery.

Past performance does not guarantee future results. For educational and informational purposes only. Not financial advice. Consult a licensed advisor.

Recovery math: why a 50 percent loss needs a 100 percent gain

Losses and the gains needed to repair them are not symmetric, because the gain must be earned on the smaller base. Lose 20 percent and $100,000 becomes $80,000; getting back to $100,000 from $80,000 requires a 25 percent gain, not 20. The required gain grows viciously as drawdowns deepen:

The last column is illustrative arithmetic at a steady 8 percent annual return, and real markets are anything but steady, but the shape of the table is the lesson: damage compounds faster than repair. A strategy that avoids the -50 percent hole does not need to be brilliant on the way up; it simply never has to climb out.

Why maximum drawdown beats volatility as a risk measure

Standard deviation, the input to the Sharpe ratio, treats a smooth 30 percent decline as low-risk because it happened gently, and it counts strong up-months as "risk" symmetrically with crashes. Maximum drawdown has neither blind spot: it measures exactly the loss an investor holding from the worst peak would have experienced, in the units people actually think in.

It also captures the behavioral risk that statistics usually miss. The practical failure mode of most strategies is not that the math stopped working; it is that the human quit at the bottom. An investor who abandoned equities in March 2009, down 55 percent and surrounded by genuinely terrifying news, turned a temporary drawdown into a permanent loss. Reading a backtest's maximum drawdown honestly means asking: would I have kept following this rule at that exact point? A drawdown you cannot hold through is not a drawdown you get to average.

The number still has limits worth knowing. It is a single worst case, so it is sensitive to the exact start and end of the data; a 10-year test that happens to dodge 2008 reports a flattering figure. It says nothing about how often smaller drawdowns occurred. And a backtested maximum drawdown is a floor estimate of the future, not a ceiling: the worst drawdown in your data is simply the worst one so far. Pair it with the Sharpe ratio and a look at drawdown frequency, and the picture rounds out.

How to read maximum drawdown in a backtest

When a backtest result lands, four checks turn the drawdown number into a decision-grade fact:

1. Locate it. Which dates produced the worst stretch? If the answer is not 2008 or 2020 for a long US equity test, ask what the rule was doing then.
2. Time it. Depth plus duration plus recovery time. A -35 percent hole with a 5-year recovery is a different life than -35 percent repaired in 8 months.
3. Compare it to buy-and-hold. Cutting maximum drawdown from -55 percent to -25 percent is real value even if returns trail; that is a trade-off worth seeing stated plainly, and it is the honest case for some trend-following rules.
4. Distrust tiny ones. A 20-year equity backtest showing a -6 percent maximum drawdown is more likely hiding a flaw (look-ahead bias, missing data, ignored costs) than revealing a miracle.

Every backtest run through an AI trading assistant shades the worst drawdown window directly on the chart and explains it in the plain-English risk panel of the investment risk analysis: how deep, how long, how long to recover, and whether the sample is big enough to trust. If you carry a rule around in your head, test it on 20+ years of split- and dividend-adjusted data and look at the worst stretch first, before the return has a chance to charm you. The return is what a strategy offers; the drawdown is what it costs.