Drawdowns Matter More Than Returns: The Risk Metric Investors Ignore at Their Own Expense

Why returns can mislead you

Most investors start with the wrong scoreboard. They ask, “How much did it make?” before asking, “How bad did it get?” That order is backwards for anyone who has to actually hold the position through a full cycle. A strategy that compounds at 12% a year but suffers a 45% drawdown may be far harder to own than one that compounds at 9% with a 15% drawdown. The second strategy can be more useful even if the first looks better in a spreadsheet.

Academic work on drawdowns has long argued that the path matters, not just the destination. Chekhlov, Uryasev, and Zabarankin formalized drawdown measures as risk statistics, showing that investors can evaluate not only the worst peak-to-trough loss but also the shape and persistence of losses ^[1]. Hurst, Ooi, Pedersen, and Stojanovic later showed that drawdown-based risk management can improve the practical robustness of trend-following and managed-futures style portfolios because it aligns risk control with the investor’s experience of pain ^[2].

That is the core idea here: returns tell you what happened over the full period. Drawdowns tell you whether you could have stayed invested long enough to earn them.

For readers who want a broader framework, this sits naturally alongside risk measurement basics, Sharpe vs. Calmar, and position sizing. Those topics are different lenses on the same problem: how much pain is acceptable for a given expected reward.

The math of recovery: losses are not linear

The simplest drawdown lesson is also the one investors forget most often: losses require larger gains to recover. If a portfolio falls 10%, it needs an 11.1% gain to get back to even. A 20% loss requires 25%. A 50% loss requires 100% ^[3]. That asymmetry is not a slogan; it is arithmetic.

Here is the formula. If a portfolio falls from 100 to 50, it has lost 50%. To return from 50 to 100, it must double. The gain is calculated on the reduced base, not the original one. This is why deep drawdowns are so damaging: they force the portfolio to work from a much smaller capital base, and they often arrive with investor confidence already broken.

Worked example: suppose two strategies each start at 100.

• Strategy A rises 20%, falls 25%, then rises 20% again.
• Strategy B rises 10%, falls 10%, then rises 10% again.

After the first two moves, Strategy A is at 90, while Strategy B is at 99. After the final move, Strategy A is at 108, while Strategy B is at 108.9. The difference is not just the ending value. Strategy A spent time much deeper underwater, which is exactly when many investors would have bailed.

Why this matters: the market does not reward you for being theoretically right if you cannot survive the path. This is one reason drawdown-aware frameworks often pair well with compound growth and asset allocation. Compounding is powerful, but only if the compounding engine stays intact.

Maximum drawdown vs. drawdown duration

Maximum drawdown is the largest peak-to-trough decline over a period. It answers: how bad did it get? Drawdown duration answers: how long did it stay bad? Investors often fixate on the first and ignore the second, but duration is what turns a bad quarter into a strategy-killing experience.

Chekhlov et al. introduced a family of drawdown measures that go beyond a single worst loss and capture the distribution of losses over time ^[1]. That matters because two strategies can share the same maximum drawdown and still feel completely different. One may recover in weeks; the other may spend years underwater. The second is usually harder to own, harder to rebalance into, and more likely to be abandoned at the bottom.

Hurst et al. showed that drawdown-based risk control can be especially useful in trend-following portfolios because it reduces the chance that a strategy’s inevitable losing streak becomes a catastrophic behavioral event ^[2]. That is not a niche point. It is the difference between a process you can follow and one you will sabotage.

Common mistake: investors often compare strategies by CAGR and volatility, then discover too late that one of them spends half its life underwater. That is not a minor detail. It is the strategy.

Historical drawdowns: the market’s real stress test

Historical drawdowns are useful because they anchor expectations in actual market behavior rather than in backtest optimism. The S&P 500 has experienced several severe drawdowns, including the 2008 financial crisis and the 2020 pandemic shock. Official index history and market data providers show that the index has endured losses large enough to test even disciplined investors ^[5][6].

The table below summarizes major drawdown episodes for broad U.S. equities and a few widely discussed strategy types. Where the figures are from public index history, they are sourced from the cited data providers. Where the figures are strategy-level comparisons, they are illustrative reference points meant to show how path risk differs across styles, not audited performance.

Two lessons stand out. First, broad equity drawdowns are not rare edge cases; they are part of the asset class. Second, a strategy can have a tolerable maximum drawdown and still be miserable if the recovery is slow. That is why drawdown duration belongs in the same conversation as return.

For investors who want to understand how market structure and execution can worsen the experience of a drawdown, the mechanics matter too. Wider spreads, forced selling, and liquidity gaps can turn a paper loss into a realized one. If you want a deeper primer on that side of the ledger, see transaction costs and slippage and market orders vs. limit orders.

What investors get wrong about drawdowns

The biggest mistake is treating drawdown as a postmortem statistic. By the time you calculate it, the damage is already done. The better use of drawdown is forward-looking: it should shape how much you allocate, how much you size, and how much pain you can tolerate before you intervene.

Here are the three recurring errors.

1. Confusing volatility with pain. Volatility is dispersion. Drawdown is loss from a prior high. A strategy can be volatile without ever suffering a catastrophic peak-to-trough decline, and it can be relatively quiet right up until it breaks your heart.
2. Ignoring duration. A shallow but endless drawdown can be more damaging than a deeper one that recovers quickly. Investors do not quit because a strategy is mathematically imperfect; they quit because it feels broken.
3. Assuming their future self will be more patient than their present self. This is the behavioral trap. In calm markets, everyone says they can tolerate a 30% drawdown. In a real drawdown, that confidence evaporates.

This is where survivorship bias becomes relevant. The strategies you see celebrated are often the ones that survived. The ones that failed may have done so because drawdowns were too deep, too long, or both. If you only study winners, you underestimate the true cost of staying in the game.

Practical takeaway: if a strategy’s drawdown profile would have forced you to sell in the past, it is not your strategy. It is a backtest you admire from a distance.

A framework for finding your personal “uncle point”

Your uncle point is the drawdown level at which you would stop following the strategy. The term is blunt on purpose. It forces honesty. If you do not know your uncle point, you are likely to discover it during a panic, which is the worst possible time to make a decision.

Use this three-step framework.

Now turn that into a decision tree.

How drawdowns should affect position sizing and allocation

Position sizing is where drawdown thinking becomes practical. If a strategy can lose 30% in a bad stretch, a 5% portfolio weight is a very different proposition from a 30% weight. The expected return may be the same in both cases, but the behavioral and financial consequences are not.

One useful rule is to size positions so that a plausible drawdown does not create a portfolio-level loss you cannot tolerate. For example, if you can emotionally and financially tolerate a 10% portfolio drawdown from a single sleeve, and the sleeve itself can plausibly fall 25%, then the sleeve should not be large enough to create more than 10% portfolio damage. That implies a rough maximum weight of 40% for that sleeve before considering correlation, liquidity, and other holdings. In practice, most investors should be more conservative because losses do not happen in isolation.

Hurst et al. found that drawdown-based risk management can improve the robustness of trend-following portfolios because it reduces exposure when losses deepen and increases it when conditions improve ^[2]. The broader lesson is not “always use trend following.” It is that risk should be sized with the path in mind, not just the average outcome.

Worked sizing example: suppose Strategy X has a historical maximum drawdown of 20%, Strategy Y has 40%, and you want no single sleeve to threaten more than a 6% portfolio drawdown in a bad case. A rough first-pass sizing would be 30% for Strategy X and 15% for Strategy Y, before adjusting for correlation and liquidity. That is not a precise formula; it is a sanity check. The point is to force the portfolio to respect the strategy’s worst plausible behavior.

Illustrative comparison: CAGR is not enough

Below is a simple illustrative comparison. It is not actual performance data. It is a teaching example designed to show why a lower-return strategy can be more usable if its drawdowns are shallower and shorter. Assumptions: 10-year horizon, annual rebalancing, no taxes, no transaction costs, and a constant risk-free rate of 0% for simplicity.

Even in this simplified setup, Strategy A is not automatically better. If you abandon it during a 35% drawdown, its theoretical edge disappears. Strategy B may deliver less on paper but more in reality because you can actually own it.

This is the part investors get wrong: they optimize for the spreadsheet, then live inside the behavior. The spreadsheet does not quit. You do.

How to use drawdown metrics in strategy evaluation

When you evaluate a strategy, ask four questions in this order:

1. What is the maximum drawdown?
2. How long does recovery usually take?
3. How often does the strategy experience losses of that size?
4. Would I still hold it if the drawdown happened in year one?

If the answer to the last question is no, the strategy is too aggressive for you, regardless of its CAGR. That is not a moral failure. It is a fit problem.

For a more complete evaluation framework, pair drawdown analysis with benchmarking and backtest checklist. A strategy can look impressive against the wrong benchmark or with hidden assumptions. Drawdown metrics help expose that, but only if the data are honest and the comparison set is sensible.

Editorial judgment: the best strategy is not the one with the highest return. It is the one whose return you can actually capture. That usually means accepting less upside in exchange for a drawdown profile you can survive.

So what

Returns tell you what a strategy earned. Drawdowns tell you whether you could have stayed invested long enough to earn it. For real investors, that is the more important question. A strategy that looks superior on CAGR but forces you into a 40% drawdown and a two-year recovery may be inferior to a slower, steadier alternative that you can hold through stress. The right comparison is not return versus return. It is return versus the human and financial cost of getting there.

If you remember only one thing, remember this: define your uncle point before the market does it for you.

Closing thought

Markets do not pay you for being brave in theory. They pay you for staying invested in practice. Drawdowns are the price of admission, but they should be a price you can afford. If a strategy’s path would make you flinch, size it smaller. If its recovery would test your patience, expect that test to arrive. And if the drawdown is deeper than your plan allows, the problem is not the market. It is the plan.