After running a DCA backtest, most people's eyes go straight to the annualized return. "Oh, 12%, not bad." Then they close the tab, satisfied they've done their homework.
But have you ever stopped to ask: what did you actually live through to get that 12%?
Maybe somewhere in the middle, your portfolio dropped 35% from its peak. You put in 100K, and at the worst moment it was worth 65K. Every morning you opened the app and saw red. You started wondering if you'd made a mistake. Your spouse asked "how's that investment going?" and you mumbled something noncommittal.
That's maximum drawdown. It's not an abstract statistic. It's the most painful downhill stretch on your entire return curve.
What It Actually Measures The definition is straightforward: the largest peak-to-trough decline across all of history. Note the word "after" — it's not the gap between any two random points. It's: you climbed to a summit, then fell from that summit into the deepest valley that followed. The biggest such fall across all historical peaks is your maximum drawdown.
An example. Your account goes from 100K to 150K, then drops to 90K, then climbs to 200K. The drawdown from that 150K peak is 40%. But if earlier there was also a stretch from 120K down to 80K, that's 33%. You take the larger one — max drawdown is 40%.
The question it answers: if you had the worst possible luck and went all-in at the exact top, how much would you have lost at worst?
Why It Matters More Than Return Because humans aren't rational agents.
There's a finding in behavioral finance that's been repeated so often it's almost a cliché, but it never stops being true: the pain of a loss is roughly 2 to 2.5 times the pleasure of an equivalent gain. The happiness from making 10K doesn't offset the sting of losing 5K.
What does this mean in practice? A strategy with 15% annualized return but 50% max drawdown feels far worse than one with 10% return and 20% drawdown. The former has higher expected value, but most people bail at the 30% mark. They're not "irrational." They're just human.
One core reason DCA gets recommended is that it naturally smooths drawdowns. You buy a fixed amount each month; when prices fall, that same money purchases more shares; when prices recover, your cost basis is lower than someone who went all-in. But "naturally smooths" doesn't mean "no drawdown." If you were doing DCA into the CSI 300 throughout 2018, you still sat through 30%+ unrealized losses. The difference is that a lump-sum investor might have seen 32%, while the DCA investor saw 25%. Seven percentage points — which can be the difference between "I can hold" and "I can't take this anymore."
Backtested Drawdown ≠ Lived Experience Here's a subtle point: the backtest tells you max drawdown was 25%, but your actual experience might feel far worse than 25%.
Why? Because backtests give you god's-eye view. You're looking at a complete curve, and you know it recovered. But when you're in it, you don't know. In October 2022, the CSI 300 had fallen from 4900 in January to 3500. You'd been doing DCA for two years, sitting on 28% unrealized losses. Every financial headline screamed "economic slowdown," "confidence crisis," "foreign capital fleeing." You didn't know if 3500 was the bottom. You didn't even know if it would fall to 3000.
In a backtest, 25% is a number. In reality, 25% is six months of anxiety, self-doubt, and the recurring thought of "maybe I should sell now and buy back after it bottoms."
My suggestion: when you look at backtest results, multiply the max drawdown by 1.5 and treat that as the "psychological pressure you'll actually endure." If that multiplied number makes you uncomfortable, reduce your position size or switch to a less volatile instrument.
Drawdown Profiles Across Strategies Since this site offers both DCA backtesting and grid trading tools, it's worth comparing their drawdown characteristics.
DCA drawdowns come from a single source: the overall market declining. You have no "active operations" that amplify or mitigate it. The only lever is to keep buying (or use a moving-average deviation approach to buy less at highs and more at lows, shaving a few points off). DCA max drawdown roughly equals the underlying index's max drawdown over the period, multiplied by a discount factor (since you weren't all-in at the peak).
Grid trading drawdowns are more complex. First, there's the floating loss on your holdings when the market drops (similar to DCA). Second, there's "grid breach" — price falls below your range's lower bound, you're fully loaded with a losing position, the grid stops turning, no new profits are generated, and there's no stop-loss mechanism. That second scenario is unique to grids, and it's why grid trading max drawdowns are often larger than DCA drawdowns.
An interesting asymmetry: in ranging markets, grid drawdowns can be smaller than DCA (because the grid automatically reduced position at highs). But in sustained downtrends, grid drawdowns can be larger (because the grid kept adding positions as price fell — buying more the lower it went). The two strategies' "worst cases" show up in different market environments.
How to Use This Number for Decisions A few practical suggestions.
First, treat max drawdown as your "admission ticket." How much drawdown you can stomach determines how large a position you can take and how aggressive your strategy can be. If you can tolerate 20% account losses at most, don't go full-size on an instrument with 40% annualized volatility.
Second, when comparing strategies, don't just look at return — look at the return-to-max-drawdown ratio. A strategy with 12% annualized return and 15% max drawdown may be superior on a risk-adjusted basis to one with 18% return and 45% drawdown. The former earns 0.8% of return per 1% of drawdown endured; the latter only earns 0.4%.
Third, pay attention to drawdown duration, not just magnitude. Dropping 25% and recovering in two months is a fundamentally different psychological experience from dropping 25% and grinding sideways for two years before getting back to even. If your backtest tool shows drawdown recovery time, look at it.
Fourth, always remember: the backtested max drawdown is "the worst thing that has already happened." The future can absolutely produce something worse. Pre-2015 backtests wouldn't have told you how brutal the 2015 crash would be. Pre-2019 backtests wouldn't have captured the 2020 pandemic shock. On top of the historical max drawdown, give yourself 30%-50% extra margin.
One Last Thing The most expensive mistake in investing isn't picking the wrong asset or miscalculating your return. It's selling on the night when the drawdown is at its deepest.
Understanding max drawdown isn't about eliminating fear — fear is normal. It's about making sure that when the fear hits, you can tell yourself: "this magnitude of decline is within my plan." With a plan, you don't panic. Without panic, you hold. And if you hold, that beautiful return curve in the backtest has a chance of becoming a real number in your account.