The One Number That Predicts Whether You'll Panic-Sell: Max Drawdown Explained
5 min read
Loading...
5 min read
On 20 March 2020, IHSG closed at 4,194.94. That was down 33.41% for the year and the lowest the index had been since 2015. Trading halts had already been triggered that month.
By 21 December the same index was back around 6,100.
Anyone who held on had a flat-to-decent year. Anyone who sold in March locked in a third of their money gone. Same index, same nine months, opposite outcomes. Max drawdown is the number that describes the gap between those two people.
Max drawdown is the worst peak-to-trough fall over a period. Not the fall from where you bought. The fall from the highest point your portfolio ever reached, down to its lowest point after that.
(Trough value − Peak value) ÷ Peak value
If your portfolio hit Rp 200,000,000 and later dropped to Rp 140,000,000, your max drawdown is −30%. It stays −30% on the record even if you fully recover, because the point of the number is not where you ended. It is what you had to sit through.
Volatility tells you how bumpy the ride was on average. Max drawdown tells you how bad the single worst stretch got. People tolerate averages. They panic at extremes.
A drawdown and its recovery are not symmetrical, and this is the part most people get wrong.
| Drawdown | Gain needed to get back to even |
|---|---|
| −10% | +11% |
| −20% | +25% |
| −33% | +50% |
| −50% | +100% |
| −70% | +233% |
Losing a third of your money does not need a third back. It needs half. That asymmetry is why avoiding a deep hole matters more than catching a fast rally.
Here is the uncomfortable part. Max drawdown is a statistic about your assets, but its practical value is as a prediction about your behaviour.
Ask yourself honestly: if you opened your app tomorrow and the total was down 35%, what would you actually do? Not what you think a disciplined investor should do. What would you do.
If the honest answer is "sell most of it," then a portfolio with a historical max drawdown of 35% is too aggressive for you, no matter how good its long-run return looks on paper. You will never receive that long-run return, because you will not be holding when it arrives.
This is why the March 2020 example matters more than any formula. The people who got the recovery were not smarter. They were holding something they could stomach.
There is no universal answer, but as rough guidance, and matching the bands the dashboard uses:
Better than −15%. Comfortable. Most people can hold through this without acting.
Between −15% and −30%. Normal for an equity-heavy portfolio. Uncomfortable but survivable if you were expecting it.
Worse than −30%. You need to have decided in advance that you will not sell. A single-asset crypto position can exceed this routinely.
The word "expecting" is doing real work there. A 25% fall you anticipated feels completely different from a 25% fall that blindsided you, even though the number is identical.
Max drawdown on its own can mislead in both directions.
A portfolio measured only across calm years will show a flattering max drawdown simply because it has never been tested. Check what period the number covers. If it does not include a real market shock, it has not told you much.
Equally, a single deep drawdown from a one-off event does not necessarily mean the portfolio is fragile now, particularly if the position that caused it is gone.
Read it next to volatility, beta and your Sharpe ratio. Volatility describes the typical week. Beta describes how much of that week tracked the wider market. Sharpe describes whether you are paid for the turbulence. Max drawdown describes the worst night. You want all four before you decide whether an allocation suits you. None of the four answers a separate question, though: whether the return underneath was actually good. For that you still need something to measure it against.
Your max drawdown sits in the Portfolio Health section of your dashboard, computed from your real holdings history rather than a hypothetical lump sum invested on day one. That's the same reason the number your broker shows isn't your real return either: both only mean something when they're computed from what you actually did, not a shortcut.
Open it, look at the figure, and do the honest test: picture your current total falling by exactly that percentage, and decide now what you would do. If the answer is "sell," you have learned something useful before the market forces the lesson on you. Adjusting an allocation while calm is considerably cheaper than adjusting it in a March.
For where the broader market sits today, the market page tracks current sentiment and volatility.