Volatility numbers are abstract. The practical test is what your holdings would lose in a real crash, in dollars, and whether you would hold on. Here is how to run it.
What is your own portfolio exposed to? Check it with Exent →
A portfolio is too risky when its likely worst drop is bigger than you can hold through. Risk is personal. The same portfolio can suit a 30‑year‑old with a steady salary and be far too risky for someone who needs the money in three years.
The damage is rarely the drop itself. It is selling near the bottom and missing the recovery. So the useful question is not “how volatile is my portfolio?” but “what would it lose in a bad crash, and would I sell?”
Both, for different reasons. Volatility is the typical size of yearly swings around the average. Maximum drawdown is the largest fall from a peak to a later low. Volatility tells you how bumpy the ride is. Drawdown tells you how deep the worst hole was.
| Holding (Exent’s data) | 5-year volatility | Beta | COVID crash, 2020 | 2022 bear market |
|---|---|---|---|---|
| Johnson & Johnson (JNJ) | 16.8% | 0.21 | −25.9% | −13.9% |
| Dividend ETF (SCHD) | 14.4% | 0.69 | −32.7% | −18.9% |
| S&P 500 (SPY) | 17.0% | 1.00 | −34.1% | −25.4% |
| Nasdaq-100 (QQQ) | 22.3% | 1.22 | −28.6% | −34.6% |
| NVIDIA (NVDA) | 51.1% | 2.21 | −37.6% | −61.8% |
| Tesla (TSLA) | 58.6% | 1.90 | −60.6% | −47.6% |
Crash figures are peak to trough, as of Sep 30, 2026. Two lessons stand out. First, low volatility did not mean no loss: SCHD fell 32.7% in the COVID crash. Second, the order changes by crash. QQQ fell less than SPY in 2020 and much more in 2022. That is why a single number cannot answer the question, and why a replay across several crashes can. Per‑holding pages such as NVIDIA and QQQ show these figures.
Multiply your portfolio value by each crash loss. Seeing dollars, not percentages, is the most honest risk test there is. An illustrative $250,000 portfolio that simply tracked the S&P 500:
| Crash | S&P 500 (SPY), peak to trough | Paper loss on $250,000 (illustrative) |
|---|---|---|
| Global financial crisis, 2007–09 | −56.5% | −$141,175 |
| COVID crash, 2020 | −34.1% | −$85,250 |
| Inflation bear market, 2022 | −25.4% | −$63,400 |
| Trade-war selloff, 2018 | −20.2% | −$50,450 |
| Tariff selloff, 2025 | −19.0% | −$47,500 |
Index drawdowns are Exent’s data; the dollar column is illustrative arithmetic, not a forecast. A stock‑picking portfolio usually needs a bigger buffer. In each of these five crashes, the median large US stock fell more than the index, and in 2022 338 of 464 large stocks fell more than SPY. Replay your own holdings with the crash test or by hand with the method in portfolio stress test.
Losses and gains are not symmetric. A 25% loss needs a 33% gain to get back to even. A 50% loss needs 100%. An 80% loss needs 400%. This is plain arithmetic, explained in loss recovery math. It means a portfolio that falls less in crashes can end up ahead of one with a higher average return, and it is easier to stay invested in. Recovery time matters too: a portfolio that takes years to regain its peak tests your patience as much as your nerves.
Start with the positions that drive most of the crash loss, measured as weight times loss. Then consider what behaved differently in stress. In the same data, in the 16 months since mid‑2016 when SPY fell more than 4%, it averaged −6.7%. T‑bills (BIL) averaged +0.2%, gold (GLD) −0.4% and 7–10 year Treasuries (IEF) −0.9%. In 2022, though, IEF lost 15.2% for the year, so bonds are not a guaranteed cushion.
Mixing assets changed the ride in a backtest on the same data. A hypothetical 60/40 mix of SPY and the AGG bond fund, rebalanced monthly from Jul 2016 to Sep 2026, had 10.1% volatility and a 20.0% maximum monthly drawdown, against 15.2% and 23.9% for SPY. It also returned 9.5% a year against 15.2%. Lower risk came with lower return. Approaches to cutting risk without a big sale are in how to reduce risk without selling.
Look at three numbers: volatility for everyday swings, maximum drawdown for the worst past fall, and a crash replay of your current holdings through several past crises. Turn the crash loss into dollars.
In Exent’s data the S&P 500 fell between 19% and 56% from peak to trough in the five crashes since 2007. A portfolio of individual stocks usually falls more than that.
It depends on when you need the money and whether you could hold through a fall of a third to a half without selling. For long horizons many investors accept it; for near-term needs it is often too much.
Not by itself. Volatility shows how bumpy the ride is. What matters more is the size of the worst likely fall compared with what you can tolerate.
Trim the positions that drive most of the crash loss, cap clusters of holdings that move together, and add assets that behaved differently in past sell-offs. Re-run the crash test after each change.
Exent replays your holdings through five past crashes and shows which positions drive the loss.
Educational content, not investment advice. Examples are illustrative; past performance does not guarantee future results.