People use "volatile" and "risky" as if they were the same word. They are two different things, and confusing them leads to two opposite mistakes.
Volatility is how much returns vary from year to year. It is the admission price for owning assets that earn more than cash.
Drawdown is how far an asset falls from a peak before it recovers. It is what you actually feel, and what can force you out.
The real risk is not that prices move. It is that a move happens when you are not able, or not willing, to wait for the recovery.
You are paid for bearing volatility. You are not paid for being forced to sell in a drawdown.
The arithmetic of losses
Losses are not symmetrical with gains. A fall of 50% needs a rise of 100% to get back to even.
| Fall from peak | Gain needed to recover |
|---|---|
| 10% | 11.1% |
| 20% | 25.0% |
| 30% | 42.9% |
| 50% | 100.0% |
| 70% | 233.3% |
That asymmetry is why the first rule of long-run investing is to avoid the very large drawdowns, and why a portfolio's worst outcome matters more than its average.
Volatility drag
Even without a single disastrous year, volatility itself costs you. An asset that rises 50% one year and falls 50% the next has an average return of zero and an actual result of -25% (1.5 × 0.5 = 0.75).
A rough rule for the long-run compound growth rate is:
compound return ≈ average return − half the variance
An asset averaging 7% a year with a volatility of 18% compounds at roughly 7% − (0.18² ÷ 2) = 5.4%. The difference, about 1.6 points a year, is the price of the swings. Lower the volatility and, for the same average, the compound return rises.
This is one reason diversification pays. Combining holdings that do not move together reduces the portfolio's volatility without lowering its average return. The saving shows up as faster compounding.
What April 2025 showed
In April 2025, the S&P 500 fell more than 12% in seven days after a tariff announcement, then rose 9.5% in a single day on a pause. Taiwan's market fell a record 9.7% in one day that same week, and Bloomberg reported that leveraged positions were being unwound.
For an investor with a long horizon, steady income and no leverage, those were uncomfortable weeks. For an investor with borrowed money, a short deadline or a plan that depended on a rebound by a date, they could have been permanent. The market was the same. The difference was the structure around it.
For more on what to decide before a shock, see the plan you wrote before the crash.
Strategies that work, and why
1. Match the asset to the horizon. Money needed within a few years belongs in assets whose worst plausible drawdown you can absorb. Money not needed for a decade can bear more. The danger is not holding equities. It is holding them against a near-term liability.
2. Rebalance by rule. A rebalancing rule sells what has risen and buys what has fallen. Over time, that tends to add return in volatile, mean-reverting markets and, more reliably, keeps the risk you intended. It also gives you something to do in a crash other than selling.
3. Hold a cash buffer for spending. The aim is to avoid selling risk assets in a downturn to pay for life. Two years of spending in low-risk assets is a common rule of thumb, not a law.
4. Avoid leverage you can't hold through a drawdown. Borrowing turns a temporary loss into a margin call. The April 2025 unwinding of leveraged bets in Taiwan is a clear recent example of the risk.
5. Diversify by cause of risk, not by number of funds. Holdings that depend on the same thing move together in a crisis. See your biggest position was never a decision.
6. Add money on a schedule. Regular contributions mean you buy more units when prices are low. A caution: in published studies of historical US data, investing a lump sum immediately has beaten spreading it out more often than not, because markets rise more often than they fall. Phasing in is a way of managing regret and behavior, not a way of earning more.
7. Keep costs low. A cost is a guaranteed drag. A 0.8-point difference in annual costs on a 6% return is about US$708,000 on US$1 million over 25 years. See wealth management is five decisions, not five tips.
What not to do
- Judge a plan by one year. A single bad year says little about a 20-year plan.
- Change the plan in the middle of a drawdown. Review after, in calm.
- Treat a recovery as proof you were right. Markets that recovered did not have to.
- Confuse falling prices with a falling business. The two are related but not the same, and the difference is the point of research.
Update, October 2026: what happened next
According to the Bank for International Settlements, the S&P 500 recovered its April 2025 losses by the end of that month and passed its previous peak in just under 20 weeks. Indonesia's market lost more than 11.5% in two days in late January 2026, and Taiwan's brokerage loans were reported up about 160% in the year to mid-2026. The episodes differed. The structure around the investor was again what decided how they felt.
Note on the examples
The recovery table is exact arithmetic. The volatility-drag formula is an approximation that works best for modest volatility. The market episodes are from public reporting; sources are listed in the plan you wrote before the crash. This is education, not investment advice.