Volatility, Drawdown and Time Horizon: Measuring Portfolio Risk
"This investment is risky": the phrase is common, but what does it actually measure? The fear of seeing a portfolio lose value at the wrong moment can be broken down into measurable quantities: volatility, drawdown (maximum loss) and the often-forgotten time horizon. This article defines them, illustrates them with fictional worked examples and recalls their limits.
Volatility: The Size of the Swings
Volatility measures how much an asset's price fluctuates around its trend. Technically it is the standard deviation of returns, usually expressed as an annual percentage. The higher it is, the wider the swings, up and down alike.
A deliberately simplified rule of thumb: with annual volatility of 15 %, roughly 2 years out of 3 the return lands within 15 points of its average, in either direction. This assumes returns are regularly distributed, which markets only imperfectly respect.
- Volatility describes size, not direction: a stock can be very volatile while rising, or very steady while slowly falling.
- It differs across asset classes: equities are generally more volatile than bonds.
- It is not constant: it climbs during periods of market stress.
It is the raw material of many indicators, such as the Sharpe ratio, which sets the return obtained against the risk taken. Our article on the Sharpe ratio covers this use.
Drawdown: The Pain Actually Felt
Volatility is a statistical indicator; drawdown is an experience. It measures the fall from a peak to the trough that follows, before the previous level is regained. Maximum drawdown is the worst such fall over the observed period.
Illustrative example: a portfolio reaches € 10,000 at its high, then slips to € 7,000 before recovering. Its drawdown is (7,000 − 10,000) / 10,000 = − 30 %.
Drawdown has a property many discover too late: the arithmetic of recovery is asymmetric.
| Loss suffered | Gain needed to get back to the starting point |
|---|---|
| − 10 % | + 11 % |
| − 20 % | + 25 % |
| − 30 % | + 43 % |
| − 50 % | + 100 % |
| − 70 % | + 233 % |
These are plain mathematics: after a 50 % loss, you must double to break even. Large falls therefore weigh far more than their percentage suggests.
Also look at the recovery time: how long it takes to get from the peak back to that level. A − 20 % drawdown regained in 4 months is not experienced like a − 20 % that lasts 4 years.
Time Horizon: Time Changes the Risk Profile
The same portfolio does not carry the same risk for someone who needs the money in 2 years and someone who will not touch it for 20. Horizon works in 2 ways.
First, it changes what you can bear. A temporary fall becomes a realised loss only if you must sell during it. The longer the horizon, the less likely you are forced to sell at the worst moment. Conversely, a near-term goal (a property down payment in 3 years, for example) makes a drawdown far more dangerous, because there is no time to recover.
Second, it changes the spread of outcomes. The following is purely illustrative, assuming annual volatility of 15 % and an average return of 6 % a year. These are working assumptions, not forecasts.
| Horizon | Average annualised return | "Usual" range of annualised return |
|---|---|---|
| 1 year | 6 % | about − 9 % to + 21 % |
| 5 years | 6 % | about − 1 % to + 13 % |
| 20 years | 6 % | about + 3 % to + 9 % |
The range narrows with time: extreme gaps smooth out in annualised terms. Do not read this as a promise, though. It guarantees neither a positive return nor the absence of losses, and markets have gone through long periods of decline. Past performance does not predict future performance.
Volatility and Drawdown: Two Complementary Measures
| Criterion | Volatility | Maximum drawdown |
|---|---|---|
| Measures | Average size of the swings | Worst peak-to-trough fall |
| Nature | Statistical, ups and downs | Lived, falls only |
| Limit | Treats + 10 % and − 10 % alike | Depends on the period observed |
A portfolio can show moderate volatility and a deep drawdown if the losses are concentrated in a single episode: it is better to read both together.
How Portfolio Composition Affects These Measures
Diversification acts on both quantities: if holdings do not all rise and fall together, the swings of the whole are smaller than those of its parts. A portfolio concentrated in one sector takes the full force of episodes that hit it.
To go further: the lessons Risk and return, Diversification and Sharpe, Sortino, drawdown, and the article Diversifying an equity portfolio. To see where a PEA (French equity savings plan) stands, the PEA diversification score tool gives a descriptive overview of its allocation.
Limits to Keep in Mind
- The past is not the future. These measures rely on historical data; an episode more violent than anything observed remains possible.
- Correlations change. In a crisis, normally uncorrelated assets can fall together.
- Risk is not only numbers. Liquidity, currency, company-specific risk: some threats escape volatility.
- Tolerance is personal. A − 25 % that looks bearable on paper can be hard to live through.
Key Takeaways
- Volatility measures the size of the swings; drawdown measures the worst fall from peak to trough.
- A − 50 % loss requires + 100 % to be recovered: large falls are asymmetric.
- Time horizon conditions the risk you actually bear: the same portfolio has a different profile at 2 years and at 20.
- The two measures are complementary, and neither guarantees anything about the future.
- Diversification directly influences these quantities, but never removes risk.
This article is for information and education only; it is not personalised investment advice. Figures in the examples are illustrative. Past performance does not predict future performance.