Does Time Really Reduce Investment Risk
A popular claim in investing is that stocks become safer the longer you hold them, a belief that shapes how millions of people allocate retirement savings. The claim is only partly true, and misreading which part is true can lead to an overconfident bet on a "sure thing" that time is supposed to guarantee but does not.
The core principle: narrower percentage range, wider dollar range
Part of the "time reduces risk" claim is genuinely true: the range of plausible annualized returns narrows as the holding period lengthens. If year to year returns are roughly independent of each other, the standard deviation of the annualized return over an n-year holding period shrinks with the square root of time: annualized standard deviation ≈ single-year standard deviation / √n. This is often called the square root of time rule, and it explains why a portfolio's average annual return over a 25-year stretch is far more likely to land close to its long-run expected value than its return over any single one of those 25 years is.
The part of the claim that is false, or at least badly incomplete, is the leap from "the annualized range narrows" to "your risk of a bad outcome disappears." The standard deviation of your total dollar outcome does not shrink with time the way the annualized percentage range does; it actually grows, because a narrower range of annualized outcomes gets applied, through compounding, across a much larger and longer-growing base of dollars. A portfolio that compounds for 25 years has far more absolute dollars riding on even a small percentage-point miss than a portfolio held for a single year has riding on a much larger percentage-point miss, and multiplying a small range by a large, compounding base can still produce a wide dollar range at the end.
The math: two worked examples
Example 1: the annualized range narrowing with time. Suppose a stock portfolio has an 18% annual standard deviation. Over a single year, a one standard deviation range spans plus or minus 18 percentage points around the mean. Applying the square root of time rule over a 25-year holding period: annualized standard deviation ≈ 18% / √25 = 18% / 5 = 3.6%. The range of plausible annualized outcomes over 25 years, one standard deviation wide, has shrunk from roughly plus or minus 18 points down to roughly plus or minus 3.6 points. This is the genuinely true part of the claim: an investor is far more likely to end up close to the long-run average annualized return over a 25-year stretch than over any single year within it.
Example 2: the dollar range widening despite that narrowing. Now translate that narrower annualized range into dollars. Suppose the expected annualized return is 9%, so the one standard deviation annualized range over 25 years runs from roughly 5.4% (9% − 3.6%) to roughly 12.6% (9% + 3.6%). Starting from $100,000: at the low end of that range, 100,000 × (1.054)^25 is approximately $381,000; at the high end, 100,000 × (1.126)^25 is approximately $1,808,000. The one standard deviation dollar range spans roughly $1,427,000, from about $381,000 to about $1,808,000. Compare that to a single one-year holding period on the same $100,000: a one standard deviation range of 9% ± 18% runs from negative 9% to positive 27%, or roughly $91,000 to $127,000, a dollar range of only about $36,000. The 25-year dollar range is roughly forty times wider than the 1-year dollar range, even though the annualized percentage range narrowed by a factor of five over the same stretch. Both statements are simultaneously true, and they are not in tension: the annualized range shrank, and the dollar range grew, because the shrinking percentage range was applied to a vastly larger compounding base.
What market history shows about long horizons
Historical data on rolling multi-year holding periods for diversified U.S. stock portfolios generally supports the narrowing-annualized-range side of the claim: the spread between the best and worst annualized returns across rolling 20 or 30-year periods has historically been meaningfully tighter than the spread across rolling 1 or 5-year periods, and the historical incidence of a diversified stock portfolio underperforming a risk-free benchmark over a rolling 20-plus year period has been low relative to shorter windows, though not literally zero across every market and every era examined globally. This is the empirical basis for the common and broadly reasonable advice to lean toward equities for genuinely long-horizon goals.
What the historical record does not support is the conclusion that a long horizon eliminates the risk of a disappointing ending balance in absolute terms, or that the specific timing of returns within the horizon does not matter. Two investors who both held stocks for 30 years and experienced the identical long-run average annualized return can end up with meaningfully different actual outcomes, and different real-life financial security, depending on precisely when within those 30 years the worst returns happened to occur, an effect closely related to what is often called sequence of returns risk.
There is a further subtlety in the square root of time rule itself worth flagging: it assumes annual returns are independent from year to year, an assumption that empirical evidence suggests is only approximately true. To the extent stock returns show any tendency toward long-horizon mean reversion, a modest, debated pattern documented in some studies of very long U.S. and international return histories, the true narrowing of the annualized range over long horizons could be somewhat greater than the simple square root of time formula predicts; to the extent returns instead show momentum or trending behavior over some horizons, the narrowing could be somewhat less. The square root of time formula is a reasonable, standard approximation used throughout the industry, not an exact law of nature, and treating its output as more precise than the underlying independence assumption warrants is its own small, easily overlooked pitfall.
How this applies in real portfolios
The correct takeaway for portfolio construction is not "a long enough horizon makes risk disappear," but rather "a longer horizon can reasonably justify more risky-asset exposure, provided the horizon is genuine and the investor still plans, diversifies, and rebalances rather than relying on time alone to do the work." A 30-year-old saving for a retirement decades away can reasonably hold a heavily equity-tilted portfolio, not because time has erased the risk of a bad outcome, but because the horizon is long enough to ride out the volatility without a forced sale, and because the higher expected return over that genuine horizon has historically been worth the wider dollar-outcome range it comes with.
The place this distinction matters most is near a goal date, particularly retirement. An investor with a 30-year horizon who has stayed fully allocated to equities the entire time and is now five years from retiring is, at that point, effectively a short-horizon investor for the portion of the portfolio needed in the near term, regardless of how long ago the investing career began; the annualized-range-narrows logic that justified the earlier heavy equity tilt no longer applies to money that will be spent within a few years. This is the entire rationale behind gradually shifting a portfolio toward bonds and cash as a goal date approaches, commonly implemented through a target-date or glide-path structure.
Actionable breakdown
- Treat "time reduces risk" as narrowing the annualized range only.
- It does not shrink the range of ending dollar outcomes.
- Both statements can be true at the same time.
- Size equity exposure to your true remaining time horizon.
- A genuinely long horizon can justify more risk.
- Recalculate the true horizon as a goal date nears.
- Do not assume a long horizon guarantees a good ending balance.
- A bad multi-year stretch can occur inside any horizon.
- Diversify and rebalance rather than relying on time alone.
- Manage sequence of returns risk near a spending date.
- Shift toward bonds and cash as the goal approaches.
- Protect near-term spending needs from a bad final stretch.
For a high income professional whose earning years and investing horizon may run three or four decades, the practical implication of the dollar-range math in Example 2 above is to build a plan around a wide, honest range of plausible ending outcomes rather than a single point estimate. Two colleagues who save and invest identically for 30 years, differing only in the specific sequence of market returns they happened to experience, can retire with genuinely different portfolio sizes purely as a result of that sequence, a form of risk no amount of career success or savings discipline fully eliminates. Planning around a range, and building in flexibility on the spending side for whichever end of that range materializes, is a more realistic response than assuming the long-run average return is what will actually show up.
Common pitfalls
The first pitfall is using "stocks are safe over the long run" to justify taking on excessive risk with money that will actually be needed within just a few years, ignoring that a genuinely bad multi-year stretch can occur inside any window, long or short, and that a short window carved out of a long career is still a short window for the money that needs it.
The second pitfall is ignoring sequence of returns risk: a portfolio that experiences its worst losses right before a planned withdrawal, such as the years immediately before or after retirement, suffers a kind of damage that later average returns cannot fully repair, since withdrawals during a depressed portfolio lock in losses that a portfolio without withdrawals would eventually have recovered from.
The third pitfall is conflating a narrower annualized range with a smaller absolute dollar risk, a subtle but consequential confusion that this article's second worked example above shows to be mathematically backward for the dollar figure, even though it is correct for the annualized percentage figure.
A fourth pitfall is treating the square root of time formula as an exact, guaranteed law rather than a reasonable approximation built on an independence assumption that real markets only roughly satisfy, which can lead to overconfidence in exactly how narrow a very long horizon's annualized range will actually turn out to be once real, imperfectly independent market history unfolds.
The most useful practical habit this distinction encourages is separating two questions that are easy to blur together: "how confident should I be in my average annual pace over the long run" and "how confident should I be in my exact ending dollar amount." The historical record supports real, if not absolute, confidence in the first question over genuinely long horizons. It supports considerably less confidence in the second, and a plan that acknowledges that gap openly, rather than quietly assuming it away, tends to age far better once real market history actually arrives.
The bottom line
A longer horizon narrows the range of likely annualized returns but does not shrink, and typically widens, the dollar range of possible ending outcomes, so time reduces relative risk, not absolute risk.
Related reading: withdrawal strategies guide, risk guide, what stocks and bonds have actually delivered, comparing returns across holding periods, drawdown.