How Different Were Historical Outcomes Over 5, 10, 20 and 30 Year Investment Horizons?
One of the most common investing lessons is that a longer investment horizon reduces risk. Most investors have probably heard some version of the advice: "stay invested for the long term." But how much difference does time actually make?
One way to think about the question is whether longer holding periods increased average returns. Another way is to ask whether longer holding periods reduced the gap between good and bad historical outcomes.
For an investor saving for a house in five years, the range of possible outcomes may matter a great deal. For an investor saving for retirement in thirty years, that same range may look very different.
Historical Outcomes
To explore this question, I analyzed every rolling historical investment period available in the dataset using a 100% S&P 500 portfolio. Rather than focusing on a single average return, the figures below show the full distribution of inflation-adjusted outcomes across different investment horizons.
5-Year Horizon
- Worst outcome: -10.3% real annualized return
- 5th percentile: -6.5%
- 25th percentile: +0.4%
- Median outcome: +8.4%
- 75th percentile: +11.8%
- 95th percentile: +20.7%
- Best outcome: +25.3%
10-Year Horizon
- Worst outcome: -3.8% real annualized return
- 5th percentile: -2.4%
- 25th percentile: +3.1%
- Median outcome: +6.7%
- 75th percentile: +11.1%
- 95th percentile: +15.0%
- Best outcome: +17.9%
20-Year Horizon
- Worst outcome: +0.6% real annualized return
- 5th percentile: +2.0%
- 25th percentile: +4.1%
- Median outcome: +7.0%
- 75th percentile: +9.2%
- 95th percentile: +12.5%
- Best outcome: +13.2%
30-Year Horizon
- Worst outcome: +4.3% real annualized return
- 5th percentile: +4.6%
- 25th percentile: +6.1%
- Median outcome: +7.3%
- 75th percentile: +8.1%
- 95th percentile: +9.6%
- Best outcome: +10.1%
The most obvious observation is how dramatically the range of typical outcomes narrowed as the investment horizon increased. Over five years, 90% of historical outcomes fell between -6.5% and +20.7% real annualized returns. Over ten years, that range narrowed to -2.4% and +15.0%. Over twenty years, it narrowed further to +2.0% and +12.5%. By thirty years, 90% of historical outcomes fell within a relatively narrow range of +4.6% to +9.6%.
Interestingly, the median outcome changed much less than the range itself. Median returns ranged from 6.7% to 8.4% across all four investment horizons. The larger effect of time was therefore not higher returns, but greater predictability.
Another observation is how quickly poor outcomes became less severe. Over five years, investors in the 25th percentile barely stayed ahead of inflation, earning just 0.4% real annualized returns. Over thirty years, the 25th percentile outcome increased to 6.1%.
This suggests that the most important benefit of a longer investment horizon may not be maximizing returns. It may be reducing uncertainty and making outcomes less dependent on starting year.
Key Takeaways
- The median outcome changed relatively little, ranging from 6.7% to 8.4% across all investment horizons.
- The range containing 90% of historical outcomes narrowed from -6.5% to +20.7% over five years to +4.6% to +9.6% over thirty years.
- The primary benefit of time was not higher returns, but greater predictability and less dependence on starting year.
How Long Did You Need to Hold 100% Stocks Before They Started Beating 60/40 Even in Poor Historical Outcomes?
One of the most common arguments for a 60/40 portfolio is that investors receive a smoother ride and better protection against poor outcomes. That trade-off is easy to understand over short periods. Bonds have historically reduced volatility and cushioned declines when stocks struggled.
But how long did that advantage persist? One way to think about the question is whether investors received better downside protection from a 60/40 portfolio. Another way is to ask whether there was a point where the higher expected return of stocks became large enough to offset that protection.
In other words: at what investment horizon did 100% stocks start outperforming 60/40 not only in the best historical outcomes, but also in the less favorable ones?
Historical Outcomes
To explore this question, I compared every rolling historical period available in the dataset using Portfolio A (100% S&P 500) and Portfolio B (60% S&P 500 / 40% 10-Year U.S. Treasury Bonds). All returns are inflation-adjusted and expressed as real annualized returns. Each line shows Portfolio A vs Portfolio B.
3-Year Horizon
- P95: 24.9% vs 16.3%
- P75: 13.7% vs 9.5%
- Median: 7.3% vs 5.7%
- P25: 1.9% vs -0.2%
- P5: -11.7% vs -6.8%
Over three years, the traditional diversification story largely held. The 60/40 portfolio provided meaningful protection in poor historical periods. Investors gave up some upside potential, but experienced less severe downside when markets performed poorly.
10-Year Horizon
- P95: 15.0% vs 10.8%
- P75: 11.1% vs 8.3%
- Median: 6.7% vs 5.2%
- P25: 3.1% vs 3.2%
- P5: -2.4% vs -1.6%
By ten years, the trade-off looked much smaller. The downside advantage of 60/40 remained, but only at the extreme lower end of the distribution. Meanwhile, 100% stocks delivered higher returns across most historical outcomes.
30-Year Horizon
- P95: 9.6% vs 7.1%
- P75: 8.1% vs 6.3%
- Median: 7.3% vs 5.4%
- P25: 6.1% vs 4.0%
- P5: 4.6% vs 3.4%
Thirty years produced a very different picture. Not only did 100% stocks outperform 60/40 in median and favorable outcomes, they also outperformed in poor historical outcomes. Even investors in the bottom 5% of historical 30-year periods achieved higher real returns with 100% stocks than with 60/40.
One way to summarize the shift is to look at the 5th percentile outcome. Over three years, the 60/40 portfolio improved the 5th percentile outcome from -11.7% to -6.8%. Over thirty years, the relationship reversed: the 5th percentile outcome increased from 3.4% in the 60/40 portfolio to 4.6% in the 100% stock portfolio.
This suggests that diversification provided meaningful protection over shorter horizons, but that the long-term return advantage of stocks eventually became large enough to dominate even in poor historical outcomes.
Key Takeaways
- Over short investment horizons, 60/40 provided meaningful protection against poor historical outcomes. The 5th percentile improved from -11.7% to -6.8% over three-year periods.
- By ten years, most of that downside advantage had disappeared while 100% stocks continued to deliver higher returns across most outcomes.
- By thirty years, 100% stocks outperformed 60/40 across the entire distribution, including the bottom 5% of historical outcomes (4.6% vs 3.4% real annualized returns).
If You Retire in 15 Years, What Return Should You Assume?
If you plan to retire in 15 years, you need a return assumption for your savings between now and then. A 7% annual return is a common choice. But how often did investors actually earn that over a 15-year period, and how much should you rely on it?
One way to think about the question is whether 7% was achievable historically. Another way is to ask how often it was achieved. A third way is to ask what level of confidence you need before an assumption feels reasonable for your plan.
For some investors, an assumption may be reasonable if it occurred in roughly half of historical periods. Others may want it to hold in three quarters of periods before building a retirement plan around it.
Historical Outcomes
To explore this question, I analyzed every rolling 15-year period available in the historical dataset using a 100% S&P 500 portfolio. Each period represents an investor who stays invested for 15 years from a different starting year. Rather than focusing on a single average return, the figures below show the full distribution of historical outcomes. All figures are inflation-adjusted (real) annualized returns.
- Worst outcome: -0.5% real annualized return
- 25th percentile: +3.2%
- Median outcome: +7.4%
- 75th percentile: +10.5%
- Best outcome: +15.1%
Half of all historical 15-year periods produced real returns above 7.4% per year. However, one quarter of historical periods produced less than 3.2% per year.
The more interesting observation may be that the answer depends on what "reasonable" means to you. If you are comfortable with an assumption that was achieved in roughly half of historical periods, then 7% appears broadly consistent with the historical record. If you prefer an assumption that was achieved in at least 75% of historical periods, a figure closer to 3.2% may be more appropriate.
Fifteen years is long enough for compounding to matter, but still short enough that starting year had a meaningful impact on outcomes. For someone with a fixed retirement date, that spread matters: the same savings could have grown very differently depending on when the 15 years started.
Two Things to Keep in Mind
First, these are real (inflation-adjusted) returns. If your plan uses 7% as a nominal figure, it corresponds to a lower real return after inflation, so the two should not be compared directly.
Second, an annualized return describes a single amount invested at the start and held for the full 15 years. Someone who saves regularly each month has money invested for different lengths of time, so their outcome can differ.
Key Takeaways
- A 7% real return assumption was achieved in roughly half of historical 15-year periods.
- With 15 years until retirement, outcomes varied substantially depending on starting year, from -0.5% to +15.1% per year.
- Whether 7% is a reasonable assumption depends largely on how much uncertainty you are willing to accept in your plan.
Explore Your Own Scenarios
Historical data can help illuminate trade-offs, but it cannot determine which outcome an investor should prefer. These analyses looked at only a few portfolios (100% S&P 500, and a 60/40 stock-bond allocation). Different conclusions may emerge with other asset allocations, bond maturities, treasury bills, gold or small-cap stocks.
Some investors may be comfortable accepting a wider range of possible outcomes in exchange for higher upside potential. Others may prefer to narrow the range of outcomes, even at the cost of lower returns. Use the tool above with your own assumptions.
More analyses
Data and Methodology
The analysis is based on rolling historical periods rather than selected examples. All returns are adjusted for inflation using U.S. CPI data, so comparisons are in terms of purchasing power. Portfolios are rebalanced annually to their target weights. The underlying market data comes from long-term historical datasets compiled by Aswath Damodaran at the NYU Stern School of Business.
- Historical market data rather than forecasts
- Actual historical periods rather than simulations
- Investable asset classes (stocks, bonds, treasury bills, gold) rather than theoretical factors
- Inflation-adjusted returns rather than nominal returns
- Full distributions of outcomes rather than a single average return
Data period: 1928–2025. Fees and taxes are not included. Past results do not guarantee future outcomes. This site is for educational purposes only and is not investment advice. Last updated: November 1, 2026. Read the full methodology or learn more about the project.