Why the Order of Returns Matters More Than Average Returns
Why the Order of Returns Matters More Than Average Returns
Most people assume that what matters in investing is the average return over time. If the stock market averages 10% annually, they reason, my portfolio should roughly match that growth. But this logic breaks down in retirement, when you stop adding money and start removing it. The order of returns—which years have gains and which have losses—can make the difference between a portfolio that sustains your lifestyle for 30 years and one that depletes in 15.
Quick definition: The order of returns is more important in retirement because withdrawals lock in losses when the portfolio is shrinking, reducing the capital base available for future recovery and compound growth.
Key takeaways
- A portfolio with +30%, −20%, and +10% (average +6.67%) will end lower if the negative year comes first, vs. if it comes last.
- Withdrawals amplify the damage of early losses because you're removing capital from a shrunken base.
- Two retirees with identical $1 million portfolios and identical 5% average returns can differ by hundreds of thousands of dollars depending on return order.
- Early-retirement losses are harder to recover from than late-retirement losses because there's more time and more withdrawals ahead.
- This is why sequence risk is often called the "order of returns problem."
The mathematical reality: withdrawals make order matter
To understand why order matters, compare what happens when you withdraw money from a shrinking vs. growing portfolio.
Example 1: Bear market first (−20%), then bull market (+20%)
- Start: $1,000,000
- Year 1: Fall 20% to $800,000, withdraw $50,000 → $750,000
- Year 2: Rise 20% to $900,000, withdraw $50,000 → $850,000
Example 2: Bull market first (+20%), then bear market (−20%)
- Start: $1,000,000
- Year 1: Rise 20% to $1,200,000, withdraw $50,000 → $1,150,000
- Year 2: Fall 20% to $920,000, withdraw $50,000 → $870,000
Over two years, the average return is 0% in both cases (−20% and +20% average to 0%). But Example 1 ends at $850,000 and Example 2 ends at $870,000. The difference is $20,000, and it comes purely from the order, not the average.
Why? Because in Example 1, the bear market hit your $1 million portfolio, cutting it to $800,000. Your withdrawal then came from this smaller base. The subsequent bull market (20% gain) was on $750,000, giving you only a $150,000 recovery. In Example 2, the bull market hit your full $1 million first, growing it to $1.2 million before the bear market cut it. The 20% loss then only reduced a larger base, and your remaining capital ($920,000) is 8% bigger than in Example 1.
Why accumulation doesn't have this problem
During your working years, order doesn't matter nearly as much because you add money continuously. If the stock market drops 30% and you have a 401(k), you typically keep contributing $20,000 annually. You buy stocks when they're cheap, and over time, the cheap buys and expensive buys average out. This is dollar-cost averaging, and it's one reason why advisors say "don't time the market; just keep investing."
But in retirement, the equations reverse. You stop contributing. Instead, you withdraw. Now, a 30% drop means:
- Your portfolio shrinks from $1 million to $700,000.
- You still need $50,000 to live on, so you withdraw from the $700,000.
- Your base is now $650,000.
- When the market recovers 30%, it grows the $650,000, not the original $1 million.
You've permanently lost compound growth on the capital you withdrew during the downturn.
A three-year example to see the real impact
Consider two retirees, both age 65, both with $500,000, both needing $20,000 annually. They'll experience the exact same three returns: +12%, −15%, and +18% (average +5%). The only difference is the order.
Retiree A: Good sequence (+12%, −15%, +18%)
| Year | Portfolio | Return | Withdrawal | Remaining |
|---|---|---|---|---|
| 1 | $500,000 | +12% | $20,000 | $540,000 |
| 2 | $540,000 | −15% | $20,000 | $439,000 |
| 3 | $439,000 | +18% | $20,000 | $498,620 |
Retiree B: Bad sequence (−15%, +12%, +18%)
| Year | Portfolio | Return | Withdrawal | Remaining |
|---|---|---|---|---|
| 1 | $500,000 | −15% | $20,000 | $405,000 |
| 2 | $405,000 | +12% | $20,000 | $433,600 |
| 3 | $433,600 | +18% | $20,000 | $491,892 |
Both experienced the same average return (5%), the same withdrawal rate, and the same time period. Yet Retiree A ends with $498,620 and Retiree B ends with $491,892—a difference of $6,728 (1.4% of starting capital). Over a 30-year retirement, the difference compounds. A single year of bad sequence early on can cost six figures in lost compounding.
Why early losses hit so much harder
A key insight: the earlier a loss occurs, the more years remain for you to withdraw from the smaller base. If a bear market cuts your portfolio from $1 million to $700,000 at age 65, you're withdrawing from $700,000 (adjusted for growth) for the next 20+ years. But if the same bear market hits at age 85, you're withdrawing for maybe 5 more years. The impact is proportionally smaller because there's less time ahead.
This is why the first 5–10 years of retirement—the "red zone"—are so critical. A 40% loss at 65 might be unrecoverable. A 40% loss at 80 is painful but survivable, because you need fewer years of returns to make it through.
Real historical example: 2008 vs. 2009
Consider two retirement start dates:
- January 1, 2008: The S&P 500 fell 37% that year. A $1 million portfolio dropped to $630,000, and you still withdrew $40,000 (4% rule). You were left with $590,000 to recover.
- January 1, 2013: The S&P 500 rose 30% that year (after already recovering in 2009–2012). A $1 million portfolio grew to $1.3 million, and after a $40,000 withdrawal, you had $1.26 million to build on.
Both retirees faced identical market returns in the long run (the same 2008–2024 market history). But the 2008 retiree spent 4–5 years underwater, watching their portfolio shrink despite positive years, while the 2013 retiree watched their portfolio flourish. The order of returns created a vastly different retirement experience.
How volatility and sequence interact
Sequence risk is worse in volatile markets. If the stock market goes up 8% every single year for 30 years with no variation, return order doesn't matter—you get the same outcome regardless of when you start. But the market swings wildly (some years +25%, others −20%). These wide swings mean that the order of good and bad years has a huge impact.
A portfolio of stable-value bonds (3% return, 0.5% volatility) has almost no sequence risk because returns are predictable. A stock portfolio (8% average return, 15% volatility) has severe sequence risk because the swings are so large.
The diagram shows the same two-year return sequence in different orders. The bear market followed by bull market (top path) leaves you with $806K and a depleted base. Reverse it—bull market then bear—and you end at $806K with it easier to maintain. Same outcome mathematically, but the path matters because of withdrawals.
The "bell curve" of sequence outcomes
If you run a Monte Carlo simulation of 10,000 possible return sequences, you'll find:
- Some sequences (lucky early gains) produce 95th-percentile outcomes: your portfolio grows to $2+ million.
- Some sequences (unlucky early losses) produce 5th-percentile outcomes: your portfolio runs out at age 85.
- Most sequences (average early returns) produce 50th-percentile outcomes: your portfolio lasts exactly as expected.
Your actual retirement will be one of those 10,000 sequences. You don't know which one. That uncertainty is sequence risk. The best you can do is plan for a realistic range and build a withdrawal strategy that survives the lower percentiles (e.g., 80th or higher).
Why this breaks conventional planning
Many retirement calculators ask: "What average return do you need?" They assume a constant 7% annual return and extrapolate growth. But real markets don't return 7% every year. In reality, some years are −20%, some are +20%, and the order matters intensely.
This is why sophisticated retirement planning uses Monte Carlo simulations or historical backtesting rather than simple average-return calculators. A Monte Carlo simulator runs thousands of return sequences—some with early gains, some with early losses—and tells you how often your plan succeeds across all of them. A historical backtest shows you how your plan would have fared if you retired in any year from 1950 to today.
Common mistakes
Mistake 1: Assuming last year's returns predict next year. If stocks were up 20% last year, many people assume they'll be up 15% this year. But returns have no memory. A down year can follow an up year at any time, regardless of history.
Mistake 2: Waiting for "the right time" to retire. Some people delay retirement hoping to time the market—retire after a big gain, avoid retiring before a crash. But you can't predict crashes. If you need to retire at 65, the optimal decision is usually to retire and then manage sequence risk through asset allocation and withdrawal strategy, not to guess when the market will be kind.
Mistake 3: Misunderstanding "average return." A portfolio with +30%, −20%, and +10% returns (average +6.67%) is not the same as a portfolio with +6.67% every year. The first portfolio will end lower if the losses come early.
Mistake 4: Ignoring the impact of withdrawals on sequence risk. Some people think sequence risk only applies to passive buy-and-hold investors. But withdrawals create sequence risk by reducing the capital base when it shrinks.
Mistake 5: Planning for a single best-case or worst-case scenario. If you plan assuming +8% returns every year, you're ignoring downside. If you plan assuming −30% in year one, you might be too pessimistic. Instead, plan for a range of scenarios and ensure your strategy survives at least the 80th percentile.
FAQ
Can I reduce sequence risk by choosing investments carefully?
Partially. Stocks have higher long-term returns but higher short-term volatility, increasing sequence risk. Bonds have lower volatility, reducing sequence risk but offering lower long-term growth. You can't eliminate sequence risk entirely, but you can reduce it by holding bonds early in retirement, rebalancing, and using a lower withdrawal rate.
Is sequence risk different for everyone?
Yes. Your sequence risk depends on your withdrawal rate, your life expectancy, your asset allocation, and (most importantly) the inflation-adjusted spending you need. A retiree withdrawing 2% is far less exposed to sequence risk than one withdrawing 6%, because a market downturn has less impact relative to a growing portfolio.
How do I plan if I don't know future returns?
Use historical simulations (backtest against 1950–today) or Monte Carlo analysis (simulate 10,000+ random return sequences based on historical distributions). Both give you a range of outcomes and a success rate, letting you see how robust your plan is.
If sequence risk is so important, shouldn't I own mostly bonds?
Not necessarily. An all-bond portfolio has minimal sequence risk but delivers 2–4% returns, forcing you to withdraw from capital in ways that still create long-term depletion. Instead, balance sequence risk against longevity risk (living longer than expected). A 60% stock, 40% bond portfolio or a rising equity glide path often optimizes both.
What return sequence should I assume when planning?
Assume a range. Run your plan assuming the 2008 sequence (early bear market), the 2013 sequence (early bull market), and average sequences in between. If your plan survives 80%+ of these scenarios, it's reasonably robust. Tools like FIREcalc or Vanguard's Retirement Nest Egg Calculator do this automatically.
Does rebalancing help with sequence risk?
Yes. Rebalancing forces you to sell when stocks are high and buy when stocks are low, which cushions sequence harm. A disciplined rebalancer selling 40% bonds to buy cheap stocks during a bear market is essentially selling high and buying low, reducing portfolio damage over time.
Related concepts
- The Retirement Red Zone — The first 5–10 years when sequence risk is most acute.
- Same Average, Different Outcome — Concrete examples of how order determines portfolio success or failure.
- Bear Markets Early in Retirement — Specific strategies for surviving an early downturn.
- Withdrawal Strategies — How to design a withdrawal plan that accounts for sequence risk.
Summary
The order of returns matters far more than the average return in retirement because withdrawals lock in losses when the portfolio is shrinking. A bear market at age 65 is far more damaging than one at age 80, because the retiree at 65 has 20+ years of withdrawals ahead from a reduced capital base. Real-world examples (2008 vs. 2013 retirements) show that identical long-term returns produce vastly different outcomes depending on the sequence. Protecting against sequence risk requires conservative early allocation, measured withdrawal rates, and ongoing rebalancing.