A retirement plan can survive mediocre returns and still fail under perfectly respectable ones. The difference is often not the average return. It is when the bad years arrive.
This matters most near the point when a portfolio changes jobs. Before retirement, a falling market is unpleasant but contributions continue buying assets. After withdrawals begin, the same decline becomes more expensive: shares must be sold while prices are down, leaving fewer shares to participate in the recovery.
The 2026 Social Security actuarial tables give the issue its proper scale. A person reaching 65 this year has a projected unisex cohort life expectancy of 20.5 more years—19.2 years for men and 21.8 for women. Those are averages, not planning ceilings. A household with two people must also allow for the real possibility that one life extends well beyond them.
Retirement is therefore a long-horizon problem with a short-horizon vulnerability. The first five years deserve their own test.
The average can conceal the damage
Consider two hypothetical retirees. Each begins with $1 million and withdraws $50,000 at the start of every year. Each experiences the same ten annual returns: -20%, -10%, 5%, 10%, 15%, 20%, 8%, 6%, 4% and 2%. Only the order changes.
When the two losses arrive first, the portfolio finishes year 10 at about $670,000. When the same returns arrive in reverse order, it finishes at about $867,000. The return set is identical. The year-10 difference is $197,000.
The arithmetic is not mysterious. After the first withdrawal and a 20% decline, the weak-first portfolio ends year one at $760,000. The next $50,000 withdrawal is then taken from a smaller base before another loss. By the time positive returns arrive, less capital remains to compound.
An investor who is still accumulating can wait and keep contributing. A retiree withdrawing from the account is simultaneously reducing the number of assets that can recover. This is sequence-of-returns risk: the path matters because cash is leaving the portfolio.
Why the first five years carry extra weight
A poor return in year 18 may still be painful. It is usually less destructive if the portfolio has already had years to grow and the spending plan has adjusted to reality. A poor return in year one can alter every withdrawal that follows.
This does not make the first five years uniquely predictable. It makes them uniquely consequential. The practical response is not to forecast the next bear market. It is to keep an early drawdown from forcing a permanent sale of depressed assets.
That distinction is important after a strong market. Rising balances can make a plan appear safer while also increasing the share of the portfolio exposed to equities. Investor.gov notes that market gains can push an allocation away from its intended risk level and identifies rebalancing as the process of restoring that mix. The relevant question is not whether the market has risen too far. It is whether the portfolio now holds more risk than the spending plan can absorb.
Separate spending from return assumptions
Many retirement plans begin with one expected return and one expected inflation rate. Those assumptions may be useful for a long projection, but they are too smooth for the transition into withdrawals.
A better first-five-year test begins with cash flows:
- How much spending is essential after Social Security, pensions and other reliable income?
- Which expenses are discretionary enough to reduce for a year or two?
- How much must come from tax-deferred accounts, taxable accounts and cash?
- Which large expenses—home repairs, health care, family support—are plausible but irregular?
Once those figures are visible, a household can decide how many years of net withdrawals should sit outside volatile assets. There is no universal number. Holding too little creates forced-sale risk; holding too much can weaken long-run growth and purchasing power. The correct reserve is tied to the household’s spending flexibility, guaranteed income and tolerance for changing plans after a decline.
Build a bridge, not a bunker
The useful structure is a bridge between current spending and long-term assets. Cash and short, high-quality bonds can fund near-term withdrawals. Intermediate bonds can cover later planned years. Equities can retain the longer job of supporting a retirement that may last two or three decades.
This is not an argument for abandoning stocks at retirement. The Social Security horizon argues against that. A portfolio designed only to avoid short-term volatility may struggle to support decades of inflation-sensitive spending. The purpose of the bridge is narrower: it gives the equity allocation time to recover without immediately being asked to fund every bill.
Asset location matters as well. Selling from a traditional IRA, a taxable account and a Roth account can produce different tax consequences. The IRS requires most traditional IRA owners to begin required minimum distributions at age 73, and the annual amount is generally based on the prior year-end balance divided by a published life-expectancy factor. Those rules can reduce withdrawal flexibility later, which is another reason to map account types before the first difficult market year rather than during it.
Use guardrails that can actually be followed
A plan becomes sturdier when it states in advance what will happen after a decline. “We will spend less if necessary” is not yet a rule. A guardrail might specify that a discretionary inflation increase will be skipped after a negative portfolio year, that travel spending will fall if the withdrawal rate crosses a chosen threshold, or that equities will not be sold for ordinary spending while a designated reserve remains.
The rule should be simple enough to use when markets are loud. It should also distinguish essential spending from optional spending. Cutting both by the same percentage may look neat in a spreadsheet and prove impossible in a household.
Rebalancing rules deserve the same clarity. If equities fall below the target allocation, the plan should say whether maturing bonds or excess cash will replenish them. If equities rise well above target before retirement, the plan should say how gains will refill the spending bridge. Decisions made in advance are less likely to become market predictions in disguise.
The five-year retirement test
Run the plan once with its normal assumptions. Then run it again with an intentionally uncomfortable opening: two negative equity years, no inflation relief on essential expenses, and one large unplanned cost. Do not ask whether that exact sequence will occur. Ask what it would force you to sell, cut or postpone.
A useful test produces three answers:
- Liquidity: which account funds the next 12 months without selling depressed assets?
- Flexibility: which spending can change, by how much, and for how long?
- Recovery: what rule moves the portfolio back toward its intended allocation?
If those answers are vague, the plan is relying more heavily on return order than the headline projection admits.
The Log: protect the transition
The risk is not that average returns are useless. It is that averages do not describe the order in which a retiree must live through them.
The 2026 longevity data point to a retirement horizon measured in decades. The illustration shows why the opening years can still dominate the outcome. A good plan must respect both facts at once: enough long-term growth to fund a long life, and enough near-term structure to avoid selling that growth engine at the wrong time.
Before changing the whole portfolio, identify the first five years of net withdrawals. That is where the retirement plan stops being an estimate and becomes an operating system.
What would change the view
- Guaranteed income that covers a larger share of essential spending.
- A later retirement date or continued earned income that reduces early withdrawals.
- A materially different asset allocation, tax position or required-distribution schedule.
- A spending plan flexible enough to absorb a prolonged decline without forced sales.
Jasper Kellan is editor of The Perspective Log. This article is general information, not personal investment, tax or legal advice.
Primary sources: Social Security Administration, Actuarial Note No. 2026.2, June 2026; Social Security Administration, 2026 Trustees Report, mortality assumptions and life expectancy; Investor.gov, Asset Allocation and Diversification; Internal Revenue Service, Publication 590-B, Distributions from IRAs.
