Milevsky Ch. 1: You Will Not Die on Schedule — Planning to a Distribution, Not an Age

"Life expectancy is a terrible planning horizon. Roughly half of you will outlive it, and those are exactly the people whose plans have to work." — the actuarial premise behind Moshe Milevsky's retirement mathematics

Investment Background

Moshe Milevsky is a finance professor at York University in Toronto who spent his career at the seam between insurance mathematics and portfolio theory. His central complaint about the retirement planning industry is not that its arithmetic is wrong. It is that the industry answers a question nobody actually faces.

Ask a financial planner to build a retirement plan and you will almost always be asked one question early: to what age should we plan? The client says eighty-five, or ninety, or "my father died at seventy-eight." The spreadsheet then runs a deterministic projection to that age, the portfolio lands somewhere near zero, and everyone declares the plan sound.

That entire exercise is built on a category error. You do not have a retirement date of death. You have a probability distribution over dates of death, and the distribution is wide, asymmetric, and — this is the part intuition never gets right — it moves as you age.

Consider the numbers a 65-year-old American couple actually faces. Period life tables put remaining life expectancy for a 65-year-old man near 18 years and for a woman near 21 years. Planners hear "age 83 and 86" and plan to 90 to be safe. But for that same couple, the probability that at least one of them is still alive at 90 is roughly 45%. At 95, it is still around 18%. Those are not tail scenarios you can dismiss; they are close to a coin flip and a one-in-five outcome respectively. A plan that fails in those states is a plan that fails almost half the time for the household that matters — the one that is still alive to experience the failure.

The mistake compounds because life expectancy is conditional. A 65-year-old woman expects to live to about 86. But a woman who reaches 86 expects to live to about 93. Survival keeps buying you more survival. The finish line recedes as you approach it, and every plan calibrated to a fixed date silently assumes it does not.

The Wall Street Translation

The Decision You Are Actually Making

The decision is not "what age do I plan to." It is: how much of my spending am I willing to leave exposed to the possibility that I live an unusually long time?

Framed that way, the plan splits into two structurally different problems that most people jam into one portfolio:

Market risk and longevity risk are not the same hazard and do not respond to the same medicine. Holding more equities does nothing about the possibility that you live to 99. Holding more bonds does nothing about it either — it just makes the shortfall arrive slower.

Why the Single-Age Plan Fails in Both Directions

Plan to age 85 and you underprovision. Roughly half of healthy, affluent, non-smoking 65-year-olds — the exact demographic that hires financial planners — will still be alive. Affluence is itself a life-extender: the gap in life expectancy at 50 between the top and bottom income quintiles in the U.S. runs to roughly a decade. The person reading a retirement book is systematically not the average person in the mortality table.

Plan to age 100 and you overprovision. You spend the entire retirement at a 2.8% withdrawal rate instead of 4%, deferring travel, grandchildren's tuition and the good years of health for a scenario that has maybe a 3% chance of occurring. That is not conservatism; it is a 30% permanent pay cut purchased to insure a low-probability state. Milevsky's point is that this is a terrible way to buy longevity insurance — you are self-insuring a pooled risk at enormous cost.

The Third Option Intuition Skips

Neither underprovisioning nor overprovisioning is required, because longevity risk is the rare risk that is genuinely diversifiable across people. You cannot diversify a market crash — it hits everyone at once. But you can diversify lifespan, because when one person in the pool dies early, the capital they no longer need funds the person who lives to 101. That is what Social Security delay does. That is what a deferred income annuity does. It is why the answer to "what age should I plan to" is often "none — pool the tail and plan the middle."

The core portfolio stays where it belongs: broad, low-cost, indexed, and doing the job it is good at, which is beating inflation over decades. Pooling is not a replacement for that core. It is a separate instrument aimed at a separate risk, and Chapters 4 and 5 examine both its real value and its real costs in detail.

Execution Rules

1. Stop asking your planner for a planning age. Ask for a survival probability. Replace "we plan to 90" with "show me the plan's outcome at the 25th, 50th, 10th and 5th percentile of joint survival." If your advisor's software cannot produce that, it is running a deterministic projection dressed as a plan. For a healthy 65-year-old couple, insist that age 95 joint survival be treated as a live scenario, not a footnote.

2. Adjust the table for who you actually are. Population mortality tables include smokers, the uninsured, and people in poor health. If you are a college-educated non-smoker with normal blood pressure and enough assets to be reading this, add roughly three to five years to the naive table before you plan. Underestimating your own longevity is the single most common input error in retirement planning.

3. Plan for the couple, not the individual. Household plans must run on joint-and-last-survivor survival, because the money has to last until the second death. The joint number is dramatically longer than either individual's. A couple whose individual expectancies are 86 and 88 has a meaningful chance of needing income into the late nineties.

4. Re-run the horizon every three to five years, not once. Because life expectancy is conditional, the horizon lengthens as you survive. A plan built at 65 that assumed 25 more years should be rebuilt at 75 assuming roughly 14 more — not 15 fewer. Treat the plan as a rolling forecast, not a one-time projection.

5. Split spending into essential and discretionary before choosing any product. Write down the annual number below which your life materially degrades — housing, food, insurance, medical, utilities. That floor is the only part of your spending that needs protection against extreme longevity. Everything above it can and should ride on the index portfolio, where it is flexible and can be cut in a bad decade.

Retirement Application

For a typical retiree the practical output of this chapter is a two-line balance sheet, not a product purchase.

Line one is the essential floor: what you must spend, matched against what you already receive for life regardless of how long you live. For most American households that means Social Security, and for some a defined-benefit pension. If your essential floor is $52,000 and your combined Social Security is $46,000, your uncovered longevity exposure is $6,000 a year — small, and manageable inside the portfolio. If your floor is $75,000 and Social Security is $38,000, you have $37,000 a year of spending that depends entirely on your portfolio surviving both markets and an unknown lifespan. Those are radically different situations, and no single withdrawal rate serves both.

Line two is the flexible layer: travel, gifts, hobbies, the things that make retirement worth having. This layer belongs in the low-cost indexed core, where it can grow, where fees are 5 basis points instead of 200, and where you can throttle it down in a bad year without eating.

The single highest-value action available to most American retirees follows directly. Delaying Social Security from 62 to 70 raises the inflation-indexed benefit by roughly 76%, and it is priced off population mortality — meaning a healthy retiree with above-average longevity gets a genuinely favorable deal. It is inflation-linked, backed by the federal government rather than an insurer, and costs no commission. Before considering any commercial longevity product, this one should be exhausted.

Risk Management

• The overprovisioning trap is a real cost, not a safe default. Cutting spending from 4% to 2.8% for thirty years to insure a 3% scenario is a poor trade. Underspending is the most common way retirees damage their own retirements, and it is invisible because nothing bad appears to happen.
• Do not confuse a long horizon with a need for more equities. Extending the planning horizon from 85 to 95 does not automatically justify a higher stock allocation. It justifies more inflation-protected lifetime income, which is a different fix.
• Health-based optimism cuts both ways. People with genuinely shortened life expectancy — active cancer, advanced cardiac disease, ESRD — should plan shorter and claim Social Security earlier. The mistake is applying "my father died at 72" to yourself when your own health profile is unremarkable; one ancestor is not a mortality table.
• Cognitive decline is the horizon risk nobody models. Roughly a third of people over 85 have significant cognitive impairment. A plan that requires annual tactical decisions at 88 is a plan that fails. Simplicity in old age is a risk control, not an aesthetic preference.

What This Chapter Cannot Do

This chapter cannot tell you how long you will live, and no chapter can. It also cannot tell you the right amount of longevity insurance for your household — that depends on your essential floor, your existing pension income, your health, your bequest intentions and your spouse's situation, and Chapters 4 and 5 deal with the costs and the crossover.

What it does not do at all is argue against indexed investing. Nothing here is a case for replacing a diversified low-cost portfolio with insurance products. The portfolio core remains the engine; the argument is only that the engine was never designed to solve the problem of not knowing when you die, and that trying to make it do so — by hoarding, or by planning to an arbitrary age — is where retirements quietly go wrong.

Key Takeaway: You will not die on schedule. Half of the healthy 65-year-olds who plan to age 85 will still be here, and life expectancy lengthens as you survive. Stop planning to a date; plan to a distribution, cover the essential floor with income that lasts as long as you do, and leave the flexible spending in a low-cost index core where it belongs.