Milevsky Ch. 5: The Annuitization Crossover — When Guaranteed Income Beats More Equity

"There comes a point where the extra return you might earn is worth less than the extra income you can be certain of. That point has an age, and it is later than the salesman says and earlier than you think." — the crossover problem

Investment Background

Every retiree eventually faces a choice between two ways of trying to make their money last: earn more, or share more.

Earning more means holding growth assets — a broad, low-cost equity index — and accepting volatility in exchange for a higher expected return. Sharing more means pooling with other retirees so that mortality credits supplement your income. Both raise the sustainable spending rate. They do it through completely different mechanisms, and — this is the crossover insight — their relative value flips at a predictable point.

Equity's advantage is largest when the horizon is long. Over thirty years, the equity risk premium compounds into an enormous edge, and volatility has time to average out. Pooling's advantage is largest when the hazard rate is high, because that is when mortality credits are large. Chapter 2 established that the hazard rate compounds. It follows mechanically that there exists an age — different for each household — at which the marginal dollar is better deployed into pooled income than into more equity.

The industry gets this wrong in both directions, and the errors are not symmetric in cost.

The insurance channel pushes annuitization at 60–65, when the mortality credit is worth roughly a single percentage point a year, the horizon is thirty-plus years, and inflation is the dominant threat. That is early, and it is expensive: you pay decades of fees, surrender flexibility during the years your health and spending are least predictable, and buy protection against a risk that has not yet become the binding one.

The Boglehead channel refuses annuitization entirely, at every age, on the grounds that stocks beat bonds. That is true and irrelevant at 88, when a fifth of the pool will not survive the year, when mortality credits exceed any plausible equity premium, and when the cognitive cost of managing a portfolio is rising steeply.

The Wall Street Translation

The Two Variables That Locate the Crossover

The crossover is not a single universal age. It moves with two things.

Age, through the hazard rate. The mortality credit is roughly 1% a year at 65, 2% at 72, 4–5% at 80, and above 8% at 88. The long-run equity risk premium over bonds is generally estimated around 3–4% a year, and that estimate is uncertain and unrealized in any particular decade. Compare the two series and the crossing occurs somewhere in the mid-to-late seventies for most people — earlier for those in poor financial position, later for those with strong health or ample assets.

Funding ratio, through need. Define funding ratio as guaranteed lifetime income divided by essential spending. A household at 130% — Social Security and pension already exceed essentials — has no crossover at all; every marginal dollar belongs in growth assets, because there is no longevity exposure to close. A household at 55% has a large, permanent exposure and should cross earlier. The crossover is a function of vulnerability, not of age alone.

Funding ratio	Situation	Marginal dollar belongs in
Above 110%	Essentials fully covered for life	Equity index — the exposure is already closed
85–110%	Small uncovered gap	Equity index; revisit pooling at 78–82
65–85%	Meaningful gap	Delay Social Security first; consider pooling mid-seventies
Below 65%	Large permanent exposure	Close the gap as a priority; pooling likely justified sooner

What the Crossover Does Not Mean

It does not mean that at the crossover age you annuitize the portfolio. It means the marginal dollar changes destination — and only up to the point where the essential floor is covered. Once the floor is closed, the crossover is irrelevant again and the answer reverts to the index core, permanently.

This is the point at which the retirement framing mandate matters most. Nothing about the crossover argues for a majority-annuitized balance sheet. Discretionary spending, inflation protection over a possible thirty-year retirement, and any bequest are all funded better by a low-cost global equity and bond index portfolio than by an insurance contract backed by an investment-grade bond book with fees layered on top. The crossover applies to a bounded slice, typically well under a third of investable assets, and it applies once.

The Honest Case Against Crossing At All

There are households for whom the correct answer is never to annuitize, and it is worth stating them clearly rather than treating pooling as inevitable.

• Already floored. Two people with $70,000 of combined Social Security and $58,000 of essential spending have no gap. Buying an annuity here converts equity into bonds with a fee, and loses the bequest.
• Bequest-driven. A household whose central financial goal is transferring assets to children or a charity is optimizing a different objective. Pooling directly destroys it.
• Impaired health. Standard payout pricing assumes above-average longevity. If yours is genuinely below average, you are subsidizing the pool.
• Large relative wealth. A household spending 1.5% of assets a year cannot run out of money in any realistic scenario. Longevity insurance solves a problem they do not have.
• Rate environment and product quality. In a very low-rate environment, or where the only accessible products are high-fee index-linked contracts with living-benefit riders, the honest answer may be to wait or to decline entirely.

Execution Rules

1. Compute your funding ratio annually and let it, not your age, trigger the decision. Guaranteed lifetime income divided by essential spending. Above 110%, do nothing and stay invested. Below 85%, put a pooling review on the calendar for your mid-seventies. Below 65%, treat it as a priority to be addressed within a few years.

2. Cross with Social Security before crossing with an insurer. Delaying to 70 is the highest-value crossover move available and should be exhausted first: inflation-indexed, government-backed, commission-free. In many households this alone moves the funding ratio above 100% and eliminates the commercial decision entirely.

3. If you cross, cross in the mid-to-late seventies and cross partially. The efficient window for most healthy retirees is roughly 75 to 82, where mortality credits are material but health is still good enough to enjoy the income. Buy in two or three tranches across those years rather than one transaction, to diversify interest-rate timing.

4. Cap the crossing at the floor gap and never above 30% of investable assets. Size the contract to close the essential shortfall and stop. Beyond that, additional pooling is converting growth capital into fee-laden bonds and forfeiting liquidity for no additional protection.

5. Do not reduce equities to pay for the crossing — reduce bonds. The pooled contract is functionally a bond with a mortality credit, so it should replace bond allocation, not equity. A household that annuitizes out of its stock sleeve ends up with less inflation protection than it started with, which defeats the purpose.

Retirement Application

Consider three households, all 76, to see how differently the crossover applies.

The Nguyens have $58,000 of combined Social Security, $52,000 of essential spending, and $700,000 invested. Funding ratio 112%. There is no crossover for them. Their entire $700,000 belongs in a low-cost global index portfolio funding travel, grandchildren and a bequest, with a two-year cash buffer. Any annuity pitch they receive is solving a problem they do not have.

The Brennans have $41,000 of Social Security, $68,000 of essential spending, and $850,000 invested. Funding ratio 60%. Their $27,000 annual gap consumes 3.2% of the portfolio just to eat, before any discretionary spending. They are at the crossover now and past it. Closing roughly $18,000–$20,000 of that gap with a joint-life payout split across two A-rated carriers might cost $220,000–$260,000 at their age, leaving roughly $600,000 in the index core, and converts a fragile plan into a robust one. The remaining core can then hold more equity, not less, because the essentials no longer depend on it.

Margaret, widowed, has $34,000 of Social Security, $46,000 of essential spending, $310,000 invested, and stage-3 kidney disease. Funding ratio 74%. On age and funding ratio alone she is a candidate. On health she is not — standard pricing would have her subsidizing a pool she is unlikely to outlast. Her better answers are a larger cash reserve, careful spending, and a hard look at housing costs.

Same age. Three different correct answers. Which is precisely why "should I buy an annuity at 75" has no general answer.

Risk Management

• Crossing too early is the more common and more expensive error. Annuitizing at 63 pays decades of embedded fees, forfeits liquidity during the most uncertain health years, and captures a mortality credit worth almost nothing at that age.
• Crossing too late has a hard ceiling. Many carriers will not issue payout contracts after 85, and pricing and availability thin out. Do not defer indefinitely on the theory that the credit keeps improving.
• Nominal payouts erode the crossover's benefit. A crossing made at 76 with a nominal contract delivers roughly 60% of its initial purchasing power by 93. Either buy the inflation-adjusted version, or deliberately keep more equity in the core to compensate — and state which you are doing.
• Never cross without a liquidity floor. Two years of spending in cash plus a dedicated long-term care reserve must sit entirely outside any annuitized capital. An irreversible contract plus a thin cash position is how a manageable emergency becomes a forced sale.
• Beware the crossover argument used as a sales frame. The genuine crossover logic supports a bounded, late, partial purchase of a plain contract. If a presentation uses it to justify an early, large purchase of a rider-laden product, the logic is being borrowed rather than applied.

What This Chapter Cannot Do

This chapter cannot compute your crossover age. It depends on your funding ratio, health, current annuity pricing, prevailing interest rates, bequest intentions and your spouse's situation, and several of those move every quarter.

It also cannot resolve the deeper uncertainty in the comparison: the equity risk premium is an estimate, not a constant. If realized equity returns over your retirement are 2% rather than 5% real, the crossover arrives earlier than any of these numbers imply. If they are 7%, later. Anyone presenting a precise crossover age is overstating what the mathematics supports.

And it does not license replacing the portfolio. The crossover governs a bounded slice of the balance sheet, once, to close a specific gap. Above that gap the answer remains what it was at 55 and will be at 95: a broad, low-cost, globally diversified index core, held cheaply and left alone.

---

Key Takeaway: The crossover is set by your funding ratio, not your birthday. Above 110% coverage of essentials there is no crossover at all; below 85%, delay Social Security first, then consider a partial, plain, joint-life purchase in your mid-to-late seventies, funded out of bonds rather than equity, capped at the floor gap and under 30% of assets — with everything above that gap staying in the low-cost index core.