Skip to main content
Annuities

The Annuity Puzzle: Why Smart Money Avoids Them

Pomegra Learn

The Annuity Puzzle: Why Smart Money Avoids Them

Annuities are a textbook solution to a fundamental retirement problem. Academics, financial planners, and even the Social Security program itself (which is fundamentally an annuity) all agree: an inflation-adjusted lifetime income stream is the gold standard for retirement security. Yet fewer than 10% of American retirees buy immediate annuities. Financial advisors rarely recommend them. Wealthy investors actively avoid them. This paradox—the annuity puzzle—reveals a gap between what the math says and what real humans want. Understanding this puzzle is essential because it helps you separate genuine annuity drawbacks from behavioral biases, and decide whether you're rejecting them for the right reasons.

Quick definition: The annuity puzzle is the observation that fewer retirees buy annuities than economic theory predicts should, even though longevity insurance provides measurable value and mortality credits (shared risk from others who die younger).

Key takeaways​

  • Academic models say annuities should represent 50–75% of a retiree's portfolio; actual adoption is under 10%, suggesting either the models are wrong or humans systematically misjudge risk.
  • Annuity costs are real: mortality credits (insurance profits), sales commissions, and expense ratios create a 1–3% annual drag that bond ladders don't carry.
  • Loss aversion is powerful: retirees overweight the pain of giving up control and forfeiting principal more than they underweight the comfort of guaranteed income.
  • Bequest motives (desire to leave money to heirs) are rational reasons to avoid annuities, but many retirees reject them for emotional rather than financial reasons.
  • The "best" approach for most is partial annuitization: annuities for essential, non-negotiable expenses; liquid assets for discretionary income and legacy.

The academic case for annuities​

Start with the theory. In 1969, economist James Tobin showed that for a risk-averse retiree, complete annuitization (turning your entire portfolio into a lifetime income stream) is optimal. Why? Because annuities eliminate longevity risk—the fear that you'll outlive your money—through mortality credits: insurance companies pool longevity risk across thousands of customers. When you buy an annuity, you're not just receiving your own principal back slowly; you're also receiving a small share of the money from customers who die earlier than expected.

To see this concretely: if 1,000 retirees each give $500,000 to an insurer, that's $500 million pooled. The insurer calculates that the average recipient will live 25 more years, so they can safely distribute about $20 million per year ($20,000/recipient). But some of those $500,000 customers die in year 5—their unexpended principal is redistributed to the survivors. By age 90, the remaining retirees are receiving far more than their actuarial share, financed by those early deaths. This is the mortality credit: a return on longevity insurance that you can't get anywhere else.

Early research suggested this mortality credit was worth 1–3% annually on your annuity. So a $500,000 annuity not only guarantees you income; it also implicitly gains 1–3% per year from risk pooling. For a risk-averse retiree, that makes annuities seem almost free money.

Yet people don't buy them. Dozens of studies confirm this. In a 2008 survey by the National Bureau of Economic Research, only 8% of retirees held immediate annuities (excluding Social Security, which is itself an annuity). For households with over $1 million, the rate was only 12%. Even in academic exercises where researchers offer subsidized annuities at favorable terms, only 30–40% of experimental subjects choose them.

Why retirees reject annuities: the explanations​

1. Loss aversion and the illusion of control​

Humans are loss-averse: we feel the pain of losing $1 about twice as strongly as the pleasure of gaining $1. When you buy a $500,000 annuity and give your money to an insurance company, your mind registers a loss. You've lost $500,000 in liquid capital. Yes, you gain a monthly check, but that check feels like income received, not principal preserved. Psychologically, they're not equivalent.

This is compounded by the illusion of control. Even a bad bond ladder—one where you run out of money at 95—feels more controllable than a good annuity, because you decide when to spend and how much. Delegating that decision to an insurance company feels reckless, even if statistically you're safer.

2. Bequest motives (wanting to leave money to heirs)​

Some retirees reject annuities because they want to leave their children an inheritance. This is rational. If you have $500,000 and you annuitize it, your heirs receive nothing if you die at 70. If you don't annuitize and you live to 100 spending only 4% annually, your heirs might inherit $200,000+. The desire to build legacy is genuine wealth preference, not a cognitive bias.

But here's where the puzzle deepens: many retirees claim bequest motives but don't have the wealth to justify them. A retiree with exactly $500,000 and no other income source cannot both guarantee themselves 30 years of comfortable retirement income and leave a meaningful inheritance—unless they live on so little that they're miserable in their 70s. Yet they reject annuities anyway, acting as if their bequest motive is larger than the math supports.

3. Mistrust of insurance companies​

Insurance companies have a reputation problem. Some of it is fair (bad product terms, deceptive sales tactics), and some of it is inherited fear from horror stories (e.g., insurance companies denying claims, or Equitable Life's 1980s scandal). Retirees hear "annuity" and think "Suez canal investment" or "Enron-style collapse." This fear is rational in degree (insurance companies do fail occasionally), but irrational in magnitude—they fail less often than banks, and the federal government insures bank deposits.

4. Mortality risk concerns (I might die early)​

Some retirees reject annuities because they're convinced they'll die young and waste their money. "Why should I lock up $500,000 if I'm going to die at 75?" This is a real risk, especially for people with serious health conditions. But most retirees underestimate their own longevity—men who reach 65 live on average to 84; women to 87. If you're healthy at 65, you have a 25–30% chance of reaching 95.

Ironically, people most worried about dying young are often the ones who should buy annuities least (or only partial annuities)—but they use this concern to justify avoiding them entirely, even when their actual life expectancy is long.

5. Regret aversion​

Buying an annuity is irreversible. If you annuitize, rates then rise, and you realize you could have gotten 30% more if you'd waited, you feel deep regret. This regret—even if irrational (you made a good decision with the information you had)—is painful. So retirees avoid annuities to avoid future regret, prioritizing flexibility over certainty.

The real costs of annuities (why the puzzle isn't entirely behavioral)​

But the annuity puzzle isn't only psychology. There are genuine economic costs to annuities that the academic models sometimes downplay.

Mortality credits have shrunk. Early research in the 1970s–1980s estimated mortality credits at 1–3% annually. But as insurance companies consolidated, markets became more competitive, and commission-hungry advisors began pushing annuities to people who didn't need them, costs rose. Modern research (particularly studies by Shiller and others) shows mortality credits are closer to 0.5–1.0% annually—substantial, but much less than once thought.

Insurance company margins are real. Immediate annuity providers (insurance companies) pay sales commissions of 3–6% of your premium. They also charge expense ratios (0.5–1.0% annually). These margins come out of what could otherwise be your income, so they're not "free" insurance.

Sequence risk in annuity purchase. Annuity payouts are set by interest rates at the moment of purchase. If you buy when rates are low (2021), you lock in poor payouts. If you wait for rates to rise and they never do, you've wasted time. This timing risk is real and asymmetric: you regret buying at peaks (can't undo it) but forgetting to buy at troughs doesn't hurt as much (you just got lower returns).

Inflation erodes purchasing power. A $2,600/month annuity purchased at 65 is worth about $1,850/month in purchasing power by 85 (assuming 2.5% inflation). Without an inflation rider, annuities are secretly a poor deal in high-inflation environments. Inflation riders are expensive (0.5–1.5% of your payout), so many retirees skip them, betting on low future inflation. This works until it doesn't.

The partial annuitization solution (and why it's underused)​

Given the puzzle, a sensible middle ground emerged in academic research: partial annuitization. Buy enough annuity to cover essential, non-negotiable expenses (food, housing, healthcare, property tax), then ladder or invest the rest for discretionary income and bequest.

For example:

  • Essential expenses: $36,000/year (75% of total spending)
  • Discretionary expenses: $12,000/year (25%)
  • Buy annuity for $36,000 at age 65: costs ~$700,000 (at mid-2020s rates)
  • Ladder or invest remaining $300,000 for discretionary income + legacy

This approach gives you the mortality credit on the portion where it matters most (essential income), while preserving flexibility and legacy value for the rest. Research suggests partial annuitization is optimal for most retirees.

Yet it's rarely done. Why? Partly because it's more complex to explain and implement than "buy an annuity" or "stay fully invested," partly because financial advisors earn higher commissions on one big annuity sale, and partly because retirees are drawn to simple narratives (either "annuities are bad" or "annuities are good") rather than nuanced tradeoffs.

Decision tree​

Real-world examples​

Example 1: Robert, 68, struggling with the annuity decision

Robert retired with $800,000 and $24,000/year in Social Security. His essential expenses are $48,000/year, leaving a $24,000 gap. A financial advisor suggests he buy a $400,000 immediate annuity, which would produce $19,200/year—covering most of the gap. The remaining $400,000 would be invested for discretionary income.

Robert is torn. He worries that if he dies at 72, he'll have "wasted" $400,000. He also doesn't like the idea of giving up principal. But his advisor points out: if he doesn't annuitize and lives to 95, he needs his $800,000 to last 27 years on a 4% withdrawal rule, giving him only $32,000/year total. The annuity top-up guarantees him $43,200/year, forever. After the advisor explains the mortality credit and shows him that $400,000 is partial, not complete annuitization, Robert decides to buy the annuity. He still owns $400,000 outright.

Example 2: Susan, 72, already retired, reconsidering annuities

Susan retired 7 years ago with $1 million and has spent conservatively, leaving her portfolio at $950,000. She's never bought an annuity, and now the annuity puzzle has caught her attention. She's 72, has excellent health, and her parents both lived to 95+. Her advisor suggests she buy a $500,000 deferred annuity, which at 72 would pay out at age 85 at a very attractive rate (because only 13 years of payment are expected). This would give her peace of mind that even if she lives to 100, her essential expenses are covered from age 85 onward.

Susan buys it. The $500,000 deferred annuity will pay her approximately $3,200/month starting at age 85. From now until 85, she'll withdraw from her remaining $450,000 at a conservative 3% rate ($13,500/year), plus her Social Security ($22,000/year), totaling $35,500/year spending. At 85, her annuity kicks in, and even if her portfolio is depleted, she still receives $3,200/month ($38,400/year).

Common mistakes​

Mistake 1: Rejecting annuities entirely because of early-death fear

A retiree at 65 with good health worries: "What if I die at 75? I'll have wasted $500,000 on an annuity that only paid me $26,000 for 10 years." This is a valid concern, but the math is wrong. If you live to 85, the same $500,000 would have paid $52,000 (and you'd still be receiving payments). The real question isn't "what if I die early?" but "what is my probability of living to various ages?" For a healthy 65-year-old, that probability is 50% chance of reaching 85+. Rejecting annuities entirely isn't rational unless your health is poor.

Mistake 2: Buying annuities when interest rates are at lows

Interest rates determine annuity payouts. In 2021, when 10-year Treasuries yielded 1.5%, buying an annuity was terrible value (your payout would be compressed). In 2024, when rates rose to 4.5%, annuities became far more attractive. Many retirees buy annuities without checking the rate environment, locking in poor payouts.

Mistake 3: Annuitizing with an insurance company in poor financial health

Not all insurance companies are equal. Before buying, check the issuer's rating from A.M. Best. If it's below A or A-, avoid it. A percentage-point difference in payout (2.6% vs. 2.8%) matters less than the insurer's ability to pay for 30 years. A few insurance companies have failed; check the history.

Mistake 4: Buying annuities without inflation riders, then regretting the lost purchasing power

A $2,600/month annuity sounds good at 65. At 85, when inflation has run 2.5%/year, that $2,600 is only worth $1,850 in today's dollars. Many retirees skip the inflation rider (costs 0.5–1.5% of payouts) and regret it. For retirees under 75, an inflation rider is usually worth its cost.

Mistake 5: Annuitizing in an IRA, creating forced RMDs after age 73

Some retirees buy annuities inside IRAs, not realizing that Required Minimum Distributions (RMDs) can conflict with annuity payouts. If you buy an annuity at 60 in a Traditional IRA, you still have to take RMDs starting at age 73, which might exceed your annuity income, forcing you to withdraw more and pay extra taxes. Annuities work best in taxable accounts (where all payouts are ordinary income anyway).

FAQ​

Why do academics think annuities are optimal, but most retirees don't buy them?​

Two camps of thought: One view is that retirees are irrational and succumb to behavioral biases (loss aversion, mistrust, bequest regret). Another is that academic models oversimplify: they assume retirees have no other income (forgetting Social Security), no bequest motives, and rational expectations about longevity. In reality, retirees often have Social Security (which is an annuity), do care about heirs, and may rationally distrust insurance companies. Partial annuitization is probably the "true" optimal, which is easier to justify psychologically.

Is an annuity safe if the insurance company fails?​

Mostly yes, but not 100%. Each state has a guarantee corporation (PGIC, Illinois Guarantee Fund, etc.) that backs failed insurance companies' annuity obligations, typically up to $250,000 per contract per person. So a $300,000 annuity in a failed insurer would be protected for the first $250,000. To be safe, buy from A-rated or better insurers, and don't annuitize more than $250,000 with a single company.

Should I buy an annuity if I have a terminal illness?​

No. If you have less than 10 years to live, the mortality credit disappears and you're essentially selling your principal at a discount. Even a generous immediate annuity assumes 20+ years of payouts; a shorter time horizon makes it unfavorable. Stay invested or spend principal steadily instead.

Can I buy an annuity and still pass money to my heirs?​

Yes, with riders. A "period certain" rider (guarantees payouts for 10, 15, or 20 years, even if you die) or a "joint-survivor" rider (continues paying your spouse after your death) both leave money to heirs—but they reduce your monthly payout by 10–20%. A "return of premium" rider guarantees your heirs receive any unspent principal—but it cuts your payout by 15–25%. These riders defeat much of the annuity's value by rebuilding longevity risk.

What if I buy an annuity, then want to exit it later?​

You can't, unless the contract allows surrenders (most don't in the first 5–10 years). Some contracts allow a 1–2% annual surrender (withdrawal), but it locks in surrender charges. This is why partial annuitization is safer: you only lock away what you're confident you won't need, leaving the rest accessible.

Are immediate annuities or deferred annuities better for retirement?​

Immediate annuities pay you starting next month. Deferred annuities start paying 5–10+ years later, but at much higher rates (because fewer years of payouts are expected). Immediate annuities make sense if you need income now. Deferred annuities (bought at, say, 65, starting at 80) are clever for people who expect to live long and want insurance on deep old age without disrupting spending now.

Summary​

The annuity puzzle—why retirees buy fewer annuities than theory predicts—reveals a tension between academic models and human behavior. Annuities are mathematically sound for eliminating longevity risk, but loss aversion, bequest motives, mistrust, and regret risk make them psychologically unappealing. The real costs (insurance margins, inflation risk, sequence risk) are also higher than early research suggested. The most rational solution is partial annuitization: use annuities to cover non-negotiable essential expenses, leaving flexibility and legacy value in liquid assets. This approach is optimal in theory but underused in practice, mostly because it requires more nuanced decision-making than simple all-or-nothing thinking.

Next​

→ Taxation of Annuities and Tax-Deferred Income Strategies