The Case Against Annuities Is About a Different Product From the One Worth Buying

Annuities have the worst reputation of any product in personal finance, and most of that reputation is earned. The problem is that the word covers two things which share a name and almost nothing else.

One is a single premium immediate annuity: you give an insurer a lump sum, they pay you a fixed amount for as long as you live. No annual fees, no subaccounts to pick, no market risk.^[1] It is structurally insurance, not an investment.

The other is a variable or indexed annuity: your money goes into subaccounts, the value moves with markets, and mortality and expense charges plus subaccount fees plus optional income riders commonly total 1.5% to 3.5% a year when the riders are stacked.^[2]

The case against annuities is a good case against the second one. It gets applied to the first, which is a different product answering a different question.

Mortality Credits, Which Are the Whole Point

This is the mechanism that makes a SPIA more than an expensive bond, and it is almost never explained.

A SPIA payout is made of three things: return of your principal, interest on it, and a mortality credit. That third component is the actuarial subsidy from people in the pool who die early, funding continued payments to those who live long.^[1]

The consequence is the important part: no do-it-yourself bond ladder can safely match a SPIA's lifetime income.^[1]

The reason is asymmetric information about your own death. Building your own ladder, you must fund to your worst case, which means planning to 100 or beyond even though you will probably not get there. That over-provisioning is expensive: capital sitting idle against a scenario that mostly does not happen. An insurer funds to the pool average, because it does not need every individual to be predictable, only the aggregate. It can promise payments past 100 precisely because not everyone in the pool will reach it.^[1]

That gap is not a fee you are avoiding by doing it yourself. It is a return source you structurally cannot access alone, and it is the same logic as every other kind of insurance: pooling converts an individually unbearable variance into a manageable average.

What a SPIA Is Actually For

Not returns. Not beating the market. It solves longevity risk: the possibility that you live long enough to run out of money.

Which is the risk that makes withdrawal-rate maths so conservative in the first place. The reason safe withdrawal rates land in the high 3s and low 4s rather than the 6s is that the plan has to survive a long retirement and a bad sequence of returns arriving early. That is the argument in the 4% rule, and a SPIA attacks it from a completely different direction: instead of spending less to survive an unknown horizon, you transfer the horizon problem to someone who can pool it.

Annuitising enough to cover essential expenses also changes the character of the remaining portfolio. Once the floor is guaranteed, the rest is money you can genuinely afford to hold in equities through a bad decade, which is an argument that shows up nowhere on the fee comparison.

The Inflation Objection, Answered Properly

The strongest criticism of a fixed SPIA is real: a fixed payment buys less every year, and over a 30-year retirement inflation is brutal to a nominal income stream.

The right response is not to avoid the product. It is to right-size it: annuitise only enough to cover essential expenses, which tend to be more inflation-stable than discretionary spending, and keep the remaining assets in growth that can outpace inflation.^[3]

That framing is worth dwelling on because it dissolves the usual argument. The choice was never annuity versus portfolio. It is how much of a floor you want under the portfolio, and the answer is almost never zero and almost never everything.

Where the Criticism Is Correct

Being clear, because none of the above defends the products that earned the reputation.

Variable annuities put your premium in subaccounts that fall with markets, then layer on mortality and expense charges, subaccount fees and rider costs, reaching roughly 1.5% to 3.5% all-in with riders stacked.^[2] Compounded over decades, a 2.5% annual drag is enormous, and the guarantees people buy them for are usually available more cheaply elsewhere.

Indexed annuities are a different thing again and the two are not interchangeable: indexed products typically put a floor under market losses, variable products do not.^[2] That floor is real, and it is paid for with caps and participation rates that are easy to misread and easy to change.

Surrender charges apply across the complex products and are the part that turns a bad purchase into a trapped one. A product you cannot exit without penalty for years deserves considerably more scrutiny than one you can sell on a Tuesday.

And the commission structure is the reason you hear about the expensive products and not the simple one. A SPIA is a small one-time commission on a transparent quote. A variable annuity with riders pays substantially more. Which product gets recommended is not a mystery, and it is worth knowing that the simplest annuity is also the one nobody is incentivised to sell you.