By Christopher B. Tobe, CFA, CAIA
Fred Reish and Chris Heye identify a real retirement problem in their Summer 2026 Journal of Pension Benefits article, “The Importance of Annuities for Retired Workers: Why 401(k) Plans Should Include Guaranteed Income.” Retirees need dependable income, and managing investments and withdrawals can become harder with age.
But identifying a problem does not establish that the insurance industry’s preferred solution is the best answer. The article explains the risks of managing retirement savings far more fully than the risks of handing those savings to an insurance company.
Its omissions matter because a product that simplifies monthly cash flow can simultaneously introduce insurer credit risk, opaque pricing, contractual restrictions, and conflicts of interest.
“Guaranteed” by whom?
A general-account annuity replaces some investment and longevity risks with a contractual promise from an insurer. The policyholder owns contractual rights, rather than direct ownership of the insurer’s underlying investment portfolio.
The insurer may hold thousands of investments. That does not mean the retiree has thousands of independently enforceable promises. The retiree’s claim remains against the issuing company.
Other annuity structures can hold assets in separate accounts or diversify guarantees among insurers. Those differences deserve analysis. Treating “annuities” as a single solution conceals precisely the distinctions a fiduciary needs to understand.
The article gives no meaningful examination of what happens if the insurer weakens, what backs the promise, or what protection the participant actually has.
Cognitive decline strengthens the case for safeguards
Reish and Heye devote considerable attention to cognitive decline. Their concern is legitimate. Their policy conclusion is incomplete.
Someone who may struggle to rebalance a portfolio may also struggle to evaluate an annuity contract, compare payout options, recognize a deteriorating insurer, or understand what happens after a rollover.
The answer cannot be merely to educate participants about the advantages of annuities. Education must explain the disadvantages with equal clarity.
A monthly payment can simplify budgeting. It does not prevent a scammer from stealing money after it reaches the retiree’s bank account. Nor does it establish that a complex insurance purchase was suitable in the first place.
If diminished capacity justifies stronger intervention by plan fiduciaries, it also justifies stronger protections against inappropriate insurance products.
Where is the downgrade clause?
The article recommends commitments that can last for decades without discussing a basic contractual protection: what happens when the insurer no longer meets the financial-strength standard that justified its selection?
If a committee selected an insurer because it was AA-rated, what remedy does the participant have after serious deterioration?
In [Lifetime Income Annuities Need Downgrade Clauses to Work in 401(k)s](https://commonsense401kproject.com/2026/08/12/lifetime-income-annuities-need-downgrade-clauses-to-work-in-401ks/), I argued for enforceable transfer or replacement rights after specified credit deterioration.
That is a proposed protection, not a universal requirement of current law. It would need to be negotiated, priced, and funded. Replacing an already-annuitized payment obligation is more difficult than moving an investment account, and a promise to find a replacement insurer is not enough.
A workable contract must identify the trigger, the replacement mechanism, who bears the cost, and what happens when several insurers weaken together.
Those difficulties are reasons to address the issue before purchase. They are not reasons to omit it.
The costs do not disappear because there is no expense ratio
The article provides no meaningful account of annuity pricing.
Depending on the product, compensation can appear through explicit charges, commissions, embedded spreads, guarantee fees, or the terms of the payout itself. An apparently fee-free contract may still be expensive.
An insurer’s entire investment spread is not pure profit or a mutual-fund-style expense ratio. It also funds insurance risks, capital, administration, and other costs. But that is an argument for explaining the economics, not leaving them invisible.
Fiduciaries should demand competing quotes on comparable terms: the same premium, age, commencement date, survivor protection, refund features, inflation treatment, and credit quality. A payout percentage is not an investment return; it can include repayment of the participant’s own principal.
As I argued in [DOL’s Biggest Blind Spot](https://commonsense401kproject.com/2026/07/20/dols-biggest-blind-spot-404a5-fee-disclosures-ignore-annuities-the-largest-fees-in-many-401k-plans/) and [Annuities Break ERISA’s Disclosure Rules](https://commonsense401kproject.com/2026/06/24/annuities-break-erisas-disclosure-rules/), a disclosure system that leaves material insurance economics obscure makes comparison harder.
The safe-harbor discussion leaves out a crucial statutory requirement
The article describes the SECURE Act insurer-selection safe harbor as obtaining specified insurer information for fiduciary protection, then associates cost review with the separate regulatory safe harbor.
That presentation is materially incomplete.
[ERISA section 404(e)(1)(B)(ii) and (C)(ii)](https://www.law.cornell.edu/uscode/text/29/1104) also require consideration of contract costs, including fees and commissions, and a conclusion that the relative cost is reasonable in relation to the benefits and services. Insurer representations address specified financial-capability considerations; they do not replace the entire selection process.
More fundamentally, protection for a selecting fiduciary does not supply additional capital to the insurer or improve a retiree’s contractual recovery after failure.
A safe harbor for the buyer is not a guarantee for the beneficiary.
Outsourcing selection does not eliminate oversight
The article also points to DOL Advisory Opinion 2025-04A as a way to shift annuity-selection responsibility to an investment manager.
But the [opinion](https://www.dol.gov/agencies/ebsa/about-ebsa/our-activities/resource-center/advisory-opinions/2025-04a) retains the appointing fiduciary’s responsibility to prudently select and monitor that manager, and preserves potential co-fiduciary liability.
The opinion also describes features that deserve more attention: multiple insurers, quarterly competitive bidding, and continuing insurer review. These are representations about the particular program, not DOL factual findings or a blanket endorsement of annuities.
If fiduciary expertise is outsourced, the committee still needs to understand the manager’s incentives, methods, compensation, and conflicts.
Where are the prohibited-transaction questions?
The article does not examine whether an insurer, recordkeeper, adviser, or affiliate is a party in interest, or how compensation moves among them.
An insurer recordkeeper recommending its own annuity presents a different conflict from an independent adviser evaluating competing providers. Proprietary products, affiliated services, commissions, and platform payments deserve scrutiny.
As I argued in [Annuities as Prohibited Transactions in Retirement Plans](https://commonsense401kproject.com/2026/03/01/annuities-as-prohibited-transactions-in-retirement-plans/), fiduciaries need a separate analysis of the transaction and any applicable exemption.
Not every annuity purchase is automatically unlawful. But an insurer-selection safe harbor is not itself a prohibited-transaction exemption. Fiduciaries should identify the exemption they rely on and document compliance with its conditions.
State regulation is part of the product
A decades-long insurance promise cannot be evaluated without examining the system supervising it.
Which legal entity issued the contract? Where is it domiciled? How much risk has been reinsured, to whom, and under which jurisdiction? What affiliated investments, private-credit exposures, or valuation assumptions support the promise?
These are the questions behind my pieces on [who regulates the annuity](https://commonsense401kproject.com/2026/08/10/annuities-who-is-your-regulator-and-does-your-401k-fiduciary-even-know/), [regulatory shopping](https://commonsense401kproject.com/2026/07/21/annuities-cherry-pick-the-weakest-state-regulator/), and [New York versus Iowa](https://commonsense401kproject.com/2026/08/11/new-york-vs-iowa-annuities-where-does-the-extra-spread-come-from/).
A higher quoted benefit does not establish superior value if the difference reflects additional risk or weaker protection. Conversely, domicile alone does not prove a particular insurer is unsafe. The committee needs evidence.
State guaranty-association protection also requires contract-specific analysis. Coverage depends on governing law and eligibility rules; it should never be presented as an unlimited federal backstop.
That is why I support [a federal insurance watchdog](https://commonsense401kproject.com/2026/09/25/congress-finally-asks-whether-america-needs-a-federal-insurance-regulator-the-answer-is-yes/) with authority to examine financial groups, affiliated transactions, reinsurance, capital, and liquidity before retirees lose access to their money.
A regular payment is not necessarily an adequate payment
The article invokes healthcare and long-term-care expenses without adequately examining inflation or emergency liquidity.
At 3% annual inflation, a fixed payment loses approximately 45% of its purchasing power over 20 years. An income check that continues for life can still become inadequate.
An annuity can help pay care bills. It does not, simply by producing income, insure the cost of long-term care. An unexpected large expense may require liquid assets precisely when an annuitized premium is unavailable.
Fiduciaries must also examine survivor benefits, early-death outcomes, portability, and any loss of guarantees when participants change employers or move their accounts.
Compare alternatives before declaring the solution
The article moves too quickly from the need for retirement income to the need for annuities.
Automatic withdrawals, diversified low-cost portfolios, bond or TIPS ladders, and strategies that coordinate withdrawals with Social Security deserve comparison. These alternatives do not all insure against extreme longevity. Annuities do not solve every liquidity, inflation, or credit problem either.
Likewise, diversified synthetic stable value may provide a useful comparison for capital-preservation money, as I discussed in [my stable-value critique](https://commonsense401kproject.com/2026/03/11/stable-value-industry-admits-synthetic-funds-are-safer-and-why-that-makes-general-account-annuities-a-prohibited-transaction/). It is not a substitute for lifetime mortality pooling.
The fiduciary task is to identify the need and compare the available ways to meet it. As I argued in [Lifetime Income: The Gateway Drug for Insurance Products in 401(k) Plans](https://commonsense401kproject.com/2026/06/09/lifetime-income-the-gateway-drug-for-insurance-products-in-401k-plans/), endorsing retirement income should not become automatic approval of the insurance products marketed under that label.
Reish and Heye make a persuasive case that retirees need help. Their article does not establish that the annuity products offered to them provide that help at a reasonable cost, with adequate transparency and enforceable protections.
Before a fiduciary asks an aging participant to trust a lifetime guarantee, the fiduciary should be able to explain who stands behind it, what it costs, what can go wrong, and what the participant can do when the promise weakens.
