The default, not the annuity shelf, now decides retirement income

A participant who retires and leaves her money in the plan does not, in most 401(k)s, get offered an income product; she gets an account, a fund menu, and eventually a required minimum distribution. Vanguard's new paper on decumulation traces that outcome to plan design rather than to participant behavior, arguing that the withdrawal flexibility a sponsor builds into a plan matters more than what the plan puts on the shelf, and it landed in the same stretch of days that BlackRock moved the target-date contest to construction and SS&C widened Black Diamond's annuity platform to seven carriers.

Read separately, those three developments are a framework from an asset manager, a research paper, a platform expansion. Read together, they describe where the retirement income decision has moved—not onto the annuity shelf, where a retiree weighs one contract against another, and not into the fund menu, where a committee picks managers, but inside the default itself: the glide path a sponsor adopts, the QDIA that absorbs the contributions, the withdrawal mechanics that decide whether a balance ever becomes income. Whoever builds the default owns the decision, and the carriers buying shelf access are competing in a channel the default mostly bypasses.

BlackRock's LifePath Solutions is the tell. The framework lets a sponsor assemble a pension-style glide path inside the 401(k) default, relocating the contest from which target-date series a committee selects to who controls the construction of the path—a bigger shift than a new share class. A target-date fund is a packaged answer to a design question the sponsor has effectively already answered by choosing the series, and it gets selected, benchmarked, and priced like the packaged thing it is. The construction layer—the mix of exposures, the schedule on which risk comes down, the point at which the path stops accumulating and starts paying—is where the fee and the relationship sit, and it is the layer a sponsor does not currently buy as its own decision.

The buyer changes with it, and the sale gets slower and more technical: choosing a series runs through a committee, a consultant, and a scorecard of returns and fees; assembling a path runs through assumptions about longevity, spending, and risk that the committee owns and the recordkeeper administers. That is a sale the incumbent target-date providers are not organized to make, because their product answers the design question before a plan gets around to asking it.

Everyone else in the DC supply chain should read that as pressure: if the glide path becomes an assembly rather than a sealed product, the sleeves that go into it get repriced as inputs, and firms that spent a decade building target-date franchises end up defending a wrapper. The annuity carriers have the sharper exposure: a path constructed to deliver pension-style outcomes turns an income guarantee into a component inside the default, and components do not carry destination pricing. An industry that has argued for three years that retirement income is something a participant selects is now watching a large asset manager argue that it is something a sponsor installs.

Alight's August numbers show why installation beats selection: trading ran at 0.008% of balances, the quietest August Alight has recorded in 30 years, while the QDIA pipeline absorbed 74% of contributions. In the aggregate, participants are not choosing; they are accepting, which is how defaults behave—why the fund holding a plan's default holds the plan's market, and why a product not inside the default has to be sold rather than received. Any income strategy that must be chosen is competing against an option that asks nothing of anyone: no participant action, no committee vote, no new recordkeeping.

Those two figures describe the same plan from opposite ends: contributions concentrate going in and activity goes quiet coming out, meaning the default does the work on both sides of a participant's life in the plan. For a sponsor, that makes the default selection the decision carrying the most assets—more than any single fund on the menu—and the one least likely to be revisited, since nothing about a silent August puts it back on the agenda.

Vanguard's paper names the incumbent: the required minimum distribution, tracing retiree drift into RMDs back to plan design and putting withdrawal flexibility ahead of the product shelf—a courteous way of reporting that most plans have already installed a decumulation default and it is the tax schedule. A sponsor that has never built an income option has still made an income decision, the one that arrives without anyone approving it.

Sponsors should sit with that. A default that produces withdrawals only when the code demands them is not a decumulation strategy, and the fact that it looks adequate while markets rise says more about the markets than about the design. Flexibility inside the plan—a systematic withdrawal, an amount sized to spending rather than to a formula, money that stays invested while it pays out—is the part of decumulation a plan committee actually controls, and Vanguard's argument is that plans have underweighted the one lever in reach.

The reason the RMD persists is probably structural: building an income default means a committee selects a payout method, takes on the work of monitoring it, and answers questions the plan has never had to answer about what happens when someone leaves work but stays in the plan. Doing nothing carries a similar outcome with none of the exposure, which is what most plans have done.

The pressure to pull the other lever is building: two forecasters have converged on a 3.5% Social Security cost-of-living adjustment for 2027, and TSCL's polling suggests the people receiving it will call it too small, which pushes the income gap back into plan design rather than into a product illustration. The case for a guaranteed income floor is now argued in whether a plan's default can be built to produce income that lasts, which is a contest the annuity business should be winning and is not, because the venue moved.

Which is what makes the annuity industry's own week instructive: SS&C widened Black Diamond's annuity platform to seven carriers, with Jackson National and Protective Life joining a roster that increasingly determines which fee-based annuities reach RIA clients. The logic is sound on its face: a shelf is how a carrier reaches advisors who will not run their own due diligence on twenty contracts, and in the RIA channel, where an advisor holds the decision, a shelf is the market.

A shelf is also a gate: a carrier that is not on one does not get considered, whatever the contract's terms, and a carrier that is on one gets considered by advisors who are not going to read the prospectus. The seven-carrier roster does two jobs at once, deciding which contracts reach RIA clients and setting the terms on which everyone else can reach them.

But the RIA shelf is not the plan default: Black Diamond's carriers reach advisors' clients through a choice a human makes, while the 74% of contributions moving into a QDIA reach participants through a recordkeeper and a committee, a path on which an annuity is either part of the default construction or is not encountered at all. Carriers expanding shelf access are buying into the channel that still asks someone to choose, which is a real business and a defensible one, but it sits next to the migration rather than on top of it.

LIMRA's final second-quarter tally makes the timing awkward: it slipped, with fixed-rate deferred carrying the miss—the product that drove the sales boom now the one shrinking, and the money replacing it behaving differently. A category that grew on rate levels is a category exposed to them, which is why carriers want fee-based shelf placement where revenue does not depend on a spread. Read that way, the platform building is an attempt to convert a spread business into a fee business while the spread still pays, rather than a bet on annuity demand.

Two quieter moves point the same way: Groom's new hire arrives with a résumé that runs from EBSA enforcement through in-house product counsel at three providers, a path that only makes sense if retirement's legal risk has migrated to the plan menu, where product design, disclosure, and fiduciary selection meet. Edelman separately pushed 3(38) fiduciary duty down to the participant account with a phone-and-digital advice bundle sold through ADP, moving the relationship, and the liability, from the committee to the saver.

The account is where permission gets granted, regardless of what the rulebook says. A former EBSA chief said on a Pontera webinar that the law never barred participant-chosen 401(k) advisers, and the practical answer sits with whoever controls the login, which is why Fidelity's December access cutoff carries more weight than the legal permission does.

Both moves treat the participant account as the unit where the next contest gets fought, which is consistent with the rest of the week: the decision is being made closer to the saver, through defaults and through advice, and the firms positioning for it are building around the account rather than around the product.

Whoever ends up owning the income decision will have captured it the way every other retirement outcome gets captured, by being the default. That is an uncomfortable conclusion for a business built on retirement income as something a person chooses after weighing alternatives. The alternatives still exist, but they sit off the path the money actually travels, and few participants step off it.

Watch whether the first glide-path construction to reach the DC market carries an income component inside the default with a carrier sitting on it—if it does, the seven-carrier shelf is a distribution channel for a product that has already lost the default, a good RIA business with a plan option attached. If it does not, the required minimum distribution keeps doing the work, and the plan's de facto income product remains the one nobody chose.