The 401(k) private-asset push lost its benchmark

CFA Institute's modeling has stripped the private-asset pitch to 401(k)s of its empirical cover: only two of the five private asset classes it tested beat a plain stock-and-bond baseline, and the other three cost retirement wealth. That arithmetic decides whether a fiduciary can put a private-markets option on a self-directed menu or must fold it into a target-date glidepath, where the committee, rather than the participant, absorbs the comparison.

The institute splits 401(k) private markets into return and ballast, and the result is more precise than a blanket rejection: a small allocation can pay off, but only in the two asset classes that beat the baseline, while the other three destroy retirement wealth. The burden shifts from 'can private markets help?' to 'which private markets, against what benchmark, and who decides when the allocation has stopped helping?' — fiduciary questions a standalone menu option cannot answer.

That arithmetic is now catching up to the Department of Labor's private-assets safe harbor, after three ranking House Democrats asked the FBI and the department's inspector general to test whether the comment file supporting the rule includes thousands of comments nobody actually filed. The process attack lands harder because the underlying substance has a benchmark hole: a rule that makes it easier for sponsors to offer private assets cannot fix the fact that, measured against a plain stock-and-bond baseline, a majority of the classes modeled come out behind.

The Invesco survey numbers show participant demand cannot fill that hole. Sixty-five percent of participants said they want a private-markets sleeve in their target-date fund, and 58% said they accept the fee trade-off, but the institute's math shows the asset classes those sleeves would hold mostly trail a plain stock-and-bond baseline, and participant enthusiasm is not a substitute for fiduciary evidence. The 58% fee tolerance figure is the one the shelf will actually turn on, because it tells you participants will pay for private exposure; it says nothing about whether the sponsor has satisfied ERISA.

The benchmark problem is sharper in the self-directed menu than in the default, because a standalone option must be justified on its own merits, monitored against a comparable index, and removed when it fails. The industry has not built the benchmark a plan committee needs to isolate private-markets performance inside a 401(k), and CFA Institute's own work uses a plain stock-and-bond baseline that three of five classes fail — a threshold problem for a menu option that must stand on its own.

Fold the same allocation into a target-date fund and the comparison changes: the private sleeve inherits the whole glidepath's objective and risk budget, the committee owns the allocation, and the recordkeepers who build the defaults decide whether the sleeve appears. MFS's Jeri Savage has said concentration has reopened the active-passive question in 401(k) lineups, and the recordkeepers who build the defaults will settle it. That is precisely the venue where private markets can still enter, because the default can absorb a small allocation as part of a diversified whole rather than as an option that must beat the baseline on its own.

The comment-file probe compounds the difficulty, because if the safe harbor's administrative record includes comments that were not filed by real participants, the rule's procedural legitimacy is open to challenge at the same moment its economic legitimacy is being questioned. Sponsors do not build plan menus on a contested foundation; they wait, or they use the default, and the default is the target-date fund where private markets can still enter.

That matters more now because the retirement pool is at a record, with total retirement assets at $51.2 trillion and the fastest-growing pool being IRAs, where mutual funds have the least reach. The 401(k) default remains where workplace wealth is formed, and if private assets cannot clear the benchmark bar for a standalone menu slot, the target-date sleeve becomes the only route into the defined-contribution system, and the recordkeepers who control those sleeves control the private-markets pipeline.

CFA Institute's split between return and ballast sharpens the argument: a sleeve whose job is ballast may be judged by a different standard than one expected to deliver return, but the institute's modeling treats them against the same stock-and-bond baseline, and three of five fail. A target-date manager running a small allocation inside a diversified glidepath can absorb that failure; a sponsor trying to put a standalone option on the menu cannot.

The active-passive question Savage raises is the same fight by another name, because private assets in a 401(k) are active decisions: someone is choosing which funds, which managers, and which vintage years to include in the sleeve. If the target-date default is where active management will be decided, as Savage suggests, then the private-markets question becomes whether recordkeepers will build it into the glidepath and accept the monitoring burden — participant demand is beside the point.

That is the real risk transfer. The math says the choice cannot sit with participants, because a participant cannot run the benchmark comparison the fiduciary must, while the default can absorb private markets because the committee owns the risk transfer. The Invesco survey does not capture that trade: 58% of participants accepting a fee trade-off is not the same as a committee accepting the monitoring, documentation, and benchmark burden.

The pace of private-markets entry now depends on the industry's next benchmark. If a target-date manager constructs a private sleeve with a defined role and a defensible comparison, sponsors will have a way to document prudence; if no one builds that benchmark, private assets will continue to enter 401(k)s only as small, committee-owned allocations inside the default, and the standalone menu option will remain a product pitch without a fiduciary home. The next move belongs to the recordkeepers and target-date managers who can supply that comparison, and until they do, the default is the only fiduciary home the arithmetic leaves standing.