Milliman published the April edition of its Pension Buyout Index on April 15, 2026. Average competitive-bid annuity purchase cost fell to 101.1 percent of GAAP accounting liability from 102.5 percent in March, with retiree-only competitive cost near 99 percent, the first sub-100 reading of the year.

Read as a buying window, that is a 140 basis point improvement. Read against the other two series in the same release, it is something else.

101.1%
April 2026 Milliman PBI competitive-bid cost, down from 102.5% in March
3.0 pts
Competitive vs. average annuity purchase cost spread, widest in nearly a year
~$10B
Q1 2026 US PRT deal volume, on pace for $49B+ full-year 2025 run rate

Key Takeaways

  • The three series moved by different amounts. Competitive bid cost fell 140 basis points, retiree-only competitive fell 180, and the all-insurer average fell only 80. A rate-driven move takes all three in lockstep.
  • The competitive-to-average spread widened to 3.0 points from 2.4 in March, the widest in roughly twelve months, which is the signature of price concentrating in one or two carriers rather than the field sharpening.
  • $10 billion of Q1 volume absorbed roughly 20 percent of the $49 billion the US market cleared in all of 2025, in three months, ahead of a summer termination queue.
  • $7 million of premium saving on a $500 million accumulated benefit obligation from the April move, against an ASC 715 settlement charge that can run to a high single-digit percentage of the same liability.
  • 108.9 percent on the Milliman 100 Pension Funding Index, the first decline in twelve months, arriving in the same release window as the pricing move.

Three Series, Three Different Moves

The index reports three cost series each month, and they answer different questions. The all-insurer average is the arithmetic mean of indicative quotes from the seven to eight major US carriers in the survey, a proxy for single-carrier pricing before auction dynamics. The competitive bid cost is the best quote in the surveyed set, the clearing price a broker-run auction should achieve. The retiree-only series restricts to retiree blocks, which carry tighter mortality bands and shorter duration and therefore almost always price lowest.

All three fell in April. They did not fall together.

PRT metricMarch 2026April 2026Move
Average annuity purchase cost (all insurers)104.9%104.1%-80 bps
Competitive bid cost102.5%101.1%-140 bps
Retiree-only competitive cost~100.8%~99.0%-180 bps
Competitive spread (avg minus best)2.4 pts3.0 pts+60 bps
Milliman 100 PFI funded ratio (prior month)~110%108.9%First decline in 12 months

That asymmetry is the finding. A credit spread rally or a long-end move would compress the annuity cost for every carrier at once and the three series would track. Competitive falling 140 basis points and retiree-only 180 while the average falls 80 says one or two carriers sharpened pricing on specific block profiles, most plausibly retiree-heavy blocks with favorable asset-sourcing, and the survey caught those quotes.

The Pension Buyout Index is distinct from the 100 Pension Funding Index it is published alongside. The funding index tracks aggregate funded ratio for the hundred largest US corporate plans; the buyout index measures the cost of transferring retiree obligations to an insurer, expressed against the sponsor's ASC 715 accounting liability.

What the Widening Spread Costs a Sponsor

The spread between average and competitive is the direct measure of what auction process is worth, and it widened to 3.0 points from 2.4.

Capacity explains most of it. PRT capital budgets for 2026 were set at year-end 2025, and roughly $10 billion of Q1 flow absorbed about 20 percent of the $49 billion the market cleared across all of 2025, per the Aon and PwC monitors.

Carriers that deployed heavily in the first quarter hold firmer on Q2 quotes to preserve room for the Q3 and Q4 termination window, which pulls the average up while the one carrier with remaining appetite bids down. The April 1 close of Brookfield Wealth Solutions' acquisition of Just Group, covered in the Brookfield-Just close analysis, reallocates capacity across a combined US and UK book in the same window.

Put the headline against a balance sheet and the saving is smaller than the decision it sits inside. Take a frozen plan with a $500 million accumulated benefit obligation, 108 percent funded, so $540 million of assets. At 101.1 percent the premium is $505.5 million, leaving $34.5 million of residual surplus. Against the March print the April move saves roughly $7 million on the same liability.

That $7 million is real and it is not the variable that decides the transaction. Under ASC 715 a full buyout clears the entire unrecognized actuarial loss in accumulated other comprehensive income to the income statement in the settlement quarter, which for a plan carrying a large balance runs to a high single-digit percentage of the accounting liability.

The surplus side is larger still. IRC Section 4980 taxes surplus reverting to the sponsor at 50 percent, falling to 20 percent only if at least 25 percent goes to a qualified replacement plan or a pro-rata benefit increase. The April pricing improves the surplus by a few million; Section 4980 decides whether any of it survives as shareholder value.

Why the Window May Be Closing Rather Than Opening

A wider spread is not a cheaper market. It is the same market with the best price available from fewer counterparties.

A mid-market block that drew five to seven quotes when capacity was abundant draws three or four when marginal carriers stop competing on price, and a sponsor capturing the 101.1 percent figure is capturing it from a narrower panel. That trades price against selection risk, and it means the index reading is least representative for exactly the sponsors reading it hardest: those still choosing a structure rather than already in market.

The lump sum interaction cuts the same way. A plan running a 2026 window on August 2025 stability-period segment rates offers lump sums calculated on lower-rate inputs, so they are rich against current annuity economics, and election rates on rich lump sums run 35 to 55 percent. That strips the longest-duration participants out, and the residual retiree block prices nearer the retiree-only series at roughly 99 percent than the blended 101.1.

A plan whose stability period picks up the April 2026 rates from Notice 2026-33 instead offers lump sums close to current market. Election drops to 20 to 30 percent, the deferred vested population stays in the block, and the annuity purchase costs more. The April 2026 417(e) analysis walks that rate move in full.

The funding side moved at the same time. The Milliman 100 index recorded 108.9 percent, its first decline in twelve months, so the funded-ratio cushion that pays the premium thinned in the same window the premium got cheaper. A sponsor with a low liability-driven hedge ratio faces both variables moving together, and the price improvement does not offset a funded ratio that keeps sliding into the quarter the deal has to close.

Further Reading

Sources