A 50% annuity purchase beats the 4% rule in AEI-ACLI models

The 4% rule has an underspending problem. It is a quiet finding, buried in new research from the American Enterprise Institute and the American Council of Life Insurers, released July 29 and first reported by InvestmentNews. In a utility-based comparison of four retirement income strategies, the classic Bengen rule came in last.

Mark Warshawsky, an AEI senior fellow, and Gaobo Pang, an independent researcher, start from a baseline: a single woman retiring at 65 with $1 million. They send more than 10,000 stochastically simulated life paths through JPMorgan Asset Management's 2026 capital-market assumptions, with federal income tax, Medicare premiums, required minimum distributions, and Social Security timing folded in. The authors describe the result as one of the most comprehensive utility frameworks yet applied to the question.

The model tests four strategies: the classic 4% withdrawal rule, a full annuitization into a nominal immediate life annuity, a one-time 50% annuity purchase with systematic withdrawals from the remainder, and a gradual ramp from 20% to 60% annuitization between ages 65 and 75. The yardstick, average certainty-equivalent consumption, combines income, balances, longevity risk, and bequest preferences. After taxes and Medicare premiums, the one-time 50% purchase comes out ahead, with a score of 59.12. The gradual ramp trails by a hair at 58.94. Full annuitization is further back at 54.58. The 4% rule brings up the rear at 53.79.

The income gap explains the ranking. The 4% rule delivers the lowest average annual net income, roughly $58,100, because it spends too slowly. Full annuitization pays about $65,700 but exhausts the account in every simulated path. The hybrids produce comparable income while keeping average balances between $470,000 and $537,000. The 4% rule's weakness is not just depletion risk late in life; it is also the underspending that shows up early.

The failure risk is stark on its own: by age 90, assets ran out before death in 12% of scenarios. By age 95, that share reaches 24%. By age 100, it is 38%. And the edge from partial annuitization does not depend on the starting point. It holds for a retiree who starts at 62 or 67, who has $250,000 or $2 million, and who is in poor or excellent health. For a rule long described as a safe withdrawal rate, the record is hard to ignore.

Full annuitization makes the opposite bet. It produces more income, but it converts the entire portfolio into one check, with nothing left for an unexpected expense or a bequest. The hybrids give up a little income and keep about $500,000 in the account on average. That is the case for partial annuitization: a split portfolio beats both extremes.

For advisers, the study is a reminder that the 4% rule was a starting point, never a complete plan. An annuity does something a portfolio cannot: convert a lump sum into a check the retiree cannot outlive, while the invested remainder stays available for the unplanned expenses an annuity cannot cover. That split, a steady stream plus a liquid pool, is the practical definition of partial annuitization.

The paper is a modeling exercise, not a prediction. It rests on a single framework, one set of capital market assumptions, and one annuity type, a nominal immediate life annuity. Utility scores are not guarantees, but the model's answer to the old all-or-nothing debate is a split that keeps money available for heirs and emergencies. Advisers translating the results into a plan will have to decide how much income floor is enough. The 4% rule may live on as a spending benchmark. The AEI-ACLI paper just argues it should not be the entire portfolio.