Retirement · October 6, 2026 · 7 minute read

What a Big Four study found about insurance in a retirement plan

When an insurance agent says a policy belongs in your retirement plan, you are right to discount it. So here is the case made by somebody with nothing to sell you.

Ernst & Young publishes an ongoing analysis on whether insurance products improve retirement outcomes when combined with a traditional portfolio. The most recent instalment was authored by Philip Ferrari, a managing director in their Insurance and Actuarial Advisory Services practice.

I am going to tell you what they found, and then I am going to tell you what the study does not say, because that second part matters more than agents usually admit.

How the study was built

The analysis uses Monte Carlo simulation, which means rather than assuming one average return, it generates a thousand different possible futures. Each one carries its own path of interest rates, inflation, equity returns and bond returns across the planning horizon.

That matters because averages hide the thing that actually breaks retirement plans. A portfolio can average a perfectly healthy return and still fail if the losses arrive in the first few years, while you are drawing income from it.

Each simulated future was then judged on two outcomes. First, how much after-tax income the plan could sustain at a 90 percent probability of success, meaning it held up in nine hundred of the thousand scenarios. Second, the median value left over at the end for heirs.

Those are the right two questions. Most people care about both, and they pull against each other.

The six strategies compared

The baseline was an investment-only portfolio following a conventional moderate glide path with annual rebalancing, prioritizing tax-qualified accounts before taxable ones. The others added insurance in different combinations: an indexed universal life policy alongside investments; a fixed index annuity alongside investments; an immediate annuity purchased at 65; and blends of these.

The two insurance vehicles were chosen because of where the market actually is. By EY's figures, indexed universal life sales ran around $3.8 billion and fixed index annuities around $125 billion, the latter up roughly 31 percent year over year. These are not exotic products; they are what a large share of the market is buying.

What they found

• Adding indexed universal life to a portfolio beat the investment-only approach on the measured outcomes.
• Adding a fixed index annuity significantly beat investment-only on sustainable retirement income, with a small reduction in what was left to heirs. That trade is the whole story in one line: more income you can count on, slightly less legacy.
• The blended strategies were more efficient than investments alone, meaning they produced more of what retirees want from the same starting assets.
• They gave more control over the trade-off, letting a household lean toward income, toward legacy, or toward a balance, rather than accepting whatever the market delivered.
• They held up even for investors comfortable with risk, which cuts against the assumption that guarantees are only for the cautious.
• They addressed risks a portfolio does not, including living much longer than expected.

The study also modeled a scenario in which future Social Security benefits are cut by half, and looked at which strategies absorbed that shock better. Whatever you believe about the politics, it is a sensible stress test for anyone retiring in the next twenty years.

The finding is not that insurance beats investing. It is that the two together handled uncertainty better than either did alone.

What the study does not say

This is where I want to be careful, because this research gets quoted badly.

It does not say these products are right for you. It models representative products with specific assumptions. Real contracts vary enormously in cost, caps, participation rates and surrender schedules, and a poorly designed policy will not behave like a modeled one.

It does not promise a result. Monte Carlo output is a distribution of possibilities, not a forecast. The 90 percent success figure explicitly means some scenarios failed.

Indexed universal life in particular depends on consistent funding and sound design. A policy funded inconsistently, or built to maximize the agent's commission rather than the client's cash value, can underperform badly and in the worst cases lapse. That is a real failure mode and it is not rare.

It is also worth knowing the audience. EY's research is written primarily for insurance carriers, discussing their growth opportunity alongside a projected global retirement savings gap of $400 trillion by 2050. The modeling is rigorous and the authors are actuaries, but it was not commissioned as consumer advice, and you should read it knowing that.

Why I find it useful anyway

Because the mechanism it identifies is the one I see in practice.

People do not run out of money because they picked the wrong fund. They run out because they had to sell assets during a downturn to pay for groceries, or because they lived eight years longer than the plan assumed, or because one spouse died and a Social Security check stopped while the bills did not.

Those are risks a portfolio is not built to solve. They are the specific risks insurance exists to transfer, which is the point the research keeps arriving at from the numbers rather than from a sales script.

The honest version of the conversation

A model is not your situation. The useful version runs your actual accounts, your actual expenses and your actual timeline, and asks whether a portion allocated to guaranteed income improves the picture or just costs you flexibility you would rather keep.

Sometimes the answer is no. If Social Security and a pension already cover your essential expenses, you may not need to buy certainty you already have.

We will show you both versions of the arithmetic, and we will tell you plainly what any product gives up as well as what it provides. The invested side of a plan sits with our securities-licensed partner, and anything touching tax treatment goes to your tax professional before you act.

You can read the original analysis on EY's site, including the full report.