Bonds vs Whole Life: Rebuilding the 60/40 Retirement Plan

August 15, 2026 · Brandon Roberts

For forty years, one mix has been the default for retirement savers: 60% stocks and 40% bonds. It earned its reputation. What almost nobody asks is why it worked, and the reason matters, because it will not hold for the next forty years. The bond half carried far more of the load than its calm reputation suggests, and it did so for a single cause that has now run out.

Pull the bonds out and test them on their own against a properly designed whole life policy, and the half everyone treats as safe turns out to be the weak link. Our position, up front: for money you are saving to turn into retirement income, whole life is the stronger and safer holding for the stable part of the plan. Aim for 60% stocks and 40% whole life, not 60/40 stocks and bonds.

The rest of this post makes the case, one comparison at a time.

One assumption sits under the whole 60/40 idea: bonds are the calm, dependable half. They were, for one stretch of history, and that stretch is over.

Where the 60/40 Returns Came From

Every clean 60/40 back-test leans on the same decades, the '80s, '90s, and 2000s, and those decades held the largest bond rally in modern history. The 60/40 is not a law of finance. It became gospel during one specific window. In September 1981 the 10-year Treasury yielded almost 15.84%. By 2020 the same note yielded about 0.5%. Rates fell, with bumps along the way, for nearly forty years straight.

Here is the part most people skip, and it explains everything that follows. A bond pays you two ways. The first is the coupon, the fixed interest you collect every year. The second is price. When market rates fall, a bond you already own, locked in at the old higher coupon, becomes worth more than a freshly issued bond paying the new lower coupon. A buyer will pay a premium to take your richer stream of interest off your hands.

So an investor who bought bonds in 1982 collected a fat coupon and watched the market value of those same bonds climb, year after year, as rates kept sliding. Paid twice, on income and on price, for four decades.

Stack that up and the track record looks spectacular. Vanguard's Total Bond Market Index Fund (VBMFX), the oldest fund tracking the broad U.S. bond market and about the only one with a history long enough to run this test, compounded near 9% a year from the early 1980s to 2015. Nine percent a year, out of bonds. Investors saw that number and filed bonds away as a quiet 9% engine they could count on. They misread it.

The 9% was not bonds behaving like bonds. It was bonds catching a one-time, four-decade tailwind off the steepest rate decline in modern history. The run minted reputations. A fair case says it turned Bill Gross and his PIMCO Total Return fund into the bond king. Right place, right time, at a scale no one arranges twice.

The Rate Math No Longer Works

The next part is close to certain, because it rests on arithmetic rather than opinion. To repeat the last stretch, rates would need to fall another 15 points from here. The room does not exist, because they no longer start at 15%. They start near 4%. Even a drop all the way to zero, which nobody expects, moves rates down about 4 points, roughly a quarter of the runway bonds had in 1981.

The price gain a bond throws off depends on how far rates fall, so a 4-point runway produces a small fraction of the appreciation a 15-point runway did. The engine is out of fuel, and no policy choice refills the tank.

Strip the price gain away and look at what a bond hands you from today. Buy a 10-year Treasury at 4.3% and hold it to maturity. You collect 4.3% a year and get your principal back at the end. That is the entire return. In a world where rates hold steady, the coupon is all you get, and 4.3% becomes the ceiling, not the floor. Earning more than the coupon requires rates to keep falling, and we just covered how little room is left for that.

The honest bear case does not claim bonds are worthless. Bonds still pay a fair coupon, and they still soften a stock drawdown a little. The point is narrower. A fairly priced bond that returns its modest coupon is a far different asset from the growth engine that quietly carried every 60/40 study through 2020. Same name, same ticker, a different job and a much lower ceiling.

The Last Five Years Already Ran the Test

Set the theory aside for a moment, because real money already ran this experiment. Between 2020 and 2023 the 10-year yield climbed from about 0.9% to over 4%, and rising rates do to bond prices exactly what falling rates did in reverse. Bond values dropped. VBMFX fell about 13% in 2022, one of the worst single years the U.S. bond market has ever recorded.

For a retiree pulling income, that 13% did more damage than the headline number suggests, and the reason has a name: sequence-of-returns risk. Picture two retirees with the identical average return over thirty years. One meets the bad years early in retirement, the other meets them late. The one who takes the loss early ends up far poorer, even though the averages match. Here is why.

During your saving years, a down market almost helps you, because your steady contributions buy in at lower prices and ride the recovery up. During your spending years, that logic flips. You are selling, not buying. A 13% loss in a year you also withdraw 5% means you sold a slice of principal at the bottom to fund your check, and that slice is gone for good. It never joins the recovery, because you already spent it.

The loss and the withdrawal stack on top of each other. This is why your account balance is the wrong number to watch in retirement. In our own replay of a 100% bond retirement, that exact dynamic dropped a balance from $301,000 to $210,000 in a single year.

One brutal year you could shrug off. This was closer to a lost decade. VBMFX, the same Vanguard fund, sits down about 16% over the past five years and roughly 14.5% over the past ten on price. Years of holding a supposedly safe asset and getting nothing back, while the people who owned bonds purely for the coupon at least kept collecting their interest.

The fund's price peaked in 2020, at the lowest rates on record, and has not climbed back since. Anyone who built a plan expecting bonds to behave like the 1990s found out what they do when they behave like the late 1970s instead.

The road ahead looks lopsided too. Ten-year Treasuries and high-grade corporates pay somewhere in the mid-4s to high-5s today. From that level, a further rise in rates is at least as likely as a fall, and a rise punishes total return, because a bond does not snap back the way a stock does. When stocks drop, they tend to rebound on a timeline you recognize. When a bond loses value to higher rates, no rebound mechanism exists.

You either sell at the loss or hold your low-coupon bond to maturity and simply earn less than you could have. Take the loss and you stay in it.

Whole life carries the opposite trade-off, and we will say so plainly. A brand-new policy needs five or six years before the cash value clears what you paid in and starts showing a clear gain. Across those exact same years, though, you skip the multi-year loss bond holders have been eating. For anyone with a long horizon, that trade favors the policy. So the bond problem is real, it is arithmetic, and it already arrived.

The stable seat in the plan still needs filling. Whole life is what we put in it.

Head-to-Head #1: One Million Dollars, Lump Sum

Start with the version everyone pictures, a single lump sum. A retiree holds $1 million in 1996 and wants $40,000 a year to spend. Run each asset through the real market from 1996 to 2025 and watch what happens to the balance.

The bond retiree hits a tax wall first. Interest from a bond fund is taxable as ordinary income, so to keep $40,000 after tax at a 22% effective rate, the retiree has to pull about $51,300 out and hand roughly $11,300 to the IRS. That larger gross withdrawal leaves the portfolio every year, on top of whatever the market is doing.

Figures come from the illustration and index backtest behind this episode. VBMFX and the S&P 500 stand in as market proxies, not recommendations. Whole life numbers are illustrated values from an accumulation-focused policy, not guarantees. Whole life also leaves a death benefit the two market rows do not.

Look at the middle row before anything else, because it keeps us honest. The full 60/40 mix earned 8.28% a year and ended near $4.8 million. That is a great result, and the S&P 500 leg produced all of it. The bonds inside that mix did not drive the return. They rode along as ballast while stocks did the work. So read the row for what it says: whole life is not trying to beat stocks.

Whole life is trying to beat the bond sleeve, the stable money you draw on when stocks are down, and on that comparison the bonds delivered a 4.85% dollar-weighted return and nothing else.

Now set the two stable options next to each other. The all-bond portfolio survived on that 4.85%, but it ended below the $1 million it started with, bottoming near $812,874 in 2024. Thirty years of $51,300 withdrawals ground it down.

The whole life policy funded the identical $40,000 of spending, and it needed no tax gross-up, because policy income comes out first as a return of your basis and then as loans against the cash value, neither of which the IRS treats as taxable income when the policy is built and managed correctly. So the policy withdrew $40,000 to deliver $40,000, while the bond account had to withdraw $51,300 to deliver the same $40,000.

Same lifestyle, smaller drain. The policy finished past $1.7 million, more than double the bond result, with a death benefit still standing and no rebalancing required along the way.

The conclusion starts to show here. If the stable money does this much better inside a policy than inside a bond fund, the plan you want reads 60% stocks and 40% whole life. One caution before we bank on that: a lump sum flatters bonds, because you buy them all at once at yesterday's yields and hold. Most people do not retire with a pile. They build it over a working career, so that is the harder test, and it is the one that separates the two.

Head-to-Head #2: The Contribution Test

To pin down the whole life side, we solved for the premium behind the policy above: $21,389 a year for 25 years, $534,725 in total. Then we ran a fair fight. We put the same $21,389 a year into VBMFX on the same schedule, let both the bonds and the policy grow for 25 years, and drew equal income out of each for the next 30. The chart shows where each one stands at the two moments that matter, the day the saving stops and 25 years into retirement.

Walk the four bars. After 25 years of identical contributions, the bond account holds $777,328 and the policy holds $1,047,712 in cash value. The bond saver never even reached $1 million, because the money grew at only 2.77% a year across that stretch. Then income starts, and the two lines pull apart for good. Thirty years of withdrawals leave the bond account at $78,418, nearly drained.

The policy still holds $1,722,707 in cash value, and it carries a $1,962,739 death benefit the bond column has no answer for.

Two mechanics open the gap, and both are worth understanding. First, dollar-cost averaging into bonds during a stretch of rising rates works against you. Every new contribution buys fresh bonds, but the bonds you already hold get marked down as rates climb, so your steady buying never compounds the way the raw yield suggests. That is how a bond fund quietly delivers just 2.77% over a full accumulation.

Second, once you switch from saving to spending, the bond fund gets no room to heal, because the bad years land on a balance you are already draining. The same sequence-of-returns problem from the last section, now applied to a lifetime of contributions. Whole life runs the identical years on completely different plumbing, which is worth its own section.

The Whole Life Ledger, Step by Step

None of this works with an ordinary policy, so here is exactly what produced these numbers. The illustration is a Penn Mutual accumulation whole life contract on a male age 40, Preferred Non-Tobacco.

It is built base-minimized and paid-up-additions-maximized, which is the design detail that matters most: the death benefit is set as low as the tax code allows for the premium, and the extra money flows into paid-up additions that build cash value fast, all kept inside the Section 7702 limits so the policy stays tax-advantaged.

In plain numbers, a $152,734 base face plus a $305,468 first-year paid-up rider gives a $458,202 starting death benefit. Funding runs $21,389 a year for 25 years, $534,725 in all, and then premiums stop for good. Income of $40,000 a year runs from year 26 through year 55, taken as basis withdrawals first and policy loans after.

The table below is the payoff, the year-by-year ledger on the current dividend scale. Three columns tell the story: the cash value you can spend, the death benefit your heirs receive, and the return on that cash value. The return column carries the weight, because it is already net of every cost-of-insurance charge inside the contract. That matters.

A common knock on whole life is that insurance costs quietly eat the return, so a return figure that already subtracts those costs is the honest one to compare against a bond.

The illustration reports a return only at certain checkpoints. A dash marks a year with no reported figure, not a gap in value. Every number is net of internal policy charges.

Read what the ledger is telling you, because one line looks impossible at first glance. The cash value climbs the entire way, from $1,047,712 at year 25 to $1,722,707 at year 55, and it climbs while the policy pays out $40,000 a year for the last 30 of those years. How does a balance grow while you spend from it?

Because the early income comes out as withdrawals of your own basis, and the later income comes out as loans, and a policy loan does not subtract from the cash value. The full cash value stays inside the policy earning dividends, and the loan sits against it. So the account keeps compounding on the whole balance even as you draw a check.

The death benefit holds near $1.9 million the entire time, so your heirs stay covered no matter how long you live. And the return settles right around 5%, 4.99% at year 30 and 5.07% at year 40, and holds there.

Anchor on that 5%, and hold it up against the forecasts from Section 2. J.P. Morgan expects about 4.8% from bonds going forward. The policy returns roughly 5%, already net of every insurance cost, with no tax owed on the income and no year where the value marks down. That is a like-for-like win on the number, before you even count the death benefit. Then we stressed the figure to be sure it holds.

Cut the dividend scale by 0.25% for the entire life of the policy, a permanent haircut no carrier has imposed, and the return still passes the bond investor's 3.94% by about year 22 and settles in the mid-4s. Illustrations are projections, not promises, so that stress test is the point: even under a permanent cut, whole life stays at or above the bond result.

Why Whole Life Fills the Bond Seat

The bond seat was never about growth. Its job was stability, holding the money you spend from when stocks are down so you are not forced to sell shares at the bottom. Whole life does that job, and it does it with two structural advantages a bond does not have. Both deserve a careful explanation, because this is exactly where people oversell the product.

It is non-correlated, which is not the same as inversely correlated. Whole life cash value does not rise when stocks fall. It simply pays no attention to the stock market at all. That distinction matters, because bonds were sold to a generation as inversely correlated, the safe harbor money would rush into whenever stocks dropped, pushing bond prices up as stocks fell. Reality stopped cooperating.

Across the last five years, when investors fled stocks they mostly ran to cash, not bonds, and bonds fell right alongside stocks in 2022. Non-correlated is the honest and useful property. You do not need an asset that jumps when stocks fall. You need one that holds its ground so you can spend from it during the drop, and cash value does exactly that.

Cash value never marks to market. This is the mechanical heart of the whole argument, so it is worth being precise. A bond fund is repriced every single day against current interest rates. Rates tick up, the fund's share price ticks down that afternoon, and your statement shows the loss immediately. Whole life cash value works on a different engine.

It grows on the insurer's general-account portfolio and the dividend the company declares, and both of those move slowly and only in one direction on your ledger. There is no day, and no year, where your cash value falls because rates spiked. When rates rose hard in 2022 and VBMFX posted one of its worst years on record, a correctly built whole life policy simply kept crediting and posted no loss at all.

Over time, higher rates feed higher dividends, on a lag, because the insurer reinvests its portfolio at the new richer yields. So the same rate move that punishes a bond fund eventually helps the policy.

Outside Research Says the Same Thing

We sell this for a living, so our word alone should not settle it. What should carry weight is that researchers who sell none of it keep reaching the same conclusion. The mechanism they study has a name, the buffer asset, and it works like this: in a year your stock portfolio is down, you draw income from the policy's cash value or take a loan against it instead of selling stocks at a loss.

You give the stocks room to recover, then shift withdrawals back to the portfolio once they do, and future premiums and dividends restore the policy. Because the order of your returns matters more than the average, a bucket you can draw from in the bad years protects the whole plan, and it protects it most in the first years of retirement, when a loss does the most damage.

Wade Pfau, who is no cheerleader for the insurance industry, published exactly this result in the FPA Journal in February 2019, with Michael Finke, showing that a properly funded whole life policy improved both income sustainability and legacy value against a bond-heavy plan over a long retirement. His post-2022 work is fair about the other side.

Higher bond yields have helped the fixed-income case at the margin, he says, and he still argues for keeping an insurance buffer in the plan. LIMRA adds a behavioral finding worth as much as any return number: retirees who own guaranteed products spend their other savings with more confidence and report higher satisfaction, because a floor they cannot outlive frees them to spend it.

Behavior, not spreadsheets, decides how most retirements go.

And it is not only academics. Ernst & Young, who sell no insurance, modeled the combination and found that a 35-year-old couple pairing investments with permanent life insurance and an annuity reached about 3.5% more retirement income and left 16.3% more legacy at age 95 than an investment-only plan, with up to 30% of annual savings flowing to permanent life insurance.

When the people selling the product and the people selling nothing run the numbers separately and meet in the same place, the result is worth taking seriously. We line up more of that research in our piece on whole life insurance vs. bonds.

The Honest Objections

We would rather take on the strongest critiques head-on than pretend they do not exist. Three deserve real answers.

Kitces on the LIRP. Michael Kitces argues that permanent life insurance, used for retirement income, trails a low-cost taxable brokerage plus a term policy. His strongest point is about product design, not product hate: most whole life sold at retail is built for protection, with a big death benefit and minimal cash value, and agents rarely disclose how much the cost of insurance drags on the return.

On that median retail policy, he is correct. On a high-PUA policy from a strong mutual carrier, the design in this post, he is not, because the whole point of minimizing the death benefit is to shrink that cost-of-insurance drag. The 5% return in the ledger above is real and already net of those costs. We agree with him about the average product. We part ways on what the right product can do.

Pfau on higher yields. With 10-year Treasuries above 4% and 30-year TIPS real yields near 2.4%, you can now build an inflation-protected income ladder that the zero-rate world of 2019 could not offer. That is a genuinely useful tool, and we will not wave it away. It solves a different problem, though.

A TIPS ladder throws off no growing death benefit, gives you no rebalancing-free bucket to draw from, and offers no loan mechanism for buffering a stock drawdown. It also spends its principal down to zero by year 30 by design, which forces you to bet on your own lifespan landing inside a narrow band. For the right client the ladder and the policy work together.

Pfau's own updated position reads as insurance plus improved bonds, not insurance instead of bad bonds.

The demoted 60/40. Grant that 4.8% bonds and 6.4% for 60/40 hold up. Those are not disaster numbers, and someone could argue a slightly higher savings rate covers the gap. Two problems. That 4.8% sits before fees, before taxes, and before sequence damage, and ordinary income tax on bond interest drops the after-tax figure closer to 3% for many households in a taxable account.

The forecast also assumes disciplined rebalancing that real retirees abandon in a scary year. Whole life is not promising a better tax-deferred bond return. It is delivering a comparable after-tax return, protecting the principal from a mark-down, and adding a death benefit at the same contribution rate. Bonds bundle none of that together.

A handful of shorter objections will hit the inbox the day this airs. Quick answers to each.

One last objection deserves a direct answer, because it is the sharpest one available. Are we pitting a friendly bond back-test against a friendly whole life illustration? Partly, and the conclusion holds anyway. The bond track uses 1996 to 2025, a window kind to bonds in its first half and brutal in its second.

The whole life illustration uses today's dividend scale, which reflects the same roughly 4% yields the insurer earns right now. Both sides start in one rate world. If rates fall from here, the bond track gains on paper, but the insurer's new-money yields fall too and pressure future dividends. If rates rise, bond back-tests get worse, but insurer yields improve and support dividends on a lag.

Rates move both sides, so this is not a one-way bet against bonds. And if you want the conservative read, use the stress-tested line where we cut the dividend scale for good, and whole life still comes out ahead.

What We Are Not Saying

We are not telling you to dump your bond fund tomorrow and buy whole life. A few honest limits keep this from turning into the kind of overclaim we spend our time correcting.

With those limits on the table, here is where the argument lands. For the money whose job is to sit stable and non-correlated until you turn it into income, whole life is the stronger and safer holding, and the bond seat is where it belongs. Said in one line: if you are accumulating toward retirement income, aim for 60% stocks and 40% whole life, not the old 60/40 of stocks and bonds.

Bonds keep their real jobs, the coupon and near-term cash. The total-return job they inherited by default is one whole life now does better. Our guide to using whole life insurance for retirement income walks through how to build the structure.

Common Questions

This article is general education, not a recommendation for any specific product or investment. It is not investment advice, and we do not sell or advise on securities-regulated products, including stocks, bonds, mutual funds, ETFs, or variable annuities. We specialize in cash value life insurance and fixed annuities. Modeled comparisons use historical index data as a market proxy and illustrated policy values, which are not guarantees. Individual results vary with product design, funding, timing, health, and personal circumstances. Decide your equity or securities allocation with your own advisor.