Costs
Why IUL Is a Bad Investment for Retirement: The Cap Math
Why IUL is a bad investment for retirement comes down to three numbers: a capped index return, no dividends, and a cost of insurance that rises every year.
IUL is a bad investment for retirement for a reason you can write on one line: an indexed universal life policy credits you the index's price change, without dividends, capped somewhere near 9% or 10%, while the cost of insurance inside the policy rises every year you age. The floor keeps a bad index year from showing a negative crediting rate. It does not keep the charges from coming out.
That is the whole case, and it is a narrow one. Indexed universal life is a real product doing a real job: permanent life insurance with a cash value that grows tax-deferred. The problem starts when it is sold as a substitute for a 401(k) or a Roth IRA, under names like "your own private pension," a phrase that borrows the language of a benefit whose investment and longevity risk sits with the employer. Those are two different products, and only one of them is priced as insurance.
Indexed universal life, explained
Follow a dollar of premium.
It hits the policy and a premium load comes off the top, often somewhere in the 5% to 10% range. What survives goes into the cash value. From there, every month, the insurer deducts a policy fee and the cost of insurance, which is the mortality charge for the death benefit you are carrying. What is left is the balance that gets credited.
Now the part most people miss. You own no shares of the index. The insurer holds bonds in its general account, spends part of the yield buying options tied to the S&P 500, and passes through whatever that option budget can buy. That is why there is a cap. The cap is not there to annoy you. It is the size of the option budget, written as a percentage.
Three dials govern what lands in your account:
- Participation rate: the share of the index move you are credited. 100% is common, but it can be lower.
- Cap rate: the ceiling on the credit in any single year.
- Floor: usually 0%. Index losses do not produce a negative crediting rate.
And one thing that is not a dial at all: dividends. Index crediting is measured on the index level, and the index level excludes dividends. That gap never closes.
Why the cap and the floor are not a fair trade
Look at the shape. The floor pays off only in years the index drops more than the dividend yield. The cap binds in every year the index climbs past 9%. Stocks rise in most years, and when they rise they often rise a lot. The two sides of that trade are not sampled equally often, and they are not the same size.
What the cap costs when the index runs hot
Take the fifteen calendar years from 2010 through 2024. Contiguous window, no years picked and none dropped. For each one, compare what an S&P 500 index fund returned against what a 9% cap and a 0% floor would have credited.
Three numbers fall out:
- Index fund, dividends included: 13.9% a year
- Index price change alone, dividends stripped out: 11.7% a year
- Credited with a 9% cap and a 0% floor: 6.5% a year
Dividends cost 2.2 points a year. The cap cost another 5.2 points on top of that.
One detail is worth pausing on. Across those fifteen years the credited rate was 9% eleven times and 0% four times. It never landed in between. The formula was not tracking the index at all. It was a switch with two settings.
And in two of the four floor years, 2011 and 2015, the index was actually up on a total-return basis, by 2.1% and 1.4%. The floor protected nothing in those years. Stripping out the dividends turned a small gain into a zero.
The honest caveat: this was an unusually strong stretch for US stocks, so the cap binds more often here than it would in a flatter decade. A weaker sample narrows the gap. It does not remove the asymmetry, which lives in the shape of the crediting rule, not in the sample.
The drag stack
The crediting rule is only half the story. The other half is what comes out of the cash value before anything gets credited at all.
To keep the two effects apart, the next chart hands the policy the same 7% the index fund earns. No cap, no dividend haircut, no penalty for the crediting formula. The only thing separating the lines is charges.
Twelve thousand dollars a year for twenty years is $240,000 of premiums. Under these assumptions the charges add up to roughly $52,600: a 6% load on each premium, $120 a year of admin, and a cost of insurance that starts at $700 and compounds. But the ending gap is $101,800, close to double the charges themselves. Every dollar pulled out also surrendered its own compounding for the rest of the term. That doubling is not specific to insurance. It is the same arithmetic that makes a one-point difference in fund fees cost about a third of a portfolio over forty years.
Put as a rate, the cash value grows at about 5.2% a year on a money-weighted basis against the 7.0% credited. Charges take 1.8 points annually, before subtracting anything for the cap.
Why the cost of insurance keeps rising
Mortality charges track mortality. Adult death rates roughly double every eight years, which works out to about 9% a year, the Gompertz regularity first described in 1825. There is a real offset: the cost of insurance applies to the net amount at risk, meaning the death benefit minus the cash value, so a growing cash value shrinks the base it is charged on. Whether that offset wins depends on how heavily the policy is funded and how long it is held. But in the decades when a retirement account should be doing its heaviest compounding, your sixties and seventies, the mortality charge is compounding too.
Then there is the surrender charge. In the chart it starts at $9,000 and grades to zero over fifteen years. Its practical effect is that the amount you could actually walk away with does not exceed the premiums you have paid until year six. Deferred annuities carry the same kind of schedule, and for the same reason: it protects the insurer's up-front commission, not the buyer.
IUL vs 401(k)
Different products, but this is the comparison the sales pitch is actually about.
An employer match has no analogue anywhere inside an IUL. A 50% match on the first 6% of pay is a 50% return on those dollars the day they land, before any index does anything at all. No crediting formula competes with that. The one condition attached is the vesting schedule that decides when the employer's share is actually yours.
Traditional 401(k) contributions are also pre-tax. In 2026 the elective deferral limit is $24,500. That deduction is worth your marginal rate immediately, 22% or 24% for a lot of households, and the entire balance compounds. IUL premiums are after-tax, and 6% of each one is gone before it starts working.
Where IUL genuinely wins: no IRS contribution limit, and no income phase-out. For someone already deferring the maximum who also has a permanent death benefit need, that flexibility is worth something. For someone not yet maxing the 401(k), the comparison is not close.
IUL vs Roth IRA
This is the tighter matchup, because a Roth removes IUL's headline tax argument.
Both are funded with after-tax dollars. Both can produce tax-free money later, the Roth through qualified distributions and the IUL through policy loans, with the lapse risk noted above. So "tax-free retirement income" is not something IUL offers and a Roth does not.
| Feature | 401(k) | Roth IRA | IUL |
|---|---|---|---|
| Money in | Pre-tax | After-tax | After-tax |
| Money out | Taxed | Tax-free | Policy loans |
| 2026 limit | $24,500 | $7,500 | None |
| Employer match | Often | No | No |
| Upside | Uncapped | Uncapped | Capped |
| Dividends | Included | Included | Excluded |
| Annual cost | Fund fee | Fund fee | Load + COI |
| Death benefit | No | No | Yes |
What actually separates them is the two charts above. A Roth holding a broad index fund receives the full total return at a fund expense ratio measured in basis points. The IUL receives the price return, capped, minus a load, minus a mortality charge that grows with age.
The Roth's real constraints are its contribution ceiling and its income phase-out, though the phase-out can often be worked around with a nondeductible contribution followed by a conversion. Those two situations are where IUL's flexibility becomes a genuine answer rather than a sales line.
Where IUL actually fits
In fairness to the product, indexed universal life does something a Roth cannot. It pays a death benefit, generally income-tax-free to the beneficiary, whenever death occurs. That is worth real money to households with an estate liquidity problem, a dependent with lifelong needs, or a business buy-sell agreement to fund.
The profile where the math holds up looks like this: a permanent death benefit is genuinely needed, the 401(k) and IRA space is already filled, the policy is funded well above its minimum so the cash value can carry the charges, and the horizon is long enough for the surrender period to burn off.
Remove any one of those conditions and the case thins out. Remove the first one, and what is left is an expensive wrapper around a capped index.
The question the math answers
"Why is IUL a bad investment" is the wrong question to finish on, because IUL is not primarily an investment. It is insurance with a savings feature bolted on, and it is priced as insurance.
The narrower question the arithmetic answers is this: does the death benefit justify giving up dividends, giving up everything above the cap, and paying a mortality charge that climbs every year? For someone who needs permanent coverage and has already used their tax-advantaged room, it can. For someone with an unmatched 401(k) and an unfunded Roth, that same policy is doing the job of an index fund for several points a year more.