Tax figures come from the IRS and index returns from the index publisher. Insurance figures come from published carrier rate sheets and from this site’s own policy model, and every picture says which. Tap show me the detail for the sourcing and the fine print.
Before any argument about caps and taxes means anything, two questions decide whether you should be having the conversation at all.
Everything else on this page is about the layers near the top. They are the wrong argument if the bottom layer is missing.
This ordering is not an insurance idea. It is the sequence most fee-only planners and consumer-finance educators use, and it follows the Consumer Financial Protection Bureau's own emphasis on building emergency savings first: cover the catastrophes, clear the expensive debt, take the free money, then invest — and only then, if you must, speculate.
Why the order is the whole point. Every layer above protection is money you are relying on staying invested. Lose your income with no emergency fund and you sell your investments in the worst month of the worst year, which is precisely the mistake the “retiring in 2000” picture below is about. The base layer is not there to make money. It is there to stop you being forced to sell.
What this means in practice, and it is not what an insurance page usually says. If you are on the bottom two layers, the right answer is almost always term insurance and an emergency fund, because they are cheap and they cover the catastrophe. Permanent insurance is a layer-three-and-above conversation. If someone is selling you a cash value policy while you have credit card debt and no employer match, they have the pyramid upside down.
So the base has to hold first. Next: how big does that base actually need to be, and for how long?
Young: big need, little saved. Older: less need, more saved. The gap between them is the only thing you need to buy.
The falling line is what your family would still have to replace if you were gone — the income you have not yet earned, what is left on the mortgage, the years of raising children still ahead. All three shrink as you age. The rising line is what you have actually put aside. Save faster and it rises sooner; move the slider and watch the crossing point move with it.
Where they cross, you are self-insured. You no longer need to buy the gap because you are the gap. That is the goal, and it is worth saying plainly that this picture is an argument for owning less insurance over time, not more.
Which is why the shape matters when you choose the type. The need is enormous at 35 and small at 70, so the cheapest honest answer for most families is a large term policy sized to the gap and timed to close when it closes. Permanent insurance earns its place when something does not disappear at 70 — a lifelong dependent, a business that has to be bought out, an estate that must be settled in cash, or a legacy you want to be certain about rather than hopeful about. Those are real, and they are specific. "You will always need life insurance" is not one of them.
Method: illustrative shapes only, not a needs analysis. The falling line assumes a $70,000 income with 60% replaced to age 65, a $320,000 mortgage amortising over 30 years from age 30, and a child-raising cost that runs out by 52. The rising line compounds your figure at 7% a year. Your own numbers will differ — this shows the shape, not your answer.
Which is the argument for owning less cover over time. So here is the honest case against that, and it is a strong one.
Cover gets more expensive every year you age — and past about 75, most carriers stop selling term at any price.
Why the term line explodes. Insurance is priced off the chance of dying this year, and that roughly doubles every seven or eight years after 50. A level-premium term policy hides that inside an average: you overpay slightly in the early years so the price can stay flat, and when the level period ends the averaging stops and you are billed the real number. Renewal is usually guaranteed without new underwriting, which is genuinely valuable — but the price is the price.
And then it stops being available. This is the part people miss. Most carriers will not issue a new term policy past about 75 to 80, and a 20-year term is generally not sold much past 70 (Banner Life writes OPTerm 20 to age 70 non-tobacco, 65 for tobacco). So "I will just buy more term later" quietly stops being a plan — not because it is expensive, but because there is nothing to buy. Any new policy also needs you to pass underwriting again, at exactly the age when that is hardest.
What permanent insurance actually is, mechanically. It is the same rising cost of insurance, pre-paid. You deliberately overpay for twenty or thirty years, that surplus compounds inside the policy, and it is spent covering the years when the real cost is $30,000 or $50,000 a year. The level premium is not a cheaper price — it is the same price, smoothed. That is the honest core of the argument, and it is a good one: if the need genuinely lasts your whole life, the only way to afford the expensive years is to have paid for them when you were young.
Now the four qualifications, because the argument is often stated too broadly.
1. Cheaper only if you live long enough. On these figures the permanent path costs more in total until about age 85. US life expectancy for a 45-year-old man is under 80, so for the median person term would have cost less overall. Insurance is not bought for the median outcome — but "permanent is cheaper" is only true past a specific age, and you should be shown that age.
2. Indexed universal life is not the cheapest permanent insurance. If all you want is a guaranteed death benefit for life, guaranteed universal life — a no-lapse-guarantee policy — is normally cheaper per dollar of cover, because it is not trying to build cash value at the same time. Whole life is normally dearer. An indexed policy earns its place when you want the same premium to do two jobs, not when you only want the first one.
3. The permanent line is flat only if the policy is funded properly. The figure here assumes 5% credited every year. At 3% the same coverage needs about 40% more premium — and if the market underdelivers, the insurer asks for more money or the policy lapses. A no-lapse guarantee does not move; an indexed policy's requirement does. The picture on structuring further down shows what happens when this goes wrong.
4. The strongest version of this argument is not about price at all. It is about certainty and insurability. A policy bought at 45 is in force at 85 no matter what your health did in between, at a cost you already know. That is worth paying for when the need is genuinely permanent — a lifelong dependent, an estate that must be settled in cash, a business buy-out, a legacy you want certain rather than hoped-for. It is not worth paying for if the need really does end when the mortgage does.
Method: term figures from this site's own term model, calibrated to 2026 carrier surveys — $500,000, preferred non-smoker male, 20-year term issued at 45 and 10-year at 65. Renewal cost from age 75 uses 2017 CSO mortality with a 15% expense load, which approximates annually renewable term. The permanent premium is solved on this site's policy engine as the level lifetime premium that keeps $500,000 in force to age 121 at 5% credited, and is not a quote. Issue-age limits per published carrier age-limit summaries; guaranteed renewability without new underwriting per Policygenius. Real quotes vary widely by carrier and health class.
So which is it, then? Here is the whole of one life, priced both ways.
The height of the chart is one thing only: what your family is paid. Not growth, not cost — the costs are listed separately underneath, because putting both on one axis is how these charts usually mislead.
The four paths, and what each costs. A 30-year-old preferred non-smoker wants $1,000,000 covered. A level policy that pays that for life costs $5,701 a year. A 30-year term costs $580 a year, and a 20-year term bought at 60 costs $4,545 — after 80 nothing is issuable. The investing path buys exactly that term and puts every dollar of the $5,701 it did not spend to work. Those three cost the same. The fourth — the growing death benefit — costs $12,774 a year, and that is not a choice we made: it is the guideline level premium, the most §7702 will ever allow into a $1,000,000 policy at this age.
What the vertical axis is, since this is where these charts usually cheat. Every line is a payout — the cheque written on the day you die at that age. None of it is a rate of return and none of it is a premium. A chart that puts "what you paid" and "what it pays" on the same axis can make almost any product look like almost any answer, so the costs are kept in the row of tiles below instead.
Starting at 30, the investing path wins — and it is not close. At 7% it is ahead from the first year and self-insured well before 80. The reason is about age, not about the products: term at 30 costs $580 for a million dollars of cover — six-hundredths of a percent of the face amount — which leaves almost the whole premium free to compound for seventy years. At 30, time is the asset and insurance is the cheap part.
The most common misunderstanding on this whole page — so read this one twice. People assume a cash value policy pays your family the death benefit and the cash value. A level policy does not. With a level $1,000,000 benefit your family receives $1,000,000, and the insurer keeps the account value. That is not a catch: the cash value is the money that makes the death benefit affordable in your eighties, by shrinking the amount the insurer is actually at risk for. What it is good for is that you can use it while you are alive — $556,599 of it by 80 on this policy — which is a real benefit, just not a second cheque for your heirs.
You can buy the version that pays both. It costs 2.2 times as much. That is an increasing death benefit — the dashed line: your family gets the $1,000,000 plus everything accumulated, passing $2.8 million by 80. The premium is $12,774 a year rather than $5,701. Why so much? With a level benefit the insurer's risk shrinks each year as your account grows, so the cost of insurance falls away underneath you. With a growing benefit the amount at risk never shrinks — you pay full freight on the whole $500,000 for seventy years and fund the account. You cannot get the cover and the accumulation out of the same dollar.
And there is a ceiling on it that surprises people. $12,774 is not a choice — it is the most the tax code will ever let you pay into a $1,000,000 policy at this age, the guideline level premium under IRC §7702. Funded to that legal maximum, the growing design still runs out at 110, which is why the dashed line stops. If you want more than that, you have to buy a bigger policy — you cannot simply pay more into this one.
Now move the slider, because this is where it turns. The whole argument rests on what the side fund actually earns. At 6% it passes $500,000 at 75. At 5%, not until 82. At 4%, not until 90. At 3% or less it never gets there at all — and on that path the policy is the better answer, because the guarantee did what the market did not. Somewhere between 4% and 5% net is the break-even, and nobody knows in advance which side of it seventy years will land on.
Four things the chart cannot show you, and they all favour the policy. It assumes you actually invest the difference, every year, for seventy years, and never spend it, and never sell in a crash — which is where this strategy usually fails in real life, not in the arithmetic. It assumes you stay insurable enough at 60 to buy that second term policy at a healthy rate; if you cannot, the term line ends two decades early. It ignores that the policy's cash value is accessible tax-free while you are alive, whereas the side fund is taxable as it grows. And it ignores that a policy is a contract while a side fund is a decision you have to keep making.
And two that favour the investing path. The side fund is yours, in full, with no surrender charge and no lapse risk. And the taxable account gets a step-up in basis at death, so the capital gain is erased for your heirs anyway.
The honest conclusion, for a 30-year-old specifically. If the only goal is the largest sum for your family, term plus disciplined investing is very likely to win from this age, and anyone telling a 30-year-old otherwise should be asked to show this exact chart. Permanent insurance at 30 earns its place for different reasons — locking in insurability while you are healthy, a need you already know is lifelong, or an honest admission that you will not actually invest the difference. Those are legitimate. "It will grow more than the market" is not.
Method: term premiums from this site's term model, calibrated to 2026 carrier surveys; policy premiums solved on this site's policy engine as the level lifetime premium keeping $1,000,000 in force to age 121 at 5% credited, 2017 CSO mortality, GPT-compliant and non-MEC. Neither is a quote. The side fund carries a 0.30% annual drag for tax on dividends and receives a step-up in basis at death. Term assumed unavailable after 80, consistent with published carrier issue-age limits. Illustrative only — real quotes and real returns will differ.
So much for protection. Everything from here on is about the money you save on top of it — starting with where it gets taxed.
Where your money is taxed, what the rate will be, and what the market does the year you need it. None of these are yours to choose. All three decide what you end up with.
Not three products — three tax treatments. Most families have almost everything in the middle one.
Taxed now — savings, CDs, a brokerage account. You already paid tax on the money going in, and you pay again each year on interest and dividends, and on gains when you sell. No limits, no rules about when you can touch it.
Taxed later — traditional 401(k), traditional IRA, SEP and SIMPLE. A deduction today, then every dollar out is ordinary income at whatever rate applies that year. 2026 limits: $24,500 into a 401(k), $7,500 into an IRA, catch-ups of $8,000 and $1,100 from age 50 and $11,250 for ages 60 to 63. Withdrawals become mandatory at 73 — and at 75 if you were born in 1960 or later, under SECURE 2.0.
Taxed never — Roth IRA, Roth 401(k), HSA, 529, and cash value life insurance. After-tax in, no further income tax on qualified withdrawals. Roth eligibility phases out in 2026 between $153,000 and $168,000 single, $242,000 and $252,000 filing jointly. Insurance is the odd one: no income limit, but what you can put in is capped by the size of the death benefit, and access is by loan.
Why the mix matters more than the maximum. If everything is in the middle bucket, two things you do not control decide your retirement — the rate Congress sets, and what the market is doing the year you need money. Balances in more than one bucket mean you choose which to draw from. That is an argument for having a choice, not for any particular product.
Source: IRS, Notice 2025-67, 13 November 2025. Required withdrawal age under SECURE 2.0; see IRS RMD FAQs.
For most families the middle bucket holds nearly everything. Which makes the next question the important one: what will that money be taxed at?
The top federal rate has been as low as 7% and as high as 94%. Money in the middle bucket is taxed at whatever it is then.
Read this carefully, because it is often oversold. These are top marginal rates — the rate on the last dollar earned by the highest earners, not what a typical family paid. When the top rate was 91%, almost nobody paid it. And most people retire into a lower bracket than they worked in, which is the entire argument for a traditional 401(k).
What the chart does establish is narrower and still worth something: the rate is not a constant, it has moved enormously within living memory, and you are being asked to guess. Having money already taxed and money not yet taxed means you are not forced to guess. That is diversification, not prediction.
Source: IRS Statistics of Income, Historical Table 23, tax years 1913–2018 (data file). 2018 onward reflects the Tax Cuts and Jobs Act top rate of 37%. Cross-checked against the Tax Foundation.
That is the first thing you do not control. Here is the second, and it is the one people feel.
Three in a row from 2000. Then −37% in 2008. Then −18% in 2022.
If you are still working, this barely matters. You keep buying through it and the recovery arrives.
If you have stopped working and started withdrawing, it matters enormously — you are selling to live on at the same time as prices fall, and those units never come back.
Source: S&P 500 calendar-year total return, dividends reinvested. Index published by S&P Dow Jones Indices; annual series compiled at slickcharts.com. Past performance does not predict future results.
Seven losing years out of forty-one. And a fall costs more than the same-sized gain returns —
Not a projection — arithmetic.
$100,000 in the index at the start of 2008 was worth $63,000 by the end of it. At a steady 8% a year it takes six years just to get back to where it started.
There is a quieter version too. Two accounts that both average 5% a year do not end up together. One swinging +30%, −20%, +30%, −20% turns $100,000 into $108,160. One earning a steady 5% turns it into $121,551. Same average, $13,000 apart, and the only difference is the bouncing.
The condition: a floor removes the down years, but a ceiling removes the biggest up years — and the forty-year saving picture below shows what that costs. This arithmetic explains why a floor is worth something. It does not say the floor is worth its price.
Harmless while you are still paying in. Ruinous once you have started taking money out. Which is what the next picture is.
Here is the strongest case for giving up the good years, and immediately afterwards the strongest case against it. Then you can run it on your own numbers.
$1,000,000, taking $60,000 a year, through the actual returns that followed.
The retiree who stopped working in 2000 did nothing wrong. They did not pick bad funds or panic at the bottom. They started withdrawing immediately before three consecutive down years, and arithmetic did the rest. This is sequence-of-returns risk.
A floor does not make you more money. It removes one specific kind of bad luck — the kind that depends entirely on the year you happened to be born.
The five buttons are five real crediting rules. The first four are currently-sold index accounts taken from published carrier rate sheets, including two with no cap at all. The fifth is not a sold rate — it is the contractual worst case, the 15% spread Nationwide guarantees it will never exceed, and it is there so you can see the floor under the promise. Notice that the ranking is not what you would guess. On this particular 25-year run the 13.25% cap finishes ahead of both uncapped accounts — a spread is charged every single year, including the flat ones, and that costs more than a high ceiling does over a stretch with this many poor years. But the 10.50% cap finishes last of the four, well behind either uncapped account. There is no rule of thumb here: which structure wins depends on the shape of the returns, which is exactly why you should test more than one.
Method: market line uses S&P 500 annual total returns. The indexed line applies each year's S&P 500 price return — index accounts receive no dividends — under the selected rule. Withdrawals at the start of each year in both. The indexed line excludes policy charges, which are real and would reduce it; it isolates the crediting rule alone. Rates: North American 1412NM, effective 15 June 2026; Nationwide FLN-0288AO, effective 15 March 2026.
That is the case for a floor, at its strongest. Now, honestly, the case against it.
$10,000 a year, same five rules. Switch between them and watch the gap move — but never close.
The cap is not one number, and it is fair to say so. Uncapped index accounts exist and are sold today. On this history an uncapped account with a 5.50% spread credits about 9.4% a year against the 10.50%-capped account's 7.1% — a large difference, and the reason the choice of account matters as much as the choice of carrier.
It still does not beat owning the index. The S&P 500 with dividends reinvested did about 11.9% a year over the same forty-one years, and no crediting rule on this page reaches it. That is not a flaw in the arithmetic — it is what you are paying for the floor. Anyone telling you an indexed policy would have out-accumulated the market over the last forty years is mistaken or selling something.
And be careful with backtests, including this one. These are today's rates applied to yesterday's returns. Rates move, and they have mostly moved down: Nationwide's discontinued IUL Accumulator II charged a 9.50% spread where the current version charges 5.75%. Run the same history at 9.50% and the answer changes a lot. The fifth button shows the honest floor of this — the 15% spread that Nationwide's contract actually guarantees as its worst case on that account. Everything above it is a current rate the insurer may change.
So the honest argument is narrower than the sales version, and more useful: a floor is not a way to get richer. It is a way to make the date you retire matter less. Which you need depends on how close you are to needing the money.
Method: contributions at the start of each year, 1985–2024. Neither line reflects taxes, fund fees or policy charges. A real policy deducts cost of insurance and expenses; a real 401(k) may receive an employer match and pre-tax contributions this ignores. Guaranteed maximum spread from Nationwide's IUL rate guide (FLM-0937NY), which states a 0% guaranteed floor, a 3% guaranteed minimum cap on capped strategies and a 15% guaranteed maximum spread on the uncapped S&P 500 strategy. Note: that is the New York rate guide; the crediting rules shown above are from the non-New York Accumulator II 2020, which does not publish a guaranteed maximum spread. Ask for your own contract's guaranteed page — every policy has one.
That is the ceiling. But two more costs sit underneath it, and neither one shows up on a cap rate.
This is given up before any cap, spread or charge is applied. Over forty years it was nearly half the total result.
Why it works this way. An index account is not invested in the index. The insurer holds bonds and buys options on the index with part of the interest, and an option pays on where the price finishes — it carries no entitlement to the dividends the companies pay along the way. So the account credits the S&P 500's price return, not its total return. Nothing improper about it; it is simply how the instrument works, and it is why every chart on this page uses the price return for the indexed line.
Why it is bigger than it sounds. The dividend yield is a percent or two a year, which sounds trivial — and because it compounds inside the return, it cost this saver 2.44 percentage points a year. Compounded across forty-one years it was worth $3.6 million on $410,000 of contributions — about 47% of everything the index produced. And it is given up first. The cap, the spread and the participation rate are all applied to what is left after the dividends are already gone, so a "9.5% cap" is a ceiling on the smaller of the two numbers, not the bigger one.
What it does not mean. It does not mean the account is mis-sold — the floor is being paid for with exactly this money, and the dividends are a large part of how the insurer affords the option budget. It means the honest comparison is "price return with a floor and a ceiling" against "total return with neither", which is what we have used throughout. If anyone shows you an indexed account compared against the S&P 500 including dividends on the same axis, the comparison is wrong in the policy's favour.
Source: S&P 500 calendar-year price and total returns, S&P Dow Jones Indices. Contributions at the start of each year, no taxes, fees or charges on either line.
That is the first cost. The second one is bigger, and it does not care what the market did at all.
The floor protects what the index credits. It does nothing about the charges — and the cost of the insurance itself rises every year you get older.
What is actually inside the policy. An indexed universal life policy is an account with a one-year renewable term insurance charge taken out of it each month. The charge is the cost of insuring the gap between the death benefit and your account value — the "net amount at risk" — and it is priced off your age. At 35 a year of that cover costs about 62 cents per $1,000. At 75 it is $18.55. At 95, $140. That is not a fee anyone is hiding; it is the price of a 95-year-old's mortality, and no product can make it cheaper.
Two things follow, and both matter. First, in a year the index falls, your account value still goes down — the floor stopped the index from taking anything, but the charges came out anyway. Second, a policy that is not accumulating fast enough enters a spiral: as the account value falls the gap the insurer is covering gets larger, so the charge gets larger, which makes the account value fall faster. That is how policies lapse.
What to ask for, and it is the single most useful question on this page. Ask for the illustration's expense pages — the year-by-year table of the cost of insurance and every other deduction, not just the projected values column. Then ask for the same illustration run at a lower credited rate. If nobody will show you either, that tells you what you need to know.
Method: produced by this site's own policy engine on 2017 CSO ultimate mortality with a preferred non-smoker class factor, expressed per $1,000 of net amount at risk per year. Real carrier charges are proprietary, differ by product and are typically higher than the mortality table alone; these figures show the shape and the scale, not any specific policy's rates.
Which makes one decision matter more than the cap, the carrier or the market — how the policy was built in the first place.
Nothing changes here except the size of the death benefit it was written at — and one version needs two full points more every year, forever, or it dies.
The trade-off in one sentence. For a given premium, a bigger death benefit means a bigger amount at risk, which means a bigger insurance charge every month, which means less money left to compound. A smaller death benefit means the opposite. The tax code puts a floor under how small it can go — go below it and the policy stops being life insurance for tax purposes — so "as small as the law allows" is the maximum-cash-value design and "as large as the premium sustains" is the maximum-legacy design.
Neither is wrong. Being in the wrong one is. If your goal is to leave the largest certain sum, the big death benefit is exactly right and the fragility is the price. If your goal is tax-free income later, the same design quietly starves the cash value for decades and can lapse before you get there. Same product, same premium, opposite outcome — and the only thing that decided it was a number on the application.
Why this is worth being blunt about. A larger death benefit generally means a larger commission. That does not make anyone dishonest — but it does mean the incentive points one way, and you are the only person in the room whose incentive points at your own goal. So state the goal out loud before you see any numbers, and ask why the face amount is what it is.
The three questions. What is this policy built to do, in one sentence? What credited rate does it need every year to stay in force — and what happens at one point lower? And what would it cost me to get out in year five, year ten, year fifteen?
Method: both figures from this site's own engine — $12,000 a year for 20 years from age 45, preferred non-smoker, 6% premium load, $10 monthly policy fee, $0.18 per $1,000 monthly per-unit charge for 10 years, 2017 CSO mortality, GPT-compliant and non-MEC, run to maturity at 121. The $590,000 figure is the smallest face amount that accepts this premium without becoming a modified endowment contract. Real products differ; the point is the size of the gap, not the exact rate.
And the cap itself is not one number either. So what does removing it actually buy you?
Four accounts on one product, one day. The regulator makes the insurer publish what each can be illustrated at. They are within 0.71 of a point of each other.
An index account is funded by an option budget — whatever the insurer earns on its bonds, spent on options. That budget is roughly fixed. A cap, a spread and a participation rate are three ways of spending the same budget, not three levels of generosity. Remove the cap and something else has to give: a 5.50% haircut off every year, or 62 cents of every dollar the index moves.
You can see it in the carrier's own filing. On North American's Builder Plus IUL 4, the capped S&P account illustrates at 6.63% and the uncapped one at 6.20% — the uncapped account illustrates lower. Under Actuarial Guideline 49-B every account is tested the same way, so this is as close to like-for-like as the industry produces.
One thing to watch for. The eye-catching participation rates — 190%, 240%, 320% — are real, but they are almost never on the S&P 500. They sit on volatility-controlled or custom indices built for insurance products: J.P. Morgan Mercury, BNP Paribas Global H-Factor, Fidelity Multifactor Yield 5% ER. Those indices deliberately damp their own movement, which is exactly why the insurer can afford 300% of it. A 320% participation rate on an index that moves a third as much is not three times the S&P. If someone shows you a par rate above roughly 100%, the first question is what index it is on, and the second is how long that index has existed.
Sources: maximum illustrated rates from North American's illustrated-rate change notice, effective 24 March 2025 — carriers refile these roughly annually, so ask for the current sheet. Participation rates from Nationwide FLN-0288AO (effective 15 March 2026) and North American 1412NM (effective 15 June 2026). Naming these products is descriptive, not a recommendation, and rates change frequently — check the current sheet before relying on any figure here.
Enough of our assumptions. Put in your own and see what your numbers say.
Your cap or spread, your tax rates, an honest guess at charges, and any stretch of real history. It will tell you whichever answer your numbers produce — including one we would rather not print.
Yes, sometimes — and here is exactly how often, so you can check it yourself. We ran the most generous currently-sold uncapped rule (no cap, 5.50% spread, 0% floor) against a traditional 401(k) in the S&P 500, matched on out-of-pocket cost, taxed at 24% going in and 24% coming out, and — to give the policy every possible advantage — with no policy charges at all. Across every rolling window from 1985 to 2025 the policy came out ahead in 6 of 22 twenty-year windows, 2 of 17 twenty-five-year windows, and none of the 12 thirty-year windows. The twenty-year wins all started between 1989 and 1994 — the windows that end inside the 2008 crash rather than after the recovery. Set the controls above to those years and you will see it.
Now put the charges back. At a modest 1% a year, which is optimistic for a real policy, the wins fall to 4 of 22 twenty-year windows and none at all beyond twenty years. That single change — from a free policy to a cheap one — is the difference between "sometimes" and "hardly ever". It is why the charge box above is the most important number on this whole page, and why you should ask any agent to show you the policy charges year by year rather than only the projected value.
So the claim is true, and the claim is rare. Both halves matter. Anyone who tells you an indexed policy cannot beat a 401(k) is wrong. Anyone who tells you it usually does is also wrong. The floor wins when the market has a bad decade at the wrong moment and loses when it does not — and you do not get to know in advance which one you are about to have. That is the entire honest case: not a better return, an insurance policy against a bad sequence.
The three things that decide it, in order. First, tax — the 401(k) figure above is pre-tax and the policy figure is not, which is worth roughly a quarter to a third of the balance and is the single strongest argument in the policy's favour. Second, charges — at 0% the policy is competitive, at 1.5% a year it is not; over 40 years a 1% drag turned $3.95M into $3.00M in our own run. Third, the market's path, which nobody controls.
Two things this comparison is not. It is not a comparison with a Roth, which is also tax-free coming out, has no cap, no spread and no cost of insurance — on almost every run of history the Roth wins, and if you qualify for one it is the harder benchmark to beat. And it ignores an employer match, which is an instant 50–100% return that no crediting rule on earth competes with. Take the match first. Always.
Why this is reconstructed rather than looked up. You may be wondering why we do not simply plot thirty years of real IUL returns. There is no such public series, and it is worth understanding why: what a policy credits depends on its own cap, spread or participation rate, each carrier sets those separately for every product, and they change them — often annually. Carriers publish only short trailing snapshots; North American's index performance sheets, for instance, shows 1-, 3- and 5-year figures, not decades. So a long IUL history has to be rebuilt the way we have rebuilt it: take the index returns from the index publisher, apply a published crediting rule, and show both sources. Anyone presenting you a thirty-year "actual IUL return" chart has done the same reconstruction — the honest ones tell you so.
Method: S&P 500 annual total return for the 401(k) and Roth, and S&P 500 annual price return under your chosen rule for the policy, since index accounts receive no dividends. Index returns from S&P Dow Jones Indices' U.S. Equities Market Attributes commentary, which reports 2025 at 16.39% price and 17.88% total. Crediting rules from the carrier rate sheets cited in the two pictures above. Illustration limits per NAIC Actuarial Guideline 49-B; for background on how insurers applied its predecessor see the Society of Actuaries' survey of 28 insurers. Traditional contributions grossed up by your current rate so the out-of-pocket cost matches, and the ending balance taxed at your retirement rate. Policy charges applied as a flat annual percentage, which understates the early years and overstates the middle ones. No surrender charges, no loan interest, no fees on the investment side. Educational illustration, not a projection or a quote.
That is money you intend to spend. But what if you never spend a penny of it?
The structure agents actually sell, followed year by year — so you can see which line is the payout, which is your money, and which is what you paid.
The design. A 35-year-old buys $750,000 with an increasing death benefit and funds it at $8,750 a year for twenty years — $175,000 in total, and deliberately just under the seven-pay limit of $11,095 so it never becomes a modified endowment contract. While the benefit is increasing, your family would receive the $750,000 plus whatever has accumulated. So as the account grows, the payout grows with it: $787,526 at year five, $907,801 at year fifteen, and $999,914 at the end of year twenty. That is where the million comes from. It is not interest on the death benefit — it is the account value being added to it.
Then it switches, and this is the part worth understanding. At year twenty the policy moves to a level benefit. The insurer raises the face amount by whatever has accumulated, so the death benefit does not drop — it locks at $999,914 — and from then on the amount the insurer is at risk for starts shrinking as the cash value keeps rising underneath it. That is the whole reason to switch. Leave it increasing and you keep paying full freight on the entire $750,000 into your nineties; on this same funding it runs out at 99. Switch it, and it never lapses.
What the money does after you stop paying. Nothing more goes in after year twenty, and the account keeps compounding anyway: $249,914 at year twenty, $405,553 at 65, $652,045 at 75, $1,093,401 at 85. Of the $249,914 at year twenty, $175,000 is your own money and $74,914 is growth. By 85 the growth is $918,401 against the same $175,000 paid in. That money is yours to use while you are alive — normally by borrowing against it, tax-free, provided the policy stays in force for life.
And then the death benefit starts rising again. Look at the far right of the chart: from about 83 the payout climbs above $999,914. That is not the policy being generous — it is IRC §7702 forcing it. To stay life insurance for tax purposes, the death benefit must keep a statutory margin above the cash value, so once the account gets large enough the benefit is dragged up with it. Useful to know, and a real feature — but it is the tax code doing it, not the insurer.
The two years that actually matter, and neither is on the sales page. The first is age 48 — policy year fourteen — which is when the cash value finally passes what you have paid in. Before that the charges and the surrender charge are ahead of you, so cancelling in year eight means walking away with less than you put in. The second is the year of the switch. If nobody sets it, or it is set too late, the increasing benefit keeps charging on the full face and the policy can run out decades early. Ask when the switch happens and what happens if it does not.
Method: solved on this site's engine — age 35, male, preferred non-smoker, 2017 CSO mortality, 6% premium load, $10 monthly policy fee, $0.18 per $1,000 monthly for ten years, ten-year surrender charge, GPT-compliant and non-MEC, run to maturity at 121. Credited at 6% every year, which is at the ceiling AG 49-B permits and is not a rate anyone is promised — a real policy credits whatever the index does under its own cap or spread, and several of those years will be 0%. Not a quote and not an illustration; a real carrier's charges differ.
That is one policy in detail. Now the practical question most people arrive with.
This is the strongest everyday reason people are shown a policy — the legal tax-free buckets run out. They run out much later than most people think.
The argument, stated fairly. In 2026 you may put $24,500 into a 401(k) and $7,500 into an IRA. Roth IRA contributions phase out between $153,000 and $168,000 single, $242,000 and $252,000 filing jointly. A policy has no equivalent dollar cap. So for a high earner who has filled everything up, it is the one remaining way to add to the taxed-never bucket. All of that is true. It is also incomplete in two ways that matter.
First: three doors most people do not know are open. A Roth 401(k) takes the same $24,500 and has no income limit whatsoever — the phase-out that blocks the Roth IRA does not apply to it. A backdoor Roth IRA — contribute to a non-deductible IRA, then convert — adds $7,500 at any income. And a mega backdoor Roth, if your plan allows after-tax contributions and in-plan conversion, adds up to $47,500 more, because the total that may go into a 401(k) from all sources in 2026 is $72,000. Add the HSA at $8,750 for family cover and that is roughly $88,250 a year into the taxed-never bucket, at any income, before a policy is needed at all. Most people shown an insurance illustration have not been shown this list.
Second: a policy is not unlimited either — its ceiling is just written differently. There is no dollar cap, but there is a ratio. Under IRC §7702A the seven-pay test caps how fast you may fund it, and under IRC §7702 the guideline limits cap it in total — both as a function of the death benefit. On this site's engine, a 45-year-old wanting to shelter $50,000 a year has to buy roughly $1,224,000 of death benefit to be allowed to. That is about $24 of cover for every $1 a year you want to put in — and you pay the cost of insurance on all of it, for life. Overfund past the line and it becomes a modified endowment contract, at which point withdrawals are taxed gains-first with a 10% penalty before 59½, and the tax-free-access argument disappears entirely.
So the honest version is an order, not a shortcut. Employer match, then HSA, then Roth 401(k), then the backdoor and mega-backdoor routes if your plan allows them, then a taxable brokerage account — which has no limit at all, no lapse risk, and whose capital gain is erased at death anyway. A policy comes after those, for someone who has genuinely filled them, wants more tax-free room, and also wants the death benefit — because they are buying $24 of insurance per dollar either way, so it only makes sense if the insurance is worth something to them. If someone reaches for a policy before working down that list, ask them why.
Sources: 2026 retirement limits from IRS Notice 2025-67 — §402(g) deferral $24,500, IRA $7,500, §415(c) total additions $72,000, Roth phase-outs as stated (IRS summary). HSA limits of $4,400 self-only and $8,750 family from IRS Rev. Proc. 2025-19. Mega-backdoor space is the §415(c) limit less your own deferral, and requires a plan permitting after-tax contributions plus in-plan conversion or in-service distribution — many plans do not. Policy funding limits computed on this site's engine under §7702 and §7702A for a preferred non-smoker; every product differs.
That is the practical case. Now the parts that hold up on the evidence, and the parts to refuse.
What holds up when you check it, what is narrower than it sounds, and what you should walk away from.
A death benefit is income-tax-free, and it is there from the first premium. That is a real advantage — and it has an expiry date.
The true part, and it is genuinely strong. Life insurance proceeds paid because of the insured's death are excluded from the beneficiary's gross income under IRC §101(a)(1). And because the gain inside the policy is never realised, holding to death means the growth is never income-taxed either — so "tax-deferred growth" does become "never taxed", on condition that the policy stays in force for life. Surrender it instead and the gain above your basis is ordinary income that year. Same policy, opposite tax answer, decided entirely by how it ends.
The leverage is the actual argument. Die at 60 in this example and $192,000 of premium has become $984,000 — because a death benefit does not have to compound its way up, it is the full amount from year one. Nothing else on the chart can do that. This is why insurance is the answer for a legacy that must be certain and certain now.
Left out #1: the other tax-free transfers. A Roth passes to heirs income-tax-free too. And a plain taxable brokerage account gets a step-up in basis at death under IRC §1014 — a lifetime of unrealised capital gain is simply wiped out. So "tax-free to my heirs" is not unique to insurance. The account it genuinely beats is the traditional 401(k) or IRA, where the SECURE Act makes most beneficiaries empty it within ten years and pay ordinary income tax on every dollar, often stacked on top of their own peak-earning salary (IRS Publication 590-B).
Left out #2: income-tax-free is not estate-tax-free. If you own the policy on your life, the death benefit is pulled back into your taxable estate under IRC §2042 — "incidents of ownership" means the right to change the beneficiary, borrow against it, or surrender it. Keeping it out of the estate normally means an irrevocable trust owning it from the start. For 2026 the federal exclusion is $15,000,000 per person (IRS 2026 inflation adjustments), so this affects very few families — but several states levy their own estate or inheritance tax at far lower thresholds. That is a question for an estate attorney, not a calculator.
The structural point, which is the one that matters to you. A policy built to leave the most money is the opposite of a policy built to hold the most cash. Maximum legacy means the largest death benefit the premium will sustain. Maximum cash value means the smallest death benefit the tax code will allow for that premium. Same product, same premium, and the two designs can differ by several times over. If you want both, you are choosing a compromise — and you should be shown it as a compromise, not as a free lunch.
Method: $12,000 a year of after-tax money for 20 years from age 45. The $984,000 death benefit is what that premium sustains to age 121 for a preferred non-smoker on this site's own model at the AG 49-B maximum illustrated rate, net of typical charges — a real insurer will quote something different. Investment lines assume 7.0% a year; the brokerage line carries a 0.30% annual drag for tax on dividends. Traditional contributions are grossed up at a 24% bracket so the out-of-pocket cost matches, and heirs are taxed at 32% — if their rate equals yours, traditional and Roth land on exactly the same number, which is the whole traditional-versus-Roth question in one sentence. Illustrative only.
Two more arguments hold up when you check them. Both are narrower than they usually sound.
Both real. Both narrower than they usually sound.
The buffer research. Wade Pfau ran 10,000 Monte Carlo simulations on a couple aged 40 with a $500,000 insurance need. Living on policy cash value in down years, instead of selling investments at a loss, produced 23% higher median spending and a 53% larger median legacy than investments alone. Two conditions. The study modeled whole life, not indexed universal life — the products differ and the finding does not transfer automatically. And the author disclosed that related earlier work was funded through an insurer that sells whole life. Neither makes it wrong. Both are why it is a serious argument worth examining rather than a settled proof.
The FAFSA point. Cash value is not a reportable asset for college aid. A 529 or ordinary savings is, assessed at about 5.64% of a parent balance each year, or 20% if held in the student's name. The condition usually left out: retirement accounts are excluded too. A 401(k) and an IRA are just as invisible as a policy. So the advantage is over a 529 or savings — not over a Roth, and not a reason on its own to choose insurance over either.
Sources: Wade D. Pfau, “Investigating the Role of Whole Life Insurance in a Lifetime Financial Plan,” Journal of Financial Planning, February 2019. FAFSA treatment as summarized by Saving for College; confirm current rules at studentaid.gov.
And three you should refuse outright.
You will hear all three. Each is wrong as usually stated.
Which leaves one question, and it is not about the product.
nineteen pictures and not one of them says “buy this”. Every line here won somewhere and lost somewhere else. What decides it is not the product — it is five things about you.
You can work that out on your own first — both tools below are free, take a few minutes, and sell nothing. Or book a free session and talk it through. No cost, no obligation, and if the honest answer is to keep doing exactly what you are doing, that is what you will hear.
Or email contact@financeforeveryfamily.org and say what you are trying to work out.