Stress-testing an IUL illustration with lower assumed rates

An indexed universal life illustration shows one future. It picks a single crediting rate, applies it to every year, and prints the account value that results. The picture is smooth because real index returns, caps and charges are not. A policy that looks safe under one steady rate can behave very differently if the rate is lower than shown, or if the charges taken out of the account are higher than the illustration assumes.

A stress test is the practical response. You keep the policy and the premiums the same, change one assumption at a time, and see what the account does. This post builds a small model with round numbers I invented, runs it at five crediting rates and with costs raised by a quarter, and shows the arithmetic so you can repeat the exercise on a real ledger. None of the figures is a product's charge schedule, a cap or a forecast of index returns. The behavior, not the numbers, is what carries over.

What an IUL account does each year

An Indexed universal life (IUL) explained policy has an account value. Each year, three things happen to it:

1. Your premium goes in, less any premium load (a percentage taken from each payment for expenses and taxes).
2. The policy takes its charges out: the cost of insurance for the death benefit, administrative fees and the cost of any riders.
3. What remains is credited with interest according to the index strategy, subject to a cap rate and a floor, often 0%.

The ledger prints the result of those steps year after year. The sensitive parts are the charges, which usually rise as the insured gets older, and the credited rate, which the illustration assumes is the same every year. Real policies compute these monthly, with charges deducted each month and index credits applied at the end of each strategy period. The annual model below is a simplification that keeps the logic visible. The direction of the results is reliable even though a real ledger will differ by some dollars.

A word about the assumed rate. Rules for IUL illustrations limit how high a rate an insurer may show, as IUL illustrations and Actuarial Guideline 49 explains, but they do not limit what the policy credits. The illustrated rate is a ceiling on the sales projection, not a prediction.

The hypothetical policy

Say a 40-year-old buys an IUL with a level $500,000 death benefit and plans to pay $6,000 a year for 25 years (age 40 through 64), then stop. The assumptions of the model, all invented:

• Premium load: 6%, so $360 of each $6,000 payment is taken and $5,640 enters the account.
• Annual charges taken from the account: $1,800 in years 1 through 10, $3,200 in years 11 through 20, $5,500 in years 21 through 30 and $9,000 from year 31 on. Real charges rise every year; steps by decade keep the arithmetic readable.
• Illustrated crediting rate: 6.5% every year.
• Each year: account value = (previous account value + net premium − charges) × (1 + rate).

Total premiums paid are 25 × $6,000 = $150,000. The question the stress test asks is whether the policy still stands when the rate is below 6.5%.

Step 1: Rebuild the first two years by hand

Do this on the real ledger first. If you cannot reproduce the insurer's numbers from the stated assumptions, ask for the missing assumption before going further.

Year 1 at the illustrated 6.5%. Net premium is $6,000 × 0.94 = $5,640. Subtract the year-1 charge of $1,800: $5,640 − $1,800 = $3,840. Credit 6.5%: $3,840 × 1.065 = $4,089.60, about $4,090.

Year 2 at 6.5%. Start with $4,089.60, add $5,640, subtract $1,800: $7,929.60. Credit 6.5%: $7,929.60 × 1.065 = $8,445.02, about $8,445.

Year 1 at 3.5%. The same $3,840 times 1.035 is $3,974.40, about $3,974. The gap with the illustrated case after one year is only $116. That small early gap is why a lower rate does not alarm anyone in the first years.

Year 2 at 3.5%. $3,974.40 + $5,640 − $1,800 = $7,814.40; times 1.035 is $8,087.90, about $8,088.

The difference grows because each year's gap is carried forward and also earns less. The way to see the effect is to run it for decades.

Step 2: Rerun the full period at lower rates

Repeating the calculation for 50 years at each rate gives the table below. Values are the account value at the end of the policy year, rounded to the nearest dollar. "Lapse" means the account could not cover that year's charges, so the policy ends unless more premium is paid.

Read across a row, then down a column. Three things stand out.

The early years look similar. At year 10 the account ranges from $38,400 to $55,187, a spread of $16,787. Someone who reviews the policy after a decade would see an account that is lower than illustrated but still growing and might conclude all is well.

The gap widens with time. At year 25, when premiums stop, the 6.5% account holds $190,825 and the 3.5% account holds $114,078, a difference of $76,747. Nothing about the policy has changed except the rate.

The later years are where the policy fails. After premiums stop, the charges keep coming. At the illustrated rate the credited interest more than covers them. At 5% the account peaks near year 30 and then declines. At 3.5% it declines faster and lapses in year 45, at age 84 in the hypothetical. At 2% the lapse comes in year 39, and at the 0% floor in year 35.

Step 3: Do the 0% case by hand

The floor case is easy to compute without a spreadsheet and shows exactly how the policy dies.

• Years 1 to 10: each year adds $5,640 − $1,800 = $3,840. Ten years: $38,400.
• Years 11 to 20: each year adds $5,640 − $3,200 = $2,440. Ten years: $24,400. Total $62,800.
• Years 21 to 25: each year adds $5,640 − $5,500 = $140. Five years: $700. Total $63,500.
• Years 26 to 30: no premium; each year subtracts $5,500. Five years: $27,500. Total $36,000.
• Years 31 to 34: each year subtracts $9,000. Four years: $36,000. Total $0.
• Year 35: no account left to pay the $9,000 charge. The policy lapses.

This is the case in which an insured paid $150,000 over 25 years and the policy ends at age 74 with nothing paid out. It is an extreme test, since a 0% credit every year for 35 years is unlikely. Its purpose is to show that the policy depends on credited interest to carry the charges in later years, and to put a floor under your understanding of the risk. A 0% floor protects the account value from index losses in a given year; it does not protect it from charges.

Step 4: Raise the costs by 25%

The illustration's charge schedule is also an assumption. In many policies, the insurer may raise the cost of insurance within a stated maximum, and it has done so in some cases. universal life cost-of-insurance increases and what to do describes how that happens and why lawsuits have followed. To test it, multiply every charge by 1.25. Charges of $1,800 become $2,250; $3,200 becomes $4,000; $5,500 becomes $6,875; $9,000 becomes $11,250.

Compare the 5% rows. With the original charges the policy survives past year 50. With charges 25% higher it lapses in year 43. A policy that survives one stress may not survive two combined, which is exactly how real problems occur: a period of lower credits and a rise in charges, in the same years.

At the illustrated rate with charges up 25%, the account at year 25 is $150,102, which is $40,723 lower than the $190,825 shown at the base costs. Even in the best case, higher charges shift the whole path down.

Why a steady rate is itself a simplification

The tests above hold the rate constant at each level, which is the easiest thing to compute but not what an indexed policy does. In real life the credit changes from year to year, with a cap on the upside and usually a 0% floor on the downside. Two invented four-year runs show how the pattern, not just the average, matters. Assume a 9% cap and a 0% floor, and ignore participation rates and spreads for simplicity.

Run A, a bumpy index: +14%, −10%, +14%, +2%. The policy credits 9%, 0%, 9% and 2%. Compounded, the index itself ends at 1.14 × 0.90 × 1.14 × 1.02 = 1.193, a gain of about 19.3%. The account is credited 1.09 × 1.00 × 1.09 × 1.02 = 1.212, a gain of about 21.2%. The floor kept the account out of the loss year, so the policy beat the index even with a cap. This is the case sellers point to, and it is accurate: the floor has real value in a volatile stretch.

Run B, a strong run: +20%, +15%, +18%, +12%. The index compounds to 1.20 × 1.15 × 1.18 × 1.12 = 1.824, a gain of about 82%. The policy credits 9% each year: 1.09 to the fourth power is 1.412, a gain of about 41%. The cap gave up roughly half the index gain.

Neither run resembles a steady 6.5%. The first averages about 4.9% a year in the account and the second about 9%, both different from the flat line on the ledger. What matters for the stress test is that the policy's charges do not care about the pattern. They are taken every year whether the credit is 9% or 0%. A long run of years near the floor hurts an account that is also paying rising charges, even if the next run is generous. So test a flat low rate, and also ask the insurer for a ledger using a historical sequence of index returns, which many companies can produce. Compare the account value at the end of the worst stretch, not only at the end.

Also keep in mind that caps, participation rates and spreads can generally be changed by the insurer within the limits stated in the contract, usually each year. IUL caps, participation rates and floors explains how those terms work and how to find the minimum guaranteed values. If the cap you were shown is lowered after you buy, the ledger you signed is no longer the plan you hold, and the stress test should be rerun with the new terms.

Step 5: Decide what the results mean

A stress test does not tell you whether to buy. It tells you how much room the plan has. Consider these readings:

• Lots of room. The policy survives at 3.5% and with higher costs. The illustration is less fragile than most, or the premium is high relative to the death benefit.
• Moderate room. It survives at 5% but not at 3.5%. This is common. It means the plan needs good performance to last to old age, and you may want to keep paying premiums longer than the ledger assumes.
• Little room. It fails at 5% or at higher costs. The plan only works if the illustration comes true, which is not a sound footing for a lifelong need.

Then consider what you can change. Paying premiums for more years, increasing the premium, or choosing a smaller death benefit all add room. So does a plan to review the policy every year and respond early. An annual how to request and read an in-force illustration shows how the account has actually performed, and stress-testing an in-force illustration at lower crediting rates applies the same method to a policy you already own.

Mistakes that make a stress test misleading

A few habits quietly weaken the exercise, so watch for them.

• Changing several assumptions at once. If you lower the rate, raise the costs and shorten the premium period together, you cannot tell which change did the damage. Change one thing, record the result, then combine.
• Stopping at the early years. A test that ends at year 15 shows a policy that looks healthy under every scenario. The failure, where it comes, shows up after premiums stop and charges climb.
• Testing only the downside you can imagine. Include the dull scenario in which the credit is only a little lower than illustrated, say 1 or 2 percentage points. That is a more probable outcome than a long run at the floor, and in the hypothetical above a drop of 1.5 points, from 6.5% to 5%, already turns an account that keeps growing into one that peaks near year 30 and shrinks afterward.
• Treating the guaranteed column as a forecast. The guaranteed column assumes the minimum credit and the maximum contractual charges, so it is a worst case written into the contract. It is useful as a boundary, but your realistic range lies between it and the illustration.
• Forgetting that you can change the plan. A policy that lapses in the model at year 39 can often be rescued by paying a larger premium or paying for more years. The test is a way to find out how much you would need to add, and when.

Using this on a real ledger

Ask the agent or insurer for these items in writing, and enter them in the blanks.

1. Premium per year: ______ for ______ years.
2. Premium load or expense charge: ______%.
3. Charges taken from the account in years 1–10, 11–20, 21–30 and 31 on, from the ledger or the policy's guaranteed and current charge tables: ______ , ______ , ______ , ______.
4. Illustrated crediting rate: ______%.
5. Assumed loans or withdrawals, if any: ______.

Then repeat Step 1 for the first two years. If your results match the ledger within a few dollars, extend the calculation with a spreadsheet and run the lower rates. If you cannot reproduce the numbers, the difference is likely to be monthly versus annual timing, a charge you have missed, or a bonus the model leaves out, and the agent should be able to identify it.

Many illustrations also include loans, often to produce tax-free retirement income. Loans change the stress test materially, because the loaned part of the account may earn a different rate than the rest and loan interest accrues. Ask for a separate ledger with the loans and lower rates, and use the policy loan calculator to look at the interest side. why universal life policies lapse, and how to prevent it explains the most common ways a flexible-premium policy ends without anyone intending it.

A fee-only planner or a fee-for-service actuary can build a more precise model from the contract. Ask your state insurance department if you suspect an illustration is misleading.

Key takeaways

• An illustration uses one rate for every year. A stress test changes one assumption at a time and watches the account over decades, not years.
• In this hypothetical, early values barely move between a 6.5% and a 3.5% credit, while the year-25 gap is $76,747 and the lower-rate policies lapse in the later decades.
• The 0% floor protects against index losses but not against charges. In the hypothetical, a 0% credit lets the account run out at year 35.
• Raising charges by 25% turns a policy that survives at 5% into one that lapses in year 43. Test credits and costs together.
• Ask for the assumptions in writing, reproduce the first two years by hand, and request a ledger at lower rates before you sign or when you review a policy you own.