Six numbers that decide whether you're comparing policies or comparing assumptions — including a real case where reading to year 10 instead of year 32 produces the opposite conclusion.
A resource for business owners and professionals evaluating large permanent insurance
Related: A Complete Guide to Canadian Life Insurance · Capital You Can't Touch — leveraging large permanent life insurance
If you are comparing two participating whole life illustrations, the year you stop reading determines which policy wins.
We recently ran two illustrations for the same case — same insured, same deposit, same payment period, same design intent. Read to year 3, the two policies were within $281 of each other. Read to year 10, one was ahead by $128,567 in cash value. Read to year 26, the other had taken the lead in death benefit — and by year 32 it was ahead by $1,340,766.
Same two policies. Three different answers. Nobody was misrepresenting anything. The illustrations were simply being read at different points on the curve.
This article explains what actually drives that curve, and gives you a checklist for reading these documents yourself.
Every par whole life illustration splits into two panels. Understanding the split is most of the battle.
This is the contractual floor. It answers: what does the insurer owe me if the dividend scale goes to zero and never recovers?
Three columns matter:
This panel is a promise. It is the only part of the illustration that is.
This projects what happens if the insurer's current dividend scale holds unchanged for the entire life of the policy.
It will not hold unchanged. Dividend scales move with the investment performance, mortality experience, expenses, lapse experience and taxes of the participating account. Over an eighty-year policy they will move many times.
The single most important sentence in this article:
When you compare the right-hand panels of two carriers' illustrations, you are not comparing two policies. You are comparing two sets of assumptions.
Two insurers with different participating account holdings, different smoothing mechanisms and different dividend scale interest rates will produce different projections from identical inputs — and neither projection is a forecast.
This does not make the right panel useless. It makes it a design document rather than a performance document. It shows you the shape the policy is engineered to produce. That shape is genuinely informative — as long as you read it as shape, not as outcome.
Here is the mechanic that explains almost every difference you will see between two par illustrations, and it is rarely explained to buyers.
When you commit $100,000 a year to a participating policy, that money splits in two:
Different carriers — and different designs within the same carrier — split that $100,000 very differently.
In our two illustrations, both funded at $100,000 a year:
| Policy A | Policy B | |
|---|---|---|
| Base premium | $58,018 | $40,126 |
| Additional deposit room | $41,983 | $59,874 |
| ADO as % of deposit | 42.0% | 59.9% |
| Guaranteed death benefit | $7,405,250 | $7,002,807 |
(Both illustrations: newborn insured, $100,000 annual deposit with ADO maximised, 100-pay, dividends switching to premium offset at year 11. Carriers not named — this is a lesson about structure, not a product recommendation.)
Read that table again, because it predicts everything that follows.
Policy A directs 58% of every dollar into base coverage. That buys a higher guaranteed death benefit — $402,443 more, guaranteed for life, regardless of what dividends do.
Policy B directs 60% of every dollar into the deposit option. ADO is a far more efficient cash-value engine — most of it lands in value rather than being consumed by the cost of coverage.
So before we look at a single projected number, we can already say: Policy A should have the stronger guaranteed floor. Policy B should have the faster cash value growth. And that is exactly what happens.
| Year | Policy A | Policy B | Difference | Ahead |
|---|---|---|---|---|
| 1 | $91,384 | $82,233 | $9,151 | A |
| 2 | $194,405 | $176,789 | $17,616 | A |
| 3 | $285,615 | $285,334 | $281 | tied |
| 4 | $390,617 | $416,340 | $25,723 | B |
| 10 | $1,110,502 | $1,239,069 | $128,567 | B |
| 23 | — | — | $606,591 | B (peak) |
| 32 | $3,208,349 | $3,626,319 | $417,970 | B |
Policy A leads for three years — its larger base coverage carries guaranteed cash value that builds immediately. Then Policy B's bigger deposit engine compounds past it, opening a lead that peaks around year 23 at $606,591.
Then something happens that a ten-year comparison cannot see: the gap starts closing. By year 32 it has fallen to $417,970 and is narrowing by roughly $35,000–$40,000 a year, accelerating.
| Year | Policy A | Policy B | Difference | Ahead |
|---|---|---|---|---|
| 1 | $8,006,083 | $7,898,026 | $108,057 | A |
| 9 | $14,825,091 | $15,592,520 | $767,429 | B (peak) |
| 25 | $17,504,765 | $17,539,610 | $34,845 | B (barely) |
| 26 | $17,837,950 | $17,709,057 | $128,893 | 🔄 A |
| 32 | $20,417,960 | $19,077,194 | $1,340,766 | A |
The death benefit crosses over at year 26. Policy B leads for twenty-four years, then Policy A takes the lead and pulls away hard.
And at the last year we can see, Policy A is compounding faster on both measures:
| Cash value growth (yr 31→32) | Death benefit growth (yr 31→32) | |
|---|---|---|
| Policy A | 6.58% | 2.51% |
| Policy B | 4.66% | 1.40% |
Stop at year 3 → the policies look identical.
Stop at year 10 → Policy B wins on both measures.
Stop at year 32 → they have split: B on cash value, A on death benefit, with A closing on both.
None of these readings is dishonest. They are all incomplete.
This is the most common failure in insurance comparison, and it usually isn't anyone's fault — ten-year summary pages are what gets emailed around.
At year 30, Policy B holds $487,463 more cash value while Policy A carries $888,515 more death benefit.
Neither policy is better. They are answering different questions:
Decide which question you're asking before you look at the numbers. Otherwise the numbers will decide for you, and they'll pick whichever column you happened to look at first.
Policy A's guaranteed death benefit is $402,443 higher — permanently, contractually, regardless of dividend performance.
That guarantee isn't free. It's funded by the $17,892 a year of extra base premium that didn't go into the deposit option. That's precisely why Policy B's cash value ran ahead for three decades.
More guarantee, less projected upside. Less guarantee, more projected upside.
Neither is "better." The question is which risk you'd rather carry.
If dividend scales underperform for a long stretch, Policy A's floor protects you. If they hold up, Policy B's engine compounds harder for a long time. You are choosing what you want to be wrong about.
Something easy to miss: Policy A's guaranteed cash value sat at exactly $365,829 for four consecutive years (years 10 through 13) before resuming growth.
Policy B's guaranteed cash value grew smoothly through the same period and overtook Policy A's at year 11 — despite Policy A leading through year 10.
This kind of plateau is a product-design artifact, not an error. But if you happened to compare guaranteed cash values in year 10 versus year 11, you'd reach opposite conclusions about which contract guarantees more.
Look at the whole column, not a row.
Both illustrations switch to premium offset at year 11 — the point where dividends are redirected to pay premiums instead of buying more paid-up insurance, so out-of-pocket payments stop.
Watch Policy A's death benefit across that transition:
| Year | Death benefit | Change |
|---|---|---|
| 10 | $15,823,762 | — |
| 11 | $15,734,891 | −$88,871 |
| 12 | $15,673,791 | −$61,100 |
| 13 | $15,642,595 | −$31,196 |
| 14 | $15,643,602 | +$1,007 |
The death benefit falls for three years before recovering. This is not a defect. Dividends that were buying additional coverage are now paying premiums, so coverage stops growing and the existing additions have to carry the cost.
Two things follow:
When you're handed two illustrations, ask for and compare these:
Confirm they're comparable at all
Read the structure, not just the totals
Read far enough
Test the assumptions
If the policy will be used as loan collateral
An illustration models a policy. It doesn't model you.
The scenarios that actually damage permanent insurance plans are rarely about which carrier's dividend scale was half a point better. They're about a business downturn that makes premiums unaffordable in year 4, or a plan built on paying premiums for twenty years by someone whose income was never that stable.
No column on the page has that in it.
Which is why the last question isn't "which illustration looks better." It's:
What happens to this policy in the year everything goes wrong at once?
A good illustration comparison narrows your choice between two reasonable products. It doesn't tell you whether you should be buying either.
Are the numbers on the right side of a life insurance illustration guaranteed?
No. The right-hand panel projects values assuming the insurer's current dividend scale continues unchanged for the life of the policy. Dividend scales change with the investment, mortality, expense and lapse experience of the participating account. Only the left-hand panel — guaranteed premium, guaranteed cash value, guaranteed death benefit — is contractual.
Why do two policies with the same premium produce such different values?
Chiefly because of how the deposit splits between base premium and the additional deposit option. A design that directs more to base premium produces a higher guaranteed death benefit; a design that directs more to the deposit option produces faster cash value growth. In the two illustrations examined here, ADO ranged from 42% to 60% of the same $100,000 deposit.
Does "premium offset" mean the policy is paid up?
No. Offset means dividends are being used to pay the premiums that are still contractually due. If the dividend scale falls, you may have to resume paying premiums. A truly paid-up policy — for example after electing reduced paid-up — has no further premium obligation, but carries a permanently reduced death benefit.
Why does the death benefit sometimes fall in the year premium offset begins?
Because dividends that were buying additional paid-up insurance are redirected to paying premiums. Coverage stops growing and can decline for a few years before resuming. In the illustration examined here, the death benefit fell $88,871 in the first offset year and took four years to recover.
How many years should I look at before comparing two illustrations?
Long enough to see whether the curves cross. In the case examined here, cash value crossed at year 4 and death benefit crossed again at year 26 — a ten-year comparison would have produced the opposite conclusion on death benefit. For a policy on a child, look at ages 65, 85 and 100.
What's the single most useful thing to ask an advisor for?
The same illustration re-run at dividend scale minus 1%. It costs nothing, takes minutes, and shows you which policy depends most heavily on the assumption most likely to be wrong.
This is general educational content, not advice, and not a recommendation of any product. The illustrations discussed are two real quotations run on identical inputs; the carriers are not identified because the purpose here is to explain structure, not to rank products. Illustration values are projections based on stated assumptions, are not guaranteed, and do not predict the performance of any policy. Any actual comparison should be made on complete illustrations, reviewed with your own advisor, accountant and legal counsel.
Wallace Wang Financial Services holds LLQP licensing (life insurance and segregated funds). This firm does not provide securities advice.