Dividends Paid Every Year for Over a Century: What a Long Dividend History Tells You

At a seminar or in a brochure, it often arrives as a single, confident line. "Insurers have paid dividends every year for more than 170 years." Through wars, depressions, inflation, pandemics and every market crash anyone can name.

It is a powerful sentence. It says less than it seems to.

A dividend paid every year is a fact about the past. A dividend scale that holds steady is a different fact. And the dividend your own policy will receive in its twentieth year is not a fact at all yet. It is a decision a board of directors has not made.

Why does this matter? Because a participating whole life policy is a commitment measured in decades, and much of the growth people point to in an illustration comes from dividends that are not guaranteed. A family that hears "dividends every year for 170 years" and concludes "so the illustration will come true" has confused a record of reliability with a promise of amount. When the scale is reduced, as scales have been many times, that family feels misled, and may stop paying premiums at exactly the wrong time. A family that understands what a long dividend history proves, and what it cannot, buys with clear eyes and keeps the policy for the right reasons.

This page takes the claim apart carefully. It shows what the record actually says, why dividend scales move, how they have moved in Canada, what the law requires insurers to disclose, and how to ask for the history and judge it.

It starts with the words of the claim itself.

What does "dividends paid every year for over 170 years" actually claim?

At its most precise, it claims that a particular insurer has declared a dividend on its participating policies in every year since a stated date. Some Canadian insurers publish such records. The claim does not say how large the dividends were, whether the scale was ever reduced, or anything about insurers in general.

The claim has a factual root. One large Canadian insurer states in its financial facts document, with figures to the end of 2025, that it has distributed participating policyowner dividends in every year since 1848. Another states in a December 2025 article that it has declared dividends every single year since 1877. From 1848 to 2026 is 178 years. Records of this length are genuine achievements of governance, and there is no reason to dismiss them.

But notice three things the sentence quietly changes when it moves from a brochure to a seminar.

1. "An insurer" becomes "insurers." A record belongs to one company and, often, to one participating account. It does not transfer to the industry, and it does not transfer to the insurer that issues a particular policy unless that insurer is the one that holds the record.
2. "A dividend was paid" becomes "the dividend was reliable." The record counts years in which a dividend was declared. It does not measure the dividend's size, and a smaller dividend still counts as a year paid.
3. "Has paid" becomes "will pay." No record, however long, is a contract term. The same documents that state the record also state that dividends are not guaranteed.

The precise version is still impressive. It simply says something narrower than the loose version.

Each statement heard at a seminar deserves the same treatment. Sixteen of them are checked on seminar claims about participating policies, checked. This page takes claim 4 in depth.

Is a dividend paid every year the same as a dividend that holds steady?

No. A dividend can be paid every year while the dividend scale behind it falls for years at a time. One large Canadian insurer's published history for a participating account shows its dividend scale interest rate falling from 9.30% in 1996 to 5.10% in 2020 and 2021, with dividends paid throughout.

This is the distinction that matters most, and the one the claim most often blurs.

Each year, an insurer sets a dividend scale: the formulas and rates it will use to calculate the dividends credited to participating policies over the coming period. One of the inputs to that scale, and the one most often quoted, is the dividend scale interest rate. When the scale is reduced, dividends are still paid. They are simply smaller than they would have been under the old scale.

Here is how that looks in one published record. The figures below come from a brochure one large Canadian insurer publishes for in-force policies that formerly belonged to one of its open participating accounts, before that account was combined with two others on 1 January 2020. The figures are reproduced from the insurer's document for selected years.

Every row says yes. And across those thirty years, the rate fell by 4.20 percentage points from its 1996 level to its low in 2020 and 2021, before partly recovering to 5.75% in 2025.

Both statements are true at the same time. A family that heard only the first one heard half of the story.

The same document reports averages: 5.44% over the ten years to 2025, 6.31% over twenty years and 7.10% over thirty years. The longer the average, the higher it is, because the earlier years had higher rates. A long average quoted today describes conditions that no longer exist.

Why do dividend scales change?

Because dividends come from the difference between what actually happens in the participating account and what was assumed when the policies were priced. The main drivers are investment income allocated to the account, mortality, expenses, lapses and terminations, and taxes. When experience is worse than assumed, dividends can be reduced; when it is better, they can rise.

OSFI's Guideline E-16 puts the principle plainly: policy dividends are generated by the differences between actual levels of experience and the levels assumed. A participating policy is priced conservatively. If the participating account then earns more, pays fewer death claims, spends less or keeps more policies in force than the pricing assumed, the surplus can be shared with participating policyholders as dividends. If not, there is less to share.

The Insurance Companies Act frames how those results reach the account. Under section 456, a federally incorporated insurer keeps its participating accounts separately from its other accounts. Under subsection 457(1), the company's investment income or losses are credited or debited to the participating account under an allocation method that its actuary, in a written opinion, finds fair and equitable to participating policyholders, that the directors approve, and that the Superintendent does not disallow. Under section 458, the company's expenses, including taxes, are debited to it under the same kind of method. Under section 460, the actuary reports to the directors every year on whether that allocation remains fair.

One insurer's May 2026 announcement lists the same set: the scale interest rate alongside mortality experience, expenses, terminations and taxes.

How the participating account's assets are invested, and how market movements reach it, is a separate subject covered elsewhere on this site. So is the difference between mutual insurers and insurers with shareholders. For this page, one point from that second subject is enough: under section 461 of the Act, a company with share capital may pay shareholders a part of a participating account's profits, capped by a formula in the Act and only in a year when it also pays dividends to participating policyholders, and the Policyholders Disclosure Regulations require its participating account management policy to state the percentage it intends to transfer.

How have scale interest rates moved in Canada over recent decades?

In general terms: down for most of the period from the mid-1990s to about 2021, then modestly up from 2022 to 2024, then mostly held, with some small increases, in 2025 and 2026. The published history above fell from 9.30% in 1996 to 5.10% in 2021. Between 2022 and 2024, most insurers in one industry survey raised their rates by 0.10 to 0.55 percentage points.

This description rests only on what can be verified from public sources today. It is not a complete industry record, and every insurer's history differs.

The long decline shows clearly in the published record above. From 9.30% in 1996 and 1997, the rate eased to 8.71% by 1999, ticked up to 8.96% in 2001 and 2002, stepped down to 7.46% by 2005, rose a little to 7.71% from 2007 to 2009, and then fell in stages through the 2010s: 7.36% in 2010 and 2011, 6.96% in 2012, 6.50% from 2013 to 2015, 6.00% in 2016, 5.50% in 2017 and 2018, 5.25% in 2019 and 5.10% in 2020 and 2021. Even inside a long decline, the scale moved up as well as down.

The turn came after 2021. An industry publication reviewing the period from 2022 to 2024 named twelve insurers. Of those, nine raised their dividend scale interest rates, by between 0.10 and 0.55 percentage points, and two kept them unchanged. In 2024, the rates it reported ranged from 5.50% to 6.40%.

In 2026, most of the announcements read for this page were about holding steady. One insurer announced in January 2026 that it would maintain its 2026-2027 scale. Another announced in May 2026 that it would maintain its rate for anniversaries from May 2026 to April 2027. A third announced in June 2026 that its scale and scale interest rate would be maintained until 30 June 2027. The insurer whose history appears above went the other way, modestly: its rate for in-force policies rose to 5.75% for 2025, and it announced 6.00% for newer policies from 1 July 2026. The May and June announcements both stated that dividends are not guaranteed.

Notice what the table does not contain: a forecast. Nobody can supply one honestly.

Notice also the pace. An insurer that smooths its dividends spreads changes over time; OSFI's guideline says smoothing should only be used to avoid undue yearly fluctuations in the scale. That works in both directions: it softens sudden drops, and it slows recoveries.

What do the law and OSFI require insurers to disclose?

A federally incorporated insurer that issues participating policies must have a dividend policy, reviewed in writing by its actuary for fairness, and a participating account management policy. Federal regulations require the dividend policy to state that dividends are declared at the board's discretion and may change. OSFI's guideline adds an annual table of scale rates and account returns over 5, 10 and 20 years.

The dividend policy. Under paragraph 165(2)(e) of the Insurance Companies Act, a company that issues participating policies must establish a policy for determining the dividends and bonuses payable to participating policyholders, before it issues any. Under subsection 165(3.1), its actuary reports to the directors in writing on the fairness of that policy when it is established or amended, and at least once each financial year. The directors must consider that report first, and a copy of the policy goes to the Superintendent within 30 days after it is established or amended.

The declaration. Under section 464, the directors may declare a dividend to participating policyholders under that policy, after the actuary reports in writing on whether it is fair and consistent with the policy. The Act says "may," not "shall." And subsection 464(3) forbids a declaration where there are reasonable grounds to believe the company is, or would be, in contravention of the capital and liquidity requirements under section 515.

What policyholders are told. The Policyholders Disclosure Regulations, in force since 1 June 2011, set out what the policies must contain and how participating policyholders can see them.

OSFI's expectations. Guideline E-16, in the version OSFI published on 18 January 2023, builds on those rules. It expects the dividend policy to say that dividends cannot be guaranteed and that, because experience can worsen, dividends may be reduced. It expects insurers to disclose how often scales are reviewed, whether they smooth dividends and how, and the approximate split of dividends between investment experience and other factors. And it expects a table showing, for the current year and as 5, 10 and 20 year averages, three figures side by side: the dividend scale interest rate, the participating account's return and the participating surplus return.

That last table is the single most useful thing a family can ask for. It puts the quoted scale rate next to what the account actually earned.

These federal rules apply to life insurers incorporated under federal law. An insurer incorporated under a province's law answers to that province's statute, so ask such an insurer which rules govern its participating account and where its disclosure is published.

What does a long dividend history honestly tell you?

That one insurer has managed its participating account with enough discipline to declare a dividend through many economic cycles. That is real information about governance, conservatism and continuity. It is evidence about a company's past conduct, not about the size of any future dividend.

It shows that the insurer priced its participating policies conservatively enough to have surplus to share in hard years as well as good ones. It shows a board willing to keep declaring dividends, even reduced ones, rather than suspending them. It shows continuity of the participating business across changes in management, regulation and, in some cases, corporate structure.

It also has a quieter, legal significance. OSFI's guideline recognises that policyholders' reasonable expectations can arise from marketing, point-of-sale illustrations and the way an insurer has administered its policies in the past. A long, consistent practice is part of what shapes those expectations.

So a long record is one piece of evidence, among several, in judging an insurer.

But it is evidence of the kind that has a built-in limit. Records are announced by the companies that kept them. An insurer whose record was interrupted does not usually advertise the fact, so the records you hear about are, by their nature, the ones that survived. That does not make them less true. It makes them less representative.

What does a long dividend history not tell you?

It does not tell you what next year's dividend will be, whether the scale will fall, how your own policy's cash value will grow, or what any other insurer will do. Nor does it change what is guaranteed in your contract. The guaranteed values come from the contract; dividends remain non-guaranteed every single year.

Here is the honest list.

1. It does not tell you the size of future dividends. A year in which a much smaller dividend is paid still counts as a year paid.
2. It does not tell you the scale will hold. The published record above shows a long period of reductions within an unbroken record of payment.
3. It does not tell you your policy's return. The scale interest rate is one input. Your cash value also reflects the cost of insurance, expenses, taxes and the policy's design.
4. It does not transfer between insurers. One company's record says nothing about another's participating account.
5. It does not transfer between accounts. Where insurers have merged participating accounts, an older record may belong to an account whose policies now share results with others.
6. It does not strengthen the guarantees. The guaranteed values are contractual obligations of the issuing insurer. They depend on its financial strength and are not backed by any government.

A record of the past is valuable when read as a record, not as a forecast.

What does the current-scale column of an illustration assume?

It assumes that the dividend scale in use on the day the illustration is printed continues, unchanged, for every year the illustration shows, often for decades. CLHIA's illustration guideline asks for at least two scenarios wherever non-guaranteed amounts appear, the second less favourable than the first. Ask to see that second scenario.

An illustration for a participating policy usually shows two kinds of columns. The guaranteed columns come from the contract. The projected columns add dividends calculated at the current scale, as though that scale never changed.

That assumption is a convention, not a prediction. Look back at the published record above: an illustration printed in 1997 at a 9.30% scale interest rate assumed a rate the account never reached again in any later year shown. The policy's guaranteed values were unaffected. The projected values, built on dividends at the 1997 scale, were not reached.

The industry's guideline on illustrations, CLHIA Guideline G6, dated April 2009, says an illustration should make clear which values or features are guaranteed and which are not, and show how the non-guaranteed ones may change. Where non-guaranteed amounts are included, it says a life illustration should display at least two scenarios: a primary scenario drawn from a range the insurer judges reasonable, and a second scenario less favourable than the primary one. The only exception is a policy whose values and features are all guaranteed. One insurer's illustration software, for example, always calculates values at three dividend scales, one of which is the current scale less 1% unless a custom scale is entered in its place.

So a lower-scale view is usually available. The question is whether anyone shows it to you.

How to read each column of a policy's table is set out on reading the cash value table. How dividends become paid-up additions, and so compound inside the policy, is what gives a lower scale its weight over time.

How much can a lower scale change the picture?

More than the small difference in rates suggests, because the effect compounds. In this page's hypothetical example, $1,000 a year accumulating at 4% instead of 5% ends about 9.9% lower after 20 years and about 15.6% lower after 30 years. These are illustrative numbers, not any policy's values.

Consider a hypothetical stream of $1,000 credited at the end of each year and left to accumulate at a constant rate. It is not a policy, a dividend or an illustration. It isolates one thing: what happens to a compounding amount when the rate applied to it is lower.

After 20 years, the 4% line is $3,287.88 lower than the 5% line, about 9.9% less. After 30 years, the gap has grown to 15.6%. One percentage point, held for decades, removes nearly a sixth of the result.

Now consider a hypothetical change partway through. The same $1,000 a year accumulates at 5% for ten years, reaching $12,577.89, and then at 4% for the next ten. After twenty years it reaches $30,624.46, which is $2,441.49 lower than the steady 5% path, about 7.4% less.

A participating policy is more complicated than this. Only part of its value depends on dividends, the guaranteed values do not move with the scale, and the scale interest rate is not the rate at which a policy grows. But the lesson carries over. A modest reduction in the scale, sustained for years, makes a visible difference to the projected columns. That is why the illustration at a lower scale deserves as much attention as the one at the current scale.

How do you ask for an insurer's dividend scale history?

Ask the insurance professional, or the insurer, for five things: the dividend scale interest rate year by year for at least 20 years; the participating account's actual return and surplus return over the same years; the dividend policy and the participating account management policy; and the same illustration at a lower scale.

Much of this is already published, because the regulations and OSFI's guideline expect it. Here is a request you can use as written.

1. The scale interest rate, year by year, for at least the last 20 years, for the participating account that would hold this policy. Ask whether that account was combined with others, and when.
2. The account's actual return and surplus return for the same years, or at least the 5, 10 and 20 year averages OSFI's guideline expects insurers to publish beside the scale rate.
3. The dividend policy and the participating account management policy, or their summaries, including how often the scale is reviewed, whether dividends are smoothed, and what share of distributed profits goes to shareholders, if any.
4. The illustration at a lower scale, for example at the current scale less one percentage point, for the exact policy, premium and dividend option being proposed.
5. The guaranteed values alone, year by year, so the floor is visible before any dividend is added.

How a policy dividend is calculated and credited is set out on what a policy dividend is, and the difference between participating and non-participating designs on participating and non-participating.

How do you judge what you receive?

Read the history as a pattern, not a headline. Look at how far the scale has moved, how quickly it responded to the economy, how the scale rate compares with the account's actual return, and whether the policy still makes sense for you at the lower scale and on its guaranteed values alone.

The most important test is the last two rows. If the policy only makes sense for your family at today's scale, it does not make sense. If it still serves its purpose at a lower scale, and the guaranteed values are a floor you can live with, the long history becomes reassurance rather than a foundation.

A participating policy is an insurance contract, not an investment, and it should be judged first on whether the family needs the permanent coverage it provides. The dividend history is one input into how that coverage is likely to behave. It is not the reason to buy it.

Why do past dividends promise nothing?

Because the law makes every dividend a fresh decision. Directors may declare a dividend under the company's dividend policy after the actuary reports on its fairness, and they may not declare one where capital and liquidity requirements would be contravened. The policy itself can be amended. No past declaration binds a future board.

A guaranteed value is a promise written into your contract. A dividend is a share of results the insurer's board decides to distribute, year by year, within rules designed to keep the insurer solvent and to treat participating policyholders fairly. The Insurance Companies Act gives the directors discretion and surrounds it with safeguards: the dividend policy, the actuary's written reports, the separate participating account, the filing with the Superintendent. Those safeguards are reasons to trust the process. They are not a promise about the outcome.

The insurers' own documents say the same. The financial facts document that states the record since 1848 also states that dividends are not guaranteed. The historical scale brochure describes its figures as past performance, which is not indicative of future performance. The May and June 2026 announcements maintaining scale rates both state that dividends are not guaranteed.

When the companies holding the longest records describe them this way, the claim at the seminar should be heard the same way.

Past dividends describe the past. Future dividends will be decided in the future.

What would an honest version of the claim sound like?

Something like this: "This insurer has declared a dividend on its participating policies every year since a stated date. The scale has gone down and up over that time. Dividends are not guaranteed. Here is the history, and here is the illustration at a lower scale."

The honest version is less dramatic and far more useful. It gives a family something to check, not something to believe.

Who should help you read the history?

A licensed insurance professional with years of hands-on experience servicing participating policies, who can obtain the scale history, the account disclosures and an illustration at a lower scale; and an accountant who can confirm how the policy fits the family's tax position. A lawyer or notary helps when ownership or estate questions arise.

The product without the understanding is just an insurance policy. The understanding without experience is just theory. A professional who has serviced participating policies through years of scale changes has seen what a reduction does to a real contract, and can show you, line by line, which values are guaranteed and which depend on dividends. In Quebec, the insurance professional is a Financial Security Advisor (conseiller en sécurité financière); in other provinces, a licensed life insurance agent.

I should be plain about my own position: I am paid by insurer commissions when a policy is bought. That is a conflict of interest, and it is one reason this page shows the scale reductions and the lower-scale arithmetic as plainly as the long records.

How should you read the figures on this page?

The scale interest rates are reproduced from one insurer's published brochure for one participating account, for selected years from 1996 to 2025. The ranges and announcements are taken from public sources read on 30 October 2026. The accumulation tables are hypothetical, computed for this page, and are not dividends, illustrations or forecasts.

The hypothetical tables assume $1,000 credited at the end of each year and accumulated at a constant annual rate, compounded yearly, using the standard future value of an annuity formula; the mid-period example applies 5% for ten years and 4% for the next ten. They ignore the cost of insurance, expenses and taxes, and they do not represent how any policy grows. A dividend scale interest rate is not a policy's rate of return. Dividends on participating policies are not guaranteed and are declared each year at the discretion of the insurer's board. Guaranteed values are contractual obligations of the issuing insurer and are not backed by any government. Amounts are in Canadian dollars.

What this page will not tell you

This page explains what a long dividend record means and how to read a dividend scale history. It does not tell you what any insurer's future dividends or scale will be, which insurer to choose, or whether a participating policy suits your family. It does not compare insurers' records or products. Those answers depend on your needs, your health, your budget and the specific contract, and they come from your insurance professional and your accountant.

Who this does not suit

A participating whole life policy does not suit a family that would buy it only because it believes dividends will match today's illustration for decades, or because a long record sounded like a promise. It does not suit a family that would be unable to keep paying premiums if dividends came in lower than shown, or that has no need for permanent coverage.

It can suit a family that needs permanent coverage, has read the guaranteed values first, has seen the illustration at a lower scale and still finds the policy serves its purpose, and treats a long dividend history as one encouraging piece of evidence rather than a guarantee. For that family, a record of dividends through many decades is reassurance about the process, not a forecast of the result.

A long record is worth respecting. It is not worth more than that.