Are tontines a solution to Canada’s decumulation challenges?

With the aging of Canada’s population, the move to defined contribution pension plans in the private sector and increases in life expectancy, the issue of decumulation is growing in importance.

Decumulation deals with converting retirees’ savings into periodic income that must last for the rest of their lifetime. That can be a challenge for Canadians who don’t have a defined benefit pension, as they’re subject to a number of risks, including longevity and the requirement to invest their own savings, especially as they age.

Read: Employers, government must step up to address decumulation dilemma

An individual guaranteed life annuity is one product that’s available to retirees to help reduce the risks associated with decumulation. By using a portion of their retirement savings to buy an annuity from an insurance company that promises to pay a monthly guaranteed amount for the remainder of their lifetime, retirees can transfer the longevity and investment risk over to the insurer.

However, most retirees don’t purchase an annuity at retirement. With the exception of annuities and defined benefit plans, products that provide longevity risk protection are rare in Canada. There’s a significant need for innovation in this area, which is why there has recently been an increased interest in a very old concept: the tontine.

Traditional tontines

A tontine is an investment vehicle that provides a periodic income to participating investors. However, what’s unique about the option is that when a participant dies, that person’s share of the investment pool remains there for the benefit of the others in the group, rather than for a beneficiary.

Read: New report proposes national pooled longevity insurance program

Here’s an example of how a traditional tontine would work: A tontine for 60-year-old male investors is established at the beginning of 2019 with 1,000 people each investing $10,000 into it. The tontine investments, which total $10 million, are used to purchased bonds that pay four per cent (i.e., $400,000) at the end of each year, at which time the living investors share the proceeds.

For example, if all 1,000 investors remain alive at age 61 at the end of 2019, they’ll each receive $400 of income ($400,000 divided by 1,000). If 780 of the original investors live to age 80 (the end of 2038), they’ll each receive $513 of income ($400,000 divided by 780). And if only 55 of the original investors live to age 100 (the end of 2058), they’ll each receive $7,273 of income ($400,000 divided by 55).

So a traditional tontine can prove very beneficial to investors who live to a very old age. And even if they don’t live a long life, at least they had protection against longevity risk.

Read: Why a little bit of retirement planning knowledge can be a dangerous thing

The concept of tontines isn’t new. The governments of England and France, as well as municipalities in Germany, used them to raise money in the late 1600s. They were also popular in the United States in the 1800s. However, they fell into disrepute in the United States due to cases of embezzlement and concerns about the potential windfalls that accrue to a tontine’s last survivors.

Natural tontines

While it’s unlikely that the traditional tontine as described above will ever become popular again, there’s growing interest in modified versions. A paper published by the Society of Actuaries in March 2018 discusses some of the characteristics of a natural tontine and how it could work in practice.

A similar concept to a traditional tontine, the natural variety is structured to provide its investors with a more even lifetime income stream. It does so by investing the funds in a way that provides more total income to distribute to investors in the earlier years, when more of them are living, and less in later years, when fewer are alive.

Returning to the example above, assume the tontine is a natural one. Instead of distributing $400,000 in income each year, the total payable from the tontine would be higher than that in the first year and would decrease annually so that the expected amount paid to each of the investors remains level over time.

Read: Plan sponsors to consider new mortality improvement scale

For example, if the total income payable from the tontine was $610,000, $475,800 and $33,550 in 2019, 2038 and 2058, respectively, the annual income payable to each investor at ages 61, 80 and 100 are as follows:

  • Age 61: $610 of income ($610,000 divided by 1,000), compared to $400 for the traditional tontine;
  • Age 80: $610 of income ($475,800 divided by 780), compared to $513; and
  • Age 100: $610 of income ($33,550 divided by 55), compared to $7,273.

Although one of the objectives of a natural tontine is to provide a level lifetime income to each investor, actual amounts won’t remain the same because the longevity experience will differ from the assumptions made at the start.

Tontines versus annuities

Since a natural tontine is quite similar to a life annuity, the paper by the Society of Actuaries examined some of the differences between the two options, including the following:

  • While an annuity provides full protection against both longevity and investment risk, there will be a charge imbedded in the premium for the insurer to take on those risks and for its profits. A tontine can avoid most of those charges, because the investors retain some of those risks.
  • A tontine enables an investor to share longevity risk with the other investors, which is referred to as protection against idiosyncratic risk. However, it doesn’t provide protection against the risk that the overall group of investors lives longer than expected, which is referred to as systemic risk.

Read: Buy-ins and boomerangs: A look at the trends in Canada’s annuity market

  • While a tontine doesn’t provide protection against the risk that the overall group of investors lives longer than expected, the upside potential of assuming it tends to be greater than the downside. In other words, there tends to be a greater benefit to the investor if the overall group doesn’t live as long as expected than if it does live longer.
  • For the period between 1986 to 2000, the paper compared historical payouts from guaranteed life annuities with the amounts natural tontines would have paid. Tontines invested in government of Canada bonds would have provided higher initial payouts than annuities about 60 per cent of the time during that period. Tontines invested in corporate bonds would have provided higher initial payouts 100 per cent of the time.
  • If there’s a demand for retirement products that provide protection against longevity risk but still offer the opportunity to invest in return-seeking assets such as equities, it’s possible to establish a tontine with a mandate that includes that type of investment. To the best of my knowledge, no similar product exists in Canada today.
  • It’s unclear what tax and other regulatory requirements would be necessary to accommodate tontines, since they don’t currently exist in Canada. In fact, it will be necessary to deal with key regulatory questions — including the tax implications and whether they would count as insurance products — before creating tontine products.

Read: Canada to still lag behind OECD average replacement rate for typical workers: report

At a time when decumulation is becoming more important than ever, there’s a need for innovation in the area so Canadians can choose from a menu of products that fit their retirement needs. With that in mind, the development of tontine products would represent a welcome addition to the Canadian marketplace. The idea of reintroducing tontines would be to complement, not replace, the guaranteed life annuity products that are currently available.

With tontines, the revival of a very old concept could prove to be part of the solution to the decumulation challenges that an aging Canadian population will continue to face.