Monday, November 16, 2015

No. 127: Tontines Are Not Likely to Stage a Comeback

A recent article about tontines and an intriguing new book on which the article is based are not likely to revive the ancient scheme. However, they provide an opportunity to revisit an interesting but almost forgotten piece of insurance history.

The Article and Book
The recent article about tontines, written by Jeff Guo of The Washington Post, is entitled "It's sleazy, it's totally illegal, and yet it could become the future of retirement." The article was posted on September 28, 2015, as a "Wonkblog" of the "Wonkbook Newsletter." I believe that the article did not appear in the newspaper. Here is the lead sentence: "Over 100 years ago in America—before Social Security, before IRAs, corporate pensions and 401(k)s—there was a ludicrously popular (and somewhat sleazy) retirement scheme called the tontine."

Guo's article was based on a 2015 book entitled King William's Tontine. The author of the book is Moshe Milevsky, a professor at the Schulich School of Business and a member of the Graduate Faculty in the Department of Mathematics and Statistics at York University in Toronto. I read the book with great interest.

The Word "Tontine"
The 11th (2003) edition of Merriam-Webster's Collegiate Dictionary defines "tontine" as "a joint financial arrangement whereby the participants usually contribute equally to a prize that is awarded entirely to the participant who survives all the others." That definition does not do justice to the complexity of tontines.

Tontines are named after Lorenzo A. Tonti, an Italian banker who has received credit for inventing the scheme. Tonti suggested the idea to the government of King Louis XIV of France as a method by which the government could raise money. As Milevsky points out, tontines were used primarily to help governments raise money to wage wars.

A tontine investor enrolled in the scheme by buying a share for a lump sum, and in exchange received interest payments. No return of principal was involved. The investor designated himself or herself as the "nominee," on whose life the tontine was based, or the investor designated another person, perhaps a grandchild, as the nominee.

The nominees were divided into age classes. Initially the interest on a class's share of the tontine fund was divided among all the nominees in the class. As nominees died, the interest on the class's share of the fund was divided among fewer and fewer surviving nominees in the class, thus generating larger and larger interest payments to the survivors. The last several surviving nominees in the class received very large interest payments, and the interest payment to the last surviving nominee in the class amounted to a bonanza. (Milevsky points out that tontines often were modified to prevent large benefits to the last surviving nominees in a class.) When the last survivor in the class died, interest payments to the class stopped and the government did not have to repay the principal of the class's share of the tontine fund.

Relationship to Life Annuities
The expression "life annuity," as used here, refers to what is sometimes called a "pure life annuity" or "straight life annuity." The annuitant receives payments as long as he or she lives, and no payments are made after the annuitant dies. Milevsky points out that there is a close relationship between tontines and life annuities. However, life annuities involve payments that are partly interest and partly a return of principal, whereas tontine payments are entirely interest.

Furthermore, as Milevsky emphasizes, the issuer of a life annuity (normally an insurance company or a pension fund) assumes longevity risk, or the risk that annuitants will live longer than was anticipated in the original pricing of the annuity. Indeed, Milevsky believes that issuers have been underpricing life annuities significantly. His favorable view of tontines is based to a large extent on the fact that the tontine issuer does not assume longevity risk, because the tontine issuer never has to repay the principal of the tontine fund.

Full-Tontine Policies
In the U.S. in the middle of the 19th century, participating (dividend paying) "full-tontine" life insurance policies paid the death benefit on the death of the insured, but provided no surrender values. Insureds were divided into age classes. Dividends were not paid each year to the insureds, but instead were deferred and ultimately were distributed to the remaining survivors at the end of a tontine period such as 20 years. Thus the remaining insureds profited directly from the deaths of other insureds and from lapses by other insureds. Although glowing sales illustrations showed large amounts to be paid to surviving and persisting insureds, full-tontine policies did not gain significant traction with the public.

Deferred-Dividend Policies
A later and much more widely sold version of the tontine policy was the "deferred-dividend" policy, which was also called the "semi-tontine" policy. Deferred-dividend policies included surrender values. Insureds who survived and persisted to the end of a tontine period such as 20 years benefited from the forfeiture of deferred dividends by insureds who lapsed their policies during the tontine period. However, because of the surrender values, the losses suffered by insureds whose policies lapsed during the tontine period were not as large as in the case of full-tontine policies. Glowing sales illustrations appealed to the gambling instincts of the public, and deferred-dividend policies were heavily sold during the final quarter of the 19th century.

It is important to point out that insurance companies were not required to establish reserve liabilities for the deferred dividends. Thus companies selling deferred-dividend policies accumulated large surpluses that were used in part to pay agents large commissions for selling the policies and in part to create a slush fund that could be dissipated by insurance company executives.

The glowing sales illustrations did not work out, not only because the assumptions were unrealistic, but also because the large surpluses were squandered. The result was policyholder disappointment and anger, and deferred-dividend policies became fodder for the muckrakers of the day. Some insurance companies, most notably Equitable Life Assurance Society of the United States, vigorously promoted deferred-dividend policies. Some companies, notably Connecticut Mutual Life Insurance Company, strongly opposed the sale of deferred-dividend policies.

Henry Hyde was the founder of Equitable. Later, in a classic example of nepotism, his son James Hyde became a vice president of Equitable. On January 31, 1905, James Hyde sponsored, at company expense, a "Parisian Ball" at a swanky New York City restaurant. The ball was said to have cost a fortune. James Hyde had spent several years in France and had become addicted to Parisian styles. He did not realize or did not care that his ostentatious behavior would generate massive negative publicity. Nor did he understand how his squandering of company funds would be viewed by the public.

In September 1905, the famous Hughes-Armstrong investigation began in New York. There were several components to the scandal that led to the investigation, and the deferred-dividend policy was one of them. Some historians believe that, if there had been no Parisian ball, there would have been no investigation.

One of the statutory requirements that grew out of the investigation was that dividends had to be distributed annually. Thus deferred-dividend policies became illegal in New York, and several other states followed New York's lead. Thus the death of tontines was not caused by the deferred-dividend policies themselves, but rather by the marketing, accounting, and executive abuses associated with those policies.

The Milevsky Book
Milevsky has a knack for presenting complex mathematical material in an amusing and readable manner. The thrust of his book is his belief that tontines have gotten a bad rap, and that they should become the basis for retirement systems that would have significant advantages over existing retirement systems. In the end he says it is time to embrace tontines, and he assures us he would be among the first to sign up.

General Observations
I disagree with Milevsky because I have two fundamental problems with tontines. First, I think it is inappropriate and distasteful for people to profit directly from the deaths of others. Milevsky may be correct when he says there is no evidence in the history of tontines that foul play occurred in the late years of a tontine scheme, notwithstanding several novels about fictional tontines involving all manner of skullduggery. Nonetheless, I am uncomfortable with the idea of gambling on human life, and the idea of people profiting directly from the deaths of other people.

Second, tontines make no provision for family members who survive the death of a tontine nominee. Milevsky and others may believe that retirement funds should be used to care only for retirees, and that the next generation should fend for themselves. However, I think many retirees want to leave some of their retirement assets for the next generation.

It is fun to read and think about tontines. In my view, however, it is preferable to leave them on the scrap heap of history.

Finally, readers of The Insurance Forum over the years, and readers of chapter 19 (especially pages 189-191) of my new book entitled The Insurance Forum: A Memoir, know I have two fundamental problems even with life annuities. First, it is difficult if not impossible to determine the price of the protection provided by a life annuity; that is, the price of the protection against living too long. Second, as in tontines, there is no provision for those who survive the death of the annuitant, except to the extent that extra costs are incurred for temporary "certain" periods following the death of the annuitant.

I prefer systematic monthly withdrawals equal to minimum distributions required by the Internal Revenue Code, even for nonqualified retirement funds. After the death of the annuitant, such an approach is likely to leave substantial funds for the next generation. However, I recognize that, for persons without substantial retirement accumulations, the minimum distribution may not be enough to live on. Thus a life annuity, which involves invasion of principal, may be needed to provide more generous retirement benefits.

Available Material
The Milevsky book entitled King William's Tontine and my book entitled The Insurance Forum: A Memoir are available from Amazon. My book is also available from us; ordering instructions are on our website,, and I will autograph it upon request.