The tontine: how does money work when the share of the deceased passes to the survivors?

Imagine ten neighbours putting the same amount of money into one jar and agreeing: "In twenty years all the money in the jar will be divided only among those still alive". In the first years everyone wishes the others good health. At the end, though, someone may be sitting there waiting to outlive everyone, perhaps even waiting for the others to go. This is not a joke. Such a structure is called a tontine: it was used both as government debt and as an insurance product, and in the end it came under a major investigation.

This article answers one question: how did the tontine come about, why did it move into insurance, and why is it still useful for an accountant to know about it today?

What is a tontine and where does the name come from?

The essence of the tontine is simple: participants put in money, and the share of a participant who dies is not returned to the heir but is distributed among those who remain alive. As a result, every surviving participant receives more as time goes by.

The name comes from Lorenzo de Tonti. Encyclopaedias mention him as the Neapolitan financier who devised the idea of the tontine. The idea is said to have been proposed in the middle of the 17th century so that the French government could borrow, but the exact year and details differ between texts, so we do not name a year. Such a loan was actually issued in France in 1689.

Why did the state borrow through a tontine? France, 1689

When a state looks for extra money, it tries ways other than ordinary borrowing. The peculiarity of the tontine lay in its condition: if a creditor died, his share passed not to his heir but to the surviving creditors. In France, from 1689 to 1759, the Contrôle général issued such a loan ten times (the Gallais-Hamonno study). According to the author's description, in these loans the surviving subscribers received their own interest every year and, in addition, shared among themselves the interest of the deceased subscribers.

For the creditor this was a bet on a long life: whoever lived longest would receive a large share.

England, 1693: war, hope and strange statistics

In 1693 the government of King William issued the first government tontine to finance the war with France (according to Milevsky). The aim was to raise about a million pounds sterling, but Milevsky also says that not enough people bought this tontine.

Later, economists returned to the data of this tontine. Milevsky's 2014 article examines whether the recorded survival rates were unusually high compared with the mortality of the population of the time, and asks whether this should be explained by selection (more healthy people being enrolled) or by fraud. This gives us an important lesson: when a payment is tied to a person's life, checking age, health and documents becomes part of the contract too.

Across the ocean: tontine insurance (from 1868)

In the 19th century the idea of the tontine moved from government debt to life insurance. According to the Ransom and Sutch study, tontine insurance was introduced in 1868 and combined life insurance with an unusual form of saving for old age. Britannica writes about Hyde that instead of an annual dividend the benefits were paid when the policy reached a set term, that is, the payment was tied to a term fixed in advance. On the American market this product became so popular that by 1905 roughly two thirds of life insurance in force was of this type.

The logic of the offer was simple: "If you live long, you get a lot". But the subsequent investigation accused this business of corruption and extravagance, as the next section explains.

1905-1906: why did the investigation lead to a ban?

In 1905 the legislature of the State of New York began an investigation of life insurance companies. This investigation is known as the Armstrong Investigation. As Ransom and Sutch write, the sale of tontine insurance was banned in 1906 because the investigation accused this business of corruption and extravagance.

Opinion on this is not unanimous. Ransom and Sutch put forward the view that tontine insurance was actuarially sound and attractive for saving for old age, and that the ban was perhaps not necessary. This is the authors' conclusion, so we do not say categorically that the ban was right or wrong.

Then and now

A present-day example: the small savings fund of "Nur Servis"

Suppose "Nur Servis" LLC (a hypothetical name) set up a festive savings fund for its employees. Each of the twenty employees puts in a total of 2 000 000 soum. The fund amounts to 20 × 2 000 000 = 40 000 000 soum (the calculation was checked in python). Suppose that after twenty years 12 people are alive and 8 have died, and under the tontine rule the fund is divided equally among the survivors (interest and inflation are not taken into account, a simplified case).

• Each living participant: 40 000 000 / 12 = 3 333 333.33 soum.
• Their net gain: 3 333 333.33 − 2 000 000 = 1 333 333.33 soum, that is, 66.67% more than the amount they put in.
• The heirs of the 8 deceased: 8 × 2 000 000 = 16 000 000 soum lost, and this money passed to the survivors.
• Check: 12 × 3 333 333.33 ≈ 40 000 000 soum (before rounding of tiyin), that is, the fund was distributed in full.

This example shows why a tontine differs from ordinary savings: money "moves" between the participants, while the total does not change. For this very reason such a structure resembles a gamble more than savings. This example was prepared by the author and is hypothetical. We do not give accounting entries, because how it is recorded depends on the legal form of the contract, and writing without relying on a specific paragraph of a standard would be wrong.

How does insurance differ from the tontine?

In ordinary insurance the participants do not take money from one another. A person who has paid the premium receives, if the event occurs, the payout written in the contract. The insurer spreads the risk among many customers. We saw this idea in the article on Halley's table and actuarial calculation. In a tontine, who wins at whose expense depends on the participants' lifespans, so it is closer to a game of chance. One more difference: an ordinary insurance company calculates its payment obligation precisely in advance, whereas in a tontine each person's share remains unknown until the last years.

A lesson for today's accountant

1. If "how much will I receive" is not known in advance, it cannot be recorded as income. In a tontine each share depends on the number of participants. This is the author's practical advice: in similar cases it is more prudent to consider income not by the expected amount but after the payment has become certain. For the rule on recognition, open the relevant paragraph of the NAS.
2. The text of the contract matters more than the accounting policy. A product with the same name can have a completely different economic substance under different terms. Before recording a transaction, read what the contract gives to whom.
3. Survival and the make-up of the participants change the outcome of the contract. Based on the data of the English tontine of 1693, researchers raised a question about the reliability of the survival rates. In an audit too, the list of participants and the supporting documents matter.
4. A product's popularity does not make it safe. By 1905 tontine insurance accounted for roughly two thirds of life insurance in force, yet in 1906 its sale was banned. Both a report and a product are judged not only by sales volume but by transparency.
5. Be careful before drawing a partner into a structure where people "win at one another's expense". When setting up a fund or savings scheme for a business, determine in advance its legal form and how it is supervised.

The history of the tontine shows one thing: a money structure can rest on an attractive idea, but until its terms are transparent and supervised, it loses trust. This is why modern insurance values terms written in advance and supervision.