OT: Tontine for Retirement Portfolios

Stocks A to Z
#5

I wouldn’t touch this with the proverbial 10’ pole.
Tontine’s are dumb.

And I had the exact opposite reaction, it’s about time.
Not only has the person initiating this he has written a whole book covering the rationale behind the concept. The 168page book has been made free for online download so anyone can make up their own mind about weather this is a rational retirement decision.
How to Build a Modern Tontine https://library.oapen.org/handle/20.500.12657/57034
I’ll have to concede that the book is aimed toward financial geeks. All his simulations and analysis of why this is a logical way for a senior with an unknown and unpredictable life expectancy to finance their final years. He backs up his analysis with opensource programs written in R. Although only the geeks will download and verify the analysis the mere fact that analysis is there and available for critique by other finance individuals is a big plus in his favor. Have you ever seen any annuity company open sourcing their methods?
I’m going to wait and see how the final mutual fund structure evolves with overhead and all before I would actually consider putting any portion of my retirement portfolio in it. But it looks especially promising for a spouse who refuses to actively manage her funds.

RAMc

8 Likes
#6

Tontine’s are dumb.

Au contraire, mon ami.
They’re very similar to commercially available life annuities, with two main differences.

  • There is no counterparty risk in a tontine as there is with an annuity.
  • The returns are wildly higher with a tontine.
    That’s because the sponsor needs no regulatory capital buffer or asset term matching.

It remains to be seen whether the fees and structure of this particular one will make any sense.
If they overcomplicate it, it won’t fly. Nobody will trust it.

The best “KISS” ones go like this:

  • A bunch of people of the same sex and approximate age, at least a few hundred, each put money into a trust.
  • It’s 100% invested in something boring like the S&P. No investment management fees or trading.
  • Dividends first cover the (absolutely minimal) running expenses, and the remainder is sent out as regular cheques split among all current members.
  • Every time somebody dies, that person’s share is liquidated and the cash divided among the remaining folks.

How well does it work? A reasonable expectation example:
Let’s say it’s a group of males age ~65 putting in $X each.
They can all reasonably expect a real income over $X/year by age 85, rising as long as they live.

Ideally the management trustee would be a non-profit, mutually owned by the participants.
If it fails, it will be because of the grabby fees and incentives of the sponsor.

Jim

21 Likes
#7

TONTINES A PRACTITIONER’S GUIDE TO MORTALITY-POOLED INVESTMENTS Paper from the CFA Institute (Charted Financial Analst)
fullmer-tontines-rf-brief.ashx
https://www.cfainstitute.org/-/media/documents/article/rf-br…
The pension system in Sweden is explic itly tontine-like.
European Union member states now allow ton tine offerings.
Canada announced a proposal to allow tontines in retire ment accounts as part of its Budget 2019 reforms.
In the private sector, the CREF variable annuity—which appears to oper ate like a tontine in virtually every way except for its name—was first offered in 1952 and remains open for purchase even today.

TIAA’s CREF the #1 not for profit retirement provider’s Variable Annuity A Tontine has now been available for 70 years.
As of May 2020 when compared to a theoretical 4% systematic withdrawal, CREF has historically paid more; as of May 2020, the payout rate for a CREF annuity at age 65 ranges from 5.9% (Joint Annuity with 20 years guaranteed) to 6.7% (Single Life Annuity).

So why is everyone so afraid of Tontines.
RAMc

3 Likes
#8

So why is everyone so afraid of Tontines.

Two main reasons that it became a bad word:

  • a lot of tontines failed a long time ago.
    Tow main reasons: either they guaranteed a minimum payout level which could not be met (bad idea),
    or the sponsor/trustee simply stole the money. I don’t think the money in the Vanguard Tontine would be at much risk of that.

  • Some popular fiction about members bumping each other off.
    There are no known cases of this happening in real life.

Remarkably, almost everyone thinks they’re banned, particularly in the US.
But nope.
It appears that they are only prohibited in 1.5 states.
(one of the laws is a bit ambiguous, so it’s either 1 or 2).

Jim

2 Likes
#9

* Every time somebody dies, that person’s share is liquidated and the cash divided among the remaining folks.

Aye, thar’s the rub, for me anyway.

So, my best friend, a reasonably healthy 75 year old, invests $1M. The next week he goes in for a routine colonoscopy, suffers a perforated colon, develops an infection, and dies.

Or, he’s hit by a truck.

What’s the widow to do?

Asking for a friend.

Joe

1 Like
#10

* Every time somebody dies, that person’s share is liquidated and the cash divided among the remaining folks.
…
Aye, thar’s the rub, for me anyway.
So, my best friend, a reasonably healthy 75 year old, invests $1M. The next week he goes in for a routine colonoscopy, suffers a perforated colon, develops an infection, and dies.
Or, he’s hit by a truck.
What’s the widow to do?
Asking for a friend.

Simple: you don’t put all your money in something like this.
The purpose of a Tontine (as I see it) is purely for covering longevity risk.
Same with any annuity.

e.g., say you’re 65 and retiring. Congratulations.
First, any money you want to be left in your estate, do something else with. Give it away or set it aside.
Then you start the problem of funding your retirement with the remainder.
Take 10% of your money, maybe 15%, and put it into a tontine.
That completely covers you from age 85-120+, however long you last.
For the other 85-90% of your money, spend it at a rate that it’s all gone over 20 years.
That is a very much better level of income, both before and after age 85, than you get with any SWR approach…and you can’t go broke.
Plus, your income rises in real terms over time. For as long as you live you feel like a genius.

Also, remember that a tontine is age matched.
If your friend were doing this, everybody else in the group would be about age 75 as well.
He might be unlucky not get all that many years of payments before he croaks, but the payments will rise very rapidly.
(in real life, people typically join a tontine at a younger age)

Side note:
The age matching can be done with fancy statistics on payouts rather than making sure all entrants in a pool meet age range requirements.
i.e., you have people of all ages in one big pool, but the sizes of the payouts each member gets get are a function of their life expectancy.
But then the math gets nasty and people don’t understand and trust the system.
The fund starts depending on the accuracy of mortality tables, rather than merely being a mortality table.
I’m a big fan of simpler solutions that can’t go wrong.
Pool 1: you have to be male and at least 60 years of age.
Pool 2: you have to be male and at least 65 years of age…
Pool 15: you have to be female and at least 60 years of age.
Pool 16: you have to be female and at least 65 years of age…
There is nothing prohibiting being “too old” for a pool. You’re just giving a statistical edge to the other participants.

Jim

11 Likes
#11

So why is everyone so afraid of Tontines.

In detective fiction, one member of the Tontine murders all the others so as to get the assets of the Tontine for himself.

I have no idea if this has ever happened in reality.

2 Likes
#12

Or, he’s hit by a truck.

What’s the widow to do?
Another example to add to Jim’s
Tontines as pointed out in the original article aren’t an instrument designed to solve all retirement situations but are uniquely beneficial for giving one individual a larger retirement income than using something like the 4% rule.
However one way a married couple could use tontines to their advantage.
Assume that they reasoned that the survivor only needed 2/3 the income needed together.
Each Purchase a tontine with 1/3 their portfolio and use the 4% rule for the other 1/3.
Significantly more income than using the 4% rule alone. (If using a tontine similar to CREF’s variable annuity).

RAM

3 Likes
#13

However one way a married couple could use tontines to their advantage.
Assume that they reasoned that the survivor only needed 2/3 the income needed together.
Each Purchase a tontine with 1/3 their portfolio and use the 4% rule for the other 1/3.
Significantly more income than using the 4% rule alone. (If using a tontine similar to CREF’s variable annuity).

Having a separate tontine for husband and wife is the right general idea, but it seems to me that the whole idea of a tontine is to enable the couple to jettison the idea of having a 4% rule at all.

Say I’m 60 and my wife is 55, and we both want to retire on our $2m in savings. 4% a year gives us $80,000 a year, but then, if we both die in 20 years, we end up with $2m that we didn’t need and could have used. The problem being that we don’t know, today, that we will be dying in 20 years - we might live another 40 years, and we have to plan for that possibility.

But if we each buy age- and sex-specific tontines, say for $250,000 each, then we can plan to use up almost all of the remaining $1.5m by about the time when the tontine would pay out, if we live that long.

We don’t know exactly when that will be, but it increases the precision of our estimate a lot, if the group is big enough. Knowing we can go through MOST of the $1.5m in the next 20 years allows us to pay out a much higher annual amount than if we had to allow for the low-probability event that it might be 40 or 50 years.

dtb

3 Likes
#14

In detective fiction, one member of the Tontine murders all the others so as to get the assets of the Tontine for himself.
I have no idea if this has ever happened in reality.

Some researchers have looked into this.
Based on the evidence, the answer is no.
So, if it has ever happened, the perp was careful to hide the crime.
Either way, it’s vanishingly rare.

Jim

1 Like
#15

As well, I read through (some of) the tedious specifications of the R code in that document. Essentially, they would plan the Tontine NOT as a “hooray for the last man that dies” situation, but as a level payout for all attendees UNTIL the last one dies. This requires mathematical actuarial calculations above my “pay grade”, but they demonstrate it can be done.

FC

3 Likes
#16

Charlie Munger has commented on the perverse and criminal incentives created by tontines.

I cannot remember if his thinking is written in Poor Charlie’s Almanac.

But if you think of human nature and the powerful incentives that a tontine creates, I think that Charlie’s rejection of Tontine’s makes sense.

What theoretically may be interesting, should not be applied if you do not consider the reality of the subject to which they will be applied.

Joining a tontine is dumb.

And IIRC, Charlie said that they are not legal because of the murders they incentivised.

Dumb idea.

#17

And IIRC, Charlie said that they are not legal because of the murders they incentivised.

For recipients of social security payments based upon their ex spouses earnings, when the ex spouse dies, the recipients payments will double. I doubt this really incentivises murders of ex spouses to increase social security payments and I think Charlie was probably speaking tongue in cheek about tontines.

Craig who is worth more dead than alive to his ex spouses.

2 Likes
#18

Charlie Munger has commented on the perverse and criminal incentives created by tontines.
I cannot remember if his thinking is written in Poor Charlie’s Almanac.
But if you think of human nature and the powerful incentives that a tontine creates, I think that Charlie’s rejection of Tontine’s makes sense.
What theoretically may be interesting, should not be applied if you do not consider the reality of the subject to which they will be applied.
Joining a tontine is dumb.
And IIRC, Charlie said that they are not legal because of the murders they incentivised.
Dumb idea.

This is an attitude that is common among people who have heard about the fiction, but not looked into the reality : )

If it’s a pool of 1000+ people, perhaps with new people coming in depending on the design, what murders do you plan to commit?
Why bother?

If it’s a fund that doesn’t allow new participants and the count in the pool drops over time,
it would normally be set up to be liquidated as soon as the number of participants drops to some low number like 30.
So a murderer would get at most 1/30th of the fund.
Which is exactly what his/her periodic payment is at that stage anyway: the last payment they got was 1/31.
Again, why bother?

Certainly humans respond to incentives. There is no incentive to murder.
There were no recorded murders when there were tens of thousands of tontines running.
They are a brilliant idea, better than any existing product in the world for handling longevity risk.
If you read about them.
This is a very good introduction
https://hbswk.hbs.edu/item/can-a-continuously-liquidating-to…

Jim

15 Likes
#19

“This is an attitude that is common among people who have heard about the fiction, but not looked
into the reality : )”

Yes,
50 years ago this subject came up in fiction, maybe even a novel of this title, built around the obvious perverse incentive.

#20

50 years ago this subject came up in fiction, maybe even a novel of this title, built around the obvious perverse incentive.

It happens on a video story, but I cannot find it these days. Hercule Poirot? Sherlock Holmes?

In any case, the Tontine involved less than six people, and the trustee was also a member, and he killed off most of the others before the detective figured it all out.

#21

It happens on a video story, but I cannot find it these days. Hercule Poirot? Sherlock Holmes?

There is indeed an Agatha Christie story, 4.50 from Paddington, involving something like a tontine.

Not to be confused with ‘tontine murder’, an Australian term that arose, it seems, because Tontine was a popular brand of pillows…

https://tvtropes.org/pmwiki/pmwiki.php/Main/Tontine

By the way, my contribution to this enlightening debate is that the perverse incentives attributed to tontines are fairly easy to avoid, as long as contributors don’t know who the others are. On the other hand, in the common inheritance scheme whereby a fortune is to be divided among survivors, the motive is present AND the identity of the competing claimants is usually well known to all, and the opportunities to plan and execute a murder are much more frequent.

Tontines make for a good plot because they are exotic and unfamiliar, but they don’t add much to our usual temptations to murder each other.

dtb

1 Like
#22

my contribution to this enlightening debate…

dtb, your contribution shall be properly assessed and judged according to the new boards rule:

“to limit the flood of off-topic posts, the number of people who will be able to post will be limited”

(Couldn’t resist :slight_smile:

1 Like
#23

<< It happens on a video story, but I cannot find it these days. Hercule Poirot? Sherlock Holmes? >>

My fictional introduction to tontines was P.G. Wodehouse: “Something Fishy”.

https://en.wikipedia.org/wiki/Something_Fishy

-Rubic

#24

It happens on a video story, but I cannot find it these days. Hercule Poirot? Sherlock Holmes?

There is indeed an Agatha Christie story, 4.50 from Paddington, involving something like a tontine.

I suppose you could call that one a Tontine (though I would not). And that one was a Miss Marple story. Pretty good as far as Agatha Christie goes, but not the one I had in mind. I thought it was The Clocks, but it is not.