Yes. And is there some "silver lining" of additional safety for subscribers as a result of dropping that guarantee? Depending on how the tontine is structured, the subscribers could have a direct call on the value of the underlying stocks/bonds to make the annual payouts (though the value of these payouts might vary somewhat). Just as we don't count on the financial strength of Vanguard or Fido for the safety of our stock/bond portfolios they hold for us. With an annuity, since the insurance company is making the mortality and market value guarantees, the subscribers are depending entirely on the company's ability to pay.
Also, I do think these tontines would not need to be run by insurance companies at all. Regulators might find that to be more convenient, but it brings a level of costs and a bar to market entry that is inappropriate since the payments would depend only on the value of the underlying assets, not the strength of the company selling subscriptions. It's the elimination of this "pass through" that would result in the most significant cost savings.
Yes, it seems that if you take out the mortality guarantee, you don't need an insurance company.
However, I think this would require a "sponsor", probably a mutual fund company. In theory, any "group of friends" could set this up. But, look at the requirements to form a group. We would need people who all:
- think this is a good retirement income concept
- have similar mortality expectations
- agree on pure life or certain plus life?
- make the same choice regarding single life vs. joint life
- agree on a slope, do they want payouts biased to be increasing, level, or decreasing?
- Or even deferred payouts?
- have similar ideas on the "right" mix of assets
- all want to start now, not at some future date.
Then look at the practical issues of determining the "right" legal structure, assuring safekeeping of assets, getting prompt notification of deaths, doing the accounting, communicating with all the members (and their families) as questions come up, actually withdrawing funds and cutting checks.
This board is probably an ideal place to find a group of people to set one of these up. What's the odds that we could get at least nn people from this board who all answered the same way on the seven choices above?
The ideal sponsor would seem to be Vanguard, as they have a big client base and they know how to, or could learn how to, do the practical stuff above. They would be starting scores or even hundreds of tontines every quarter (I'd guess). It wouldn't be too long before they would have thousands outstanding, each pretty small. And, of course, they have to commit to continue to do this for 40 years into the future - what happens to the tontines if Vanguard decides it doesn't want to fool with them any longer? Somewhere, some regulator is going to want assurance that the sponsor isn't going to disappear with the money, or walk away from long term commitments. Looking at that list, I think the sponsor would want to charge enough to cover all those expenses and commitments.
(Note that Milevsky's concept was that the sponsor would guarantee investment performance - something that's reasonable if the investments are bonds. We've moved away from that.)