Met Moshe Milevsky Last Week...........

[bongo2]

To me, a classic "tontine" pays out nothing until the next-to-the-last death, and then gives everything to the sole survivor. That doesn't seem very useful to retirees.

Milevsky talks about a series on one-year "tontines" in which all the funds are paid every year, regardless oft he number of deaths. This type of "tontine" seems useful to me. As he points out, it is closely related to a payout annuity, which is certainly legal.

You could do that, but the one-year variety would only work for sudden deaths, as people would drop out after deteriorating health is discovered. A tontine can pay out dividends and some agreed-on percentage of capital each year (like an annuity) and then liquidate after 50 years, for example. You could avoid the investment restrictions, overhead, and risk premiums of insurance companies and still hedge some of your longevity risk. Interestingly, on Wikipedia there's a comment that they operate "underground" today. It also mentions JTWRS, but I think you'd have estate-tax issues there.

 

[ats5g]

I'm curious as to when these professors find time to teach? I understand the current theory is "publish or perish", but with Milevsky jetting all around the country, who the heck is teaching his class? I'd be a tad ticked if my kid was taking his class and sitting through lectures with a teacher's asst.
The biggest problem I have with the University of Texas is that the school is more of a publishing school, so the professors deal with the kids like they are a necessary evil. They take in the money, but teaching has become secondary. Just strikes me as something wrong here.

This was one of the reasons I left UVA for a smaller school. The profs lectured and the teaching assts [i.e. grad students] answered the questions, graded papers/tests, etc. I remember one stats class were 2/3 of the class dropped it after 2 classes. We couldn't understand a thing the prof was saying, and his board writing was terrible.

 

[barbarus]

I'm curious as to when these professors find time to teach? I understand the current theory is "publish or perish", but with Milevsky jetting all around the country, who the heck is teaching his class? I'd be a tad ticked if my kid was taking his class and sitting through lectures with a teacher's asst.

A common misconception. A school earns its real money on research, not teaching. If possible, universities would become purely research institutions.
One of the large Midwestern monoliths, a Big Ten school, has a saying.
"Students cost money, research makes money".

A professor who is teaching is neglecting his real job.

 

[saluki9]

I'm curious as to when these professors find time to teach? I understand the current theory is "publish or perish", but with Milevsky jetting all around the country, who the heck is teaching his class? I'd be a tad ticked if my kid was taking his class and sitting through lectures with a teacher's asst.
The biggest problem I have with the University of Texas is that the school is more of a publishing school, so the professors deal with the kids like they are a necessary evil. They take in the money, but teaching has become secondary. Just strikes me as something wrong here.

My friend is a wholesaler for one of the largest variable annuity distributors. He recently had Jeremy Siegel in for a lunch meeting with his best advisors. He said that in order to get him to speak for ONE HOUR he had to pay $25,000 plus first class airfaire. I wouldn't be spending much time in the classroom either!

 

[Independent]

You could do that, but the one-year variety would only work for sudden deaths, as people would drop out after deteriorating health is discovered. A tontine can pay out dividends and some agreed-on percentage of capital each year (like an annuity) and then liquidate after 50 years, for example. You could avoid the investment restrictions, overhead, and risk premiums of insurance companies and still hedge some of your longevity risk. Interestingly, on Wikipedia there's a comment that they operate "underground" today. It also mentions JTWRS, but I think you'd have estate-tax issues there.

The 50 year thing is interesting. I think that a more practical example would be 20 years. Imagine a pool of 65 year-olds who agree to all invest $x in a mutual fund. No payouts to anyone for 20 years. Then split the balance among the survivors at age 85.

There are practical problems. You don't know how many other people will die, so you aren't sure what your share will be (if you're one of the survivors). Of course you get more preditability if you have more people in the pool.

You also need to: Assemble all the people; Create a legal structure to own the assets; Do regular accounting and auditing; Find a way to assure that the "survivors" are really alive on the payout date. Pretty soon, you look a lot like an insurance company.