1. The pool is cool.
The first thing that came into my head is related to a long series of "challenging" (euphemism for unpleasant) conversations I've had with a very wise and seasoned veteran of the financial services industry. The mantra-on-a-loop from that source is "never ever annuitize" with examples given of insurance products here or there with rider this or rider that. These are products I will never ever know as well as the veteran so I am not really equipped to judge. The mantra-theory I think was that unnecessarily and permanently giving away bequest dollars to an insurance company is a moral sink hole when financial wizardry can achieve similar goals (see point three below). "Yes, but" I say: no amount of financial wizardry and repackaging of forced savings plus compounding can, by an individual investor, replicate access to a pool (unless I'm missing something which I might be) or, for that matter, pay forever...as far as I know. That, and I don't think we should willy nilly over-fetishize (a word we will see again shortly) bequests. Agree to disagree would be the better part of valor but we didn't get there yet. The "pool," though, is is a big idea worthy of consideration and exploitation especially these days in a world with waning interest in DB plans. That's why annuitization, if we were speaking the same language (which may not be true), is of interest to me, mantras notwithstanding. Tontines play the same game as annuitization and pools just in a different way but with a clouded past and hopefully a better future.
2. Academics and actuaries seem to have their payment stream and bond fetishes.
The second thing I was thinking about was that almost all the writing about alternatives to annuities or traditional tontines that I have seen seem to fetishize (there it is again) payment streams, longevity maker-taker risk, bond portfolios and duration matching etc etc. The authors of this piece even invent some new declining dividend bonds! (if I read that right) to fund the pool. My guess is that this is because the guys (usually guys; let's change that up shall we) that write the stuff are actuaries and academics and not investors or real retirees. There are other ways to do this kind of thing that might be of interest to the market where market is defined as investor/retirees.
3. Up to a particular, underappreciated point there is no such thing as bequest.
The third thing I was thinking about is that in the absence of annuitization I must reserve capital for superannuation and/or spend less. There is no way around this. Longevity is not a number it is a probability and it has a distribution and the distribution tails off to the right well past the mean (say 85), past the mode (say 90), past the third quartile (say 95), past 100, past 105 and it keeps on going to something like 120...for now. Unless I am callow or stupid or an uninformed gambler I must allocate capital today to cover that future consumption and not use it today. Given the probabilities it seems like dead capital that I'd rather use now but I can't. I could estimate it's present value but won't. If I live past 85 or 95 I will need it for consumption: nursing home or diapers or vitamins or whatever. BUT! If I die before then I shouldn't really think of that capital as a potential bequest (or "lost bequest" if I sink it into an insurance contract) because I am not thinking of the other side of the coin. If, say, I don't have the money and have consumed it before 95 then my children or their children (or the state) will pay. That is a negative bequest. If one is going to fetishize bequests one must also fetishize negative bequests. Let's say the reserve today is hypothetical 100k. It sits there dead. If I buy a DIA with it its gone, poor heirs boo hoo. But then again they are no longer on the hook either. So let's say 100k doesn't really belong to the heirs at all in any way. It belongs to me age 95+ via consumption or it belongs to the annuity pool which is the same thing. Or alternatively I'd maybe have to have an explicit agreement with my children that they can maybe get it now and if I die "late" they are on the hook for everything thereafter. That's an annuity by another name and not really a bequest. So, just to be clear, my opinion is that "there is no bequest" on this side of the place where longevity probability meets future consumption. Maybe there is bequest on the other side of that but that is a wealth management and planning problem not in this note.
To summarize some opinions:
- access to the pool can unlock or create value out of thin air in ways financial engineering can't
- academics and actuaries over-focus on payment streams and aren't thinking like investors
- our capital, one way or another whether we like it or not, belongs to the future, up to a point, and not to heirs
My Make Believe Pseudo-Tontine
So, after getting half way through the piece I decided to create my own fake partial tontine in the way I would like to see it that addresses some but not all of this. For fun let's call it an "equity-linked fixed-duration structured tontine-ish note." Rather than:
- use bonds
- make huge payments to one last survivor (the article has solutions to this too)
- have an open ended duration
- make periodic continuous payments
we will rather:
- use a return generation process from equity markets or perhaps related derivatives or swaps
- have a fixed duration length like some kind of a non div paying unit investment trust
- have one total-return payment at the end to survivors; the dead will forsake their stake
- have more modest returns of and on capital that are made to a larger survivor group
Note that it will not pay all the way to the last oldest survivor. This will just be a direct transfer of capital from the non-survivors to the survivors over a fixed time frame. It is a total-return finite-duration risk pool product. The survivors can do whatever they want thereafter but they do still have risk.
Here is how I'd describe the setup:
- we'll have 1000 60 year olds with 100k each in dead, locked up capital as described in #3 above
- The pool starts with $100M
- The note duration is, say, 25 years so to age 85
- mean longevity expectation is 85
- we have a return generating process with
o 8% arithmetic return
o 10% standard deviation
o 7.5% geometric mean return expectation over 25 years
- a conditional (and deterministic) survival probability of .57 at age 85 for the cohort
After we close the note in 25 years:
- 100k would have grown to 609,834 (somebody check my #s, I make dumb mistakes)
- The pool should have grown to 609.8M
- there should be 570 survivors give or take
- 609.8M/570 = $1,070k per survivor
- 1070 - 609.6 = 460k incremental gain from pooling to survivor
- the PV of 460k incremental discounted at some bond/inflation rate (3%?) = 219.7k
So just by engaging with the pool I have unlocked or created 220k of new value today out of thin air by way of the transfer of capital from the early diers to me...but only conditional on my survival. I can use that now (or really "then") for consumption, annuity purchases, another tontine, whatever. If I die early there is no loss to heirs because it was never really theirs anyway, it was my dead capital now more profitably used if I survive. The heirs should throw me a party because I just inoculated them with respect to ton of nursing home fees they might have had to dig deep for. Also there is no need to worry much about:
- an insurer needing big capital reserves
- longevity advancements and actuarial risks tipping the boat
- need for institutional default risk or margin-of-error loads
- perverse social effects of few survivors with big payouts
- underwhelming fixed income risk premia driving the return generation
Is this vaguely comparable to a DIA today? I have no idea. I should check that out some time. Fees, loads, frictions, taxation? Again, no idea. Also this does not solve consumption risk management from 60 to age 85 nor does it really solve much after that especially the far tail longevity risk which, if you think about it, is what we really want to solve...so I don't really think we have advanced things a ton. We just got a little boost from the pool to help us on our way. It's like a birthday gift for making that far. My advice would be to use the $1070k at this point and annuitize it.
Let's be clear though. Most of this is silly. I'm just playing around. But I am serious about the idea that access to "the pool" needs more creativity than just creating annuity-looking-thingys with weird new bond structures that will never happen. There are other ways. There have to be.
No comments:
Post a Comment