Medical algorithms applied to SWR

[donheff]

This guy has joined with his MD/scientist son to do some SWR research using medical algorithms that pull out people who die early from the stats to more accurately reflect portfolio survival rate. Kind of interesting but not, in and of itself, actionable. He also reports some related conclusions that are somewhat interesting including that "by about age 83, nearly everyone who is going to go bankrupt has, while portfolio ruin has just begun. Thus, the probability that a retiree filed bankruptcy due to sequence of returns risk is quite small – retirees go broke for several reasons, but sequence of returns risk doesn't appear to be a major contributor." I suspect that part of the reason for that is decreased spending in the 80+ years. If the savings are gone old folks just rely on SS and Medicaid to sorta survive and see no advantage to bankruptcy.

 

[W2R]

If the savings are gone old folks just rely on SS and Medicaid to sorta survive

+1 Exactly. Most of the people over 80 that I know, grew up with a values system that views bankruptcy in a negative light; they would be ashamed to go bankrupt if they can sort of manage on SS and Medicare.

I still fully plan to spend at least as much when I am over 80, as I do now. Maybe that is a pie-in-the-sky attitude, but I plan to make my declining years as easy and pleasant as possible. I can think of a number of pricey services and devices that could help and that most elderly simply cannot afford.

Of course, if I don't HAVE the money to spend, then I won't spend it and I will manage on SS and Medicare plus my small pension. I suppose that would support the conviction that many have, that somehow the desire to spend magically vanishes at age 80 or so. It may for some people but I seriously doubt that it will for me.

 

[Independent]

He references "actuarial tables". That makes me think he is trying to improve the "probability of portfolio survival" by including some sort of death function.

We normally use FireCalc by setting some fixed time horizon (eg 30 years), then running many investment scenarios over that horizon. It seems to me that we explicitly recognize that future investment returns are unknowable and can only be discussed in terms of probabilities, but at the same time we take the equally unknowable date of death and make a deterministic assumption about it.

It's possible to add a calculation to FireCalc that makes both investments and deaths probability functions. I think the process is straight-forward (I think I posted it somewhere on this forum some time ago). It doesn't require "Kaplan-Meier analysis", at least not as I understand that term from a quick read of a wikipedia article.

 

[nun]

This surely doesn't have much value for retirement planning because you have to be alive to retire. As you get into retirement and you hopefully become one of the survivors I can see it's utility if you don't follow a liability matching approach.

 

[Meadbh]

Thanks for posting. The scientist in me is delighted to see Kaplan-Meier survival curves applied in this way. They have done the modeling; the next logical step would be to follow a large cohort of retirees over their remaining lifetimes and determine which variables are associated with the best portfolio survival curve. Of course, the "population at risk" would diminish over time as people die off.

Funding such a long term study would require the resources of a wealthy organization with a vested interest in the result, most likely an insurance company or other financial institution, giving rise to concerns about conflict of interest.

 

[Lsbcal]

...
I still fully plan to spend at least as much when I am over 80, as I do now. Maybe that is a pie-in-the-sky attitude, but I plan to make my declining years as easy and pleasant as possible. I can think of a number of pricey services and devices that could help and that most elderly simply cannot afford.
...

What pricey services and devices are you thinking of W2R ?

I can imagine things like:
Gardening service
Quality taxi drivers
Dog walkers + sitters
Food market delivery
Prepared food service
Handyman guys

 

[LARS]

I still fully plan to spend at least as much when I am over 80, as I do now. Maybe that is a pie-in-the-sky attitude, but I plan to make my declining years as easy and pleasant as possible. I can think of a number of pricey services and devices that could help and that most elderly simply cannot afford.

me.

Awww, that is so cute. You think your mind will be sufficiently with it to comprehend the number of pricey services and devices that most elderly can't afford.


Sent from my iPad using Early Retirement Forum

 

[Lsbcal]

I don't know about W2R but I'm planning on being as sharp as a very sharp tack when I'm > 80. And I really mean this.

But I'll be walking probably and not running much or at all.

P.S. I'm going to start a thread on this.

 

[LARS]

I don't know about W2R but I'm planning on being as sharp as a very sharp tack when I'm > 80. And I really mean this.

But I'll be walking probably and not running much or at all.

P.S. I'm going to start a thread on this.

Just remember two things: a) it is never wise to tempt the Gods and b) the Gods laugh while man makes plans!


Sent from my iPad using Early Retirement Forum

 

[SecondCor521]

He references "actuarial tables". That makes me think he is trying to improve the "probability of portfolio survival" by including some sort of death function.

We normally use FireCalc by setting some fixed time horizon (eg 30 years), then running many investment scenarios over that horizon. It seems to me that we explicitly recognize that future investment returns are unknowable and can only be discussed in terms of probabilities, but at the same time we take the equally unknowable date of death and make a deterministic assumption about it.

It's possible to add a calculation to FireCalc that makes both investments and deaths probability functions. I think the process is straight-forward (I think I posted it somewhere on this forum some time ago). It doesn't require "Kaplan-Meier analysis", at least not as I understand that term from a quick read of a wikipedia article.

There is this article, from the original REHP site:

Combining Safe Withdrawal Rates and Life Expectancy

Bottom line: "Pretty amazing, huh? A 5% chance that your portfolio runs dry in 40 years, combined with a 5% chance you'll die in less than 40 years, still keeps your overall financial picture solvent more than 99% of the time."

 

[Independent]

There is this article, from the original REHP site:

Combining Safe Withdrawal Rates and Life Expectancy

Bottom line: "Pretty amazing, huh? A 5% chance that your portfolio runs dry in 40 years, combined with a 5% chance you'll die in less than 40 years, still keeps your overall financial picture solvent more than 99% of the time."

Thanks. I wasn't posting here in 2003 so I didn't see that.

I think he has the same idea, and his numbers go in the same direction, but I can't see exactly how he calculated them.