Trinity Study

Financial Independence/Retire Early -- Learn How!
Post Reply
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Trinity Study

Post by ataloss »

Here is a nice summary of the trinity study
http://www.zunna.com/Research.htm
includes a link to a free AAII article by Cooley

For stock-dominated portfolios, withdrawal rates of 3% and 4% represent exceedingly conservative behavior. At these rates, retirees who wish to bequeath large estates to their heirs will likely be successful. Ironically, even those retirees who adopt higher withdrawal rates and who have little or no desire to leave large estates may end up doing so if they act reasonably prudent in protecting themselves from prematurely exhausting their portfolio. Table 4 shows large expected terminal values of portfolios under numerous reasonably prudent scenarios that include withdrawal rates greater than 4%.


With current valuation 3-4% isn't as conservative (IMHO) as it once was. Still I wouldn't say that the study is invalid. I posted this for those who might want to actually read the study before declaring it invalid.
:)
Have fun.

Ataloss
peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Re: Trinity Study

Post by peteyperson »

If you're worried about current valuation levels invalidating the study, why don't you just correct the portfolio for the inflated valuation and then apply the 3-4% number deemed safe?

Seems the logical solution on this issue rather than going all round the houses for the same thing!

Petey
ataloss wrote: Here is a nice summary of the trinity study
http://www.zunna.com/Research.htm
includes a link to a free AAII article by Cooley

For stock-dominated portfolios, withdrawal rates of 3% and 4% represent exceedingly conservative behavior. At these rates, retirees who wish to bequeath large estates to their heirs will likely be successful. Ironically, even those retirees who adopt higher withdrawal rates and who have little or no desire to leave large estates may end up doing so if they act reasonably prudent in protecting themselves from prematurely exhausting their portfolio. Table 4 shows large expected terminal values of portfolios under numerous reasonably prudent scenarios that include withdrawal rates greater than 4%.


With current valuation 3-4% isn't as conservative (IMHO) as it once was. Still I wouldn't say that the study is invalid. I posted this for those who might want to actually read the study before declaring it invalid.
:)
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

Cooley Report and Simulated Annealing

Post by therealchips »

Thanks for the reference, ataloss. The results are valuable and worth pondering. It is particularly useful to me that the study included withdrawal rates for preserving the capital's purchasing power, rather than just avoiding going broke. With my retirement stash 88% in equities, the study says that I have underweighted them! :roll: For the goal of preserving capital's purchasing power over thirty years or so, the optimal allocation turned out to be 100% equities. Caveats about past results are no guarantee of future results apply, of course.

The study includes fine examples of the use of simulated annealing, a mathematical optimization technique I have applied to aircraft and ship design. The study you cited applied simulated annealing to finding optimal portfolios with three asset classes. This range of applications shows how flexible the technique is.

(There goes Chips, pushing both preservation of capital and mathematical optimization again.)
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

Chips
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

bylo has a bunch of links related to swr
http://www.bylo.org/saferetr.html
I liked the AAII article for its simple presentation. Although I have heard of simulated annealing I have only a vague idea what it is.
Have fun.

Ataloss
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

First Draft, Introduction to Simulated Annealing

Post by therealchips »


He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

Chips
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

Chips, it is going to take me a while to digest this. Especially the temperature function.
Have fun.

Ataloss
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

Post by therealchips »

Hi, ataloss,

I went very light on the temperature function. My fingers were tired. :) It took me a while to write that sketch. The only reference I consulted is my memory. Simulated annealing came to my attention with an article published in Science in 1983. I applied it to some aircraft design problems over the next ten years.

The online tutorials for simulated annealing that I found seem too technical for us right now. I will think about a better explanation of the temperature function. There is a mildly interesting technical result that says that simulated annealing will converge to the true global optimum if the temperature is lowered slowly enough. The problem is that such an ideal cooling schedule requires infinite computer time. :cry: We get useful results in reasonable amounts of computer time with cooling schedules we know are not perfect.

Comments and questions from you or other participants here are welcome whenever any of you are ready.

Did I let on about how much I enjoy this subject? :lol: I have seen experts look at results found by these search techniques and say "This solution is a good idea. Why didn't I think of that?"

(editted to change "as say" to "and say".)
Last edited by therealchips on Mon Jul 28, 2003 9:41 am, edited 1 time in total.
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

Chips
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

"This solution is a good idea. Why didn't I think of that?"


LOL, I am going to be out of town for a few days, but this is interesting
Have fun.

Ataloss
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

How to Avoid Cracking the Retirement Nest Egg

Post by therealchips »

Today's New York Times as an article entitled How to Avoid Cracking the Retirement Nest Egg at http://www.nytimes.com/2003/08/17/busin ... 7RETI.html
To tell people with $1 million portfolios that they can spend less than $4,000 a month is "absurd," said Dr. Philip Cooley, the Prassel Professor of Business at Trinity University in San Antonio, who has published several papers on sustainable withdrawal rates in the last five years.

"When you recommend such low rates, you cause the retiree to forgo a joyous life in early retirement," he said, "and they end up with a lot of money later on, when they can't take advantage of it."

He said his research indicated that over 30 years, a withdrawal rate of 7 percent of the initial portfolio value, without adjustments for inflation, is sustainable 86 percent of the time for a stock-dominated portfolio.


I would clarify that statement this way: "To tell people with $1 million portfolios that they cannot spend more than $4,000 a month is absurd" said Dr. Cooley. (That, I suppose is the intent of the quotation, but not what it says literally.) This, of course, is a 4.8% withdrawal rate and Cooley says it is absurdly low.

Planning to live on withdrawals that are fixed in nominal terms, with no adjustment for inflation, is too risky for me. I have clear memories of what inflation did to retired people in the 1970's. My own work suggests that, with a logarithmic utility function for money and using standard mortality tables, optimal policy is to adjust the withdrawals annually for a little less than the inflation rate. I'm not willing to make that my plan -- partial adjustment for inflation -- but Cooley makes a point that supports my point: The earlier years of retirement are the ones you are most likely to live through.

I don't mind the idea of having a lot of money "later on (that I) can't take advantage of." That's the price of a policy that takes only a small risk of going broke while still breathing. Further, since I inherited a part of my capital stash, leaving something to heirs seems only fair. I do not follow the "Die Broke" idea. As I have posted, my present plan is to hold my withdrawals under 2% a year for the next thirty-odd years. Like ataloss, I'm contemplating taking only the dividends and selling nothing.

(Footnote on the logarithmic utility function for money: this recognizes that people place a value on additional money according to what they already have, rather than on the absolute amount of the addition. So, an additional $10,000 means more to the man with a $50,000 stash than to the man with a $5,000,000 stash. So far as I know, all SWR studies ignore the idea of a proper utility function for money. The omission doesn't seem to trouble anyone but me. I'm waiting to pounce on somebody-or-other, to say that his SWR work is invalidated by this relevant but excluded consideration. :lol:)
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

Chips
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

"This is not a contract," Dr. Cooley said. "We're talking about a financial plan." People who retire into a bear market can cut back on their withdrawal rate, he said. Then, when things look better, why not bump it up again?


nice article

somewhere hocus has said something about swr varying depending on savings/income level but he didn't mention a utility function for money iirc :)
Have fun.

Ataloss
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

gummy and chips on these subjects

Post by therealchips »

The estimable and sorely missed (because he hasn't posted lately) gummy addresses safe withdrawal rates and life expectancy at http://home.golden.net/~pjponzo/withdrawals.htm. I get the following result from his spreadsheet, using its default values except for my age: You are a male aged 63. There is a 95% probability, that with a withdrawal rate of 5.6%, your portfolio will outlive you. When I "Press F9 for another set of random returns." repeatedly, I see going broke far too often for my tastes, and seemingly a lot more that 5%.

gummy addresses utility theory here http://home.golden.net/~pjponzo/utility-theory.htm where he says
So, choosing an appropriate Utility Function reflects the importance that humans place on money. We're more interested, NOT in the value of money, but in the value of its Utility.
He gets to a logarithmic utility function in Part III.

My idea is to look at picking thirty (say) withdrawal amounts, one for each of the next thirty years, with the goal of maximizing the total expected utility of those withdrawals. (This is a 30-dimensional optimization problem, and practical as well, just the sort of thing that tickles me so much.) The expectation reflects the planners life expectancy. For your amusement, I mention that maximizing the sum of the logarithms of a set of numbers is equivalent to maximizing the product of the numbers, the result of multiplying them all together. Clearly, with that objective function, you won't take any chance of going broke (and then being obliged to make withdrawals of zero afterward) since that would introduce a zero into the product of all withdrawals, which makes the product zero. When I try this 30-dimensional optimization in my spreadsheet, sure enough, it sets up a withdrawal schedule that generally declines with each passing year. It also calls for spending far more money than I am willing to spend. :wink: Maybe I should re-run the optimization using the acid test or worst case data I got from one of ataloss's posts.
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

Chips
User avatar
BenSolar
*** Veteran
Posts: 242
Joined: Mon Nov 25, 2002 5:46 am
Location: Western NC

Re: gummy and chips on these subjects

Post by BenSolar »

therealchips wrote: My idea is to look at picking thirty (say) withdrawal amounts, one for each of the next thirty years, with the goal of maximizing the total expected utility of those withdrawals. (This is a 30-dimensional optimization problem, and practical as well, just the sort of thing that tickles me so much.) The expectation reflects the planners life expectancy.


Thank you for your continued posts on this subject, TRC

It is very interesting and important, I think.

Ben
"Do not spoil what you have by desiring what you have not; remember that what you now have was once among the things only hoped for." - Epicurus
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

Post by therealchips »

Thanks for your comment, Ben. I know I repeat myself, but there's always a chance that someone new has come by, or that I've improved the presentation. :D
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

Chips
Post Reply