Withdrawal rates - theory vs practice

[Roger_R]

I am some what new to the withdrawal rate aspects of ER, and have routinely found discussion and information from you folks as helpful. My homework most commonly says the 4% withdrawal rate as a starting point. But I've also seen lower higher and "just live off the interest and don't mess with the principal" theories. I'd be curious of your philosophies on this and how you've established your comfort zone for withdrawal.

My current thinking on the 4% rate is that it may be a little aggressive. It seems to assume that market performance can be predictable quite a while out, that your income producing potential from principal will gradually be depleted, and that you won't live to some extraordinarily long age. While the numbers seem to work out fine, these things cause me to worry about these withdrawal rate.

I'm 52 and still working, but in the process of projecting retirement expenses and incomes that say I can be set free from wage based work in the next 0 to 5 years. The touted 4% withdrawal rate puts me closer to the zero years left, but at this point might also cause a few too many sleepless nights afterwards.

Your thoughts are helpful.

Run your numbers against FireCalc and see what it says. It will pit your portfoilo against about 100 years of U.S. Stock Market History. Better than listening to any prognosticators.

Since we cannot predict the future, we base our plans on the past.

[Michael]

Quote:

...might also cause a few too many sleepless nights afterwards.

Then don't do it. Wait until you feel comfortable.

[Roger_R]

I've run FIRE calc a number of times with a number of senarios. It says I should be just fine reitiring very soon. But as a number of folks have commented, it's just a tool to give get you in the ballpark.

Quote:

Then don't do it. Wait until you feel comfortable.

I might have to add that work in itself causes some sleepless nights, too

Whoever said "I'd rather be lucky than good." had a point. I retired with almost no thought to saving,
budgets, stc. SWR? Never heard of it. FIRECALC?
I would have guessed it was something to predict
if your house might burn down

Now, here I am at age 60, closing in on SS. My net
worth is up substantially (no, that does not include
my hard working wife's money if anyone is curious).
Plus, I have not touched my IRA. Also, with any luck at all, we will
never dip into the principle (my kids will like that).

I would like to say this was due to my superior brain
and extraordinary talents. Although those are pretty
impressive, most of it was probably luck. I would still
have found a way to stay retired, even without
any unexpected help from providential
serendipity.

JG

Quote:

I've run FIRE calc a number of times with a number of senarios. *It says I should be just fine reitiring very soon. *But as a number of folks have commented, it's just a tool to give *get you in the ballpark. * I might have to add that work in itself causes some sleepless nights, too *

It's a gamble, like everything in life. You may never be comfortable retiring until you have been retired 5 years.

No matter how much money you have, someone here could come up with a scenario where you'll have to eat cat food.

No one can make this decision for you.

[hocus]

My current thinking on the 4% rate is that it may be a little*aggressive.

At the SWR Research Group board, we use a different methodology than the methodology that produced the 4 percent number. Our methodology produces numbers lower than 4 percent at times of high valuation, and numbers higher than 4 percent at times of low valuation.

Here is a link to a post that provides an overview of the work done at the SWR board as of May 2004.

http://www.nofeeboards.com/boards/viewtopic.php?t=2505

JWR1945 and I will do our best to answer any questions that you put forward either here or at the other board.

[Nords]

... doesn't mean your SWR will work out anyway.

But one sleepless night from work stress is far more damaging than any number of sleepless nights spent worrying about SWR. Especially when ER means you could always make up for a sleepless night with a nap! So it's time to worry constructively, not negatively.

Speaking as a left-handed INTJ nuclear engineer, you could spend a few more bucks (and hours of data entry) for a FinancialEngines analysis that goes into thousands of different Monte-Carlo simulations with not much more reassurance that you'll survive. And just when your thirst for supporting data is unbearable, you'll encounter Bernstein's "Retirement Calculator from Hell".

So if you trust FIRECalc's prediction that you can retire now, then it's worth the risk. That probability (80% or 99.9% or whatever you chose) is the flip side of having to seek part-time employment in a few years. The glass is still 51% full.

I think 4% will work out fine because, like the cat-food comment, I don't have any valid information that says it won't. (There's lots of scary "it might not" stuff, but no "WILL NOT".) Living off principle & interest is even safer-- especially if you like bequeathing inheritances. But is that holy grail worth sacrificing your health now for a putative improvement? It's analogous to a bishop preaching to the medieval masses that a serf's lifetime of sacrifice will be rewarded in heaven.

Make the leap. Retire and fearlessly put a 4% 2005 SWR into a MM account while you experience ER. Your stress will drop, your health will improve, your head will clear, your creative juices will start flowing again in 3-6 months, and you'll realize that you'll be fine. It's hard to appreciate this while work stress is buzzing in your head, but in a couple years you'll regret waiting as long as you already have.

Or you could spend the next five years adding a few more digits onto that 99.9% success rate. Would that improve the sleep quality?

Quote:

Whoever said "I'd rather be lucky than good." had a point. I retired with almost no thought to saving, budgets, stc. ...

John, thank God the Calamity Fairy hasn't found you yet.

[grumpy]

It seems to me the question should be whether you can live comfortably on a 4% withdrawal rate, not whether the exactly correct rate is 4% or 3.8% or 3.1% ...?

Do you have a good handle on what it currently costs you to live? Do you have good insight into what expenses will decrease after you retire and what expenses will increase? If you have good answers to these questions and a 4% withdrawal will let you live the lifestyle you want, then go ahead and retire. Your actual expenses will not be exactly what you estimate. Your actual withdrawal rate may be slightly more or less than 4%. You have a lot of retirement years ahead of you. You will have time to make adjustments.

Good luck.


Grumpy

[Roger_R]

I have indeed come up with what I consider a decent estimate of future expenses. Of course health care and insurance are a wild card. At 55 I'll qualify for employer paid health insurance into retirement. For near immediate separation from my employer, a 3% withdrawl is slicing the bean a bit thin and 4% gives a little wiggle room.

Thanks for the thoughts. Among those giving good advice are there any of you who have a significant dependence on savings (vs pension) and have commited to the 4% withdrawal rate. (I know that John G. has more of an intuitive bent on things, which I envy but which is not inherent in my way of thinking about things).

[ESRBob]

Roger,
I am 100% savings-funded ER, and adhere strictly to the 4% rule.

Nords echoed my thoughts -- basically, Carpe Diem! *All the studies in the world won't get you closer to certainty on this or any other number.

But as Cut-Throat said, you need to get comfortable with this yourself, too.

Actually, I have a few tips for the timid:

If you can do it on less than 4%, then it will be safer. *3% looks bombproof , 3.5% is probably also bombproof. *So if you really wanted to keep working a few more years, getting to 3.5% might help you.

Your expenses may have been stable and predictible, but things do change -- up and down. *Travel, which lots of ERs like to do, can be pricey. *(there are cheap fares and renting houses/apts vs staying in hotels can help a lot, too.) *Also human nature wants a little creep up in lifestyle every year. *Even if it is only for such line items as "charity". *If I gave 10 last year, part of me wants to give 11 this year, or even 12. *That is much faster than inflation, but in many ways we are wired up for expansion of 'needs/wants' faster than inflation or faster than the zero real growth most of us are implicitly planning our ER finances around. *

When the Social Security admin folks talks about how they will balance the books by indexing to CPI instead of Wage Growth, they know what they are talking about. *One stat I read today suggested it would halve the benefit over 50-60 years -- just using the CPI instead of growth in wages.

Everybody who is still working has their consumption or disposable income implicitly tied to wage growth; yours will *be tied to CPI or again, implicitly, zero real growth. *How do you feel about that?

The technique I use, btw, is a slight modification of the 4% rule, which stems from research used by foundations who want endowments to keep up in real terms in perpetuity. *(www.zunna.com)


I make mid-term corrections every year, spending 4% of the starting value of the portfolio each year. *I get a raise in good years, a cut in bad years. *Over time, my consumption should also keep up with inflation, but there are no guarantees. *The real value of the portfolio, though stays intact in perpetuity (against historical models) better than the traditional 4% models, which can take you all the way to zero after 30 or 40 years -- this was not long enough for me and kept me up during more than a few sleepness nights when I was worrying about 4% SWR on the traditional model. That is why I changed my approach. I believe this approach gives me the best high consumption now when I need it balanced against long term risk of portfolio withering

I also have a rigorous portfolio allocation along the lines of CoffeeHouse or Bernstein in low-cost funds. And I tip toward value, small and international which I believe gives me a percent or so edge over the long run. I stay away from the SP500 itself and use other non-market-weight equities instead -- eg. small and value and even a few managed funds. Going against the crowd with a non-market-weight portfolio is a type of contrarianism that suits me well and seems to play out well in the historical modelling - - hope it does as well going forward.

[Nords]

Quote:

When the Social Security admin folks talks about how they will balance the books by indexing to CPI instead of Wage Growth, they know what they are talking about. *One stat I read today suggested it would halve the benefit over 50-60 years -- just using the CPI instead of growth in wages.

Gosh, the ECI rising faster than the CPI would imply that wages are rising faster than inflation. Or are both numbers equally invalid for different reasons?

Quote:

The technique I use, btw, is a slight modification of the 4% rule, which stems from research used by foundations who want endowments to keep up in real terms in perpetuity.

I know that the IRS requires foundations to spend 5% of their assets each year, and in the late '90s the IRS was actually considering raising that number to 6%. But does anyone know how the IRS arrives at those numbers? Is some GS-12 running scenarios on FIRECalc, or do they assume that foundations can handle the Trinity study plus another 1% for fundraising, or is there some other analysis?

[charlie]

Some of you may recall that I am now in RMD.
This may be obvious to most of you but it just
recently occurred to me that it might be better
to draw your RMD in a lump at the END of the
year rather than monthly during the year.
This would keep "Uncle" waiting on his taxes
until the last possible minute. It causes brain
farts to think about this too much but I believe
this is only a "one time" advantage and probably
does not amount to much over the long term.

Cheers,

Charlie

You know, my "shoot from the hip" approach to SWR
and budgeting may come back to bite me some day.
It's worked okay so far though, and I hate to think
about trying to nail down every penny (coming or going).
Takes some of the fun out of ER. That said, if I screw up,
going back to work or living on tuna fish has little appeal. Certainly, since full ER in 1998, my confidence
and comfort with the whole business has increased a
lot. After I drew my last paycheck, I moved into a
$330/mo. apartment in town. Obviously not what I wanted but the price was right. When an opportunity
came up to move into a brand new waterfront efficiency in the country, I worried about the rent going up to $425/mo.
(I know you folks on the coasts will find this amazing).
Anyway, now the extra $95.00 per month wouldn't faze
me. I was quite happy with the new place and would
still be there if I had not remarried in 2001. My bottom
line on SWR is to at least maintain my net worth where it is
now. If I do that, I don't really care what my SWR is.

JG

[wzd]

ESRbob says

Quote:

The technique I use, btw, is a slight modification of the 4% rule, which stems from research used by foundations who want endowments to keep up in real terms in perpetuity. (www.zunna.com) I make mid-term corrections every year, spending 4% of the starting value of the portfolio each year.

Thanks for the link - Somehow I had missed it in my searching and reading last year. After reading thru one of the reports, http://www.zunna.com/Research/Choice...Retirement.pdf, I picked out the following numbers that I thought were interesting:

Table 1 - Don't Go Broke, Spend to Maintain Standard of Living, which is the firecalc algorithim/philosophy, shows 4.03% SWR for 30 years at 100% survival. This is a historical simulation, like firecalc, so it's not surprising that the numbers come out the same.

Table 4, Maintain Buying Power, Spend to Maintain Standard of Living, which means the ending portfolio should be the beginning portfolio scaled up by inflation and withdrawals are scaled up by inflation each year. It shows a SWR for 30 years at 100% survival of 2.98. However the other interesting number (see below for why) is the SWR for 30 years at 75% survival of 4.16.

Table 5, Maintain Buying Power, Spend Percentage of Account Value, which means the ending portfolio should be the beginning portfolio scaled up by inflation, and a percentage of the portfolio is spent each year (not a fixed, inflation scaled percentage). It shows a SWR for 30 years at 100% survival of 4.16.

Comparing these last two strategies, it tells me that if I adopt the table 5, 100% survival strategy, that 75% of the time my withdrawals will, on the average, equal or exceed the fixed 4.16 percent scaled for inflation amount. 25% of the time, where the fixed amount scenerio fails, the percentage of portfolio method will kick in and enable it to survive.

As more than one person has said in the past, maintaining fixed, inflation adjusted withdrawals when the portfolio is shrinking, is just against the nature of most of the ER's. The percentage of portfolio is probably a more comfortable model - the risk is lifestyle instead of running out of money.
My comparison of the two 4.16% withdrawals is just one more way to think about, and compare strategies.

One final number to leave you with from the paper: If you want to withdraw a percentage of portfolio, and end with the portfolio amount equal to the starting amount, without inflation adjustment, the 30yr, 100% survival SWR is 7.81%!! Using my above logic and other tables, the withdrawal amount will trail inflation slightly more than 50% of the time, but you will not run out of money, nor will you eat into your pricipal (after 30 yrs, not true each year).

[Roger_R]

Just wanted to say thanks to ERSBob for good advice and the link. I'm still sorting through some of the Zunna info., but it is interesting and maybe alittle more perspectives than the FIRE calc.

We gave those Zunnas and the other tribes a bad deal.
Don't understand why they'd want to help us out with our investments.

JG

[ESRBob]

Quote:

Gosh, the ECI rising faster than the CPI would imply that wages are rising faster than inflation. *Or are both numbers equally invalid for different reasons?

Nords,
Yeah, it definitely has been the case that wages have risen faster than inflation, though not evenly across the spectrum of salary earners. *Those at the top seem to have walked away with the lions share of the gains, while those at the bottom and middle seem to have been closer to breakeven. *Without finding chapter and verse on that, it is what I remember.

The rising real wages are what have made all the mcmansions and fancy cars and electronic goodies and restaurant meals out, plane trips, expensive hobbies etc etc a fact of modern life -- 50 years ago they would have all seemed like hopeless extravagances for the average middle class person.

I think about this because it is just possible we will be living a 2005 lifestyle in 2055, the equivalent of living a 1950s lifestyle today, if we plan around zero real growth in portfolio and everybody else edges slowly up by even 1% a year in their real earnings/spending.

'course the 50s weren't bad! *(i'm told... I only arrived for the tail-end of them and was mostly having diapers changed, not really scoping the lifestyle all that carefully-- I think I would have preferred Pampers to the cotton & pin variety, though, given the choice!).

It is also true that the govt's inflation figures try to incorporate some measure of quality improvements. *So (not sure if this is a perfect example) today's cars are better, and they cost more, than cars 30 years ago. *Govt tries to set those two facts against each other, and say that for the equivalent car, price didn't really go up at all. *So inflation for the same absolute standard of living will be at the official CPI rates, but if, for example, you bought an avereage new car in 1975 and an average new car today, you'd find the cost would be up a lot more than CPI. *Cars may or may not be a good example, but this shows how creeping improvements in lifestyle can stealthily get into the budget despite what seem like tame overall inflation, and make your personal actual inflation rate higher than CPI. *Put another way, the same real income won't produce the same relative lifestyle lin 50 years as it does today.

The reverse scenario would also apply-- if strong productivity gains bring falling prices, govt is going to try to reflect that in CPi, too, by saying in effect that we have $3,000 worth of 1990 computer today for $500, offsetting other rising prices elsewhere in the CPI calculation. But that $500 computer won't power the game your kid wants and that everybody is playing now. I'm doing a lousy job explaining all this, but bottom line it might make CPI a lowball estimate of what it will take to keep up an average lifestyle going forward.

on the IRS rules about foundations disbursing 5%, I attribute it to nothing more than historical guesswork, perhaps built up in a time when interest rates were a lot higher than they are today? *It certainly looks aggressive today with what we know about SWR. *Maybe fundraising was supposed to make up the difference.

Roger -- glad you found the zunna study useful.

Wzd -- interesting analysis on the zunna study. *If anyone wants to learn more, keith marbach, the author, has answered questions etc in the past for me and has his email posted with the article.

Hey ESRBOb! I was wearing diapers in the early 40s, before Pampers, etc. My mother was cross eyed
and was not very accurate with those pins (Ouch!).
My spouse things the trauma I suffered is what is wrong with me today

JG