Another Study Challenging the 4% SWIP Rule

[chinaco]

Another Study that questions the efficacy of the 4% rule.

Our results indicate that a 2.52 percent withdrawal rate could be sustained in each of our randomly generated scenarios, but a 4 percent withdrawal rate is associated with an 18 percent probability of portfolio failure.

A Safer Safe Withdrawal Rate Using Various Return Distributions


Seems to be a fad!

I think someone could come up with a legitimate study to challenge any guidelines.

MB said it in another post: (Paraphrase) pick your target WR% (one that has some reasonable chance of success) and manage to reality as the future unfolds.


[Conspiracy Theory --- Sarcasm] Gotta wonder.... after all the years boomers were fed "Market Good", now its time to prepare for the big finale for boomer retirement... Time to Harvest the Commissions... Sell Annuity (Income) and Life Insurance (Estate) "Guarantee Good"!

 

[Dawg52]

My studies show a 2.51 w/d rate is best. Don't ask me for any data though.

 

[chinaco]

My studies show a 2.51 w/d rate is best. Don't ask me for any data though.

I have concluded that 0% is the safest... but there is still a small chance of a catastrophe that would cause of failure.... like a mega comet impacting the earth!!!

Maybe I should continue to w*rk and feed the pot!

 

[rescueme]

Maybe I should continue to w*rk and feed the pot!

No. You should continue to w*rk to ensure my (future) SS ...

 

[HFWR]

My studies show a 2.51 w/d rate is best.

I thought 2.52 sounded high...

Maybe I should continue to w*rk and feed the pot!

Maybe you should quit work, and smoke some pot...

 

[Independent]

Another Study that questions the efficacy of the 4% rule.


A Safer Safe Withdrawal Rate Using Various Return Distributions


Seems to be a fad!

I think someone could come up with a legitimate study to challenge any guidelines.

MB said it in another post: (Paraphrase) pick your target WR% (one that has some reasonable chance of success) and manage to reality as the future unfolds.


[Conspiracy Theory --- Sarcasm] Gotta wonder.... after all the years boomers were fed "Market Good", now its time to prepare for the big finale for boomer retirement... Time to Harvest the Commissions... Sell Annuity (Income) and Life Insurance (Estate) "Guarantee Good"!

The poorer results here result from using a variety of probability distributions (Beta, Extreme, Gamma, Laplace, Logistic, Lognormal, Pert, Rayleigh, Wakeby, and Weibull), while the standard Trinity study uses only lognormal. I didn't see that they reported results separately for each distribution.

Most of us would say these studies require that you take a wild-___ guess at the future mean return (they use 5.1%) and the date of death (they use exactly 35 years after you start withdrawals). Once you've made assumptions on those two items that could be off by 20%, it's hard to see why you would want to push real hard on a theoretical probability distribution.

It's more important to have some flexibility after you retire.

 

[arebelspy]

I'm curious if 2.53% failed in one of the scenarios.

Since one has the ability to see issues coming and do things about it, I'm not too worried about having a slightly higher SWR (though I will probably aim for 3%)

 

[bbbamI]

My studies show a 2.51 w/d rate is best. Don't ask me for any data though.

My WR was 2.71 percent last year. Iza huzzy.

 

[REWahoo]

Last year, our 5th year of retirement, our WR was 5.4132539 percent. We did some kitchen and bath remodeling which really blew the budget.

This year we are on track for a WR of 4.4825018 percent, much closer to my goal.

 

[Dawg52]

My WR was 2.71 percent last year. Iza huzzy.

No comment.

 

[Animorph]

There's a lot of w*rking and waiting to reach 2.52% from 4%, given a fixed level of expenditures. That's not worth the risk reduction to me. I agree, some flexibility should go a long way towards maintaining portfolio viability. I'm not aiming to be a billionare when I die, I'm trying to maintain a constant lifestyle.

 

[W2R]

Last year I spent 2.58%, including everything I can think of - - living expenses, remodeling the whole house for the move-that-never-happened, taxes, and so on.

This year I am on track to spend 1.91% including living expenses and taxes, but no remodeling this year. Can't remodel every year.

I think a sensible withdrawal rate for me is between 3.0% - 3.5% so stay tuned - - my actual withdrawal rates are going to increase.

 

[REWahoo]

I think a sensible withdrawal rate for me is between 3.0% - 3.5% so stay tuned - - my actual withdrawal rates are going to increase.

You realize if you helped fund my WR it would be a win-win situation!

 

[bbbamI]

No comment.

'Nuff said.

 

[mickeyd]

My 0% WR is still holding up pretty well, so 2.51% looks steep to me.

 

[jacob]

Yikes! Giving a result like that with 3 digits of precision (2.52) shows that the authors may not quite understand the difference between precision and accuracy.

 

[haha]

Does anyone still think that these "studies" prove anything? In the last couple days we have heard that 7% is good, and that 2.53% (2.53!) may be too much.



Ha

 

[Meadbh]

This year I am on track to spend 1.91% including living expenses and taxes, but no remodeling this year. Can't remodel every year.

Sweet!! Sounds like you are in great financial shape W2R.

I have been scrimping and saving like mad for the past couple of years but I've spent a lot of money this year what with the move and everything. I will get some of it back, e.g. some moving expenses will be reimbursed and most of the expenses (e.g. new furniture) are nonrecurring. I have a suspicion that if you are established in a low cost location the ongoing expenses can be kept quite modest and hopefully fairly consistent. That facilitates better expense and SWR planning and, I think, is a good reason to move to your chosen ER location while still working, if possible. Which is exactly what I've just done.

 

[W2R]

Sweet!! Sounds like you are in great financial shape W2R.

Yes - - I had to work two more years after FI to get health benefits, so I padded my meager nestegg with that money. Then about a year before I qualified to officially retire (and gain retiree health insurance and pension), I was fortunate enough to inherit unexpectedly. I haven't spent any of that yet, though. I am gradually expanding my lifestyle but only at a rate that I feel I can handle.

I have been scrimping and saving like mad for the past couple of years but I've spent a lot of money this year what with the move and everything. I will get some of it back, e.g. some moving expenses will be reimbursed and most of the expenses (e.g. new furniture) are nonrecurring.

Moving can be pretty expensive, even with reimbursment. New furniture seems less expensive if you consider the cost per year that you will be using it.

I have a suspicion that if you are established in a low cost location the ongoing expenses can be kept quite modest and hopefully fairly consistent. That facilitates better expense and SWR planning and, I think, is a good reason to move to your chosen ER location while still working, if possible. Which is exactly what I've just done.

I think your comments in the last paragraph are very perceptive. Ongoing expenses are lower in a low cost location and if you move while still working, so much the better. Also it is cheaper for a long time resident to live in a location than it is for a newcomer. I have lived here for I guess fifteen years, I know where to go and who to talk to in order to get things done cheaply and well. Also, I chose my present home with the idea of living in it until I draw my last breath. So, it is single story, low maintenance, and in a quiet neighborhood.

 

[Nords]

I think this article is a mathematical solution process in search of a problem.

They analyzed all of those different types of distributions despite saying that stock returns don't exactly match any of them. Yet somehow the sophistication of their "almost right" distributions is more worthy of analysis than the "not quite right" normal & lognormal distibutions. But maybe they were hoping to discover that stock returns actually do follow some sort of distribution pattern.

Financing a constant 4 percent consumption plan using the volatile returns from a traditional retirement portfolio is fundamentally flawed. Generally, a constant consumption plan is best supported by a portfolio that generates corresponding constant returns.

So the researchers knew that a fixed withdrawal scheme was flawed yet went ahead and beat the crap out of the fixed-withdrawal data anyway. Of course they used really cool math methods so their work has value... even if their logic does not.

In other words, if your returns are going to be variable then maybe your spending should be too. This sounds like Bud Hebeler's "Analyze Now" or Otar's red/yellow/green light or Bob Clyatt's 4%/95%. But since those guys didn't publish in peer-reviewed financial journals, the researchers didn't study variable withdrawal schemes.

What we need is a good financial analysis of variable withdrawal schemes.

You realize if you helped fund my WR it would be a win-win situation!

Hey, maybe you guys could sell annuities to each other!

 

[chinaco]

I guess professionals need to publish.

With all of the unknown variables that can affect the value of securities (in a portfolio).... I think it is unrealistic for anyone to believe that they make a plan for 30 to 40 years into the future and think there will not need to be some sort of adjustments... whether they find they can spend more or need to spend less... say 10, 15 or 20 years into it.

 

[Independent]

What we need is a good financial analysis of variable withdrawal schemes.

I worked on a simple one, and think I posted a little bit here.

My thought was that any variable scheme has to allow you to decrease your withdrawal rates.
But, any "successful" retirement withdrawal pattern maintains at least some minimum withdrawal in every year.

Putting those two sentences together, a withdrawal scheme that could be analyzed has to have some ability to go down to some floor, but must insist that you never go below the floor.

The traditional 4% Trinity scheme essentially says that your floor is 4% of the original assets, because it has no downside flexibility.

IIRC, I came up with something like "the maximum of 6% of current balance or 3% of initial balance" has a survival rate of about the same as a 4% level.

Of course, starting at 6% means 50% more spending in the first year than a 4% rate, but it justifies that by allowing for spending that would go down 25% below the 4% rate if the portfolio balance dropped.

Using such a scheme means sitting down at the beginning and saying "How low would I be willing to go?" That's probably a good exercise. I felt more comfortable with that approach than saying 95% of prior year, because the 95% rule has no bottom.

 

[JmfromTx]

I used Otar as my guide (3% max, taking less) but I think it really boils down to common sense. If you have saved enough to voluntarily retire early, hopefully you are smart enough to go with the flow and hunker down if you need to and spend less if the situation arises. I left myself a lot of hunker room and won't hesitate to use it if I need too.

 

[obgyn65]

I don't.

Does anyone still think that these "studies" prove anything?


Ha

 

[arebelspy]

IIRC, I came up with something like "the maximum of 6% of current balance or 3% of initial balance" has a survival rate of about the same as a 4% level.

I'd love to see links to any posts you have made on this, especially regarding your methodology. I'm curious what similar rule would correspond to a 3% SWR.