USA Today article on SWR

[siamond]

http://cornerstonewealthadvisors.com/files/08-06_WebsiteArticle.pdf

Now we're speaking... I read it once (with an open mind), need to read it again (with a skeptical mind!), but I really like their general line of reasoning. Capital preservation rule, prosperity rule, eliminate inflation rule, try to start high and enjoy life while staying adaptive and very cautious about the final outcome, etc, this all seems VERY WELL grounded in common-sense and sensible logic.

What I also really like that it seems to auto-adjust by itself, whatever happens. So to a large extent, it should auto-correct even if the future is way different from what we've seen in the past. Now THAT is very re-assuring.

The only reservation I have is that the 20% threshold triggers and 10% rate adjustments (reduction or increase) seem a tad discontinuous. I would have expected something more incremental.

Now of course, the $1M question is... Did they get it right with their financial model, was it validated enough by peer reviewers, are the conclusions (which seems amazingly rosy) correct? 5% to 6% initial WR, that seems a lot...

Anybody aware of further analysis on this paper? Seems rather ground breaking to me? Or... is it a fake Holy Grail?

 

[Independent]

http://cornerstonewealthadvisors.com/files/08-06_WebsiteArticle.pdf

Basically creates "financial guard rails" for withdrawals, capping Inflation increases at 6%, lowering withdrawals if the WR rises more than 20% above the original and increasing when WR drops more than 20% of the original. This is some heavy going, a fairly technical paper, but worth reading.
As I understand it if you start with a 4% WR and sometime later the inflation adjusted amount and the size of your portfolio makes the actual WR either over 4.8% or under 3.2% would mean you would NOT take that amount but would,lower it by 10%, or increase it, respectively.

I skimmed the article. I was looking for statistics on withdrawal amounts. How often did the real withdrawal amount fall below the initial amount? What was the lowest annual withdrawal amount?

I didn't see numbers like this in the tables. It seems that if I were going to use a rule like this, I'd want some idea of the probability and severity of reduced withdrawals.

 

[enjoyinglife102]

SWR's are enormously impacted by rate of inflation used yet for the most part, a government estimate based on a basket of goods that has little relevance to any individual is used out of convenience. My experienced inflation has been much lower than the CPI. The "average" healthcare costs often cited also scares a lot of people into not spending. In a forum dominated by probabilities and likely scenarios, the chances that one is going to experience these health costs is dramatically reduced just by walking 20 minutes a day. So take a walk, manage you individual circumstances and live the life that makes you happy.

 

[ducky911]

I think that the article should have touched on the expenses. Someone could mistakenly think that they could hire an advisor and still take 4%.

fire calc is .18 expense(pretty low) on a 75/25 portfolio

 

[siamond]

I skimmed the article. I was looking for statistics on withdrawal amounts. How often did the real withdrawal amount fall below the initial amount? What was the lowest annual withdrawal amount?

I didn't see numbers like this in the tables. It seems that if I were going to use a rule like this, I'd want some idea of the probability and severity of reduced withdrawals.

I'd suggest you take the time to thoroughly read it. Was very worth it to me.

In the tables at the end, you see the avg # of trigger events (cuts/freezes/raises). Should answer your first question. You'll also find the PP (purchasing power) column, which will partly answer your 2nd question. I would indeed have liked to see a quantified lowest number, as you suggested... More precisely, I would have liked to see the effect of a floor limit on such lowest withdrawal amount (and maybe a ceiling too).

I tried to find some solid discussion about this fascinating article on the Bogleheads forum, but so far, I've had limited luck, just a few references in passing.

 

[rayvt]

Whatever the starting number nowadays, I'd love to hear a self-adjusting SWR process which is better balanced, something with reasonable protection.

Also http://schulmerichandassoc.homestea..._to_Create_Retirement_Withdrawal_Profiles.pdf

Guyton's original paper was obtuse and the allocation/trading part was incomprehensible, IMHO. But when Klinger came along and changed the focus to withdrawals, it all made sense.

A cool (?) part is that you can arguably bump up the SWR to 5% and still be safe. What saves you is the pre-defined, automatic reductions of withdrawals when the portfolio gets whacked by a bear.

The reason that the standard SWR is 4% is that there is no automatic mechanism for cutting back in bad times.

 

[NW-Bound]

The problem with WR reduction is that for some people, it may mean a drastic change in lifestyle, such as going into the boondocks in an RV. And they are scared of it.

 

[Major Tom]

The problem with WR reduction is that for some people, it may mean a drastic change in lifestyle, such as going into the boondocks in an RV. And they are scared of it.

Going into the boondocks in an RV would involve an increase over my current expenses - but I still want to do it

 

[Independent]

I'd suggest you take the time to thoroughly read it. Was very worth it to me.

In the tables at the end, you see the avg # of trigger events (cuts/freezes/raises). Should answer your first question. You'll also find the PP (purchasing power) column, which will partly answer your 2nd question. I would indeed have liked to see a quantified lowest number, as you suggested... More precisely, I would have liked to see the effect of a floor limit on such lowest withdrawal amount (and maybe a ceiling too).

I tried to find some solid discussion about this fascinating article on the Bogleheads forum, but so far, I've had limited luck, just a few references in passing.

Thanks. You're correct, those provide number of times but not severity. I still think that severity is important. Maybe I feel strongly that any year with spending below $D is a "failure".

Some years ago I tested a much simpler rule that, like this one, traded flexibility for initial spending (i.e. the more downside flexibility I have, the more more money I can take in the first year).

The rule was just "Take the greater of x% of my current balance or an inflated y% of my initial balance". I remember x=6 and y=3 were reasonable for 95% and 30 years.

The advantage of describing the rule that way is that I know that in my testing, any "success" scenario had spending of at least my floor in every year.

 

[siamond]

Also http://schulmerichandassoc.homestea..._to_Create_Retirement_Withdrawal_Profiles.pdf

Guyton's original paper was obtuse and the allocation/trading part was incomprehensible, IMHO. But when Klinger came along and changed the focus to withdrawals, it all made sense.

A cool (?) part is that you can arguably bump up the SWR to 5% and still be safe. What saves you is the pre-defined, automatic reductions of withdrawals when the portfolio gets whacked by a bear.

The reason that the standard SWR is 4% is that there is no automatic mechanism for cutting back in bad times.

Thank you! You're right, this paper is easier to read, and presents the concepts in a very logical order, while reaching very actionable conclusions.

This is EXACTLY what I was looking for. Not only starting around 5% is indeed attractive, but the auto-adjustment baked in the process should make it very future-proof. No more doom-saying baked in the equation. No more fear of market timing. No more implicit assumption that the future will be somewhat akin to the past, except in choosing the starting point, but then it will auto-correct as needs be. If the future is grim, your withdrawals will be too, but this is fair game I guess. Same thing with a rosy future, which is very cool. Or with a grim/rosy/grim/rosy future. Etc.

(Mr Independent, you'll also find the answers to your questions, btw)

Now something nags me... Maybe it's just my perception, but it seems that this general model is NOT widely publicized and used. Why would that be? This does seem close to the Holy Grail...

 

[rayvt]

Now something nags me... Maybe it's just my perception, but it seems that this general model is NOT widely publicized and used. Why would that be? This does seem close to the Holy Grail...

Can you imagine a Financial Advisor trying to explain the Guyton-Klinger withdrawal rules to a typical client? Yeah, me neither.

Nor can I see the typical FA just doing the details himself and then telling the client each year how much to withdraw that year. I personally had a little experience at this once -- I mentioned to my in-laws that they should let their dividends automatically reinvest instead of spending the dividends. Wow!!! She accused me of trying to take food out of her mouth and put them in the poorhouse.

When you're trying to make a living (as a FA), it's better to stay conventional and simple -- which means tell your clients to take constant 4% or less.

 

[tuixiu]

The danger is for anyone about to retire we are in uncharted waters.

We are always in uncharted waters.

Otherwise things would be that much easier to plan.

 

[NW-Bound]

Going into the boondocks in an RV would involve an increase over my current expenses - but I still want to do it

Oh man, I've read about your low rent and living expenses. Forget about RV'ing!

RV'ing is supposed to be cheap, but if you want to move about, gas cost is high. And when your vehicle breaks down, it hurts like the Dickens in the pocketbook.

You should stay put, my friend. It's better to live vicariously through the blogs of the other guys. Or wait till you get SS, and indulge yourself then.

 

[daylatedollarshort]

I wonder if any models of any financial experts 30 years ago accurately predicted SWR 30 years into the future. I doubt there were no more percent that got it right beyond the odds of getting it right through random chance.

 

[bmcgonig]

I wonder if any models of any financial experts 30 years ago accurately predicted SWR 30 years into the future. I doubt there were no more percent that got it right beyond the odds of getting it right through random chance.

Were they even predicting this stuff 30yrs ago? I thought Bengen was an original in all this in the 90s, but not sure. But then again Bengen etc., was not making a prediction, just stating a fact of history.

Question for those in the know: in the Bengen et al studies, do they choose judiciously from where they take the 4% each year? E.g., take it out of the bonds not the stocks if the stocks had a bad year. I think that would make a difference. Also, many of us would have a few years of cash to buffer against a bear, so would that affect the Bengen calculation? Just wondering if anyone had given these studies that kind of thought ?

Thanx

B

 

[bmcgonig]

Thank you! You're right, this paper is easier to read, and presents the concepts in a very logical order, while reaching very actionable conclusions.

This is EXACTLY what I was looking for. Not only starting around 5% is indeed attractive, but the auto-adjustment baked in the process should make it very future-proof. No more doom-saying baked in the equation. No more fear of market timing. No more implicit assumption that the future will be somewhat akin to the past, except in choosing the starting point, but then it will auto-correct as needs be. If the future is grim, your withdrawals will be too, but this is fair game I guess. Same thing with a rosy future, which is very cool. Or with a grim/rosy/grim/rosy future. Etc.

(Mr Independent, you'll also find the answers to your questions, btw)

Now something nags me... Maybe it's just my perception, but it seems that this general model is NOT widely publicized and used. Why would that be? This does seem close to the Holy Grail...

One problem I would have with this study is the assumption they make regarding stock, bond and cash returns. They may be right or they may be a tad optimistic

"Stocks are represented by the total returns of the S&P 500, with a lognormal return relative mean of 9.62 percent and a standard deviation of 19.5 percent. For bonds, total returns of long-term Treasuries are used and the values are 4.99 percent and 6.96 percent. Three-month T-bills are the basis for cash returns and the values are 3.74 percent and 2.98 percent".

Just getting to 3.74 in treasuries again would be nice.

 

[Ronnieboy]

We are always in uncharted waters.

Otherwise things would be that much easier to plan.

It would be so much easier for me to plan my retirement if I just knew when I was going to kick the bucket (without my own interference)

 

[bmcgonig]

It would be so much easier for me to plan my retirement if I just knew when I was going to kick the bucket (without my own interference)

Not to start an annuity discussion, but I remember reading that the greatest advantage of annuities was that they take that question out of the retirement equation.

 

[bmcgonig]

One problem I would have with this study is the assumption they make regarding stock, bond and cash returns. They may be right or they may be a tad optimistic

"Stocks are represented by the total returns of the S&P 500, with a lognormal return relative mean of 9.62 percent and a standard deviation of 19.5 percent. For bonds, total returns of long-term Treasuries are used and the values are 4.99 percent and 6.96 percent. Three-month T-bills are the basis for cash returns and the values are 3.74 percent and 2.98 percent".

Just getting to 3.74 in treasuries again would be nice.

I'd like to be able to rerun the simulations in that paper using Wade Pfaus lower projected stock and bond numbers to see would we get back to 4% ?

 

[siamond]

Question for those in the know: in the Bengen et al studies, do they choose judiciously from where they take the 4% each year? E.g., take it out of the bonds not the stocks if the stocks had a bad year. I think that would make a difference. Also, many of us would have a few years of cash to buffer against a bear, so would that affect the Bengen calculation? Just wondering if anyone had given these studies that kind of thought ?

Guyton/Klinger assumed an annual rebalancing. And this includes the money being withdrawn (view it as reverse rebalancing if you wish, sell first what's high compared to your asset allocation). I think they included a side comment that this was a good thing to do, but didn't change the needle very much.

Can't remember what assumption Bengen made, but this was documented in his original study. Something very simple.

As the pile of cash, well, you should view it as part of your capital, it's just one especially liquid asset category.

 

[siamond]

One problem I would have with this study is the assumption they make regarding stock, bond and cash returns. They may be right or they may be a tad optimistic

"Stocks are represented by the total returns of the S&P 500, with a lognormal return relative mean of 9.62 percent and a standard deviation of 19.5 percent. For bonds, total returns of long-term Treasuries are used and the values are 4.99 percent and 6.96 percent. Three-month T-bills are the basis for cash returns and the values are 3.74 percent and 2.98 percent".

Just getting to 3.74 in treasuries again would be nice.

As far as I understand, they used the most factual data available... the past returns! Whether this is optimistic or pessimistic is crystal-ball reading. Up to each of us to add their layer of caution about the future.

Now, if the withdrawal rate you choose at the beginning was a bit off, the beauty of the model is that it should get back in line based on recent events (i.e. how the market behaves as time goes by). Which is NOT at all the case with those 4% (or 3% or whatever) hard-wired models.

So I guess that if you perceive that the future market returns will be lower than in the past, then yeah, their 5.3% suggestion should be revisited downward. Personally, I started to plug 5% in my own models, with better peace of mind than my previous 4.5% assumption, as it will correct itself if I follow their rules.

Of course, the question would then become... will the auto-correcting withdrawal rate satisfy your needs...

 

[siamond]

I'd like to be able to rerun the simulations in that paper using Wade Pfaus lower projected stock and bond numbers to see would we get back to 4% ?

Would be great indeed. Does anybody know if there is an online tool of sorts allowing to rerun those simulations?

Maybe we should request to the Firecalc experts to include this Guyton/Klinger rule model in the spending model tab... This would be cool.

 

[Huston55]

Would be great indeed. Does anybody know if there is an online tool of sorts allowing to rerun those simulations?

Maybe we should request to the Firecalc experts to include this Guyton/Klinger rule model in the spending model tab... This would be cool.

FireCalc already models this pretty closely with the "work Less, Live More" rule in the spending tab. You can adjust the % of subsequent year's spending with this tab. It's not exactly Guyton/Klinger but, it's probably close.

 

[siamond]

FireCalc already models this pretty closely with the "work Less, Live More" rule in the spending tab. You can adjust the % of subsequent year's spending with this tab. It's not exactly Guyton/Klinger but, it's probably close.

Yes, I know, but still quite different... To the point that a 5% WR is acceptable for one model, and only 4% is acceptable for the other, if the academical guys got it right...

 

[bmcgonig]

Klingers website

Klinger has a website (do you trust a guy called Klinger?

Publications

And he sells the simulator for 50 bucks.