What Berstein says about swr- via email

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

What Berstein says about swr- via email

Post by ataloss »

I couldn't coax him to say anything about mathematical certitude but he is clearly supporting a lower withdrawal rate. I sent him a link to gummy's
swr/life expectancy page which he said was "nice"

i *do* like the trinity study--it's groundbreaking work, it's just that the
embedded return is all wrong going forward . .

all depends on just how long you're going to live, how safe you want to
be, and how much you want to retire on. the three trade off against each
other. if you're 50 and you can't live with even a 2%-4% risk of failure, then you're stuck with a 2% withdrawal rate. if you don't mind a 10% risk (and i can live with that) then 3%-4% is likely ok. and if you're 70, you can probably get away with 5%. the trinity study?? fuggedaboudit. the methodology is sound, but remember, there's a 7% real equity return embedded into it. who in the peanut gallery wants bet their retirement on that?

I have a question about 4 Pillars. I am part of the group of investors that read your books, anxiously await each Efficient Frontier issue and discuss the implications endlessly. Obviously we would all like to know the maximum safe inflation adjusted withdrawal rate from a portfolio. We have been looking at statements in 4 Pillars, especially page 234. You indicate that because of lower projected stock returns (Gordon equation) and the risk of adverse sequences one "might not be entirely safe" withdrawing more than 2% (based on the market at that time). I take this to be a cautionary warning that we can't always rely on the 4% range number from the Trinity study and others when valuations are extreme. You go on to mention reasonable withdrawal rates of 2-5% with a small risk of a need to reduce spending if you hit an adverse sequence of returns. My friend feels that since this 2% number was "produced as the result of a mathematical calculation" and based in data that you were "able to state (not predict) that the safe withdrawal rate for investors retiring in 2000 was 2 percent, not 4 percent."
Although I accept that a 2% inflation adjusted withdrawal based on
initial, portfolio value is always safer than 4%, I don't think we can actually call the 4% rate unsafe.

We have a mutual friend who thinks the Trinity study result will be a
fine guide to the future. Can you help us? Was any inflation adjusted withdrawal greater than 2% "unsafe" at the time the book was written?
Have fun.

Ataloss
raddr
*** Veteran
Posts: 265
Joined: Mon Nov 25, 2002 3:25 am
Contact:

Re: What Berstein says about swr- via email

Post by raddr »

the trinity study?? fuggedaboudit. the methodology is sound, but remember, there's a 7% real equity return embedded into it. who in the peanut gallery wants bet their retirement on that?


In fact, looking back at my Gordon equation thread, a 7% real return going forward would be about 3 SD's above the mean for 40-60 years going forward - using overlapped data.
Last edited by raddr on Sun Nov 14, 2004 4:13 am, edited 1 time in total.
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

I wanted him to address 2000 but I think he is referring to the present. If so, 2% in 2000 may have (in retrospect) been optimistic. :roll:

Since I understand these to be projections based on estimates I will refrain from emailing some diatribe about inconsistency.
Have fun.

Ataloss
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

i *do* like the trinity study--it's groundbreaking work, it's just that the embedded return is all wrong going forward .

Bernstein makes both of the two most critical points in this one sentence.

The first critical point is that SWR analysis is a powerful tool, "groundbreaking" in inpact. Why is it so powerful? Because it is determined by making reference to data. Thus, it does not tell you the same old thing that you can get from listening to talking heads. It tells you what is, it identifies for you the factors that really matter in investing.

There was a day when most people did not understand how much volatility mattered. Scott Burns has material on his web site where he shows that Peter Lynch once argued that it was safe to take a 7 percent withdrawal from stocks because 7 percent is the real long-term return. The data didn't support that argument. The data said that volatility requires that you bring the withdrawal down a few points to be safe. Peter Lynch was wrong about this, and the data was right.

We learned what the data said about this through SWR analysis. It is a groundbreaking tool because it tells us these sorts of things. But the volatility question was not the last question that the data wanted to help us with. Today, the data is saying something different and new and exciting. The data is saying that changes in valuation levels matter too.

That's the second point that Bernstein makes in the sentence above. As much as he likes the conventional methodology, he would not trust it to tell him what is safe. Why not? The conventional methodology makes an implicit "prediction" that the return going forward will be about 7 percent. Bernstein knows this is crazy. He's looked at the data, and he knows this. Raddr has looked at the data too. He says that the chances of this "prediction" coming true are about 1 in 740.

It's not a reasonable prediction. But this prediction influences the results generated by the conventional methodology. It throws them out of whack. You cannot come up with a reaosonable assessment of the SWR using this methodology.

I hope that there is no one on this board who would argue today that a SWR study not counting for the effect of volatility is valid. It is not reasonable not to count something that you know darn well has a big effect. It's the same way with valuation. It affects the result to the question being asked. So it should be considered in analyses of that question.

Bernstein is not going to sign a statement this week saying that the conventional methodology is invalid. It is not reasonable to expect that he would. He has not even participated in our discussions! We need to make the case to him. That's one matter of pressing business for next year.

Once we make the case, there is no reason to think he would not sign such a statement. He clearly has a great appreciation for the value of the SWR tool, as do I. He clearly sees that the conventional methodology fails to account for changes in valuation, and that that is an important factor, as do I. Why wouldn't he want to make the tool better? Why wouldn't he want to make the results of the analysis accurate? I think he will sign on, not in a week, perhaps in a year.

We are far ahead of the curve here. We are not using this board to learn what the rest of the world thinks about how to retire early. We are developing the insights needed to tell the rest of the world how to do it. We are at the center. We are what's happening. This place is where the action is right now.

There's a responsibility that goes with that. There's a responsibility to try to get it right.

Those who are interested in this debate and who have not yet seen JWR1945's post on PE/10 and SWRs (posted at the SWR Research Group board) need to look at it now. here is a link.

http://www.nofeeboards.com/boards/viewtopic.php?t=1173

The key words in the thread were spoken by JWR1945. He says: "What is new is that we have a reliable indicator to convert current prices to intrinsic value." Think for a moment about what that means for the FIRE community. In my assessment, it means more than what has been said in all the other posts on all the other questions that have been examined at the two boards all added together.

Some others may not see the matter as quite that important, but it's hard for me to imagine that anyone can not see that it is incredibly important. We should be celebrating, not attacking each other. I did not do this alone. I had a lot of help. I couldn't have done it without JWR1945 obviously. But I also couldn't have done it without raddr or without BenSolar. Datasnooper added something significant at one point, Biggaloot added something significant at one point. There were lots of others who made important contributions.

This sort of thing was not possible in the days before internet technology made the formation of discussion board communities possible. We have used this new technology to change the world in a not insignificant way.

Can we take a one-day break from tearing each other to pieces to celebrate the extraordinary work we have accomplished here? A lot of work remains, but what we have done thus far is staggering in its potential impact. Can we give a little recognition to ourselves for what we have done?

Can we begin putting what we have learned thus far to good use? Please?
User avatar
BenSolar
*** Veteran
Posts: 242
Joined: Mon Nov 25, 2002 5:46 am
Location: Western NC

Post by BenSolar »

ataloss wrote: I wanted him to address 2000 but I think he is referring to the present. If so, 2% in 2000 may have (in retrospect) been optimistic. :roll:

Since I understand these to be projections based on estimates I will refrain from emailing some diatribe about inconsistency.

:lol: LOL

Bernstein's statement within the book is inconsistent. He talks about needing to knock off 2% from real returns to be safe, and in the next breath talks about 3% or 3.5% real return going forward and how that means that to be entirely safe, investors may not be able to take more than 2%. Hmm 3-2 = 1 if I'm not mistaken. :D

And there is the problem of the growth rate he was using being possibly optimistic. 2% real growth in earnings/dividends is thrown around a lot, but I've seen 1.5 % used frequently to. Dividends were down to 1% in 2000 + 1.5% growth - 2% safety factor = .5% SWR :shock:Crazy numbers! :shock:No wonder he fudged, it looks too ridiculous.
"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
peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Post by peteyperson »

What you have to remember here is we're in the periodic down cycle of the seasons of the market. Predicting without emotion at such a time is very difficult. People become naturally pessimistic and forecast low.

I've seen no data to suggest that the long term growth rate of American or British business has changed. What we're seen is growth in pricing too far, a reduction in dividends to reflect the lower performance of business, a slow down that happens every decade or so and this is all a natural cycle.

Trying to forecast now is troublesome for those reasons. It looks like we'll have a few years of sub-par performance while business improves gradually and market pricing corrects gradually too. Has anyone actually seen data on what is expected in the second decade from now? Buffett and others have said that the next decade will be poor but no comment on anything beyond that. Why? That's too far out to predict.

Petey
BenSolar wrote:
ataloss wrote: I wanted him to address 2000 but I think he is referring to the present. If so, 2% in 2000 may have (in retrospect) been optimistic. :roll:

Since I understand these to be projections based on estimates I will refrain from emailing some diatribe about inconsistency.

:lol: LOL

Bernstein's statement within the book is inconsistent. He talks about needing to knock off 2% from real returns to be safe, and in the next breath talks about 3% or 3.5% real return going forward and how that means that to be entirely safe, investors may not be able to take more than 2%. Hmm 3-2 = 1 if I'm not mistaken. :D

And there is the problem of the growth rate he was using being possibly optimistic. 2% real growth in earnings/dividends is thrown around a lot, but I've seen 1.5 % used frequently to. Dividends were down to 1% in 2000 + 1.5% growth - 2% safety factor = .5% SWR :shock:Crazy numbers! :shock:No wonder he fudged, it looks too ridiculous.
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

Is it possible to know the intrinsic value of the market?

Post by therealchips »

The key words in the thread were spoken by JWR1945. He says: "What is new is that we have a reliable indicator to convert current prices to intrinsic value." . . . Can we take a one-day break from tearing each other to pieces to celebrate the extraordinary work we have accomplished here?

OK, nothing personal, just a quotation from Four Pillars page 54,
Princeton economist Burton Malkiel famously stated that "God Almighty himself does not know the proper price-earnings multiple for a common stock." In other words, it is impossible to know the intrinsic value of a stock or the market.

I detect a difference of opinion. (and remark upon it in the mildest possible tone :wink:)
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

Chips
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

Bernstein's statement within the book is inconsistent.

I think that may well be a technically correct statement, but also that it is unfair to Bernstein. Bernstein was not issuing a formal SWR study. It is not reasonable to hold him to the same standards as you would someone who was doing that.

He was writing for the general investor, and the points he makes are of great importance to the general investor. The things he said are light years ahead of what anyone else was saying at the time. The Bernstein book was published before my May 13, 2002 post was put up. I do not know of anyone who had faulted the conventional methodology for failing to consider the effects of changes in valuation prior to the time that Bernstein did so (I understand that he did so only implicitly and not directly).

He advanced the world's knowledge a great deal with that book. It is not right to look back now when we have had the chance to do dozens of analyses and find fault with him for speaking loosely on some of the details. He got the essentials right before anyone else did, and he deserves our gratitude for doing that, in my opinion.

Dividends were down to 1% in 2000 + 1.5% growth - 2% safety factor = .5% SWR Crazy numbers! No wonder he fudged, it looks too ridiculous.

I think you are absolutely right on why he fudged a bit. I think he looked at those numbers and was shocked and tried to state things in as moderate terms as he possibly could. The mesage that the data is telling us for the SWR at the top of the bubble is shocking.

I think it would help if you would consider what a SWR really is. It is an extremely conservative number. It tells you what happens in the worst-case scenerio. You should expect that number to be low. When you think carefully about what that number is, it is not quite so surprising to see how low it drops at times of extreme valuation.

What the SWR analysis is doing in essence is comparing the stock asset class to a 100 percent safe and inflation-adjusted asset class, like TIPS. Conceptually, that is what is going on. You are asking "if I need to absolutely safe, need I go with TIPS or can I get away with going with stocks?"

But the stock asset class will in most circumstances provide a return a whole bunch better its SWR, while the TIPS class will not. The stock class has a large potential growth component tacked on. So in ordinary circumstances, you should intuitively expect the SWR for stocks to be lower than that for TIPS. I'm not saying this is true in reality, just that stocks would be more appealing than TIPS at a lower SWR than TIPS. Stocks can provide a lower SWR and still remain a more attractive asset class. Stocks do not need a high SWR to possess a lot of appeal because they offer so much appeal through the growth opportunity they provide.

Thinking about it that way can help in understanding why the number should not always be as high as we have been led to believe it always is. That doesn't mean that you would often expect it to fall to 2.0 percent or less. But the overvaluation at the top of the bull market was the greatest we have seen in 130 years, and not by a small amount. Those circumstances were extraordinary in the extreme.

You are shocked by the SWRs you are reading from the data. You really should be shocked by the level of overvaluation we experienced. The low SWR numbers are just a reflection of that. Because the SWR is a data-based tool, it doesn't have the option of allowing its shock to influence its reading. It just reports to you what the level of overvaluation in effect at the time tells you about what is safe.

The SWR informs you. It doesn't tell you how to invest. It provides information in one neat compact number that permits you to gain perspective as to what sort of deal is being offered through investment in stocks at a particular point in time. It helps you make illuminating comparisons. But it only does this if you insist that the number be solely a reflection of what the data says.
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

I detect a difference of opinion. (and remark upon it in the mildest possible tone )

The point you are making here is a good one, in my view. Bernstein did indeed say that, and I would like to understand better why he said it and whether (and how) it is consistent with some of the other things he said. We need to make sense of this business, not prove one side right and one side wrong.

We have a tiger by the tale. It may be that, after we put together a report stating what we have come up with and show it to Bernstein, he will change the way he stated that a bit. That would be a big deal, I think.

JWR1945 noted that not all details have been tied down, that there is much work to be done. There is. But the insights we are developing may someday provide us with an investing tool of incredible power, one that Bernstein himself did not know about at the time he wrote that book.

Bernstein offered some powerful insights. He did not put all the pieces together. He did not formally put forward an alternative to the conventional methodology. That is one of the things that we need to try to do. JWR's post from yesterday brings us a good bit closer to being able to do that than we were two days ago, in my opinion.
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

I detect a difference of opinion. (and remark upon it in the mildest possible tone )


Hi Chips, I think that part of the issue is the level of certainty one attributes to predictions. Even Bernstein's statement:

current retirees may not be entirely safe withdrawing more than 2% of the real starting values of their portfolios per year!

is subject to misinterpretation. I think he means that he finds it inconcievable that a 2% withdrawal rate would fail. Others have interpreted this to mean that he is quite literaly saying that the chance of success is 100% for a 2% swr

In any case Trinity is outdated :wink:

I still like this graphical representation of swr and p/e by intercst
(I think gummy had somethig similar)
Have fun.

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

Post by BenSolar »

ataloss wrote: I still like this graphical representation of swr and p/e by intercst
(I think gummy had somethig similar)


The server is set up such that the image can't be linked to directly. :(The good one is the graph at the bottom of this page: http://rehphome.tripod.com/pestudy1.html
"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
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

very clever, I didn't know this was possible (and I saw it when I previewed)
Have fun.

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

Post by ataloss »

now I see it!
Have fun.

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

Post by BenSolar »

ataloss wrote: now I see it!


You can see the graph on this board? I can't. Maybe your computer is drawing from your Temporary Internet Files, not from the server ...

B
"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
WiseNLucky
** Regular
Posts: 84
Joined: Tue Nov 26, 2002 3:59 am
Location: Florida

Post by WiseNLucky »

And there is the problem of the growth rate he was using being possibly optimistic. 2% real growth in earnings/dividends is thrown around a lot, but I've seen 1.5 % used frequently to. Dividends were down to 1% in 2000 + 1.5% growth - 2% safety factor = .5% SWR Crazy numbers! No wonder he fudged, it looks too ridiculous.


We're getting awfully close to that "true" swr of zero. :wink:
WiseNLucky

I just wish everyone could step back and get less car and less house then they want, and realize they don't NEED more. -- NeuroFool
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

:oops:
You can see the graph on this board? I can't. Maybe your computer is drawing from your Temporary Internet Files, not from the server ...


I did see it on this board but....
I did hop over to rehp and probably got it as a temporary file
now it is gone
Have fun.

Ataloss
Post Reply