Ataloss takes advice from hocus

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

Ataloss takes advice from hocus

Post by ataloss »

I read page 234:
A particularly bad returns sequence can reduce your safe withdrawal amount by as much as 2 percent below the long-term return of stocks.


Thus he takes the 3.5% Gordon equation long term stock return projection and subtracts 1.5%. That is his sophisticated methodology! Probably explains why he calls it "Portfolio Theory for Poets."

The preceeding couple of pages deal with variability to lead up to why annualized average return = annual withdrawal does not work

I can see how hocus misinterpreted this
Have fun.

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

Post by hocus »

That is his sophisticated methodology!

Ataloss:

I am not a numbers guy. So I am not the one to go to for an explanation of the details of what Bernstein did to come up with his number.

I recalled JWR1945 having put up a post dealing with the question, so I looked it up. Here it is:

http://www.nofeeboards.com/boards/viewt ... ght=#p6953

Here is the entire text of the post:

"I have finally figured out how William Bernstein came up with his numbers. (Reference: the Four Pillars of Investing by William Bernstein.) He relies entirely on the Gordon Equation and Monte Carlo simulations. He validates his Monte Carlo simulations by using the results from the Trinity Study and other studies that use the historical sequence methodology.

He causes a whole lot of confusion because he likes to come up with plausible rules of thumb. He throws around percentages all over the place. There are some logical errors hidden within the details behind these plausibility arguments.

Now let us take advantage of some quotes that hocus has extracted.
Quote:

Page 56: "Going forward, it looks like stock and bond returns should be in the 6 percent range, not the 10 percent historical reward. Don't shoot me, I'm only the messenger. [Comment: This translates into an SWR of 2 percent]....

"On an intellectual level, most investors have no trouble understanding the notion that high past returns result in high prices, which, in turn, result in lower future returns. But at the same time, most investors find this almost impossible to accept on an emotional level.....

Page 57: "At the end of the day, the Fisher DDM method of discounting interest streams is the only proper way to estimate the value of stocks and bonds. Future long-term returns are quite accurately predicted by the Gordon Equation....

What Bernstein has done is this: He used the Gordon Model to estimate the long-term growth rate of stocks at recent valuations. His estimate was 3.5% real growth. Then he used his Monte Carlo simulation and calculated a Safe Withdrawal Rate. The answer was 2%.

raddr makes similar calculations assuming 3.5%. I suspect that his Monte Carlo simulator is much more accurate than William Bernstein's.

I (JWR1945) believe that there needs to be an adjustment to the Gordon Equation/Dividend Discount Model. It is because payout ratios are much lower now than they were previously. See the following thread (by JWR1945) dated Tue Feb 11, 2003 at 10:27 pm CST about stock market return projections.
http://nofeeboards.com/boards/viewtopic.php?t=494

With the help of BenSolar and others, my final numbers are in the neighborhood of 4.7% for the real return of the stock market for the next decade or two.

We now have three numbers. The lowest is 3.5% and it is favored by William Bernstein and possibly raddr. The intermediate projection is 4.7% that I favor and possibly BenSolar. The long-term historical real rate of return is 6.5% to 7%.

If I were to use Bernstein's approach, I would have come up with a larger estimate of the Safe Withdrawal Rate for 2000 than he did. If would have been lower than that of the historical sequence method because I would have projected lower real stock market returns. I actually used a different approach. (I scaled the answer back into the historical range of valuations.) My estimate was 2.3%.

Have fun.

end of JWR1945 post

I am OK with all what JWR1945 says.

I am also OK with all that Bernstein says. Most of what I know about the Gordon Equation, I know from reading Bernstein. What Bernstein says about it persuades me that Bernstein was taking a perfectly reasonable approach in using the Gordon Equation to incorporate valuation into his SWR calculation.

My sense is that there is more than one reasonable way to incorporate valuation into the analysis, and I'm OK with all of those that have been discussed. The only thing that I consider unreasonable is the idea of making no adjustment whatsoever.
raddr
*** Veteran
Posts: 265
Joined: Mon Nov 25, 2002 3:25 am
Contact:

Post by raddr »

I think that hocus is right about Bernstein taking the Gordon number and plugging it in to a MC simulator and getting a 2% SWR from this. I have done much the same but my MC simulator adjusts for mean reversion tendency which IMHO makes for more realistic numbers. FWIW, Bernstein knows that conventional MC methodology gives results that are too pessimistic so he puts in a smaller number for volatility than what has been seen historically. This gives better numbers but two wrongs don't make a right, IMO.
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

FWIW, Bernstein knows that conventional MC methodology gives results that are too pessimistic so he puts in a smaller number for volatility than what has been seen historically. This gives better numbers but two wrongs don't make a right, IMO.

This observation is the sort of thing that we should be thinking about when discussing my claim that the conventional methodology is invalid.

I accept that you have to be cautious about calling a methodology "invalid." Raddr is above expressing a negative opinion on a methodology employed by another researcher. This happens all the time. It is good that this sort of thing happens because it is through the conflict of viewpoints expressed on a question like this that people come to advance knowledge of the best ways to do the research. You start saying that there is only one way to do things or else the studies are invalid, and you shut down the exploration of avenues that might lead you to good stuff down the road.

The normal way to handle a difference of opinions on methodology is for the two researchers to each explain in their respective studies how they did things and to make the case for their respective approaches. It's a good thing that that is the normal rule, in my view.

That said, there are exceptional circumtances. The word "invalud" means something and there are circumstances in which using it serves a purpose. My claim is that this SWR matter is such a circumstance.

There are two things that I believe justify my invalidity claim: (1) valuation always affects SWRs; and (2) the effect is sometimes great. If valuation were a factor only every now and again, or if it were a factor that never made all that much of a difference, I would say that it is the sort of thing where different researchers could hold different valid viewpoints. Some could elect to include a valuation factor and some could elect not to.

A factor that always affects the outcome of a question under study is too important a factor to ignore, in my view. When it comes to SWRs, I see four big factors that are at play: (1) nominal returns; (2) inflation; (3) volatility; and (4) changes in valuation. My view is that the fourth is roughly as important a factor as the other three, and that it would be invalid to do a SWR study ignoring any of the four.

Would we consider a SWR study that failed to consider the effects of inflation to be valid? Say that someone issues a study that says "the long-term annual return on stocks is 10 percent, and, thus, as long as the future is no worse than the past, you can have 100 percent confidence that a 10 percent withdrawal will work." Is the study valid? I say no. I say the researcher doesn't know what he is talking about, and the study is dangerous.

Say that a study says "the long-term real return is 6 percent, so you can have 100 percent confidence that a 6 percent withdrawal will work, presuming that the future is not worse than the past." Valid? I say no. The researcher has not considered volatility. Volatility affects SWRs. This is a foul ball.

It's the same with valuation, in my book. You can't get the right number without considering changes in valuation levels, so you should incorporate data on them into your analysis. As to which particular data to use to address this factor, I think we need to leave that to the researchers.

Look at how inflation is handled in the intercst study. My understanding is that he calls for an inflation adjustment based on one of the government-issued inflation estimates. Are those estimates perfect? They are not. I have heard people raise various questions about them. But so what? Are you going to nail intercst to a cross because someone made some niggling comment about an inflation report? I say no.

It is legitimate to question the inflation adjustment he uses. It's OK to issue a different study that uses a different means for making an inflation adjustment. But his approach is reasonable. So the study employing it is valid.

I see a difference with someone not making any adjustment whatsoever for valuation. In that case, it's not because the researcher is choosing one of several reasonable approaches over another. The researcher is being outright unreasonable. The researcher is acting as if valuation didn't matter! I think that's dangerous. I think it is deliberately misleading to issue a study like that (presuming that you knew at the time you did it that valuation always affects the result).

I see no justification for doing something like that. I see no good that could come out of a study like that (once the valuation effect became known). So I see no downside in calling it what it seems to me it really is, invalid.

It's a question of standards. Some things you can have reasonable differences of opinion on. Some things are so basic that a person claiming to be informed in the field simply must be aware of them.

The fact that valuation affects SWRs is not so well known today, because we are just starting to figure out where we stand on this thing. I think that one of our next steps should be to publicize what we come up with to a broader community. After that is done, I think it is fair to insist that researchers who want their work to be viewed as valid to live up to the reaonable demands of the community making use of their studies.
[KenM]
*** Veteran
Posts: 133
Joined: Tue Mar 04, 2003 12:54 am

Post by [KenM] »

FWIW, Bernstein knows that conventional MC methodology gives results that are too pessimistic so he puts in a smaller number for volatility than what has been seen historically


That just about sums up my opinion of Mr Bernstein. I respect his views and agree with much of what he says, but it appears he's no more averse to adjusting things or presenting data or wording his statements so that they fit his hypothosis than many others.
KenM
Never try to teach a pig to sing. It wastes your time and annoys the pig.
User avatar
BenSolar
*** Veteran
Posts: 242
Joined: Mon Nov 25, 2002 5:46 am
Location: Western NC

Invalid?

Post by BenSolar »

hocus wrote: my claim that the conventional methodology is invalid.

I don't agree with this claim. In science of all sorts, researchers frequently chose to study one part of a problem and basically ignore everything else. I see it all the time in food and health studies. Just because a study doesn't present the whole picture doesn't mean it is invalid. Just that it is limited. Authors usually include some wiggle words to indicate that they haven't presented the entire picture. The REHP study does that.

Message board posts from the past may present a different picture, with less use of wiggle words and a greater appearance of certainty. But message boards aren't peer reviewed journals, either, and, in fact, are notorious for spreading all kinds of bad information and advice. Reader beware! We may not like it that intercst doesn't acknowledge the evidence we have that high valuations put the 4% SWR into not-so-safe territory, but ... c'est la vie. Most people on that board acknowledge it now, with a few noted exceptions.

I would drop the 'invalid' campaign. I don't think it is very supportable. But that's just my opinion, FWIW. :?
"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
[KenM]
*** Veteran
Posts: 133
Joined: Tue Mar 04, 2003 12:54 am

Post by [KenM] »

hocus
I may be the only one dumb enough not to understand you, but I still don't know what you mean when, for studies based on historical data, you say
someone not making any adjustment whatsoever for valuation


Most people on this board probably agree that PE10 is a reasonable measure of valuation. The analysis in all the historical studies, including REHP, included the following historical worse-case situation on valuations of the 30 year period starting in 1929.

PE10 of 32 in 1929 which was the highest in history prior to the recent bubble.
PE10 dropped to 6 in 1932.
PE10 finished at 18 at the end of the 30 year period in 1959.

As jwr's Big Swings and Valuations thread shows, for a particular asset allocation, the SWR was 4% for 30 years starting at the highest valuation in previous history which was 1929.

Therefore (and I know I'm repeating myself) all the historical studies have covered historical highs/lows/changes in valuation in the same way they have used historical highs/lows/changes in inflation, highs/lows/changes in volatility etc. in the historical period covered by the study. They then indicate an SWR while stating something similar to, as REHP, "This study is based on historical data...past performance does not guarantee future results". What we have for valuations, however, is the classic "the-future-has-turned-out-to-be-worse-than-the-past" scenario because the 2000 bubble valuations of PE10 of 44 were higher than the previous highest historical valuations in 1929 included in the period covered by the studies.

So I still don't understand what you mean when you say the studies based on historical data didn't include valuations and/or changes in valuations. :?
KenM
Never try to teach a pig to sing. It wastes your time and annoys the pig.
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

It is possible that Bernstein got the 2% number from Monte Carlo analysis. It doesn't say that anywhere in the book , but I can't exclude the possibility that that is where it came from.

Although I will agree that statements by Bernstein are authoritative, the hocusian point seems to be that Four pillars is inerrant or at least infallible. That is, intercst is wrong because he conflicts with the "word of Bill,"￾ a superior or perhaps supreme authority. This is the only reason I can see to find that intercst is wrong without knowing the reasoning and assumptions (hocus would say data) behind his number. If I planned to write a book saying that A was wrong and B was right it would be important to me to know exactly how I could prove B. I would not be satisfied with "it says so on page 234 of Four Pillars." This is where I differ from hocus. Hocus says he doesn't "know all the details"￾ and I say that we "have no idea"￾ but in essence we are saying the same thing. Bernstein makes a strong case for lower stock returns and gives a swr without explanation.

I would contrast this with what I see as raddr's approach. He has entered well reasoned estimates for a defined tsm/fixed portfolio into a Monte Carlo program and concluded that safe withdrawal rates will likely be much lower that 4%. Since the process is deemed reliable, the only way to get a higher withdrawal rate would be to start with what he feels are unreasonable assumptions. This seems highly rational to me. If you (or intercst) want to disagree with raddr you have to make a case for higher stock returns (or errors in Monte Carlo analysis or implementation.)

If hocus wants to claim that he absolutely knows that 2% is right and 4% is wrong on this basis it is fine with me.

I would be comfortable with something similar to what Bernstein said about historical analysis "sometimes being unreliable"￾ coupled with reasoning about lower future expected returns. Applying a historically derived swr when valuations are outside the historical range is obviously a weakness. I am not writing the book but I don't think anyone wants to read a chapter about what was said on some internet board or unprovable charges that some study was deliberately misleading.
Have fun.

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

Post by ataloss »

KenM and bensolar,
You are engaging in a FRH (frequent response to hocus.) the issue of valuation and the "traditional" swr approach and the "invalidity" of "traditional" approaches have been explained to hocus dozens of times on this board and the REHP. Too many of those posts have been written by me. IMHO it is futile to go on.

The reason why I say that the 2.0 percent number is valid and the 2.3 percent number is valid, but the 4.0 percent number is invalid, is that the first two numbers both are the product of methodologies which consider the effect of changes in valuation, while the latter is not.

This is the hocusian definition of a valid study. He won't be swayed.

I admit that this wasn't a very nice response

I see that my problem is that I don't speak hocus.

Valid SWR methodology: explicity uses input related to valuation, nature and reliability of the input data and calculation process - not important

Invalid SWR methodology: no explicit use of valuation (even if valuation data is implicit in the data used)

Perhaps we could formulate a FRH statement?

Personally, I have reservations about a 4% swr taken when the market is at unprecedented valuation and I have no general interest in defending intercst but I get sucked into these pointless discussions. If several of us had a mutually agreed upon statment on this matter we would only need to update it if hocus actually came up with something new. What do you think?
Have fun.

Ataloss
User avatar
ben
*** Veteran
Posts: 231
Joined: Mon Feb 17, 2003 4:00 am
Location: The world is not enough

Post by ben »

Ken; you said it SO well! :D- exactly what I don't get either!

You will all have noticed that I am refraining from much participation in these ongoing SWR threads as FRH simply risks that I will be looking for emoticons not found on the standard templates...
Normal; to put on clothes bought for work, go to work in car bought to get to work needed to pay for the clothes, the car and the home left empty all day in order to afford to live in it...
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

In science of all sorts, researchers frequently chose to study one part of a problem and basically ignore everything else

There's nothing wrong with a study that studies only part of the question. If someone puts forward a study called "A Look at Some Aspects of the SWR Question," I am OK with them using the conventional methodology.

The issue for me is, do they purport to calculate the SWR? The SWR is a defined concept. Either you have attempted to examine the defined concept or you have not. I think it is fair tob say that any study called "The SWR Study" is purporting to tell you what the SWR is.

Authors usually include some wiggle words to indicate that they haven't presented the entire picture. The REHP study does that.

If you are saying that there is language saying saying something to the effect of "this study does not examine the data that would need to be considered to determine the SWR," then I would like to take a look at that language.

Message board posts from the past may present a different picture

There is an implicit criticism of intercst being made here that I believe is unwarranted. I believe that intercst does a reasonably good job of presenting the results of his research when asked to do so. I cannot think of any claims he had made about the research that are not valid in the event that the methodology is valid.

It is only when the methodology is questioned that intercst goes off the rails, in my view.

We may not like it that intercst doesn't acknowledge the evidence we have that high valuations put the 4% SWR into not-so-safe territory,

Why should he acknowledge this if the analysis he did shows that it is not true? The analysis he did shows that the SWR is 4 percent. What is there not to like in the act of an analyst accurately presenting to the public the results of his research?

I would drop the 'invalid' campaign. I don't think it is very supportable.

My plan is not to drop it, but to etake it to a much bigger audience. I am thinking of writing articles for personal finance magazines, going on radio interviews, publishing a report of my own, pulling in experts better known than those that I have referred to previously, and so on. I do not want to do these things if I am wrong, obviously. If I am wrong, you are doing me a big favor by telling me so.

But I need data. All the data I have seen in the first 14 months points in one direction--that the conventional methodology is invalid. If there is data out there indicating that changes in valuation do not affect the SWR, I very much want to see it. Some data coming from that direction could resolve this thing in 10 minutes. I just haven't seen anything like that yet.
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

So I still don't understand what you mean when you say the studies based on historical data didn't include valuations and/or changes in valuations.

Here is what Bernstein says on Page 73. "Although the Discounted Dividends Model informs us well about expected returns, it tells us nothing about future risk. We are dependent on the pattern of past returns to inform us of the potential risks of an asset. And in this regard, I believe that the historical data serve us well. Although anything is possible in finance, it is hard to imagine the stock markets of the next century throwing anything our way that would surpass the 1929-1932 bear market."

I agree with that statement. He is saying some important things with that statement. I believe you need to read those sentences over a few times to get the full impact.

One thing he is saying is that the "future worse than the past" thing is highly unlikely. Yes, it is possible that the U.S. government could collapse, but in that case you are probably cooked anyway, so there's not much good done in worrying about it too much. As far as volatility goes, the 1929 situation was plenty bad, so it is unlikely that the future will be worse than the past in that regard. (I'm ignoring here the raddr research on the "lucky draw" effect not because I do not think it is important, but because discussion of it adds complexity that I am trying to avoid for the moment).

So Bernstein is saying that the "future worse than the past" caveat is unlikely to kick in. We are probably not going to see a worse effect from volatility than the effect experienced by retirements beginning in the year 1929. So he has full confidence in the results of the conventional studies, right?

No, he does not. Look at the earlier part of the statement above. He is describing the things he calls "risk" (I refer to this effect as "volatility") as something independent of the thing he refers to as "expected returns" (I refer to this effect as changes in valuation). He is saying that these are two separate effects. You can't assess just one, you need to look at both.

The conventional methodology looks at only one. The conventional methodology does a perfectly good job of looking at the volatility question. It handles that aspect of the question well. That only gets you halfway to knowing the SWR, however. To know the SWR, you also must factor in the effects of what he refers to as "expected returns" and what I refer to as "changes in valuation." Is he sure that this effect matters? He says elsewhere that he is sure as a matter of "mathematical certainty." That's pretty darn sure.

Is 4 the SWR for a retirement beginning in 1929? No. It does not meet the definition of "SWR" used in SWR analysis. The data indicates that there is a not insignificant chance that that retirement will fail.

It turned out that it succeeded. But that is something that the 1929 retiree knew only afterwards. A SWR must be calculated in advance. The data does not give grounds for saying that a 4 percent withdrawal is safe in the circumstances that applied in September 1929. Valuation levels were too high to justify a claim that that withdrawal level was safe.

The fact that something works does not mean that it was safe all along. There are lots of cases when people take risks and they end up not suffering any negative consequences. That's what happened to someone who retired in 1929 with a 4 percent withdrawal. He took a risk and he got away with it. That's nice for him. But it does not justify us proclaiming in studies that that rate somehow was safe all along. It was not. Not according to the historical data (for purposes of making this point, I mean all of the data available to us today, not just the data that was available in 1929).

The 4 percent number is not the safe number. It is just a number that happened to pop up as a result of the way the conventional methodology was constructed. There had to be some number that popped up. The fact that that number happened to be the one that popped up does not give it some great significance.

People are giving great significance to the fact that it happened to be 4 that popped up, but it could just as easily have been 3.5 or 4.5 or something else. There is no great significance in a single data point. People are looking at the withdrawal rate that happened to work for a retirement beginning in 1929 and drawing unwarranted conclusions from a single lucky case for a 4 percent withdrawal.

What the conventional analysis tells you, I believe, is the average SWR over a long period of time. If you properly calculated all the SWRs for each of the past 100 years, added them together, and then divideded by 100, I believe that the number you would get would be something close to 4. I guess it's good to know that number. But that number is not the SWR as defiined for purposes of SWR analysis.

You are not able to go back in a time machine and retire a little bit in each off the past 100 years. All retirements take place at a particular point in time, and the valuation levels in effect at the time at which those retirements commence has a critical effect on whether they survive or not, according to the historical data. Before you hand in your resignation, you need to know the particular SWR that applies to your retirement, not a number that is an aggregation of all the SWRs for the past 100 years.

The premise of the conventional methodology is that SWRs remain stable with changes in valuation levels. That's why existing studies say that the SWR for a retirement beginning in Janaury 2000 is 4 percent and the SWR for a retirement beginning today is 4 percent. It's an absurd thing to say, but this absurd claim is a logical consequence of the use of an absurd methodology. It's a garbage-in, garbage-out situation.

The SWR is defined to be the highest take-out number that works in the event that the worst-case scenario pops up in your retirement, assuming that the future is no worse than the past. You cannot calculate that number without considering the effects of changes in valuation levels. Why? Because the phrase "assuming the future is no worse than the past" is a key part of the definition. How can you assess what will work if the future is like the past if you are not willing to look at a factor that in the past has always had a crtiical effect on the question posed?

Could Bernstein's 2 percent SWR fail? It could indeed. It will fail if the future is worse than the past. He is telling you the highest take-out number that works if the worst-case scenario turns up and if the future is no worse than the past. That's what he is supposed to be telling you. Bernstein's number conforms to the definition of "SWR" used in SWR analysis. The number generated by the conventional methodology does not.
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

The hocusian point seems to be that Four pillars is inerrant or at least infallible.

I don't think Bernstein is infallible. I am personally convinced that what he is saying is right (not the details, but the general thrust), but I have no problem with people saying that they disagree with him. If Bernstein is wrong in the thrust of what he is saying, then there is a strong chance that I am wrong too. If he is a little off on some small semantics point somewhere, I don't think that necessarily has any great practical consequence.

I was saying what I am saying now years before I saw the Bernstein book. All that the Bernstein book did was add to my confidence level. I thought about it the way he does way back in 1996. But it made me feel better about what I was saying to have a recognized expert come down the same on the key points.

I would be comfortable with something similar to what Bernstein said about historical analysis "sometimes being unreliable"￾....

He didn't say the conventional methodology is sometimes "unreliable." He said that it is sometimes misleading.

I am not writing the book but I don't think anyone wants to read a chapter about what was said on some internet board or unprovable charges that some study was deliberately misleading.

I've dropped the chapter on investing from the book I am working on. But I do intend to write some articles on this for personal finance magazines, and to engage in other efforts of that nature to publicize our findings.

I will not be saying that any analysts who have used the conventional methodology in the past were deliberately misleading their readers by doing so. My claim will be that, after I have achieved 99 percent certainty on my claim (I am now at 90 percent) and written it up in a comprehensive way in a report and publizcied the findings, that anyone using the coenventional methodology from that point forward is deliberately acting in a way that misleads readers of such studies in ways that could cause them serious life setbacks.
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

You are engaging in a FRH (frequent response to hocus.)

I am largely in sympathy with the comment being made by Ataloss here. We have gone over some of these points many times before. It is unlikely that an argument that failed to persuade me earlier will persuade me now, and it is unlikely that the board community is not aware of arguments that have been put forward many times before. So I am not sure that any constructive purpose is served by rehashing old stuff.

It's OK by me if we let this drop. I am not trying to push my views on anyone. If you have questions about what I am saying that you would like me to address, I would like to be as helpful as possible. But there is no law that says that anyone needs to agree with me. I don't feel any hostility towards anyone who has expressed viewpoints on this question other than my own (that includes intercst, by the way).

I am just a guy who had an insight about SWRs and who put it forward on an internet discussion board. I didn't generate the tens of thousands of posts that have resulted. Almost all of mine were in response to questions put forward by others. I had 14 thread-starters in the entire 12 months of the TMF phase of The Great Debate. There are lots of people on that board who put up more than 14 thread-starters in the course of a year.

You will not hurt my feelings even a little bit if you turn the discussions in some other direction. I am absolutely cool with that. I have other venues in which I intend to pursue this. I will be putting up stuff at my web site when I get it up next year. I will be writing articles for magazines. Perhaps I will give some speeches. Things like that.

I would like to stop in here now and again to let people know of developments and I would like to use the "SWR Research Group" board as a place to accumulate the research already done and to develop the arguments and make them sharper. So I would like to remain a member of the community here. If there is not too much interest here in participating on the SWR Research Group board, that's fine with me. I intend to use that board to develop my case, and if a few people want to help out, that would help me a lot. But there is certainly no expectation on my part that a lot of people will do that, or that a lot of people even should. It is perfectly natural that other people would have different priorities. This is a priority to me. Only you can determine what is a priority for you.

Does any of that make sense to people?
[KenM]
*** Veteran
Posts: 133
Joined: Tue Mar 04, 2003 12:54 am

Post by [KenM] »

hocus,
The folowing statements of yours are totally, completely, utterly, 100% incorrect. If you can't see that, then you have no hope of leading a campaign to improve SWR analysis.
the "future worse than the past" thing is highly unlikely
So Bernstein is saying that the "future worse than the past" caveat is unlikely to kick in

...the 2000 bubble valuations were far worse than valuations seen in the previous 130 years. The "unlikely" event has already kicked in.
the things he calls "risk" (I refer to this effect as "volatility")
Bernstein's "risk" is most definitely not your "volatility".
The conventional methodology looks at only one
Is 4 the SWR for a retirement beginning in 1929? No
What the conventional analysis tells you, I believe, is the average SWR over a long period of time. If you properly calculated all the SWRs for each of the past 100 years, added them together, and then divideded by 100, I believe that the number you would get would be something close to 4.

You have a lot of people with a math background on this board telling you that many of your assertions are wrong. Therefore, regrettably, if you can't see that the above statements are totally incorrect then you are going to be very lonely on your new board.

One further point....
He (Bernstein) is telling you the highest take-out number that works if the worst-case scenario turns up and if the future is no worse than the past.
........and if Bernstein had been writing his book in 1995 when the previous worst case valuations in history had been 1929, his SWR would have been 4% - that's what the historical data said on valuations. By 2000 the future had turned out to be worse than the past so he says 2%. Why can't you see that ?
KenM
Never try to teach a pig to sing. It wastes your time and annoys the pig.
User avatar
BenSolar
*** Veteran
Posts: 242
Joined: Mon Nov 25, 2002 5:46 am
Location: Western NC

Post by BenSolar »

hocus wrote: What the conventional analysis tells you, I believe, is the average SWR over a long period of time.


You've got to get this part straight, hocus. The historical studies cherry-pick the worst withdrawal rate that survived every period in the 130 years in the study. So, yes, the 4% survived a 1929 start, and it survived the 1966 start and every other bad start inbetween.

As far as the invalidity thing goes, we will have to agree to disagree. I'm up for dropping this part of the debate.

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

Post by hocus »

The folowing statements of yours are totally, completely, utterly, 100% incorrect.

OK, KenM I think it is constructive for you to list the statements that you take issue with. That should be a help to anyone trying to make sense of this thing.

I think the best thing for me to do in response is just to say that I stand by the statements.

You have a lot of people with a math background on this board telling you that many of your assertions are wrong.

It is a fact that there are many people with math backgrounds on both boards who have said that I am wrong.

Therefore, regrettably, if you can't see that the above statements are totally incorrect then you are going to be very lonely on your new board.

I'm coming at it from a different perspective, KenM. It's the being lonely part that excites me about this. If most people agreed with me, I wouldn't see any great profit to the FIRE community to going about the task of proving the point. It's because so many do not understand the realities of SWRs that I find it so important to get the word out on this issue.

One of the things that came up[ on the other board a lot was the request that I go back to writing about the soft side of early retirement. There's nothing I would like better! This numbers stuff makes me dizzy. The investing thing is my least favorite part of the FIRE equation. My book is not even going to mention investing, partly for that reason.

But I am a journalist by background. One thing you develop as a journalist is a nose for news. This thing has "MAJOR NEWS EVENT" stamped all over it. It is positively off the charts. The reaction to this has been phenomenal. I have never been involved in anything like this in my entire lifetime. This thing is an opportunity.

I wish that my Big Opportunity had been on some question relating to the soft side. I wish I had gotten this reaction to my "Retire Different!" post or my "The New Luxuries" post. That would have been better for me all the way around. I wouldn't need any help to develop my arguments on those sorts of questions. It's a pain that my Big Opportunity happened to be on the investing side of the equation.

But that's the way it played out. I don't ask questions. I just go with the Wave wherever it wants to take me.

By 2000 the future had turned out to be worse than the past so he says 2%. Why can't you see that ?

You are saying that the "if the future is no worse than the past" caveart had already kicked in on January 1, 2000. If that is so, do you believe that analysts who used the conventional methodology were under an obligation to inform the readers of their studes that this was the case?

You are saying that the underlying methodology was valid but that the results were invalid in that time frame. Why didn't the SWR experts let us know?
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

You've got to get this part straight, hocus. The historical studies cherry-pick the worst withdrawal rate that survived every period in the 130 years in the study. So, yes, the 4% survived a 1929 start, and it survived the 1966 start and every other bad start inbetween.

I've had that part straight from the first day, BenSolar. The existing studies do indeed cherry-pick the worst withdrawal rate that survived every period in the 130 years. What of it?

Cherry picking the worst withdrawal rate that survived every period in the 130 years does not reveal the SWR, as defined for purposes of SWR analysis. It is the fact that studies using the conventional methodology report this cherry-picked number as the SWR that makes them invalid. They go about the process of determining the SWR through use of an invalid methodology for answering the question posed.
User avatar
BenSolar
*** Veteran
Posts: 242
Joined: Mon Nov 25, 2002 5:46 am
Location: Western NC

Post by BenSolar »

hocus wrote:
I've had that part straight from the first day, BenSolar. The existing studies do indeed cherry-pick the worst withdrawal rate that survived every period in the 130 years. What of it?


Well when you make a statement like this:
What the conventional analysis tells you, I believe, is the average SWR over a long period of time.

it is certainly confusing and seems to say that you believe the study performed some kind of averaging to get the number.
"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
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

it is certainly confusing and seems to say that you believe the study performed some kind of averaging to get the number.

I do not want to leave anyone with that impression. The people who did the studies did exactly what they said they did. I have no doubts about that.

What I am saying is that the number that they came up with is not the number that they thought they were going to come up with. People are putting a great deal of emphasis on the fact that the 4 percent number survived in all the historical periods examined. It's a true statement, but not a terribly significant one for purposes of SWR analysis.

It appears to me (I am not certain) that the 4 percent number is roughly the average of all SWRs for the historical periods examined. I believe that that is the question that is being addressed when you employ the conventional methodology.

There's nothing wrong with answering that question. If someone wants to put out a study called "An Investigation Into Determining the Average of All SWRs for the Past 130 Years," I'm OK with that.

I am objecting to the idea of saying that that methodology can reveal to you the SWR for any particular year. It can't do that. It is not designed to do it. It does not produce an accurate answer to that question.

If you want to know the SWR, then you need to make use of a methodology designed to tell you what the SWR is. That means that your methodology must factor in the effect of changes in valuation. There is no other way to figure it out other than to look at all the data that bears on the question being posed.

Telling me that 4 is the number that popped up when you used an invalid methodology does not impress me. Use a valid methodology and tell me what pops up then. "4" may be the answer to some other question. It is not the answer to the question "What is the SWR?" Only by sheer coincidence would the answer to that question happen to turn out to be "4."
Post Reply