Safe withdrawal ideas, thoughts and provocations

Financial Independence/Retire Early -- Learn How!
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

IOW, you have no idea how Bernstein got this number but you are willing to accept his "reasonable way" as fulfilling your "data based" swr objective.

What Bernstein says in the book is that he used the Gordon Equation. That's a data-based tool. I am willing to take him at his word on this point until such time as someone puts forward some reason why I should not be willing to do so.

Bernstein says on Page 54 that the Gordon Equation "provides an accurate way to predict long-term stock returns" and that the Gordon Equation is "as close to being a physical law, like gravity or planetary motion, as we will ever encounter in finance."

In contrast, he characterizes the results of the historical return sequence methodology standing alone as "sometimes misleading."
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

Post by therealchips »

Hocus said
When you are purporting to rely on data to help you analyze a question, you must look at all the data that bears on that question to have any hope of generating an accurate assessment of it.

As I see it, this statement is too strong. One might very well reach a correct conclusion even after ignoring some of the data. For example, the errors that come from omitting some of the data might cancel each other out, or the data that is supposedly relevant may in fact be irrelevant. Granted, it is wiser to use all the data, but the fact that someone has omitted some of it is no guarantee that his conclusions are incorrect and everyone has to make a judicious -- but possibly mistaken -- determination of which data are relevant. :wink:
How about this for a possible rephrasing?
When you are purporting to rely on data to help you analyze a question, you must look at all the data that bears on that question; this maximizes your chances of generating an accurate assessment of it.

Am I nitpicking? I don't think so. My purpose is to encourage calmer rhetoric.
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

Chips
peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Post by peteyperson »

Chips, I have to totally disagree with you here.

Lucking into the right result by excluding some data and still coming up okay isn't a good solution. When relying on the result to plan how much you need invested before FIRE in order to supply your annual spending budget, you need to know that the methodology is accurate.

Petey


therealchips wrote: One might very well reach a correct conclusion even after ignoring some of the data. For example, the errors that come from omitting some of the data might cancel each other out, or the data that is supposedly relevant may in fact be irrelevant. Granted, it is wiser to use all the data, but the fact that someone has omitted some of it is no guarantee that his conclusions are incorrect and everyone has to make a judicious -- but possibly mistaken -- determination of which data are relevant. :wink:
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

My purpose is to encourage calmer rhetoric.

I understand, Chips.

I have a purpose too, and I am deliberate in my choice of words and phrases in this matter. I am saying that the conventional SWR methodology is invalid. I am not saying that it is "inferior" or "not my cup of tea" or "not really so hot." I am intentionally using strong rhetoric because I want people to know how serious the flaws that we are talking about here really are in my estimation.

I am challenging the board community to argue why the claim of invalidity is not on the mark. You can't just say "that's a strong word." I am perfectly well aware that it's a strong word. But "invalid" is a real word, and there must be circumstances in which it really does apply. I am saying that we are facing such circumstances in the matter before us today.

All others are permitted to hold me to what I say, of course. That's the way the game works. When I say that the conventional methodology is "invalid," I better know what the heck I am talking about, or I am asking for a lot of grief.

I am confident that I know what I am talking about in this matter. I am not pointing to some sort of rounding error problem. In the year 2000, the conventional methodology said that a retiree seeking a $60,000 annual withdrawal needed a stash of $1.5 million. The historical data says $3.0 million. That's a difference of $1.5 million. 1.5 million! This is a big, big error. This is huge. If the methodology problem is not fixed, there are going to be a lot of early retirees paying a big, big price for relying on SWR analysis in the structuring of their retirement plans.

There are people registered to this site who are facing a potentially troubled future because of their reliance on the conventional SWR methodology. Andrew61 is one that comes to mind. He retired in 2000 on a plan without a lot of slack. Should be be taking steps to return to the workforce today? Probably he should be, if what I am saying is on the mark. We really need to know for sure so that someone like Andrew61 doesn't get himself in any deeper of a financial mess due to his reliance on what studies posted on internet discussion boards tell him about the historical data, don't you think?

Andrew61 is not a cartoon characrter. He is a real person. I believe that we have an oblgiation to the Andrew61's of the world to try to get this right. People say all the time "it's only a discussion board, what difference does it make?" What if Andrew61 has kids and what if they don't get to go to college because the FIRE discussion boards gave bad advice on the SWR question? Is that no big deal? I say it is a very big deal indeed.

And we are not in this matter talking about someone offering a crazy personal opinion on the internet. People offer crazy personal opinions on the internet all the time. We are talking here about presumed :"experts" claiming to have looked at historical data and claiming to report that data to the public. If they got it all wrong, does that matter? I say yes, it matters big time.

I've been at this for 14 months now. I was tentative at the start. The post I put up on May 13, 2002, was put forward in the form of a question. I was saying to the REHP board, Do you think it would be a good idea to consider valuation levels when calculating the SWR? But 14 months have passed now, and no one has laid a glove on me on this issue yet. There are quite a few who would very much like to do so. That tells me something. It tells me that they haven't been able to find any holes in my argument.

It still might happen. I was 90 percent sure of what I was saying when I put up the first post, and I am 99 percent sure now. There's still a 1 percent chance that someone will come up with a flaw in my case. I think it's a good thing if that happens. Everyone learns if that happens.

But what if it doesn't? What if we go another 14 months and no one lays a glove on me in that time either? What does that tell you?

I think it tells you some important stuff. I think there are some implications that follow from that. I will be posting about what I think some of them are early next year. I don't think now is the time. But I believe there are big implications that follow from us reaching a consensus that the conventional SWR methodology is in fact invalid.

I believe that is where we are going to end up. Heaven knows we took the long road. I would never have chosen the road we took, but perhaps it will work out for the best. The important thing to me is that we end up with the right answer and that we are absolutely sure of it before we take the further steps that I believe we eventually will want to take. We need to be sure that we have thought of everything and it is as clear as clear can be that that conventional SWR methodology really really really is invalid before we proceed to Next Steps.

So I don't want to be coy in the way I state my claims. I want to be clear. I want you to know just how flawed I think that conventional methodology is. I think it is terribly, terribly dangerous. It hasn't been around all that long a time, and look at the damage that the conventional SWR methodology has already caused.

The other side of the story is that, properly performed, I see SWR analysis as being a tool of great power in helping people to make more effective investing decisions. I see great value in doing it right and great damage in doing it wrong. So it's important that we figure out together what is involved in doing it right.

"The methodology is invalid" is a strongly worded claim to put forward. I know it. And I stick by my words. In this case, strongly worded claims are justified.

The results generated by the conventional methodology are wrong. People should stop putting forward claims based on it. Today would be a good day for them to stop.
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

Post by therealchips »

Thanks for the response, Petey. I knew you disagreed with me, as I disagree with you, and I appreciate the mildness and directness of your response. Whenever you mention "intrinsic value," you lose me. I cannot see that efficient market theory is so defective that any of us individual investors can know the value of stocks better than the combined judgement of the many people who are studying, analyzing and, most importantly, investing in the market and actually determining current prices. Do you remember the story of Four Pillars about expert opinion of where a missing submarine was? Expert opinion was widely diverse, but the average of their opinions had only a tiny error in it.

Here are some of the factors of data and policy that I think are relevant to SWR analysis which seem to be missing entirely from the discussion (except perhaps for my raising these issues from time to time):

The effects of the new 15% tax rate on bolstering the stock market and on capital gains and dividend income and, in general, the effects of income taxes on after-tax standard of living

The extent to which the retired person's assets are in tax-sheltered accounts and his plans, if any, for incurring taxes on optional withdrawals

The basis of the retired person's assets in the taxable accounts which, of course, is necessary for computing capital gains taxes

Whether the retired person uses some utility function for money other than the simple linear one, so that he might actually plan on a declining standard of living in advanced old age as a trade-off against the probablility that he will die before achieving that age.

Personal life expectancy and the extent, if any, that someone's planning horizon should exceed his life expectancy

These factors may not strike you as relevant to SWR analysis, but you may agree with me that they are relevant to retirement planning in a broader context. My question is this: How confident are you that you can "luck into the right result" while excluding such vital personal considerations as taxes, life expectancy and the utility of money?

By the way, and for your eyes only, the word is "accumulation", where in my dialect, the third syllable is pronounced just like "you". :)
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

Chips
peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Post by peteyperson »

Hi Chips,

I have Four Pillars arriving Friday in the mail.

The intrinsic value on a P/E basis is in fact the average P/E ratio over 200 years of investing. So it in fact works along the lines you are happy to work within.

I don't myself believe in Efficient Market Theory. EMT usually is associated with the idea that all stocks are fairly valued all of the time and that it is not, on average, statistically likely (or even possible) to out perform the market. With people like Peter Lynch, Warren Buffett and others, clearly that is not the case. Also companies like Dell and Microsoft prove that not all companies revert to the mean performance over time. Bogle tries to suggest this in his book which I reject. He uses the Fidelity Magellan fund as an example of an outperformer that eventually reverted to the mean, however this ignores that it did so because Lynch retired and there are only a handful of people worldwide who can invest as well (most other mutual fund manager cannot outpace the market even before investment costs) and also the size of the fund made it more and more unlikely to overcome the restrictions of size. So I actually agree that EMT is not something I buy into at all. I do buy into the long term average performance of the market which has shown remarkable consistency over the decades. This suggests that business can only grow so fast similarly to the GNP. Within this average however, there are companies that do better and companies that do worse, which delivers that consistent average.

I agree with your latter points of the number of other issues which are relevant to financial planning.

Petey



therealchips wrote: Do you remember the story of Four Pillars about expert opinion of where a missing submarine was? Expert opinion was widely diverse, but the average of their opinions had only a tiny error in it.
peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Post by peteyperson »

Hocus,

Are you saying that the withdrawal rate is 2% and not 4% in general or are you saying that like our previous discussion it varies depending on the level of overpricing to the mean P/E? For example: a stock portfolio priced at $200,000 sitting on a P/E of 28.2 (2x the mean average). This is 2% because you can get the same US$ value withdrawal that someone sitting on $100,000 on a P/E of 14.1 who would take a 4% withdrawal but get the same US$ value withdrawal.

Or is your post indicating that returns are more like 2% w/d rate irrespective of valuation, they are half what intercst says at any given time on the basis that stock market investing doesn't deliver anything like the return that people have been led to believe. That would be an alternative interpretation of what you've said.

I'm not trying to muddle the issue, I'm trying to help you add clarity to your arguement by spelling it out using real $ examples, adding where valuation came into the calculation and where it didn't. Perhaps selecting two different valuation levels and giving real $ examples for each, so we can all see a comparitive example. That might help get your point across better and we might all benefit.
hocus wrote: I am confident that I know what I am talking about in this matter. I am not pointing to some sort of rounding error problem. In the year 2000, the conventional methodology said that a retiree seeking a $60,000 annual withdrawal needed a stash of $1.5 million. The historical data says $3.0 million. That's a difference of $1.5 million. 1.5 million! This is a big, big error. This is huge. If the methodology problem is not fixed, there are going to be a lot of early retirees paying a big, big price for relying on SWR analysis in the structuring of their retirement plans. I've been at this for 14 months now. I was tentative at the start.

The post I put up on May 13, 2002, was put forward in the form of a question. I was saying to the REHP board, Do you think it would be a good idea to consider valuation levels when calculating the SWR? But 14 months have passed now, and no one has laid a glove on me on this issue yet. There are quite a few who would very much like to do so. That tells me something. It tells me that they haven't been able to find any holes in my argument.
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

Are you saying that the withdrawal rate is 2% and not 4% in general or are you saying that like our previous discussion it varies depending on the level of overpricing

It varies according to the level of overvalution. The 2 percent number is the number that applied to a high-stock portfolio at the top of the recent bull market, according to Bernstein.
peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Post by peteyperson »

Ah. That makes sense then. Thank you for the clarification.

Petey


hocus wrote: Are you saying that the withdrawal rate is 2% and not 4% in general or are you saying that like our previous discussion it varies depending on the level of overpricing

It varies according to the level of overvalution. The 2 percent number is the number that applied to a high-stock portfolio at the top of the recent bull market, according to Bernstein.
raddr
*** Veteran
Posts: 265
Joined: Mon Nov 25, 2002 3:25 am
Contact:

Post by raddr »

hocus wrote: Are you saying that the withdrawal rate is 2% and not 4% in general or are you saying that like our previous discussion it varies depending on the level of overpricing

It varies according to the level of overvalution. The 2 percent number is the number that applied to a high-stock portfolio at the top of the recent bull market, according to Bernstein.


If you are talking about a TSM-dominated portfolio then yes, numerous data (including my simulations) would indicate a 2-2.5% number. However, if you are properly diversified and limit or preferably exclude the still frothy TSM stocks then you may be looking at a 4-4.5% SWR. Big difference. :D
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

Hocus:
What Bernstein says in the book is that he used the Gordon Equation. That's a data-based tool. I am willing to take him at his word on this point until such time as someone puts forward some reason why I should not be willing to do so.

Bernstein says on Page 54 that the Gordon Equation "provides an accurate way to predict long-term stock returns" and that the Gordon Equation is "as close to being a physical law, like gravity or planetary motion, as we will ever encounter in finance."

In contrast, he characterizes the results of the historical return sequence methodology standing alone as "sometimes misleading."



from jwr1945:
See the last sentence before the first new paragraph on page 234 of The Four Pillars of Investing. It states: Quote:
Recall from Chapter 2 that it's likely that future real stock returns will be in the 3.5% range, which means that current retirees may not be entirely safe withdrawing more than 2% of the real starting values of their portfolios per year!
http://www.nofeeboards.com/boards/viewt ... rdon#p7294



mathematical certitude? :)
SWR analysis, when it is true to what the data says, is different.
What the data says is not a matter of opinion. Data analysis involves
numbers. Data is hard, objective. You add up all the numbers that bear on the question being examined and you get a right answer to the question posed. That I find more helpful than someone's opinion of whether some investing idea is "good" or not.


as if the future growth rate is known. The Gordon equation does not give a swr.

It remains that Bernstein says that a historical swr "may sometimes be misleading" but you know this for a fact. Bernsteins guesses at a future swr but you know it
Last edited by ataloss on Thu Jul 17, 2003 4:32 pm, edited 1 time in total.
Have fun.

Ataloss
[KenM]
*** Veteran
Posts: 133
Joined: Tue Mar 04, 2003 12:54 am

Post by [KenM] »

To appreciate your point, I would need to understand what calculation has been done to generate the 2% number from the inflated figures



Ataloss
I definitely agree
I'm surprised Bob hasn't made a few judicious comments yet.
I've managed to restrain myself so far ......... but I see from the title of the thread that we're allowed to be provocative .......
KenM
Never try to teach a pig to sing. It wastes your time and annoys the pig.
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

If you are talking about a TSM-dominated portfolio then yes, numerous data (including my simulations) would indicate a 2-2.5% number. However, if you are properly diversified and limit or preferably exclude the still frothy TSM stocks then you may be looking at a 4-4.5% SWR. Big difference.

I of course understand that the SWR changes when you change your allocation.

I'd like to point out the implications of a word choice you made, raddr, because I think that choice goes to the question of why ataloss and some others are having such a hard time understanding what is going on here.

You say that numerous data "would indicate" a number between 2 percent and 2.5 percent. Why "would indicate?" The number is the result of calculations performed on data, is it not? Why not say "the data tells you" or "the data reveals" or "according to the data, the number is such and such?" Why the weasel phrase "would indicate?"

I think it is because you are concerned that someone might look at that number and mistakenly take it to be a prediction of an actual rate of return. You and I know it is not that. A SWR is not an actual return. But you and I are the sort of people who do not like the advice we use to be put to purposes in which it will do others harm. So we are cautious in how we say things, trying to anticipate mistakes that others might make and to do what we can to help others not fall into such traps.

I don't fault you for using the phrase you did. I understand the motivation. But you know what is happening in this debate? The difference in the levels of caution employed by some of those who use the conventional methodology and the level of caution used by those who use valid methodologies is causing some of the people trying to follow what is being said to form mistaken impressions about which side relies on data to inform their SWR calculations and which side does not.

There is one user of the conventional methodology who exercises none of the sort of caution that you evidenced in your statement above. This individual states things in dogmatic terms, the SWR is such and such, according to this individual, and it can never be anything but that. A lot of people have been taken in by these strongly worded assertions. A lot of people have come to believe that someone stating things so boldly and so clearly and so confidently must have some idea what he is talking about. You and I know that he does not. Others haven't followed developments so closely. Those of us trying to put forward accurate information on what the SWR is need to take this reality into account when speaking to this subject to a greater extent than we have in the past, I believe.

Look at what happened to Bernstein. He is another person who I would describe as being on the cautious side. He didn't say that "the SWR was 2 percent at the top of the bull market." He used a weasel phrase, like you did. He said it might have been as low as that. Was he right to put it that way? Arguably not. He used data to come up with the number. The number was produced as the result of a mathematical calculation. Perhaps he should have said the SWR was 2 percent.

I know why he worded things the way he did, and so do you. The two of us have struggled with what is involved with speaking accurately on this issue, and we are aware of the damage that can be done by speaking carelessly. So I can not find it in my heart to find fault with Bernstein for his caution. He doesn't want to mislead, he doesn't want to express greater confidence in the SWR tool than he believes it deserves. He knows what the data says, but he also feels a desire to be fair in how confidently he states things.

Look at what is said of him as a result of his decision to exercise this caution, however. It is said that he pulls numbers out of hats, that he guesses, that he has no idea what the SWR really is. Use invalid methodologies but state things in dogmatic terms, and the world will beat a path to your door. Use valid methodologies, but exercise due caution in your wording, and people will accuse you of playing guessing games and pulling rabbits out of hats.

I think we have been unfair to Bernstein. I think he was reasonable in the way he said things, even if not 100 percent accurate. Obviously, his calculations produced an exact number, so perhaps he shouldn't have used a phrase such as "as low as." But he did it for a good reason. And that's in big contrast to those who use the conventional methodology knowing it to be invalid and put forward their claims in the most confident manner possible.

The problem we face is that thoe who are aiming to work a con are willing to say absolutely anything to score points while those who are trying to impart informed and honest advice are cautious in their pronouncements. It would help, I think, if people parsing the words would bear in mind the motive of the people putting forward the words. When you say "would indicate," that should not be taken as a sign that you did not use data to come to your answer. And when others say "the SWR is" that should not be taken as evidence that the person saying it knows what he is talking about.

It is possible for people using invalid methodologies to use extremely confident language, and for those using valid ones to exercise caution. In fact, when you think about it a little, that's sort of what you would expect to see happen. Perhaps Bernstein and you should be given extra points for stating things carefully and we should detract points from the individuals who assert so confidently that they know the precise number even though they have made clear that they can't be bothered to incluide all of the data bearing on the question being examined in their analyses.

Personally, I am weary of the word games. I have sat at my computer for 14 months now, responding to question after question after question on this matter in an attempt to be certain that there is not one sincere one that I failed to respond to. But things have gone on too long at this point for me to believe that all of the questions being put forward are sincere ones.

William Bernstein does not play guessing games, and he does not pull rabbits out of hats. He is a serious person when the circumstances require that he be serious. There are some others who have put forward "studies" of SWRs for whom I cannot say the same. There are some others who are actively and deliberately working a con on the FIRE community.

I am reaching a point where I am not going to play this game anymore of pretending that a William Bernstein and an intercst are of the same stature and that the views they express on SWRs should be given equal weight. Bernstein is a responsible, honest, and informed analyst. intercst is a con man. When intercst says that it is "loony" for anyone to think tha the number might ever be anything less than 4 percent and Bernstein says that it could be as low as 2 percent, I have no doubt in my mind from that point forward that the true number is a whole lot closer to 2 than it is to 4.

I am looking for a community of people interested in engaging in serious and reasoned discussions of the SWR issue. I am not into semantics games, I don't have the time. If a board can be set up here where those interested in sharing with each other honest and informed insights on how to achieve FIRE early in life, I'm in. If I need to spend hours of each of my days batting back and forth nonsense claims that William Bernstein is puling rabbits out of hats, I have better things to do with my time.

Is it permissible at this site to set up a board where only those interested in engaging in reasoned discussions on a question are permitted to participate and others can be invited out of the discussions? If it is, I will shortly be setting up a board called "SWR Research Group." Those who accept that the convetional methodology is invalid will be welcome to partipate. Those who have possess desires to engage in endless semantics games can use other boards at this site or elsewhere to serve their internet posting needs.

I believe that this issue is the most important issue facing the FIRE community at this time. I believe that it must be resolved. I believe it will be. Whether that happens at this particular site or not depends on how people participating at the board feel about the prospect of going about the business of doing what it takes to resolve it in a manner that has any hope of doing any good for the aspiring early retirees who will live with the consequences of what we decide on this matter.
raddr
*** Veteran
Posts: 265
Joined: Mon Nov 25, 2002 3:25 am
Contact:

Post by raddr »

Hi hocus,
Why not say "the data tells you" or "the data reveals" or "according to the data, the number is such and such?" Why the weasel phrase "would indicate?"


I don't believe that you can say "the number is such and such" - at least not with any reasonable degree of precision. Frankly, know one knows with great precision that future market returns will be so there is some uncertainty as to the SWR going forward. Do you know precisely what that number is?
Look at what happened to Bernstein. He is another person who I would describe as being on the cautious side. He didn't say that "the SWR was 2 percent at the top of the bull market." He used a weasel phrase, like you did.


Well at least I'm in good company. I'll take Bernstein as fellow weasel anytime - LOL! :lol:
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

I don't believe that you can say "the number is such and such" - at least not with any reasonable degree of precision. Frankly, no one knows with great precision that future market returns will be so there is some uncertainty as to the SWR going forward. Do you know precisely what that number is?

This aspect of the matter trips people up more than any other.

The key here is making a distinction between the concept of a safe withdrawal rate (SWR) and the concept of an actual rate of return. There is no one who can tell you the actual rate of return that you will obtain on your investments. That's clear. So we should put that question to the side and forget about it.

A SWR analysis DOES NOT aim to make predictions as to the actual rate of return. I repeat, it DOES NOT aim to do this. People need to get that idea out of their heads if they hope to make sense of this stuff.

What does a SWR analysis do? It assesses the probabilities of various possible future events, looking to available historical data to do so.

I think that it is fair to compare what a SWR analysis does to what a weather report does. Let's say that a weather analysis uses a change in wind velocity to determine the chances of it raining tomorrow. The weather-man does the analysis and he reports that there is an 80 percent chance of rain tomorrow. The following day, it does not rain. Was the analysis invalid?

It was not necessarily invalid. The weather man was not making guesses as to what would actually happen, he was using data to assess the probabilities of various future possibilities. The fact that the one-in-five shot is the one that happened to turn up does not in any way, shape, or form show that the analytical technique used was invalid. It is not reasonable to expect that a tool that aims to reveal to you that there is an 80 percent chance of it raining to be able to predict the future. The tool is not designed to predict the future. That's not something it does.

It is possible for a weather analysis to be invalid, however, just as it is possible for a SWR analysis to be invalid. Say that the data on wind velocity shows that the wind velocity level we have today indicates that the odds of it raining tomorrow are 80 percent,but there is one particular weather man who doesn't like the idea of examining wind velocity when making his weather reports. That weather man says that he is 100 percent confident that it will be sunny tomorrow based on an examination he did of cloud patterns.

Now it may be that cloud patterns can be used to help you assess whether it will rain tomorrow. However, if it has been establshed beyond any reasonable doubt that wind velocity also plays a role, then a weather analysis that does not consider the wind velocity factor is an invalid analysis. That weather report is invalid.

It doesn't matter whether it ends up raining the tomorrow or not. If it rainsa, the analysis was invalid. If it does not rain, the analysis was invalid. The invalidty is not the result of what happens in the future, since the tool does not aim to predict the future in the first place. The invalidity is the result of failing to include factors that bear on the question in the analysis that was done. The analysis was invalid at the time it was done, not at some later data when we know whether it rained or not.

Any SWR analysis that does not consider the effect of changes in valuation is invalid, in my view. It doesn't matter whether the take-out numbers they recommend end up working or not. The tool is not supposed to predict the future, so it should not be required to do that in order to be held valid. The tool aims to assess the probabilities of various future possible events by looking to historical data relevant to the question being examined. If it fails to do that, it is invalid.

The reasons why the conventional SWR methodology is invalid is because it does not even take into account the effect of changes in valuation levels. Nothing that ever happens in the future will ever make such a methodology valid. Nothing. Such an analysis is invalid on the day it was performed, and will always be invalid from that day forward. An analysis that fails to consider factors with a critical bearing on the question can never be valid.

Studies using the conventional SWR methodology are just flat-out wrong. They do not even try to answer the question posed. It is not even trying to deliberately exclude data that is known as a matter of mathematical certainty to affect the answer to the question being posed.

Well at least I'm in good company. I'll take Bernstein as fellow weasel anytime - LOL!

It's the other way around. When it comes to SWRs, he is in your company. You set the standard of excellence in this field, raddr. There is no one on Planet Earth who understands SWRs better than you, with the possible exception of JWR1945. You and JWR1945 stand at the top of the pyramid, and Bernstein stands very near the top.

I say this not just because of your numbers abilities. It's also because you had the guts to take intercst on in Janaury 2002, a few months before I kicked off The Great Debate. The FIRE history books will make note that you planted the flag first. I have been an advocate for honest and informed posting on this issue for a long time now, but it takes something special to do it first, and it wasn't me who possessed what it takes to do that, it was you. I noticed.

Wanderer was first re a different aspect of the question, by the way. I was second on both aspects, the substance aspect and the process aspect.

Where I come in first is on the stubbornness aspect. I think it is fair to say that no one comes close to me in that department. I am the most stubborn advocate of honesty in SWR analysis on the face of Planet Earth at this time. We all make our contributions in the areas in which we are most able to give, you know?
peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Post by peteyperson »

Hi hocus,

A couple of long posts there from you.

I think your weather analogy for SWR predictions was as master stroke. We should use that in the future. It is a very visual way of explaining the purpose and limitations. It will prove especially useful to new visitors who wish to join the discussion and get up to speed quickly.

I agree that one would want to show caution when talking of a swr. You can only say that based on the past data, if we sustained no worse periods in history or if you retired in any of the last 100 years and funded retirement for x number of years from that start date, the worst scenario would have been y. It is then up to the individual once they are happy the methodology used to perform the calculation and the data were both valid and accurate, to decide how best to proceed.

To extend the point out some, if the next several decades produce 3.5% real return in stocks instead of the historical 7%, then the figure spit out would be too optimistic. It then becomes a hard job to refactor in a different rate of return over the top of the historical data. Do you cut every yearly return by 3.5% to reflect the drop in return while keeping the historical ebbs and flows of the market in tact? Is that a valid way to analyse that. Perhaps a statistician would know the answer there. I find this an interesting question. Being able to ape the past market conditions but lower the return on stocks would provide meaningful data for those who believe the market will not continue to produce 7% real or who simply wish to be more cautious in the financial planning.

I am more than satisfied that a high valuation lowers the w/d rate (of the annual budget + inflation) being that it reduces the percentage of a stock market investment that needs to be withdrawn to obtain the same $ amount + inflation each year. I am unclear what data intercst has left out which makes his swr calculations invalid, outside of perhaps not accounting for valuation issues. Care to expand on that?

Lastly, I don't think it is necessary that everyone agree on the right swr analysis or data. If a group of people between themselves manage to advance the discussion sufficient to derive lasting results, then that is enough. If the explanation covering the group viewpoint is clear and logical and mathematically accurate, it won't matter if others choose a different perspective and use other figures. What matters is who is right. I feel that we are making that progress and that we are on the right path to establish accurate definitions for key swr issues, valid data points and the best methodology.

Petey



hocus wrote: The key here is making a distinction between the concept of a safe withdrawal rate (SWR) and the concept of an actual rate of return. There is no one who can tell you the actual rate of return that you will obtain on your investments. That's clear. So we should put that question to the side and forget about it.

A SWR analysis DOES NOT aim to make predictions as to the actual rate of return. I repeat, it DOES NOT aim to do this. People need to get that idea out of their heads if they hope to make sense of this stuff.

What does a SWR analysis do? It assesses the probabilities of various possible future events, looking to available historical data to do so.

I think that it is fair to compare what a SWR analysis does to what a weather report does. Let's say that a weather analysis uses a change in wind velocity to determine the chances of it raining tomorrow. The weather-man does the analysis and he reports that there is an 80 percent chance of rain tomorrow. The following day, it does not rain. Was the analysis invalid?
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

I don't think it is necessary that everyone agree on the right swr analysis or data.

There are many aspects of the question on which there are reasonable differences of opinion. For example, Bernstein has one way of incorporating the valuation factor into his analysis, and he comes up with an SWR of 2 percent in the year 2000. JWR1945 favors an alternate approach and he comes up with 2.3 percent. Both approaches are valid. It's legitimate for us to have discussions as to which is better, but there is no need for us to achieve a consensus on this point.

I think we do need to reach consensus on whether the conventional SWR methodology is valid or not. Use of that methodology has caused losses to aspiring early retires in the millions of dollars. I will put up a post someday providing links to posts at the Motley Fool board where people make clear that claims made based on that methodology have influenced their investing decisions. We now know that what the conventional methodology says is safe is in some circumstances the opposite of what the historical data says is safe. We need to get the word out on that point.

The whole point of a board like this is to help human beings learn how to achieve financial independence early in life. You don't do that by misleading people as to what the historical data says on the question of what is safe. There are a number of things that I think we can do to get the word out on the invalidity of the conventional methodology. First, however, we have to agree that it in fact is invalid.

We are not there yet, but we are getting there. Slow and steady wins the race.
galagan
* Rookie
Posts: 5
Joined: Thu Mar 20, 2003 3:00 am
Location: Billings, MT

Post by galagan »

hocus wrote: Now it may be that cloud patterns can be used to help you assess whether it will rain tomorrow. However, if it has been establshed beyond any reasonable doubt that wind velocity also plays a role, then a weather analysis that does not consider the wind velocity factor is an invalid analysis. That weather report is invalid.


In my opinion, your emphasis on theoretical correctness is commendable. But keep in mind that many people are results-driven and may be comfortable sacrificing accuracy for simplicity.

It's important to remember that although choosing not to consider a factor may make one's analysis "invalid," it's helpful to know just how important that factor is in coming up with the end result. For instance, I've been brushing up on physics lately, and there is some discussion about the force of gravity. Depending on one's elevation and latitude/longitude on Earth, the force of gravity can differ by a measurable amount. However, for most applications, using a standard figure for gravity gets you close enough to the real answer for practical purposes. Users who focus on those practical applications may not care about the theoretical invalidity of the method they used to get the answer, as long as the error is measurable and comfortably low. However, there are other users who must have the increased accuracy of a method that takes all factors into account.

I'm pleased about this board because people here are evaluating those other factors and trying to determine how important they are. Once their research is sufficient to draw conclusions, one will be able to see how much considering other factors refines the analysis and gets you a more accurate result.

dan
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

Post by therealchips »

Hi, Petey,

You said
I have Four Pillars arriving Friday in the mail.

The intrinsic value on a P/E basis is in fact the average P/E ratio over 200 years of investing. So it in fact works along the lines you are happy to work within.

May I quote Bernstein and Malkiel without anyone supposing that I think their authority, or any authority, gives the final answer to important questions? My purpose here is to show that my doubts about the notion of intrinsic value come from widely respected men and not just my own investment experience. Page 55, middle of the page, Four Pillars of Investing has this (where the brief quotation does not, I think, distort the context of the original and readers will judge the matter for themselves) :
The discounted dividend model is a powerful way of understanding stock and bond behavior. As we've seen, it isn't of much use in accurately predicting the fair value of the market, let alone a stock. Princeton economist Burton Malkiel famously stated that "God Almighty himself does not know the proper price-earning multiple for a common stock." In other words, it is impossible to know the intrinsic value of a stock or the market.

(I can't find that "famous" statement on the internet. Too bad.) Bernstein then goes on to say that the discounted dividend model gives us the Gordon equation which is useful for estimating the long term expected return of the market. He puts that estimate at 6.5% as of the publication date, 2002, and seems to mean something like "thirty years" as long term.

Back to the point: The idea of intrinsic value of a stock or a market is dubious to many people. As you disagree with Bogle, I predict you will disagree with Bernstein, too.

Beside the point: I won't show up at hocus' forum for those who reject "conventional methodology", except in the totally unexpected case that someone invites me there. This is not because I embrace the conventional methodology and its 4% withdrawal rate, but only because I'm still asking questions that hocus is tired of answering.

Let me tell you how seriously I dissent from the 4% canon. My planned withdrawal rates are 1.5% in my sixties, 1.6% in my seventies, 1.8% in my eighties, and 1.9% in my nineties, with a real expected return in the market of 2% for the next thirty odd years, constant after-tax inflation-adjusted standard of living, and fairly flat purchasing power of the total retirement stash over the duration of the plan. I think that 2% assumption is absurdly pessimistic for such a long time period, even with today's high price-earnings ratios, but I'm being cautious. Bernstein's 6.5% is more plausible. If he turns out to be correct, I'll let my standard of living float up a bit over the years, or maybe I'll just leave that much more in the estate for the next generation. I never bought into the canonical notions of using up the stash during retirement or of freezing the real withdrawals regardless of market activity.
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 »

I'm pleased about this board because....

It's good to hear your voice, Galagan.
Post Reply