I learned of the Scott Burns column linked below from an
Ataloss post on the FIRE board.
Ataloss picked it up from an
RKMcDonald post over at the Motley Fool's Retire Early Home Pafe (REHP) board.
http://money.msn.com/articles/invest/ex ... pecial=msn
Excerpt:
Their findings, in a nutshell [Burns is referring here to the research paper linked below]: Stock returns from periods of high P/E ratios are lower than returns from periods of low P/E ratios. Stock returns are still higher, however, than competing returns on cash or bonds. In other words, stocks are less attractive as investments today, but they are still better than the common alternatives.
Here is a link to the research paper itself.
http://www.fpanet.org/journal/articles/ ... -art10.cfm
The paper appeared in the February 2002 issue of the Journal of Financial Planning. The Burns column also is from early 2002.
Here is an
excerpt from the study.
"Our results indicate that the best estimate of future average returns is no longer the long-term average returns on stocks found, for instance, in the Ibbotson's series of 10 to 12 percent a year.
An investment advisor can obtain a more accurate measure of expected returns by making expected returns conditional on the current P/E ratio. (the bolding was added by me)."
This last statement addresses the core dispute of the Great Debate. The important question has never been "Are stocks good or are stocks bad?" or "Is a personal withdrawal rate (PWR) of 4 percent reasonable or not?" The important question has always been, "Is it true, as
intercst asserts, that any move away from a 74 percent S&P stock allocation requires the aspiring early retiree to delay his retirement?"
Intercst says that this is so as a matter of third-grade mathematics, presuming that the future is like the past. And he is right, if the conventional SWR methodology is analytically valid. He applied the methodology properly in his study, his calculations are accurate. If that methodology is a valid means of determining the SWR (as defined in the studies), then
intercst is right to tell aspiring early retirees that it will take longer to achieve a safe retirement by going with alternate portfolio allocations.
If the methodology is invalid, then
intercst is wrong to say this. The excerpt from the study that I have quoted above is strong evidence that
intercst is wrong and that the conventional methodology is invalid. It does not state things in as strong terms as I would. But it makes the key point that it is "more accurate" to take into account changes in valuation levels than to not take such changes into account.
In one respect, the claim being made by the authors of the research paper is
stronger than the claim that I have made that the conventional SWR methodology is invalid. The authors are saying that changes in valuation should be taken into account in
all financial planning (they state that "financial planners can get a better estimate of future stock market average returns by using current market valuation conditions"). I have generally limited my assertions of this type to analysts preparing SWR studies, a particular type of financial planning tool that requires a particular concern re safety (conventional SWR studies purport to reveal the withdrawal percentage that is 100 percent safe, presuming that the future is like the past).
I believe that the authors are correct that all financial advisors need to take into account the message of the historical data, that changes in valuation levels affect future returns as a matter of mathematical certainty. But I also believe that SWR researchers have a special and greater responsibility to get this right. A financial advisor who gives bad advice on long-term stock returns to someone who is not going to retire for years does damage, but it may be possible for the person making use of the bad advice to recover from the effects of the bad advice by saving more in later years. Giving bad advice on SWRs to someone on the verge of handing in a resgination notice can cause the most severe life setbacks.
That said, I believe that the conclusions of this paper point to an aspect of the Great Debate that has been little considered until now but which may take on great importance in days to come. The reality is that it is
not just SWR analysts like
intercst who have misled investors as to what the historical data says about what sorts of long-term returns are likely for those investing in stocks. The basic approach used in the conventional SWR studies is used in all sorts of financial planning articles and products.
The paper states that: "Our results indicate that the best estimate of future average returns is no longer the long-term average returns on stocks found, for instance, in the Ibbotson's series of 10 to 12 percent a year. An investment advisor can obtain a more accurate measure of expected returns by making expected returns conditional on the current P/E ratio." If that is so, then the conventional SWR methodology is flat-out invalid. It is irresponsible to use anything less than the best available estimate of future long-term returns in an analysis of what is 100 percent safe for someone questioning whether he has saved enough to turn in a resignation notice or not. But that is not all that is so. If the claim made by the authors of this paper is correct, then there are all sorts of financial planning tools that need to be modified in the years ahead. SWR stuidies should be first on the list, in my view. But it may be that we can attract posters to this board by making our case for consideration of the effects of changes in valuation levels to a wider audience than those planning early retirements.
Anyone hoping to invest effectively benefits from coming to terms with the questions that have been put on the table in the course of the Great Debate.
This paper is rich in implications, too many for me to examine in one post. I hope to be putting up follow-up posts on this paper at later dates. For now, though, I would like to draw attention to a statement by Scott Burns set forth in the column linked above. Burns has made approving reference to the
intercst study in the past, and I have hopes that persuading Burns to condemn the deceptive claims that intercst has put forward citing his study as support may advance the debate considerably. At a mimimum, I hope to be able to persuade Burns that the conventional methodology is invalid, arguing in part that the uses to which
intercst has put the methodology reveals the danger inherent in it in painful-to-consider real world terms. Given Burns' broad platform and longstanding interest in numbers-rooted financial analysis, I believe that he is probably the best expert to choose for our second Special Event Discussion at the SWR board, one that I hope to schedule after getting my web site up (the target date is May 13, 2004) and bringing some new posters to this board. Burns has great credibility, and a strongly worded statement from him may be what we need to bury the conventional SWR methodology 10 feet in the ground where it can not cause financial losses to FIRE communities of the future.
The following statement from Burns is highly encouraging to me in this regard. He says:
"The differences in return can be dramatic. If you buy stocks in a period of low P/E ratios (under 10), your average annualized return five years later will be 18.57%. Buy stocks in a period of high P/E ratios (over 15), and your five-year return will be about half as much, 9.28%.
More important, the variability of your return will be nearly twice as great as in the low P/E period. (the bolding is mine).
We have seen lots of evidence during the course of the Great Debate that it is returns experienced in the early years of a retirement that determine whether the plan will ultimately go bust or not. Burns is saying here that the valuation level of stocks on the date you retire affects the variability of returns in the years to follow. It seems to me that he is saying that volatility is greater in the early years of retirement for those who retire at times of high valuation.
If that is so, that statement is the final nail in the coffin of the conventional methodology, in my view. If volatility is greater for those retiring at times of high valuation, then the SWR is
different for those retiring at times of high valuation. There is no way that I can imagine that the SWR could possibly be the same for investors retiring at two different valuation levels, if what Burns is saying here (commenting on the linked research paper) is so.
I incorporated this post into the "Pizza Hut University" thread because it sheds light on the question that I was putting on the table in my first post on this thread. Much has been made at the FIRE board of my supposed "arrogance" in being the special one who brought this issue to the attention of the FIRE community (by putting up the May 13, 2002, post that kicked off the Great Debate). The reality is that I have not been arrogant in the typical ways in which that character trait is revealed. I have not claimed any special expertise in investing matters, and I have acknowledged a great la
ck of ability to deal effectively with numbers-related issues. So it is a strange sort of arrogance on display in my posts.
What I have been is forthright in stating that I brought to the FIRE community's attention an issue of earth-shaking importance, that the conclusions of the
intercst study (a study so important in the formation of the FIRE movement that one poster at the REHP board has said that
intercst should be considered the "inventor" of the Retire Early concept) are false claims. The study did not consider a critical factor that must be considered in any analysis of what is safe and therefore it provides the wrong answer to the question it poses, I said. After the claim was met with intense and prolonged challenge and attack, I stood in there and fought the good fight to permit this insight to be explored in depth by the FIRE community, and I continue to do this today.
I am not claiming to be smarter than others or more experienced on financial matters than others. I am claiming to be better informed on the question of SWRs than any poster who has contributed a post on the subject during the course of the Great Debate. I knew that the convetional methodology was invalid in 1995, so I have had eight years now to study the benefits of SWR analyses that provide accurate assessments of what is safe according to the historical data. There is no other poster in the FIRE community who can say that.
To be sure, there are some on this site who see the importance of valuation issues. But it is equally clear that these posters have not made the conceptual leaps necessary to understand the risks of getting the SWR number wrong and the benefits of getting it right.
I will offer one example, with no intent of making any sort of disparaging remark re the poster who will be named, whom I consider a fantastic addition to the posting community here. This particular poster has not engaged in the smears against me, so I hope that no one will make the mistake of thinking that I mention his name here because of some sort of personal hostility that I feel towards him as a result of the smesrs. This poster is a fine poster, and, from all appearnaces, a fine human being as well.
The poster I am referring to is
Gummy. I am not providing an exact quote here, but he has said something to the effect that "I tell my children to plan on a 4 percent withdrawal and not to worry about the math." This is a statement that has been made in a variety of forms by dozens of posters at one time or another in the course of the Great Debate. It is this sort of statement that reveals to me the real damage that has been done by the
intercst claims and by similar claims that have been made by other SWR analysts employing the conventional methodology.
I believe that the
Gummy advice to not worry about the math re SWRs is bad advice. That is as simple as I can put it. There are times when taking a 4 percent withdrawal is a good idea. There are other times when those with a 74 percent allocation to S&P stocks need to move their withdrawal down to 2 percent. There are yet other times when those taking a 74 percent S&P allocation can achieve the level of safety that they seek with a withdrawal of 6 percent.
The SWR varies as valuation levels vary. This is a reality proven by a reasoned examination of the historical data. And it is a reality of immense importance to anyone aspiring to achieve financial independence early in life. Knowing
how this number varies from time to time opens up all sorts of doors, doors that have never been opened on any FIRE board to date. These are doors that should be opend, doors that must be opened. The growing desire on the part of middle-class workers to learn how to achieve financial independence early in life will at some time
force those doors open, in my view.
I cannot say when it is going to happen. The forces opposed to any opening of the doors are strong today. But I do see encouraging signs. The Peter Bernstein interview linked earlier in this thread was an encouraging sign. The publication of "The Four Pillars of Investing" was an encouraging sign. The publication of "Yes, You Can TIme the Market!" was an encouraging sign. The publication of the Ed Easterling research, and the comments he offered to us at this board, were encouraging signs.
There are encouraging signs everywhere mixed in with the less than encouraging signs that we here are all too aware of. I think that this thing is going to turn, not tomorrow, not next week, not next month. Perhaps next year. If not then, then certainly by the end of the 2005. The benefits to investors of knowing how the SWR changes from time to time as market valuations change are so great that the desire to learn the realtiies will in time become too great to be denied.
That's what I believe. I don't have a crystal ball, so if you want to find out for sure you will have to just wait and see. I will pursue my efforts with no hostility towards those who continue with the smears. I don't have the time, energy, or desire to get involved in that stuff. What I want to do is to lead this thing to where we can have an atmostphere that permits reasonsed debate on the realities of SWRs. When that happens, I don't think that there will be any saying that the long effort to get there was not worth it. I don't think that there will be any saying "Oh.,
hocus, your pre-SWR posts were better than your SWR posts" then. I don't think there will be any showing any reluctance to thank me for my efforts to make possible a civil disucssion of the realities of SWRs then.
We'll see. In the meantime, I hope that those who have not yet made up their minds one way or the other, those who see at least some possible value in exploring the questions of substance that have been raised in the Great Debate, will take a serious look at the study linked above and give some thought to the implications to your hopes of achieving financial independence early in life. I am here to tell you that I have been studying that question for eight years now, and the implications are huge. Charges of arrogance be damned, I will say it one more time so that there is no possibility that anyone here will not know what I think on the question--The SWR debate is the most important debate that has ever been held at any FIRE message board by a factor of at least 10. The personal attack stuff is an obvious waste of everyone's time. The issues of substance that I tried to bring to the table with the May 13, 2002, post are of great importance, and the entire FIRE community will see that clearly when the smoke clears.
The study linked above is further vinidcation of the point argued in the May 13, 2002 post. It is not possible to determine SWRs accurately without taking changes in valuation into account. The conventional methodology is invalid.
Saying so in public ain't no mistake! The Harvard guys know all sorts of things that I do not. But there are some things that they teach you at Pizza Hut University that those guys from Harvard never heard tell of. I majored in SWRs in my days at Pizza Hut, and I am glad I did; I would be unretired today if I has given any credence to the wild SWR claims made by
intercst. When you are calculating numbers, there is a right and a wrong, and
intercst happened to get it wrong. That's the way it is.
I've got my pizza-smeared notes from the classes I was taking back in the mid-1990s to prove it so.