The Great SWR Investigation - Part 1

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

Post by hocus »

If, for example, a house worth 100K today appreciates at a real rate of 4% then it would be worth about 330K in inflation adjusted dollars 30 years later. Thus a buyer would have to have greater than three times the real annual income to afford the house. This just can't go on for too many decades.

I like that way of looking at the question. It still seems to me that you could reasonably expect better than a zero rate of return, however. Productiivity increases have enhanced the value of our labor, have they not? If that is so, couldn't those dollars find their way into increases in the value of the houses we live in?

It sounds to me that you are making a distinction between "productive" assets (corporate stock) and "non-productive" ones (housing). The distinction makes sense to me. But it seems to me that some sorts of housing would become more valuable as the areas in which that housing exist become more productive. If you buy housing in an area that becomes more productive, the housing value should reflect those productivity gains (because of increasing demand for housing in that area).

Perhaps it is only "productive" housing (productive because it permits people to live near where productive corporate work is being done) that increases in value. But, then, there probably are lots of non-productive corporate entities too. The non-produtive ones go bankrupt and are removed from the field. It is only the productive ones that we hear about.

The argument you are making is one that appeals to me. But it is one that has implications for the returns expectations that are reasonable for other asset classes too. Are the long-term returns of stocks limited to the producitivty gains that result from corporate activity? I think you are suggesting here that they are. If that's so, it means that there is a cap on the level of long-term gains that can be generated by ownership of stocks.

If that is so, that makes the valuation concerns more pressing than most have considered them to be to date. It means that not only is the downside risk inherent in stock ownership higher than we once thought. It means that the upside potential is perhaps more limited than we once thought (though still substantial).

This argument pushes my thinking capacities to their limit. I am not at all trying to shoot down what you are saying. I like the argument you put forward. I am just attempting to tease through some implications that occur to me as part of a quick reaction.

It strikes me as a little odd to think of the capital tied up in residences as dead captial (producing no real return). It may be that I am entirely wrong to find it odd. I possess no expertise on this question. It just isn't what I intuitively thought to be the case.
wanderer
*** Veteran
Posts: 363
Joined: Tue Nov 26, 2002 9:33 am
Location: anytown, usa

Post by wanderer »

However If I add in the comfort and quality of life factors along with all the other hidden benefits of home ownership IMO that changes the return on the investment considerably.

and what are the odds of the Sunset Arms Apartment Complex countenancing a monkey en premises?
regards,

wanderer

The field has eyes / the wood has ears / I will see / be silent and hear
User avatar
FMO
* Rookie
Posts: 30
Joined: Mon Nov 25, 2002 4:00 am

Post by FMO »

hocus:

I am surprised to hear that the data shows this. Perhaps it is because I have lived in the DC metropolitan area for so long. I believe that housing prices in this area have done better than the overall data suggests, at least in my lifetime. That particular experience has probably informed my "sense" on this question.

At some point, I would be interested in views as to the "why" of the finding that ownership in a residence produces such a low real return. Does anyone have thoughts as to why this use of capital produces such low returns, compared to most others?

raddr:

I bet wanderer, FMO, or other RE guru can set us straight here. Is it reasonable to expect long term returns much greater than inflation for residential housing?


First let me say that it is unreasonable to attempt to draw conclusions about the long-term real returns of owner-occupied homes by looking only at appreciation data. Many people would say that a home is not an investment at all since it doesn't produce income. I personally think that a home is an investment, but to fairly evaluate it requires consideration of more than just the rate of appreciation. A home provides utility as well.

According the the Office of Federal Housing Enterprise Oversight (OFHEO) from 1975 to 2002 Washington D.C. area real estate appreciated at a compound rate of about 6.1%. The CPI-U increased at a compound annual rate of about 4.6% over the same period. So it appears that on average, D.C. area real estate outpaced inflation by 1.5% over that 27 year period. To calculate a real total return would require that you consider a host of other factors, including the degree of leverage employed, upkeep expenses and the value assigned to the shelter which an owner-occupied home provides.
FMO

"The mark of a successful man is spending an entire day on the bank of a river without feeling guilty about it."
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

I guess it is the unwarranted extension of the certitude to withdrawal rates that is my problem. Bernstein never said that it was a matter of mathematical certitude that the 4% withdrawal rate was unsafe.

I believe that this question touches on an aspect of the question that is rich in interesting implications.

When I first read the Bernstein book, I thought that the most important section was the discussion on Page 234 in which he says that the safe withdrawal rate at the top of the bull market was 2 percent. That language seemd most important because it directly refers to the subject matter of our concern.

Over time I came to see that the "mathematical certitude" language on page 12 is of greater import. That claim is a really big deal. It should influence our understanding of all sorts of stuff.

The direct Bernstein claim is that valuation levels affect stock market returns as a matter of "mathematical certtitude." That is an extremely strongly worded claim. It means that there are no exceptions. He is not saying that changes in valuation may affect returns. He is saying thay they do affect returns.

Now, changes in returns expectations always affect SWRs, do they not? Can anyone come up with a scenario in which you lower your expectation of future returns, but your SWR remains the same? I cannot imagine such a scenario which did not call for an expectation of the future being different than the past. The James Glassman scenario (a change in the risk premium for stocks) might be an exception to the usual rule, but that is a case where the future turns out different than the past.

If valuation always affects returns, and returns always affect SWRs, then valuation always affects returns. No? I see no escape from the force of that logic.

If all this is so, then it is not just in the year 2000 that the historical sequence model produces the wrong SWR. The historical sequence model always produces the wrong SWR. Because it does not factor in the effect of changes in valuation levels.

To be sure, this does not mean that the SWR is always lower than what is determined by use of the historical sequence model. It means that it is always different. It could in some circumstances be higher.

If the SWR was truly 4 percent in the year 2000, as some say, then it must be significantly higher than that today, say 6 percent. You do not see many on the other board proclaiming that the SWR is 6 percent today, but that is a logical consequence of their claim that it was 4 percent in the year 2000. You can't possess reasonable confidence in one claim without having reasonable confidence in the other as well.

If the SWR is truly 4 percent today, as some claim, then it surely was not 4 percent in the year 2000. It must have been 2 percent or so. The SWR cannot possibly have been the same at the beginning of 2000 and at the beginning of 2003 if what Bernstein is saying about valuation levels affecting returns expectations as a matter of "mathematical certitude" is true.

I think this is a very big deal. I more or less thought this was true all along. But reading Bernstein state it with such clarity increased my confidence level in the idea a great deal. I believe that many implications follow from this which at some point we should explore.

My sense is that even Bernstein is a bit reluctant to fully confront the implications of what he is saying. He see the implications. But he holds back from exploring them at some points. That is why you have people like Telegraph saying that Bernstein says that 4 percent is "OK" even though Bernstein says that 2 percent is the SWR. Telegraph is exagerrating the extent of the contradiction, but he is not wrong to draw our attention to a little bit of a contradiction. I believe that Bernstein himself is still coming to terms with the implications of his statement.

Here is the ultimate implication, in my view, just to let the cat out of the bag. If Bernstein is right, it undermines confidence in the claim from the book "Stocks for the Long Run" that the risk associated with stocks can be avoided by holding them for the long term. If Bernstein is right, that claim is right at some valuation levels and not at others.

A finding that that is so would represent a big change in what most people think about stocks as an asset class. And I firmly believe that that is a fair implication to draw from Bernstein's statement that valuation levels affect returns expectations as a matter of "mathematical certainty."

If the "Stocks for the Long Run" claim generated enough confidence in stocks as a long-term investment to pull valuations up to levels never experienced before, then the success of the book's key idea (that stock risk can be eliminated by a plan to hold for the long term) was itself responsible for creating an investing environment in which the book's key idea was no longer true.

My sense is that many people have a dim sense of this implication of the Bernstein claim, and that the cognitive dissonance that results from the suggestion that the Stocks for the Long Term claim is no longer valid (or at least was no longer valid in the year 2000) is part of the explanation for why The Great Debate has proven to be so controversial. On the surface it is a discussion about a boring number. But the implications are shocking to anyone whose thinking on stocks was developed during the era in which the "Stocks for the Long Run" insight was dominant in most investment allocation analysis.

The Bernstein statement should result in a fundamental change in the way that most people think of the risks attached to investing in stocks, in my view.
hocus
Moderator
Posts: 435
Joined: Mon Dec 02, 2002 12:56 am

Post by hocus »

To calculate a real total return would require that you consider a host of other factors, including the degree of leverage employed, upkeep expenses and the value assigned to the shelter which an owner-occupied home provides.

These things would all need to be considered in deciding whether it is a good idea to purchase a home or not. However, they are not relevant to the question of whether I should assume a 4 percent take-out from the $300,000 that I have "invested" in my home.

I am persuaded that I should at a minimum adjust my plan so that the take-out number from that $300,000 is lowered to 2 percent. I have not yet decided to go lower than that, but I will continue to consider the option of a more sweeping change in that assumption.
User avatar
FMO
* Rookie
Posts: 30
Joined: Mon Nov 25, 2002 4:00 am

Post by FMO »

hocus wrote: To calculate a real total return would require that you consider a host of other factors, including the degree of leverage employed, upkeep expenses and the value assigned to the shelter which an owner-occupied home provides.

These things would all need to be considered in deciding whether it is a good idea to purchase a home or not. However, they are not relevant to the question of whether I should assume a 4 percent take-out from the $300,000 that I have "invested" in my home.

I am persuaded that I should at a minimum adjust my plan so that the take-out number from that $300,000 is lowered to 2 percent. I have not yet decided to go lower than that, but I will continue to consider the option of a more sweeping change in that assumption.


You are probably already achieving a takeout in excess of 4% by simply living in the home, since you are consuming 100% of the "imputed income" attributable to homeownership. If you are talking about "taking out" 4% of the appreciation component only, this is unsustainable in my opinion. Based on the historical data, it appears that 2% is unsustainable as well. Of course, we are talking about averages here and actual results will vary from one location to the next.

If we were to compare an owner-occupied home to investment real estate we will discover that the total return would be less for a number of reasons. The first being that a home purchased to live in is not selected primarily for its investment attributes (ability to generate income). Second, many expenses incurred by the homeowner do not receive the favorable tax tratment afforded to income properties. That is to say that expenses such as maintenance, repairs, insurance, etc. are not deductable. There is no depreciation allowance as well.
FMO

"The mark of a successful man is spending an entire day on the bank of a river without feeling guilty about it."
User avatar
ben
*** Veteran
Posts: 231
Joined: Mon Feb 17, 2003 4:00 am
Location: The world is not enough

Post by ben »

Ataloss; I am a big fan of Bob! He makes more sense to me than you.. :wink: who is that Bob-guy?? :D
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...
therealchips
*** Veteran
Posts: 174
Joined: Sat Jan 04, 2003 4:00 am
Location: Henderson, Nevada, USA

Post by therealchips »

I am glad you brought this up, TheRealChips. This is the way that I use SWR analysis too. I aim to know not the number that will leave me with zero assets at the end of x years, but the number that I can take out and still retain the same level of financial independence.

I see this as a matter of personal preference. In this case, there is nothing invalid about doing the calculation the other way. It is really just a different way of approaching the safety question. If you are depleting your assets over time, you probably need to be more safe in other aspects of your plan. If you are retaining the same level of financial independence, you can probably afford to go with a lower safety percentage for your plan (on the thinking that you have some slack built in).

Also hidden in the original question may be the assumption that the retired person wants the same purchasing power in his withdrawals every year of his retirement. Actually, he may want his withdrawals to go up or down some, or a lot, with the value of his assets.

I update my plan every year. I started . . .


Thanks for your post where you said what I just quoted, Hocus. I am responding to it without reading all the subsequent posts.

Starting with the last point: I update my plan every week. I enjoy doing the calculations (especially when the market is rising :D), which require only entering the values of my various holdings and waiting milliseconds for the computer to come up with the answer. I'm not willing to spend as much as my analysis says I can, so the frequent computation promotes peace of mind rather than indicating its absence.

The capitalist injunction was Thou shalt not touch thy capital!. This accords with the theory that we are only stewards of our capital and will pass it along undiminished to a later generation. That concept seems to be entirely missing from SWR research, so I am surprised and pleased to hear that your approach resembles mine. Has any one else in these boards said this too? Of course, I'm not sure that I can preserve the purchasing power of the assets, although I have done just fine so far. That approach matches an ancient Naval principle: Preserve thy maneuvering room!

I include my IRA and much of the value in my taxable accounts as part of the total value of the assets whose purchasing power I mean to preserve. All of that value is due to money I saved out of my paychecks and to its increase under management that I picked. This amounts to a determination to enhance the value of the assets that came to me by inheritance, and not merely preserve that value. This is all very traditional capitalist thought or theory, so far as I know. The stewardship concept helps me deal with the moral issue of being "rich" when half the world is desperately poor.

Changing the subject to owner-occupied housing:

If you count your house as part of your retirement kitty, then you have to recognize that its yield to you is free rent plus possible capital gains minus the costs of ownership. It's easier just to omit the house on the asset side and omit rent on the expense side of the analysis.

Calculation of Gross National Product includes the fair rental value of owner-occupied housing, less something for depreciation and maintenance and interest, if any. I agree with that approach and its implicit idea that this net rental value is genuine income. In this, I flatly reject the idea in "Rich Dad, Poor Dad" that nothing counts as an asset unless it puts cash in your pocket. Keeping me from having to take cash out of my pocket (for rent) makes my house an asset too. Doing without shelter, or expecting someone else to pay for my shelter, is not part of my planning. Of course, some people buy far more house than they need for shelter (as I have :roll:) and burden themselves with mortgage payments that starve their cash-producing investments (as I have not :D). I have not paid rent since the 70's and have not paid on a mortgage since the 80's. Still, I do not count on any appreciation in my house, but just to let it sit there as part of my final estate. When I calculate how much I can withdraw from my retirement assets, I do not include the house. I simply enjoy living here, and pay all the costs of owning the house out of my withdrawals. If someone plans to sell his house when the kids are gone or when he retires, and either rent or buy some less expensive house to free up some retirement capital, my comments do not apply exactly. In my case, the house I sold in Los Angeles to buy my new house in Henderson did not produce enough money to cover the new place. I treated the difference as a current personal expense like groceries rather than an investment. That is highly conservative, I suppose, but I do not plan to sell this place ever, unless Henderson comes to resemble Los Angeles. Los Angeles is my original home town where my family has lived for a hundred years now. What has happened there is tragic, but that is a subject for another day.
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

Chips
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

ataloss
2. JWR1945 I need some help in understanding something. What do you mean by "best estimate." Is this the same as what hocus is saying?

I have started a Part 2 thread that looks once again at definitions. I have posted a version of where we left off on this thread. I have also posted a new definition, which I call Alternative A.

Both of these definitions are getting lengthy. I think that you will find that I have addressed your concern. (BTW, this does not necessarily mean that you will find that what I have posted is satisfactory.)

I have defined best estimate in words that are consistent with the point that hocus has made. My definitions (the original, as revised, and Alternative A) require a mathematical calculation that stays put. (I.e., it is independent of the actual occurrence of future events.) They require an explicit specification of the sense in which the calculation is best.

We aren't there yet as far as agreeing on definitions. We are getting close. When we do reach agreement, these kinds of discussions will be a lot easier.

Have fun.

John R.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

ataloss
I told Bob that it was plain to see that a 75% stock allocation wasn't ideal in 2000. Bob concedes this but is insisting that retrospectively timing the market is easier than doing it prospectively.

The new definitions will help in this matter. What Bob is saying is that a Safe Withdrawal Rate calculation should not depend upon the actual occurrence of future events.

[The matter of valuations is important. The fact that the study does not take that into account is important. But I would not make an adjustment of valuations a necessary feature of an SWR calculation. It is OK to talk about an SWR when valuations are ignored. It is just that this limitation should be pointed out to people and its implications (which are highly significant) should be explained.]

[There is much more to be said about valuations.]

Going back to the study and looking at allocations: BenSolar was able to get intercst to make calculations based upon starting in each of the twelve months instead of always at the first of January. Using the identical calculation procedure as in the original study, the optimal stock allocation changed to (close to) 50% as opposed to 75% to 80%. Bob will certainly take notice. That is an extreme sensitivity.

Have fun.

John R.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

hocus
If valuation always affects returns, and returns always affect SWRs, then valuation always affects SWRs. No? I see no escape from the force of that logic. [Correction added.]

If all this is so, then it is not just in the year 2000 that the historical sequence model produces the wrong SWR. The historical sequence model always produces the wrong SWR. Because it does not factor in the effect of changes in valuation levels.

See Part 2 of this thread. I am suggesting definitions that would allow for SWR calculations that do not consider valuations...but which would require that people point out this deficiency and its implications (which are highly significant).

It is true that failing to include the effects of valuations results in an inferior number.

If it were 1993, we would not be nearly as concerned about this as we are today. It is only because valuations went far outside of the historical range that we are concerned. Strictly speaking, the historical sequence model was not applicable throughout the bubble. That makes it compelling to seek a good adjustment. Otherwise, we would have to throw it out completely.

hocus
If the "Stocks for the Long Run" claim generated enough confidence in stocks as a long-term investment to pull valuations up to levels never experienced before, then the success of the book's key idea (that stock risk can be eliminated by a plan to hold for the long term) was itself responsible for creating an investing environment in which the book's key idea was no longer true.

Absolutely correct and well worded. Acceptance of the key idea, not necessarily the book itself, caused its own failure.

Have fun.

John R.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

FMO points out to hocus:
According the the Office of Federal Housing Enterprise Oversight (OFHEO) from 1975 to 2002 Washington D.C. area real estate appreciated at a compound rate of about 6.1%. The CPI-U increased at a compound annual rate of about 4.6% over the same period. So it appears that on average, D.C. area real estate outpaced inflation by 1.5% over that 27 year period. To calculate a real total return would require that you consider a host of other factors, including the degree of leverage employed, upkeep expenses and the value assigned to the shelter which an owner-occupied home provides.

Another factor that you should consider is the population density of your area.

Here in Okaloosa County, Florida, part of Eglin AFB effectively isolates the southern half (near the Gulf) from the northern half. The southern part filled up gradually throughout the years and now it is full. This area (the southern end of the northern half of the county) is growing explosively. This entire county is probably part of the same real estate area in the sense mentioned above.

Along the same lines, expect people to bid up prices of homes around you as limited by their ability to buy. As your area's population grows (i.e., bigger Government), those who are price limited will move farther and farther out. Those who remain will have higher incomes. Reports of the increase in average prices are likely to fail to adjust for the increase in the size of the DC area.

Have fun.

John R.
bpp
** Regular
Posts: 98
Joined: Tue Nov 26, 2002 6:46 am
Location: Japan

Post by bpp »


It is OK to talk about an SWR when valuations are ignored. It is just that this limitation should be pointed out to people and its implications (which are highly significant) should be explained.


It would also be nice to see a good description of the issues involved in using valuations. I'm still a relative newbie, but as far as I can see there does not seem to be any canonical definition of valuation. "Valuation" seems to be a rather vague idea with intuitive appeal, but which seems maddeningly hard to pin down.

Or maybe I just haven't stumbled across the One True definition yet in my studies. I've seen PE (and PE-10), Gordon Equation, and others, but they all seem to have some interpretation problems attached. Is there a generally agreed upon one?

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

Post by hocus »

It is only because valuations went far outside of the historical range that we are concerned. Strictly speaking, the historical sequence model was not applicable throughout the bubble.

I agree that the historical sequence model was not applicable throughout the bubble because we reached valuation levels not experienced during the 130 years examined. This reality highlights the need to factor in the effects of valuation. It dramatizes the concern.

But I believe that it is important to incorporate the valuation factor at other times as well. I do not have a calculation of the SWR by William Bernstein for a non-bubble year, so I am a bit less confident on this point. But the Bernstein comment that valuations affect returns (and thus SWRs) applies during non-bubble years as well. So I believe that the need to include a valuation factor is still present.

The argument the other way is that the effects of changes in valuation are accounted for by the historical sequence data. I do not believe that the historical sequence data adequately takes the effect of changes in valuation into account.

Take the year 1982 and the year 1996. The level of valuation was lower in 1982. But both years are within the range of valuations covered by the 130 years of historical sequence data. Is it reasonable to believe that the SWR was the same in both years? Not if valuation levels affect returns. If valuation levels affect returns, the SWR was higher in the year with the lower valuation level (1982). Someone retiring in 1982 had a reasonable expectation of better returns that someone who retired in 1996. Thus, he had a higher SWR.

I believe that you can see this by looking at the withdrawal rates in the intercst study in which the safety level designation is less than 100 percent. Take a case where it is 95 percent. That means that, of 100 possible start years for a retirement, there were five in which the retirement went bust. If valuation does not have an effect, the busted retirements should occur randomly. It appears to me that the busted retirements are not random, but bunched in start years of high valuation.

That suggests to me that you should not be mixing together data from the high valuation and normal valuation years to determine the
safety of your retirement plan. Those who are planning to retire in times of normal valuation should be checking what happened in years of normal valuation, and those who plan to retire in years of high valuation should be checking what happened in years of high valuation.

Do you not agree that there appears to be a correlation between high valuation years and busted retirement years? If there is such a correlation, isn't it proper to take that correlation into account when assessing the safety of a plan? If you take valuation into account, I believe that you will find that retirements beginning in years of normal valuation really are safer than the <b<intercst study says, and that retirements beginning in years of high valuation really are less safe than the intercst study suggests.
If you looked only at data from high valuation start years, you might find that the odds of success of plans now assessed as 95 percent safe are really 50 percent or 75 percent. That's a big difference.

I view this as a highly impotant point, and I very much would like to get it cleared up. I acknowledge that this point is more foggy that the one dealing with the bubble years. The issue is much easier to explain when you are discussing the bubble years. But I believe that valuation's effects are also important (just a bit less so) in non-bubble years,and that valuation should always be taken into account.

An easy way to appreciate that valuation is not taken adequately into account by studies using the historical sequence model is to take note that the SWR does not vary as valuation levels go up and down. It should. Valuation affects returns, and returns affect SWRs. So SWRs should go up and down as valuation levels go up and down.

The SWR calculation is not only supposed to reveal what is safe. It is supposed to tell you the highest take-out number that you can use consistent with your safety desires. If you do not factor in valuation, you are not getting the highest safe number.

I just do not see what argument there is for not counting valuation when it is a factor that always has an effect. I can see for practical reasons limiting the number of adjustments to perhaps three--one for high-valuation years, one for medium valuation years, and one for low valuation years. But I do not see the case for failing to make any adjustment whatsoever.

A small adjustment in the SWR can mean a big difference in the construction of a plan. Why not make the calculation as accurate as is possible as a practical matter? Valuation affects the SWR and the historical sequence approach reports SWRs as if valuation levels never moved in one direction or the other or else did not have an effect on returns when they did. Why?
User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss »

Wanderer:
I don't think owners of monkeys have any illusions about renting.

JWR1945:
Bob was indeed very interested in the sensitivity of the allocation to the month of withdrawal. He pointed out to me that the Trinity study selected a few allocations (Stocks 100%, 75%, 50%, 25%, 0% ) and looked at the success rate for withdrawals of 3-12%. Bob said something that sounded like "excessive curve fitting and retrospective optimization"￾ but we were at a picnic and I couldn't make out all of it. I am getting the feeling that Bob considers the swr to be a "rule of thumb"￾ and I will try to correct this idea.

BPP:
"Valuation" seems to be a rather vague idea with intuitive appeal, but which seems maddeningly hard to pin down.

I agree.

Ben:

Bob seems to think more clearly that I do. Wish he would write his own posts.

Hocus:
X-Y=Z
If X is the value of the market in the future (assuming dividends reinvested )
And Y is the current value; Z will be the difference (which can be annualized)
Bernstein says that if Y is larger Z will be smaller, This is mathematical certitude.

You notice that he doesn't say (as a matter of mathematical certitude) that future returns will be less than the 6-7% long term historical rate, just that they will be lower than they would have been if the market was lower (when the book was written)

It seems highly likely that the retrospective maximal withdrawal rate for 2003-2033 will be higher than for 2000-2030. Still it is by no means certain the retrospective maximal swr for 2000 won't be 3.7%.

FWIW, I agree that "Stocks for the Long Run"￾ or at least the understanding that stocks had low risk contributed to reduction of the equity risk premium.
Have fun.

Ataloss
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

bpp
Or maybe I just haven't stumbled across the One True definition yet in my studies. I've seen PE (and PE-10), Gordon Equation, and others, but they all seem to have some interpretation problems attached. Is there a generally agreed upon one?

BenSolar is the best person to talk about this. But he has just gotten back from vacation and he has a lot of catching up to do.

We use P/E10 because it is the One Available definition. We have had success using it to make quantitative adjustments. We do not claim that it is perfect. I will assert that it is often adequate.

The data that we use was compiled by Yale Professor Shiller. He has shown that stock market returns correlate well with earnings if you average earnings overs several years. He looked at 30 years originally and the correlation was very strong. He found that 10 years also works well. We all know that a single year's earnings have very little predictive power. Dr. Shiller credits Benjamin Graham with the idea of averaging earnings over several (5 to 10) years.

Professor Shiller's data is available at his web site. His data are used by many, including many who calculate Safe Withdrawal Rates. It covers the yeas from 1870 to within the last few months. He lists S&P 500 index values, earnings, dividends, consumer's price index and P/E10. Wherever appropriate, he provides both nominal and inflation adjusted numbers.
http://www.econ.yale.edu/~shiller/

Have fun.

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

Post by therealchips »

Have I missed something here? Hocus said
At some point, I would be interested in views as to the "why" of the finding that ownership in a residence produces such a low real return. Does anyone have thoughts as to why this use of capital produces such low returns, compared to most others?
referring to a 1% or 2% annual appreciation in residential housing values. Hocus, are you overlooking the value of free rent? The capital gain is small, but the value of living in the house is like a stream of income during all the years you own the house and live in it. As I have said before, the fair rental value of owner-occupied houses, less something for maintenance, depreciation and interest, is part of Gross National Product and, in that sense at least, very real income to the home-owner. This is an implied rent that you pay yourself, although not deductible as an expense and not taxable as income. In the real world, property taxes and insurance reduce that stream of income, but GNP calculations ignore such considerations, as I recall.

I don't see what difference it makes to you what the rate of capital gain in your house is anyway, other than the obvious satisfaction of having made a good investment. If the house appreciates 10% one year, say, and you continue to occupy it, then your implied rent has risen about 10% too. None of the increased value of the house is available to cover other expenses (unless and until you sell or take a mortgage), so what difference does it make? How does lowering your prediction of capital gains on you house increase the amount of current income you need from writing or whatever?

Short answer: It is only the capital gain benefit of home ownership that has historically been small in most places; the implied current income in the form of free rent is fairly large.
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

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

Post by therealchips »

Pardon the repetition, but I have to express my disagreement. Hocus said
It sounds to me that you are making a distinction between "productive" assets (corporate stock) and "non-productive" ones (housing). The distinction makes sense to me.
The distinction makes no sense to me. Housing, owner-occupied housing, is highly productive. It provides shelter which has a fair current market value. It is like a stock or bond that pays high current dividend or interest but produces little in the way of capital gain. Regarding housing as unproductive is the error of "Rich Dad, Poor Dad". Regarding housing as productive is the good sense of GNP calculations. GNP includes the fair rental value of owner-occupied housing. If you and I move into each other's unmortgaged houses and start paying each other rent, GNP would not increase. That value is in there already. I'm quoting my econ professor from 1968.

Owning your house outright is roughly the equivalent of having a rich uncle pay the rent you would otherwise have to pay. Surely that would count as income. Would you tell the kindly old gentleman that his contribution is not productive?

I have spoken (or rather written). :lol:
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

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

Post by therealchips »

Thanks, John, for your comment on my comment. :)

John R says
We can even specify two constraints such as satisfying a minimum balance throughout the portfolio's lifetime and staying above a minimal ending balance.)

Pardon me, but I don't think that is so. In retirement, when the value of my retirement assets falls below any particular number, there is no way I can repair the damage. (This is the voice of sad experience.) I don't have any other assets. I can reduce my spending until maybe the market repairs the damage for me, but in the meanwhile, the portfolio balance is below the minimum I meant to preserve.
He who has lived obscurely and quietly has lived well. [Latin: Bene qui latuit, bene vixit.]

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

Post by [KenM] »

Ataloss; I am a big fan of Bob!
Me too! I assume like all good engineers he can't write very well. Is that why he doesn't post himself? :)
KenM
Never try to teach a pig to sing. It wastes your time and annoys the pig.
Post Reply