Doom and gloom SWR based on 1965-1982

[cute fuzzy bunny]

As long as you look at the 10-20% times it failed, the economic conditions that prevailed at those times, and you're pretty comfortable that such a set of conditions, or ones equally unpleasant, dont occur.

99% isnt "safe" if we hit the set of conditions that invokes that 1%...

 

[sgeeeee]

. . . I am a believer in valuation.

Hi Mikey,

I'll just kinda ignore the ***** issue, and talk about valuation and SWR . . .

Of course one of the problems with valuation is that you can't develop a metric that quantifies the valuation you really care about (ie. the ratio of current price to futureearnings). Backward looking metrics like PE or PE10 are reasonable approximations of the desired valuation provided price and earnings are not changing rapidly. But it is exactly during rapidly changing financial times when we would most like to have a valuation metric.

Valuation is not an independent variable. SWR is affected by valuation but is also affected by many other financial metrics such as inflation, employment rate, fed borrowing rates, productivity, demographics . . . It is naive to think we might be able to modify SWR based only on valuation without considering all the other factors that might have an impact on the result.

True valuation (current price to futureearnings) is, of course, already considered in FIRECALC or any other historical simulator. These simulators simply find the worst case SWR over all time. The results that calculated consider all valuations that have been observed in the past. Such an analysis considers not only valuation, but how valuation was correlated to all other important economic indicators of the past. Of course today's PE and PE10 values are very high compared even to the worst historical values. But we should remember that PE and PE10 are only poor approximations for the valuations we really care about (current price to futureearnings).

JWR did do some nice work culminating in an interesting plot of SWR vs PE10(valuation). The correlation between SWR and PE10 is easily observed in the plot. But the amount of variation in SWR that is not accounted for (historically) by PE10 is also clear. Error bars for any curve through the historical data are greater than + or - 1% for the range of data represented by history. The correlation is also clearly non-linear so that any attempt to extrapolate the historical data to PE10 ranges outside of the historical range are highly suspicious.

So, you start with a poor approximation for valuation, ignore the correlation between your poor approximation and all other factors that effect SWR, ignore uncertainties (errors) evident in the historical data and you can come up with a modified SWR -- which was only a rule-of-thumb estimate to begin with.

 

[hocus]

Author Dr. William J. Bernstein ( the guy ***** continually misquotes)...

Here's a link to an article by William Bernstein from his Efficient Frontier site that sums up some of the important points made in Chapter Two of his book "The Four Pillars of Investing."

http://www.efficientfrontier.com/ef/403/fairy.htm

William Bernstein: "I’m going to discuss the single most important issue in finance—future stock market returns—in the clearest, most descriptive, least mathematical terms possible.

"....Nobody knows what the market is going to do tomorrow, next month, or even in the next five years. And in the final analysis, what the market does over such relatively short periods is irrelevant to the average investor. What is important is return over the next few decades, and we do have a pretty good idea of what’s going to happen over such long time periods.

"The biggest area of confusion among the investing public concerns just where stock returns come from. The most popular misconception is that future stock returns somehow derive from past stock returns—that is, from the Stock-Returns Fairy. In the past few decades, the packaging of historical financial returns has become an industry bigger than the GDP of some South American nations. The silliness of this approach is obvious: if you pay twice as much for an asset as you should have, that increases the return of the guy who sold it to you, just as surely as it reduces your future return.

"So just where do stock returns come from? In order to answer this question, first ask yourself this: how much would you be willing to pay for a business that distributes $10,000 each and every year?...Thus, in the very long term, stock price increases come from only one source: increases in dividend income.

"....The value of the stock market almost exactly tracks the dividends and earnings it produces; in short, it behaves just like any other business."

The methodology used by JWR1945 to calculate SWRs takes account of the earnings they produce. The approach used in the REHP study does not; the REHP study reports the same SWR regardless of whether valuation levels are low, medium, high, or extremely high. It is not possible to accurately determine the SWR without taking the valuation factor into account. It is a factor critical to the question being examined (What withdrawal rate is safe?).

 

[hocus]

It is naive to think we might be able to modify SWR based only on valuation without considering all the other factors that might have an impact on the result.

All of the other factors are taken into account both in the William Bernstein analysis and in the JWR1945 analysis. Bernstein says on Page 73 of "The Four Pillars of Investing" that "Although the discounted dividends model (DDM) informs us well about expected returns, it tells us nothing about future risk. We are dependent on the pattern of past returns to inform us of the potential risks of an asset. And in this regard, I believe that the historical data serve us well."

These simulators simply find the worst case SWR over all time.

There are only two data points in the historical record at which we experienced the level of valuation experienced in 1929. In both of these cases, an investor taking a 4 percent withdrawal barely avoided going bust. The data is showing that a 4 percent withdrawal is not safe at that valuation level. But the studies are saying the opposite, that a 4 percent withdrawal is "100 percent safe" at that valuation level. Why?

The reason is that the conventional methodology studies do not even examine the question of what withdrawal rate is safe. All that they are determining is what withdrawal rate survived.

Someone who smokes in bed and survives the experience one or two times is not thereby justified in claiming that this activity is a "100 percent safe" one. The activity is a risky activity that in any one or two particular instances may or may not result in disaster. The same is true of the activity of taking a 4 percent withdrawal at the valuation level that applied in 1929. That withdrawal rate survived twice at that valuation level, but it was not a safe withdrawal rate in either case.

 

[unclemick]

Life is not 100% safe, especially if you get killed.

 

[ESRBob]

We raised and home-schooled 2 kids. Both are college graduates, and one has a net worth closer to 8 figures than 7. Neither is yet 30 years old.

Mikey,
My hat's off to you -- the true sign of success in my book is raising happy, successful, well-adjusted kids. I'll bet a big part of their success is that you and their Mom were around for them a lot. How people think their kids will turn out fine when they are abandoned to TV and video games is beyond me...

But on your other points, it sounds like you pulled together an intuitive blend of investing (financial, real estate, liquid/illiquid, systematic/opportunistic) along and leisure and family time. You had to make regular investment decisions -- when to buy and when to sell. At that time, (in the 70s) the concept of indexing, asset allocation or SWR either didn't exist or was not supported mathematically. Curious whether we have 'quantified the way things always have been' or 'made ourselves a new false god that won't stand the test of time', but I guess time will tell. I think Bernstein's admonition on his site that success rates over 80% are meaningless in an unknowable future resonates for me.

Just curious -- were you raising your kids in California? I am an east-coaster now, being a bit of a freak for doing ER in NY, but I reflect that lots of Dads in suburban SF where I grew up seemed to be doing much the same program you describe. I wonder if this might be another case where Californians might have been blazing the trail.

ESRBob

 

[ESRBob]

 I wonder if this might be another case where Californians might have been  blazing the trail.

ESRBob

I meant to say West Coasters, but then again, people were probably doing this in Greek and Roman times,too, and a certain group in every generation just has to re-discover how to create a reasonable work-life balance.

ESRBob

 

[amt]

As long as you look at the 10-20% times it failed, the economic conditions that prevailed at those times, and you're pretty comfortable that such a set of conditions, or ones equally unpleasant, dont occur.

99% isnt "safe" if we hit the set of conditions that invokes that 1%...

Please note that intercst is saying that 80-90% safe is probably plenty safe USING MONTE CARLO.  Based on historical calculator, the SWR that was 80-90% safe under Monte Carlo, would most likely be 100% safe.  IOW, those sets of conditions that you were mentioning had never occurred in history.  Since the probability of economic and political stability of this country and the civilization for the next 40 yrs is only in the range of 80-90% per Bernstein, if you shoot for your SWR to achieve higher safety than that, you may not gain that extra safety.  IOW, even if your withdrawal rate is 0%, your chance of making it for the next 40 yrs is still 80-90%, since there might be 10-20% chance that the government may confiscate everything, or North Korea may bomb many US cities, driving everything into chaos, etc.

In the article intercst quoted, Bernstein clearly favors the withdrawal rates around 3 to 4% range.  The following is quoting Bernstein:

"Mind you, this is not a call for wild abandon. The above table constrains the retiree desiring a theoretical 97% success rate (of portfolio survival) from spending more than 3% per year of the initial real amount of his nest egg. Taking the accident propensity of the species into account would allow him to spend about 4%. But if you believe that we're about to encounter a bad returns sequence or simply wish to leave a few baubles to your heirs, you're right back to 3% again.


So live a little, and enjoy your money, for tomorrow we may be consumed by the ghosts of Hitler, Lenin, and Attila the Hun. And at withdrawals of 3% to 4% of your nest egg, don't spend it all in one place."


From  http://www.efficientfrontier.com/ef/901/hell3.htm  

 

[amt]

Please also note that the 3% SWR I mentioned in my initial post (91% success for 40 yrs and 97% for 30yrs) was based on the data from only 1965-1982, no other years; i.e., even if EVERY YEAR of the next 30 to 40 yrs will be like the years 1965-1982, at 3%, you still have 91-97% chance of making it.  To me, that's pretty reassuring, especially if you can live at 3%.

 

[cute fuzzy bunny]

Speaking of which, where is Chuck and his bazooka right now?

 

[hocus]

In the article intercst quoted, Bernstein clearly favors the withdrawal rates around 3 to 4% range.

It is true that there is language in this article and elsewhere in which Bernstein indicates that he considers a 3 percent withdrawal or a 4 percent withdrawal as acceptable for some retirees. Those statements do not necessarily contradict either Bernstein's or JWR1945's analyses finding that the historical data shows the SWR for recent years to generally be a good bit lower than that.

In an earlier post in this thread, I noted JWR1945's finding that the historical data shows that a 4 percent withdrawal has a 50 percent chance of working out at today's valuations. No one is saying that a 4 percent withdrawal absolutely will not work. It depends on what sort of returns sequence happens to pop up beginning with the retirement start date.

When Bernstein says on Page 234 of "The Four Pillars of Investing" that the SWR for a high-percentage-stock portfolio is 2 percent, he is not saying that no higher withdrawal rate will work out in actual fact. He is saying that if a returns sequence as bad as any we have seen in history (but no worse) turns up, the highest withdrawal rate that will work is 2 percent. He is not predicting that the worst-case scenario will turn up; he is telling us what will happen in the event that it does.

The REHP study reports that a 4 percent withdrawal will work even in the event that a worst-case scenario turns up. This is simply not so. The reason why the REHP number is so far off from the number that Bernstein comes up with as the number that will work in a worst-case scenario is that the Bernstein analysis includes an adjustment for changes in valuation levels and the REHP study does not.

Please take note of the language in any earlier post of mine in this thread in which I quote Bernstein as drawing a distinction between the "expected returns" question and "future risk" question. He says that the conventional methodlogy studies "serve us well" in providing valuable information re the "future risk" question. JWR1945 and I agree with Bernstein on this point.

The flaw in the REHP study is not that it does not deal adequately with the "future risk" question. It is that it does not deal at all with the "expected returns" question. You can accurately determine the historical surviving withdrawal rate (HSWR) without taking the "expected returns" question into account. You cannot accurately determine the SWR without taking the "expected returns" question into account.

 

[arrete]

Don't you think the world would be a better place without the blether of percentage points of withdrawal rates?

arrete

 

[John Galt]

Personally, I would prefer to debate the merits
and multiple uses of dryer sheets.

John Galt

 

[charlie]

I vote we abandon this thread and let *****, amt and
salaryguru hash it out to their heart's content.

I misplaced my bazooka, TH, just like I misplace everything else ....... such is the life of a senile old
phart.

Charlie

 

[sgeeeee]

I vote we abandon this thread and let *****, amt and
salaryguru hash it out to their heart's content.

I misplaced my bazooka, TH, just like I misplace everything else ....... such is the life of a senile old
phart.  

Charlie

BLOW ME!

 

[cute fuzzy bunny]

Dammit Chuck...

Nords? Anything uranium tipped from your naval days?

I vote we let ***** hash it out with himself. He's demonstrated the ability to have extensive conversations with himself over something that really isnt that important.

Heck, I was ER'ed for three years before I even heard the term "withdrawal rate", let alone a 'safe' one.

Somehow I managed.

Lord knows what would have happened if I had never heard of it though. The SWR police might have swooped in and sent me back to work.

Hey has anyone ever tried cutting out like the middle three paragraphs of one of *****'s 18 page missives and used that to respond to him...just to see if he'd argue with it or support it?

I almost wish the stock market would go to zero tomorrow, and the bond market fully default. That way we would all be reduced to unsafe withdrawal rates, go back to work, and that would effective end the discussion ...

 

[Nords]

I've said it before and I'll say it again...

... *****, please leave and go post somewhere else.

 

[hocus]

Re: Doom and gloom SWR based on 1965-1982here is

*****, please leave and go post somewhere else.

I am certainly not going to push my views on anyone, Nords. I recommend that any community members who want to explore the Data-Based SWR Tool in depth go to the SWR Research Group board (at NoFeeBoards.com) and do it there. The environment is not good at this board at this time for informed discussions of the topic. There is no constructive purpose served by generating further friction.

That said, all of the various boards need to come to terms with this issue in time. SWRs are important and there is a lot of interest in the various commnities in exploring the realities of SWRs. This has been demonstrated time and time again. There is close to zero interest in experiencing further friction, so we all should be doing what we can to keep that stuff to a mimimum. But we also need to keep in mind that the ultimate goal is to create an envrionment in which we can have reasoned discussions of this important Retire Early topic.

One thing that causes a serious problem is when someone puts forward a post that recommends that someone look at the REHP study without pointing out that serious people like William Bernstein have raised serious questions about the accuracy of that study. These are money questions we are dealing with at this board. People are going to suffer serious life setbacks if we give them bum advice. Bernstein thinks that the REHP study is bum advice. He says that the results of that study are "highly misleading." Anyone who recommends that someone look at that study in putting together his plan is also under an obligaton to let that person know about the flaws of that study and about where to find more information re those flaws.

I never have put forward a post for the post of causing friction. I never will. But it is a seriously wrong thing to do to tell aspiring early retirees that the historical data says something that it does not say. Any poster is entitled to express any opinion he chooses. None of us is permitted to put forward highly misleading claims re what the historical data says and not expect to be challenged on those claims.

 

[unclemick]

Well ***** you've certainly convinced me that the whole idea of SWR should be thrown in the trash can. The data posted over at SWR Research has reinforced my prior lifetime's suspicion of over use of numbers/calculations - a curse of engineers and certain other types. At best - a good data set/mythical portfolio can reinforce and offer insite into general principles.

Beyond that - it gets into 'debate over icons in the Byzantine Empire'.

Handgrenades(25X type), DeGaul and the Norwegian widow ride on! Yeah.

 

[hocus]

Well ***** you've certainly convinced me that the whole idea of SWR should be thrown in the trash can.

I find that an extremely interesting comment, UncleMick. It makes me feel a little good and a little bad.

I am glad if the stuff I have put forward has caused people to question the SWR concept. Questioning is good. If the SWR concept is a strong one, it will survive any questioning directed to it. If it is not, it will indeed be thrown in the trash can, and should be. My personal view is that the SWR concept is a strong one, and that it will survive the questioning. But I think that it is a positive sign that people like yourself are questioning its value.

Your comment makes me feel a little bad because it is not my ultimate goal at all to cause people to become dismissive of SWR analysis. There are a lot of people who have suggested that I have some sort of personal animus towards intercst. People who say that sort of thing just do not understand my motivations even a little bit. On the day when I discovered the intercst web site, I was jumping up and down and printing out all the articles and boring my wife about with talk about what I found. The biggest reason why I found the site so exciting is that he used SWR analysis to plan his retirement just as I did and he used a 4 percent withdrawal rate in his plan just as I did.

It was always clear to me that intercst and I did not agree on everything. But it seemed to me that we agreed on the most important things. To this day there is a sense in which that is true. There are a lot of people today who argue that the SWR is not really the product of a numerical calculation, that it is just the product of guesswork. You never hear me saying that and you never hear intercst saying that. He has said a lot of things that I wish he hadn't said, but I like it that he consistently argues that the SWR is the product of a mathematical calculation. I think he is absolutely right about that (SalaryGuru is another poster argues for the integrity of SWR analysis).

My long-term goal is not to cause people to lose confidence in the SWR concept. It is to restore people's confidence in the concept. The SWR concept is in the very early stages of its development. It's just not reasonable for people to expect that the pioneers of SWR studies would get every factor in the analysis 100 percent right. Advances in scientific knowledge don't work that way. The way it works is that someone comes up with something, it catches on and becomes popular, and then someone else makes a refinement, and so on.

The Data-Based Tool is a refinement. The fact that JWR1945 and I have refined the conventional methodology tool does not mean that we see no value in it. We see great value in it. I have always recommended that aspiring early retirees buy the REHP study and use it in their planning. JWR1945 makes use of the study in the work he does at the SWR board. We are not trying to bury the conventional studies, we are trying to make them more useful and valuable.

All of the friction that has been generated for 26 months has been 100 percent unnecessary. I am not making a personal attack on intercst when I say that the methodology in his study generates highly misleading results. He didn't even develop the methodology. He used a methodology that had been used by a good number of others before him. What I am trying to do is to enhance the REHP study, to make it better.

In the long run, the enhanced study would reflect well on all of us. We all should be working towards the goal of producing the best SWR analysis we can possibly produce. Different community members have different sorts of contributions to make. We should stop dividing up into enemy camps and instead work together to advance the community's knowledge of the subject matter.

I think we are going to find our way to doing that. I think that the friction is going to subside in time but I think that we will be talking about SWRs for a long time to come. It may be that we need to go through a stage first in which a number of community members just lose confidence in the SWR concept altogether. Perhaps people need to lose confidence in an old approach before they can develop confidence in a new one.

It is sort of like how Sleeping Beauty had to go to sleep for a while before she could realistically expect her Prince to come and awaken her with a kiss. It might be that the best thing would be to permit our interest in SWRs to go to sleep for a bit with the hope that we will be able to return to the subject with a refreshed perspective for dealing with it at some later time.

 

[intercst]

Re: Doom and gloom SWR based on 1965-1982here is

*****, please leave and go post somewhere else.

That said, all of the various boards need to come to terms with this issue in time. SWRs are important and there is a lot of interest in the various commnities in exploring the realities of SWRs. This has been demonstrated time and time again. There is close to zero interest in experiencing further friction, so we all should be doing what we can to keep that stuff to a mimimum. But we also need to keep in mind that the ultimate goal is to create an envrionment in which we can have reasoned discussions of this important Retire Early topic.

Actually many of us on the other successful early retirement boards enjoy Hoco-Mania and the "friction" that ensues. Not a week goes by when someone provides a link to one of the most humorous ***** threads from this or the nofeeboards.

We also have the benefit of several medical professionals who post on the other boards who have given us a play-by-play analysis of Hoco-Mania with prognosis and treatment suggestions.

intercst

 

[Nords]

I've said this before, too, and I'll say it again.

Stop feeding the troll, Greaney.

 

[unclemick]

18 pages breaks the existing record -- and then, and then maybe Dory shall return and say -"Jeez Louise"
Now that was cool!

 

[cute fuzzy bunny]

I vote it stops here.

Please stop poking the monster with a stick, its not dead.

 

[charlie]

Does anybuddy want to talk about P11?

Charlie