Set forth below is the text of a comment that I recently posted to the discussion thread for another blog entry at this site:
The Bengen study and the Trinity study also used the “what survived in the past” method. And not surprisingly came up with just about the same result.
If you want to see that methodology replaced you need to come up with a better way.
And convince others that your way is better.
I will leave it as an exercise for the reader to decide if you have been successful.
I agree that Greaney employed essentially the same methodolology as the Trinity study and the Bengen study. Both of those studies are in error as well. As you note, they look at what survived in the past. That’s part of the story but not the entire story. You need to look at what survived in the past in similar conditions (that is, at similar valuation levels).
If someone drives a car 150 times and gets in accidents that nearly kill him but don’t quite do so on four of those occasions and those four occasions happen to be the four times he drove drunk, those experiences do not show that driving drunk is “100 percent safe.” What they show is that driving drunk is extremely dangerous.
That’what Greeaney’s study (and Trinity and Bengen) show re retirements using a 4 percent withdrawal rate for a retirement beginning when valuations are where they were at the top of the bubble. Someone who retired in 1929 and used a 4 percent withdrawal had only $1 left in his account at the end of 30 years. At times when valuations were reasonable, the portfolio grew in size. It was super, super safe. But retirements at a CAPE of 33 barely squeaked by. Barely squeaking by indicates risk, not safety. And of course the CAPE was far higher in 2000 than it was in 1929. So those retirements were even more risky than the 1929 ones that left the retiree with $1 in his account. Those were insanely risky retirement plans, not “100 percent safe” (Greaney’s words) ones.
I want to see that long discredited methodology replaced. And, of course, working with John Walter Russell and Wade Pfau and many others, I came up with a better one. There have been thousands of our fellow community members who have expressed a desire that honest posting be permitted so that people can learn about the better methodology. Numerous big-name publications, like the Wall Street Journal and the Economist, have noted that the 4 percent rule does not apply when valuations reach the levels that we saw at the top of the bubble and that apply today. Yet there are still people promoting this long-discredited methodology today. And there are still investors who are tricked into placing their confidence in it because they do not get to hear the criticisms of it that people like me would put forward if it were not for the criminal behavior (death threats, acts of extortion. etc.) that you Goons have employed to keep people from hearing the straight story.
I believe that this will change in the days following the next price crash. I think that people will be pissed when their “100 percent safe” retirements fail. We’ll see. You obviously would not have behaved in the manner in which you have behaved if you thought that the methodology employed by Greaney could convince people in a community in which honest posting re the last 40 years of peer-reviewed research was permitted.
A con relies on deception to convince people. Something real can win support even in an environment in which challenges to its logic are permitted. Your own behavior shows that the Greaney study cannot survive challenges. You have hurt millions of people in very serious ways. I am confident that I will indeed be successful in the days following the next price crash, when the on-the-ground realities show people how dangerous it is to tolerate a con of this magnitude. Please mark me down as saying that Greaney should have corrected his study within 24 hours of learning of the error he made in it and that he never should have engaged in a single criminal act in “defense” of his “100 percent safe” retirement claims.
A failed retirement is a serious life setback.
My sincere take.
Here is a small sample of the comments that our fellow community members have advanced:
Investing Discussion Boards Ban Honest Posting on Valuations!
Rob
“You need to look at what survived in the past in similar conditions (that is, at similar valuation levels).”
The studies look at what survived in all conditions, including the highest valuation levels.
Even if you just restricted your search to the very small subset of occasions when valuations were at there highest you would find the same thing. 4% survived on every occasion.
Imagine these are the percentage withdrawal rates that survived.
6.2 5.3 7.4 9.3 6.1 8.2 (Low valuation)
4.7 6.3 7.1 7.2 5.3 6.5 (Medium valuation)
5.4 5.5 4.0 4.9 6.2 4.7 (High valuation)
As you will note 4.0 is the lowest
If you divide the numbers into 3 groups L M H valuations you will notice that the low in the L line is 5.3 and in the M line 4.7
But the low in the high valuation line is 4.0. Just looking at the high valuation subset does not get you a number below 4.0
This is why I often use the analogy of drunk driving, Evidence. Driving sober is highly safe. You could go on hundreds of drives and not get in a single accident. Driving drunk is insanely risky. You can’t mix it all together and conclude that driving drunk is “100 percent safe.”
Say that you have driven sober 146 times and never had a single accident. And you drove drunk four times and were in accidents requiring hospitalization each time. But you didn’t quite die in any of those accidents. Would you conclude that driving drunk is “100 percent safe.”
That’s not what the data is telling you. It is telling you that driving drunk is insanely risky. The next time you drive drunk, you might well end up dead. It is POSSIBLE that you will only require hospitalization one more time. But you are taking a big chance.
So it is with withdrawal rates. Take a 4 percent withdrawal in a retirement beginning when the CAPE is 44 and there’s a 30 percent chance that you will have at least one dollar remaining in your account after 30 years. But there’s also a 70 percent chance that that retirement will go bust. Having a one in three chance of seeing your retirement survive is not going with a plan that is “100 percent safe.”
Not according to this boy.
I naturally wish you all good things.
Rob
“This is why I often use the analogy of drunk driving, Evidence. ”
Which is a terrible analogy. There are thousands (millions ?) of cases where drunk driving led to death. You don’t need to restrict your analysis to one person who managed to do it without dying.
Whereas there are zero cases where a 4% withdrawal rate failed.
I think the analogy is strong.
We do not yet have a case where going with a 4 percent withdrawal at a time of super high valuations brought on a failed retirement.But we have a case where it caused the retirement account to be reduced to $1 at the end of 30 years (despite the fact that this is an asset class that generates an average long-term return of 6.5 percent real!). And we can compare the return pattern that applied in the case with the other return patterns that have applied over time and see that it was not an unlucky one. What happens if the next one is an unlucky one?
People need to know in advance when a suggested retirement plan carries this much risk. I believe that we need to open every discussion board and blog on the internet to honest posting, without a single exception. I don’t see that one as being optional, I see it as being 100 percent imperative.
My best and warmest wishes to you and yours, Evidence.
Rob
“But we have a case where it caused the retirement account to be reduced to $1 at the end of 30 years”
Because that is how the study was designed. They set out to find what would last 30 years. If you want to find out what would last 30 years and have some left over at the end you can do that.
I understand that that is the way the study was designed.
The problem is that you are making a big deal out of the fact that 4 percent always survived. It BARELY survived in a tiny number of high valuation tests. That doesn’t suggest safety. It suggests a lot of risk.
Having a huge amount in the portfolio at the end would suggest safety (and you have that for retirements starting at low or moderate valuation levels). If you had hundreds of tests and it always at least barely survived, that would suggest that it will probably survive the next time as well and you could reasonably call that “safe.” But bare survival in a tiny number of tests does not suggest safety. It suggests just the opposite. It is showing that going with a 4 percent withdrawal for a retirement starting from an insanely high valuation level is as risky as all get-out.
It MIGHT barely survive one more time. You cannot say with certainty that it will fail. But you cannot say that it is safe, For something to survive on a tiny number of occasions does not make it safe. The concepts of “survival” and “safety” are not at all the same. These studies are exercises in semantic trickery. It is a national scandal that they have not all been corrected in the 20 years since the error in them was exposed.
Rob