NoFeeBoards.com!

4 Pillars

Trinity Study page 231

1966 Acid test with various stock bond mixes page 232-233

"These are profoundly disturbing results. Since the real equity returns were 5.3% during the period, the conservative back-of-the-envelope method of withdrawing the real return every year should have allowed us to safely withdraw 5.3% annually and still have our real principal intact. In fact, such a withdrawal rate completely depleted all the portfolios no matter what their stock/bond composition. The amortization method predicts that we should have been able to withdraw 6.7% per year if we were willing to completely deplete the portfolio in 30 years. As you can see from figure 12-1 and 12-2 a 6.7% withdrawal rate would have completely depleted all the portfolios in about 15 years. This means that a penalty of about 1.5-2% was extracted by the luck of the draw."￾ In other words, a particularly bad returns sequence can reduce your safe return by as much as 2% below the long term return of stocks. Recall from chapter 2 that it is 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!

p 234

I'm not sure what the point of this is but I do recall Bernstein using conventional MC simulation along with the Gordon equation to get a 2% SWR number. I wish I could remember where I saw this, probably in one of the "Efficient Frontier" journals from way back.

As you may know, I am not a big fan of conventional MC simulation since there is no mean reversion adjustment.

I'd vote "Not at all". It could just as easily have been a guess - or am I not allowed to use that word???

:DI have been scolded for using the guess terminology. I don't doubt for a minute that you can get a 2% swr from a monte carlo simulation but from the context it still appears to me that Bernstein simply estimated (is that a safe word to use?) the lt market returns and subtracted the 1.5- 2% for "luck of the draw" effect. This precedure and his terminology leave me thinking that he is not stating that the swr in 2000 was 2%with "mathematical certitude" as hocus has said.

I guess I am responding to demands from hocus to "read the book" and his claims that Bernsteins 2% was the result of a calculation with a high degree of relaibility (although we don't know what that method was)

JWR may be right that Bernstein got 2% from monte carlo but since Bernstein says that projected stock return is 3.5% (real) and the "luck of the draw" reduces returns by 1.5 to 2% And since 3.5% - 1.5% = 2% I suspect that he used this simplified and not entirely "valid" (in the usual meaning of the word) method to calculate swr

It isn't clear to me that bernstein thinks that future swr can be defined with a high degree of precision despite the claims by houcs about "mathematical certitude"

And since 3.5% - 1.5% = 2% I suspect that he used this simplified and not entirely "valid" (in the usual meaning of the word) method to calculate swr

I like the thought that Bernstein may have used an "invalid" methodology

This precedure and his terminology leave me thinking that he is not stating that the swr in 2000 was 2%with "mathematical certitude" as hocus has said.

Bernstein never said any such thing, and I never said that he said any such thing.