Market Cycles and Safe Withdrawal Rates

[walkinwood]

The September issue of the Journal of Financial Planning has an article on how market cycles influence the SWR.

Market Cycles and Safe Withdrawal Rates

The article talks about secular bull & bear cycles that typically last around 30 years. He goes on to show that the returns during these cycles deviate significantly from the mean. He also says that it is impossible to correctly state where we are currently in a secular cycle, but points to metrics like P/E to give us clues.

Finally, he uses a 100% S&P portfolio to illustrate the different SWRs that a bear-bull-bear sequence, a bull-bear-bull, and a start midway in a sequence would produce.

I found most of it pretty obvious and would prefer to use FireCalc or the studies by the likes of Bengen than try to finesse an SWR based on a guess of where we are in a secular market cycle.

Please note that he uses "capital" returns of the S&P 500 for a lot of his calculations. That is, he does not include dividends in the return numbers. I didn't catch that at first and couldn't quite understand why he was using a 4.8% cumulative return for the S&P from 1881 to 2000.

 

[W2R]

Interesting! Thanks.

For me, it raised more questions than it answered. That's not necessarily a bad thing, though! I think I'll read a little more.

 

[Animorph]

Sounds like the same thing Ed Easterling of Crestmont Research was saying a year or two ago. He actually posted here for a while, but it didn't seem like a whole lot of the people here were too excited obout what he was saying. He was working the P/E (market cycle position indicator) predicts likely future returns angle.

Certainly seems like something that should be kept in mind.

 

[Free To Canoe]

You really have to piece together the data to get back to 1881. S&P 500 begins in 1957. S&P 90 starts in 1923. Even the Dow 30 doesn't go back to 1881. He uses the "S&P Composite Index are available from Robert Shiller".

He considers the shorter 83 year Ibbotson data inappropriate. Perhaps it is inappropriate to use his longer data set to discern between a 2.8% and a 2.7% SWR. Probably pretty good for discussing general trends.

Thanks for the article.

Free to canoe

 

[yakers]

Do we know where we are in the market going forward? I thought that PEs were back to normal/average, not a bargain but not a reason to change an AA.

 

[IndependentlyPoor]

You really have to piece together the data to get back to 1881. S&P 500 begins in 1957. S&P 90 starts in 1923. Even the Dow 30 doesn't go back to 1881. He uses the "S&P Composite Index are available from Robert Shiller".

The Shiller data is great. It is available here.
http://www.econ.yale.edu/~shiller/data.htm
It includes the monthly S&P, dividends, and the CPI so that it is easy to convert to constant dollars.
He uses an odd decimal representation for dates. I spent some time converting his dates to proper date format and was going to post it in Excel format, but it seems that Excel doesn't support dates before 1901 or something like that.
If anybody wants the iWork Numbers spreadsheet, I will be glad to post it.