Monday, August 22, 2011

Safe Savings Rates, High Withdrawal Rates, and the Retirement Decision


In the comments appearing about "safe savings rates" at The Economist, the commenter named Sharma economist makes a good point:

I am not sure I get all of this. Maybe someone can help. At the point of retirement, a person still has to determine the withdrawal rate. The article says that people retiring when the stock market is low can withdraw more. That sounds fine looking at retrospective data, but how does someone determine that the stock market is low at the time of retirement? How does having a target savings rate rather than a lump sum target alleviate uncertainty in length of life and returns from the stock market?

The shorter answer to these questions are you can't know this in advance.  "Safe Savings Rates" comes with a strong caveat that it is based on what would have worked in the worst-case scenario offered thus far by history, and of course future retirees (since I consider 30-year retirement periods, this potentially means anyone who retired after 1980) could experience new worst-case scenarios.  This same caveat also applies to "Safe Withdrawal Rates" as future retirees may find that though a 4% withdrawal rate worked in the past U.S. data, it may not always continue to work in the future.

That being said, I think this offers a good chance to emphasize that this is still a fundamentally conservative approach to retirement planning.  I do get a lot of mileage out of the 1921 retiree, whom I suggest to use a 9.06% withdrawal rate.  That is quite high, of course.  But I'm not always willy-nilly suggesting such high withdrawal rates.  The following figure can make more clear about this, as I plot the value of the cyclically-adjusted price earnings ratio (PE10) at the retirement date against the withdrawal rate needed to withdraw the desired retirement expenditures called-for by the "safe savings rate" approach. 



There are 8 years (1918-1922, and 1980-1982) that this approach calls for withdrawal rates above 7 percent.  But all of those 8 years occurred at times when PE10 was below 10 and these savers would not have been able to accumulate so much wealth. Actually, the highest withdrawal rate called for whenever PE10 was above 10 was a 5.96% withdrawal rate in 1986 when PE10 was 11.72.  Whenever PE10 was above 15, the highest withdrawal rate called for was in 1989 when PE10 was 15.1 and the suggested withdrawal rate is 5.3 percent.

Just to emphasize the point of safe savings rates again, I do think it provides two important critiques about retirement planning based on achieving "The Number" which is some sort of wealth accumulation goal. Those hoping to retire at the end of a bear market would have found it very difficult to have saved their "Number" so that they can use the "4% safe withdrawal rate" to get the retirement income they desired. But historically, markets recovered and this approach gives some hope that people facing such bad circumstances may have a chance to still enjoy their retirement.  Again, the caveat still applies which is provided through Sharma economist's question: perhaps markets never will recover.  But in that kind of scenario, unemployment would probably not get any better, and people may be forced into retirement anyway.  There is some hope for salvation.

On the other hand, this approach also has something to say about the "Number" for retirees around the year 2000, who would have enjoyed an easier time reaching their "Number" than anyone else in history. They should understand and recognize that because it was so easy to reach their "Number" from a relative historical perspective, they better be thinking about the possibility that they may really need more than their number.  In other words, they may find that the 4% withdrawal rate is not going to work for them.