Gary Burtless study on cohort risk

["Steven H. Johnson" <info@sscommonsense.org>]

Robert Reischauer points to the issue of cohort risk.  An age
group that retires when the market is at its highest high can
do as much as three times better on its PRA annuities than an
age group that retires at the depths of the market's lowest low.
And he cites the Gary Burtless study, done at Brookings, in 
support of this.

I agree with the basic finding of the study.  I've used the S&P
500 historical data and gotten the same results.

But there is one seriously flawed assumption in the Burtless 
study that ought to be addressed.  Burtless assumed a six percent
payroll investment rate, going entirely into S&P 500 stocks, with
all dividends reinvested.

Suppose we imagine the entire workforce having done exactly that.
And earning exactly the same rate of return that the S&P 500 
delivered historically - roughly 7 percent in real terms.

Within forty years, the total capital accumulated in Personal
Retirement Accounts would be equal to about 180% of GDP.

On the other hand, the average Market Capitalization-to-GDP ratio
during the time period covered by the Burtless study was roughly
65 percent.

In other words, if you extrapolate the Burtless assumptions to
cover the entire American workforce, you wind up with PRA's whose
total value is equal to three times the average size of the
entire stock market.

This raises a host of issues which I won't try to explore in this
comment.  Anyone interested in a more extensive and rigorous
exploration of the issue is welcome to visit the Common Sense on
Social Security website, at www.sscommonsense.org   The essay,
"Are Seven Percent Returns Realistic?", lays out the necessary
analytic approaches and historic data. 

-Steve Johnson