An intriguing study from the San Francisco Federal Reserve Bank has postulated that the aging of the baby boom generation will act as a depressant for equity prices for the next 15 years or so. I question the study.

First, surveys show that 25% to 30% of boomers have virtually no assets whatsoever, so it reasons that they don't have any equities to sell.

The theory is that they will sell equities to finance their retirement is not a new one. But the authors, Zheng Liu and Mark Spiegel, both Fed employees, isolate a demographic statistic and then overlay it with price/earnings ratios to identify an interesting correlation.

That statistic is the M/O ratio, or ratio of middle-aged folks between 40 and 49 years old, to the old-age cohort aged to 60 to 69 years old. Between 1981 and 2000, the M/O ratio increased from 0.18 to about 0.74, the authors found. "During the same period, the P/E ratio tripled from about 8 to 24," Liu and Spiegel write.

According to the authors, they estimate that the M/O ratio explains about 61% of the movements in the P/E ratio. How exactly this occurs they do not explain.

You and I could give a host of other reasons why P/E multiples expanded so dramatically in those two decades. Let's start with the unprecedented and secular decline in both inflation and interest rates. Throw in strong economic growth and rising productivity. Add the technology boom of the 1990s. I'd like to see the correlations, or attribution analysis, of the P/E multiple explosion with any and all of those phenomena.

In particular, I'd bet if one plugged in falling inflation or declining interest rates, either one might explain significantly more than 61% of P/E expansion. Of course, when one runs an attribution analysis against single variable, the result is likely to overstate the significance of that variable. That's why a multi-variable analysis might be more illuminating.

And what would also be interesting to see is a long-term historical comparison between falling interest and inflation rates, and increasing life expectancies. It's possible that none of these developments have anything more to do with P/E compression or expansion than the price of tea in Ceylon, but it is still a question worth asking.

Feeding Census Bureau demographic data into their M/O model, Liu and Spiegel project that P/E ratios should decline from about 15 in 2010 to 8.4 in 2025 before rising marginally to 9.14 in 2030. Again, I have some questions.

What if many boomers keep working, as they say they will, into their 60s and 70s? What happens if interest rates stay low for the next two decades, and equities offer higher dividend yields than 10-year Treasurys? That phenomenon persisted from The Great Depression through 1958 and it just emerged again during the Great Recession. During the 1940s and 1950s, people bought equities for dividend income, not because they expected much capital appreciation. Indeed, they were too shell-shocked to think stocks might actually go up.

Many folks attributed the equity boom of the 1980s and 1990s to demographics. I never bought into it. Nonetheless, Liu and Spiegel have produced a provocative study.