Grandinite Takes On The S&P 500/NAHB Thesis

Has anyone looked at the Dow or the S&P 500 lately? Kinda scary.

Long-time readers may recall how I picked up on the story about how the National Association of Homebuilders Index appears highly correlated with the S&P 500, when the S&P 500 is lagged by one year. The NAHB index measures homebuilder’s confidence levels, whereas the S&P is a measure of the stock market, of course.

In econspeak, this one year lag means that changes in the NAHB Index give an impression of what happens one year from now in the S&P 500. If the NAHB index is an accurate indicator, then we stand to witness Dow 7,000 within the next several months.

You can find a superb graphical description of what I am talking about HERE, in regards to the spurious correlation. When you look at the broad picture, the correlation disappears.

The correlation has largely been discredited, because it only seems to exist after 1995. Before then, the two indices did not track each other closely, as discussed in Business Week a while ago.

I have something to say that has largely been left out of the analyses purporting to “discredit” this theory, and it’s that the companies in the S&P, and their respective weightings have changed over time. Have a look at the latest list.

Okay, now that you’ve done that, ask yourself this: where were Apple, Intel, Microsoft and other tech companies sitting in this list back in 1980? They were barely a blip on the market cap radar. As these companies ascended throughout the 1990’s, they definitely added something to real estate values in places like Silicon Valley, and all that easy tech capital flowed into real estate, which boosted builder’s confidence.

The difference between the economy before 1995 and after is that the US economy has become more dependent on housing-related construction and consumption. All that easy money has fuelled capital gains, which people extract in the form of mortgage equity. That’s what’s been driving purchases of consumer durable goods. As that equity dries up, so does consumption, and consumption has accounted for an increasing share of total GDP through time (see chart 4).

I’m not ready to dismiss this theory outright, due to the changing structure of the economy. Economists tend to view things in a static universe, assuming away qualitative differences. To dismiss the correlation simply because it does not hold prior to 1995 is throwing the baby out with the bathwater, in my opinion. Instead, we should look to why the correlation does not hold, and if the economy prior to 1995 is fundamentally different from the economy thereafter.

Thus, we can’t throw this one out just yet, because the composition of the S&P 500 has changed over time. Tech companies and financial brokers have risen through the ranks, and the economy may actually be more dependent on housing than it was prior to ‘95.