Another way to value the stock market: Tobin's Q

Q, the all-powerful, omnipresent, omniscient, err... wrong Q... One somewhat obscure way to value the stock market is using a statistical measure developed by James Tobin called Tobin's q. It is a method that develops a ratio based on the stock market value to net worth at replacement cost (i.e. replacement value of book equity measured at replacement cost.)

I ran across an analyst firm, Smithers & Co, that keeps up-to-date charts of the q ratio. The principals of the firm also wrote a book on valuing the market using the q ratio: Valuing Wall Street : Protecting Wealth in Turbulent Markets (I haven't read the book.) On their website, you can access their q FAQ page, a chart plotting q value, or a q chart along with cyclically adjusted P/E (CAPE). (If the Smithers & Co site ever go down or is not accessible to the public for free, you can also find the q chart on this site run by an anonymous user.) The FAQ on the Smithers & Co site answer some key questions about the q ratio (including one about whether intangibles affect the ratio) so check it out if you are curious about this statistical measure.

I have reproduced the chart of q, along with CAPE (cyclically adjusted P/E--this is the 10 year P/E from Shiller) below:



The above chart plots the q ratio and the CAPE relative to their average, and contains data up to the end of 2008. A value above zero implies the market is overvalued. Here is what Smithers & Co had to say regarding valuation:

US Stock Market Value as at 31st December, 2008 (S&P 500 = 903)

Valuation method Over (+) or under (-) valued
CAPE +4.10%
q including statistical discrepancies -8.50%
q excluding statistical discrepancies +30.60%

Of the two approaches using q, we prefer the data which exclude statistical discrepancies, as these seem mainly to represent the change in accounting which comes from marking to market rather than marking to cost. We prefer the latter on two grounds. First, it means that the data are more consistent over time and, second, because q depends on valuing assets at cost, allowing for inflation, rather than at market values. Indeed, if all corporate assets were valued at market, q would always equal one and would be useless as a guide to value. The major adjustments made to market values serve to underline the imprecision of the calculations. It is reasonable to conclude that the US market is no longer clearly overvalued, but is a long way from being cheap, as has occurred in previous major bear markets.

The q value is consistent with the valuation implied by the P/E ratio and what I have pointed out before. Namely, the market is attractive but not exceptionally cheap. This, though, was the stance at the end of the year. Given how the market has sold off sharply, with the S&P 500 down 25% after the start of the year, it's not clear what the q value says right now. It is possible that the q value would indicate slight undervaluation right now.

Regardless of what one thinks the q value says, it should be noted that, as Smithers & Co point out, the q value cannot be used to make short-term decisions. Similar to the P/E ratio, just because something is undervalued doesn't mean the market will rally soon. Nevertheless, it does provide a long-term macro view of the stock market. As more and more stars point towards undervaluation, the safer it will be.