Summary
Tobin’s Q has been a reliable measure of stock market valuation for over 100 years. Depending on how you measure it, the US market is 60% to 75% over-valued.
What is Tobin’s Q and Why Does it Matter?
Andrew Smithers and Stephen Wright published ‘Valuing Wall Street’ in 2000. This excellent book showed that Equity Tobin’s Q is an excellent measure of whether the stock marked is over-valued. The book correctly predicted a long period of sub-par performance for the US stock market.
On page 121 of the book Smithers and Wright provided a simple timing rule, based on Tobin’s Q, that outperforms a buy-and-hold strategy with far less risk.
In the US at least, Tobin’s Q is easy to calculate. The required figures are available from the US Federal Reserve reports. Download the Z1 Flow of Funds report from the Federal Reserve. Divide the value in B.102 line 35 (market value of equities outstanding) by B.102 line 32 (net worth) and you have Tobin’s Q for listed stocks.
See http://www.smithers.co.uk/page.php?id=34 for more details and a graph of Tobin’s Q.
Tobin's Q up to 1Q 2010
The graph shows a number of interesting features
- Tobin’s Q is high when the market is about to crash eg 1929, late 1999.
- Tobin’s Q is low when it is a good time to invest eg 1921, 1983.
- There are clear signs of reversion to the mean. Q fluctuates around a central value of about 0.6-0.7.
- The stock market at the moment is expensive.
Similar figures are available for the UK but not for other countries. However the Cyclically Adjusted Price to Earnings Ratio is a reasonably good substitute for Tobin’s Q in those other countries. You can also double-check valuation levels using the price to book ratio of index funds that track those countries.
Tobin’s Q has held up well since the publication of this book, both in practical terms as a guide to investing and in the world of academic debate. I use it myself . When Q is high (expensive) I am much more cautious and quicker to react to market declines. What Q is low (cheap) I am more aggressive and more likely to get into the market.
Why Does Tobin’s Q Work?
If Tobin’s Q is high, this means that companies are selling for more than the price of their components. This means that it is often possible to create new companies and float them for more than it cost to create them. This is exactly what happened during the dot.com era. Companies will also sell stock to buy real assets, which are cheap in comparison. They may also pay dividends in the form of stocks, and pay staff, in part, with stocks or stock options. This excess supply of stock eventually confronts a finite supply of investors, and the result is a crash.
On the other hand, if Tobin’s Q is low, then company stocks are selling for less than the value of their components. No-one is going to create new companies in this situation. Company shares will be under-valued, so companies will buy stocks back and issue few new stocks. Companies will take over other companies for cash, because the assets of the company are worth more than the shares cost. All these factors lead to increasing stock prices.
You might expect that Tobin’s Q would average around 1.0. In fact, it averages less than that, around 0.6 to 0.7. Exactly why this occurs is a matter for furious academic debate, because it is not supposed to happen according to the efficient markets theory. One possible reason is that companies over-state their profits, resulting in an over-estimate of the value of the assets of the companies. Whatever the reason, it does not matter much because Tobin’s Q is mean-reverting and has strong predictive power.
Complication
In 2009 Andrew Smithers published ‘Wall Street Revalued – Imperfect Markets and Inept Central Bankers’. This is another excellent book which focuses economic policies that can prevent asset bubbles, which are so damaging to the economy and to people’s lives.
Smithers points out that the numbers provided by the Federal Reserve are not perfect. When companies own assets that are themselves priced on asset markets, then the value of those assets can become inflated.
It turns out that this makes quite a difference. In footnote 102 on page 172 Smithers describes the technique to take out these changes in market prices, which appear in the Fed’s Z1 table as “statistical discontinuities”. Effectively you have to accumulate the values in table R 102 Line 20 and subtract this from the net worth figure above.
The following graph shows the impact of this change. It compares the standard calculation (in red) with the revised calculation (in blue).
Perhaps even the Fed is not above a little “creative accounting”.
Conclusion
If we assume that fair value for the market occurs when the standard Tobin’s Q is near its long-term average, then during the Global Financial Crisis the US market went from overvalued to near fair value at the bottom in early 2009, and has since become overvalued by about 50% again.
If we use the adjusted value, then the recent run-up in prices has again seen the market move into a dangerously over-valued range, and we are now due for continued poor investing performance.
It is worth noting that when the market is over-valued it has never fully recovered to that level without first becoming very cheap. This would imply a fall of 60% (standard Tobin’s Q) to 75% (adjusted Tobin’s Q) in real stock prices from current levels.
The view that we are just at the end of the beginning of the Global Financial Crisis may well be correct.
Tim Josling
