Rational exuberance: The bubble that wasn’t

(May 2006) Here’s a fact: the late 1990s were a time of irrational exuberance (said Alan Greenspan, no less), with more froth than Lake Ontario in a hurricane. When the floor dropped on the markets many investors were mauled and markets are just now recovering from these historic lows.

But this flies in the face of long-term market behaviour of investing rationality and of the law of large numbers. We should reasonably expect that investors are rational over the long term. This means that they learn from adverse market performance and decision making, and that over several market cycles their average skills should be reflected in long-term market trends. And because there is a very large number of investors with diverse investment objectives, risk tolerances and skill sets, “the market is always right.”

The question of investor rationality in the face of highly volatile markets (both from an ex-ante and an ex-post-facto perspective) is intriguing. Interestingly, it is fed by some pretty serious research, which I will attempt to expound.

In Was There a Nasdaq Bubble in the Late 1990s? , professors Lubos Pastor and Pietro Veronesi of the University of Chicago say, “Not necessarily.” In fact, they start with a telling quote from a colleague: “…’bubble’ characterizations should be a last resort because they are non-explanations of events, merely a name that we attach to a financial phenomenon that we have not invested sufficiently in understanding.” And they add: “…we don’t claim that there was no ‘bubble;’ we only argue that it is not obvious that there was one.”

The fundamental question for Pastor and Veronesi, covered in their highly mathematical paper, is whether Nasdaq equities were overvalued or appropriately valued on March 10, 2000, the index high. To answer, they develop a valuation model based on the market-to-book ratio as an increasing function of uncertainty about the growth rate of a company’s book value. They then derive a level of “implied uncertainty,” to wit, the level of uncertainty that makes the model market-to-book ratio equal to the observed market-to-book ratio.

In a nutshell, Pastor and Veronesi show that the valuation of a firm depends not only on payoff expectations (as in standard valuation models) but on the variability (another word for uncertainty) of those payoffs. If there is a “reasonable” likelihood that an equity may become, say, the next Microsoft, it will be very valuable (even though it also has a probability of cratering).

The authors find implied uncertainty levels that they claim are plausible because they match both the high level and the high variability of Nasdaq stock prices at the height of the market. Also, given the inverse correlation between equity premia and price uncertainty, the decrease in equity premia over the last decades may contribute to explaining the increase in valuations, through increased valuation uncertainty. Finally, they claim that the “bubble” burst because of a significant downward revision in expected Nasdaq profits.

This is not the only paper that, in reviewing the 1990s, suggests there was greater rationality at work in valuations than was previously thought.

In Earnings Growth and the Bull Market of the 1990s: Is there a case for rational exuberance? , Jinho Bae and Charles Nelson investigate whether there was a permanent increase in the rate of earnings growth during the 1990s. By modeling S&P 500 earnings from 1951 to 2000, they attempt to find indications of a one-time structural break or of a gradual change in pattern that could explain observed high valuations; what they show, however, is that increases in earnings growth can be explained within the confidence interval of their model without resorting to a structural break explanation. In brief, consistent modeling can be developed that shows rationality as opposed to one-time, specific “mania” explanations.

In Near-Rational Exuberance, James Bullard, George Evans and Seppo Honkapohja at the St. Louis Fed develop the concept of exuberance equilibria — situations of self-fulfilling expectations in which additional subjective beliefs are added — and create an equilibrium in which a) expectations are consistent, b) agents are rational in including an extra variable, given that everyone else does (akin to herd behaviour/mentality), and c) the equilibrium is stable, i.e., the outcome is clear to everyone. Let’s call the extra variable the ‘New Economy factor’ or ‘the dot-com factor’ and you can see how this can explain very high market valuations. In fact, the authors interpret the exuberance equilibrium in the asset-pricing model as a cause of “excess volatility.”

And in yet another paper, Rational Exuberance: The Fundamentals of Pricing Firms, from Blue Chip to “Dot Com”, Mark Kamstra proposes a model designed to incorporate non-dividend-paying firms or firms with negative cash flows. Although not attempting an explanation of the bubble phenomenon, such work supplies the tools of investigation into rational explanations of presumably “irrational” market phenomena.

Where does this leave us? These efforts to incorporate unusual behaviour into economic and financial modeling are, well, rational.

We wish to obtain cogent explanations of observed phenomena, and these explanations should serve to show that these phenomena (the Nasdaq bubble) can be understood and hence, predicted. This will give no money back to the dot-com victims; perhaps the next step for analysts is to give us the bounds of their modeling so that we can better understand “just how bad it can get.”

This article originally appeared in Advisor’s Edge Report. Pierre Saint-Laurent, M.Sc., CFA, CAIA, is president of AssetCounsel Inc. He can be reached at PSL@AssetCounsel.com.

(05/26/06)