Investors stumble from one bull market to another


SERIOUS MONEY: Could investors be in need of a recovery programme similar to alcoholics?

ALBERT EINSTEIN, the 1921 recipient of the Nobel Prize in Physics, defined insanity as "doing the same thing over and over again and expecting different results".

The physicist's definition is frequently used to describe the irrational behaviour of the active alcoholic and even the second step of the well-publicised 12-step programme acknowledges that only "a power greater than ourselves could restore us to sanity".

Could investors be in need of a recovery programme similar to that practised by those with a desire to stop drinking?

Surely investors do not stumble like drunks from one bull market to the next and are far too prudent to party overly hard and suffer the painful hangover that accompanies an inevitable bear market. Think again.

Prof Vernon Smith, the joint-recipient of the 2002 Nobel Prize in Economics, is a pioneer of experimental economics and has conducted numerous controlled experiments to explore the dynamics of asset markets.

In a recent paper entitled Thar she Blows? Can Bubbles Be Rekindled With Experienced Subjects?co-authored with Reshmaan Hussam and David Porter, Smith examines the impact of learning on the formation of price bubbles.

The authors construct an experimental asset market in which investors trade a security with a finite life of 15 periods.

The security pays a random dividend each period and participants are instructed that there are four equally likely outcomes and given the payouts - zero, eight, 28 and 60 cent - that correspond to each outcome.

Determining the security's fundamental value is a relatively simple task given the information provided. The expected dividend payout in each period is 24 cent and the security's fundamental value at the start of the 15-period trading experiment is therefore $3.60 and declines by 24 cent each period.

Given that participants in the experiment have all the information necessary to determine fundamental value relatively easily, an observer might jump to the conclusion that the formation of a price bubble would be virtually impossible, yet this is exactly what happens.

The inexperienced investors initially price the asset at a significant discount to its fundamental value, but the security becomes overpriced by the fifth period and a massive bubble is created by the 10th period, with the asset often reaching three to four times its fundamental value.

This sets the stage for a crash, which takes place through the final periods as the experiment draws to an end.

Participants in the experimental market are then invited to trade the same security for a second time but, contrary to what might be expected, a further bubble, though smaller in magnitude, is generated by the same players upon returning for a second session.

The bubble gathers momentum far more quickly than in the first experiment, with prices typically moving above fundamental value in the second period and reaches its climax by the seventh period with the security peaking at roughly twice its fundamental value.

One interesting finding in Smith's work is that the participants are aware that the pricing structure is irrational as the so-called echo bubble develops during the second trading session, but are confident that they will be able to exit the market before other players.

Of course, such a belief proves self-defeating as the simultaneous exit by many participants causes prices to collapse.

Echo bubbles have been observed all too often throughout financial history. Indeed, this is the fourth secular bear market since the early years of the 20th century and the initial fall from the peak has always been followed by an echo bubble.

The 49 per cent decline from 1906 to 1907, an 89 per cent drop from 1929 and 1932, a 36 per cent fall from 1968 to 1970 and the halving of stock prices from 2000 to 2002 were all followed by miniature versions of the original that saw prices rise to excessive levels once again from 1907 to 1909, 1932 to 1937, 1970 to 1973 and from 2002 to 2007 respectively.

Echo bubbles are observed both in experimental markets and in historical data, but do investors return for a third time and make the same mistake all over again?

The historical evidence would suggest not, as stock prices have typically traded below long-run average valuation multiples for a protracted period following the deflation of an echo bubble.

Indeed, the price/earnings multiple on trend earnings dropped below average in 1938 and did not move above average until 1955. Similarly, the multiple fell below its long-term mean in 1973 and remained below average until 1987. The evidence from experimental markets corroborates historical fact and Prof Smith notes that "once a group experiences trading a bubble and a crash over two experiments and then returns for a third experiment, trading departs little from fundamental value". However, Smith et al have demonstrated that it is possible to revive a bubble under certain conditions.

The third game is virtually identical to the previous two, apart from significantly higher cash levels and a greater variability of payoffs. The extra liquidity in spite of participants' previous experience sparks a further bubble, though both the duration and magnitude are substantially less than the first two games.

The conditions in today's market environment would appear to be similar to Smith's third experimental market as massive monetary stimulus has allowed stock prices to trade at above-average valuations again.

Stock prices could well go higher, though perhaps Einstein captures market behaviour best when he quipped that "only two things are infinite - the universe and human stupidity and I'm not so sure about the universe."