WASHINGTON, Oct. 28 (UPI) -- The Efficient Market Hypothesis, first expounded in 1965 in a paper by Eugene Fama, is the capstone of modern finance theory, that has been responsible for three Nobel prizes in Economics (Modigliani, 1985, Markowitz/Miller/Sharpe, 1990, Merton/Scholes, 1997), although curiously enough not one for Fama himself.
Now that almost everyone involved has got the Nobel, it's time to drive a silver stake through the hypothesis' heart. It bears no relation to market reality, and conclusions drawn from it are in most cases flat-out false and economically damaging.
There are three forms to the EMH, which makes it difficult to argue with -- if you find an counter-example to one of the forms, proponents simply retreat to one of the others.
The weak form asserts that all past stock market data are reflected in securities prices, in other words that technical analysis is of no use.
The semi-strong form asserts that all publicly available information is fully reflected in securities prices, in other words that fundamental analysis is no use.
The strong form asserts that all information is fully reflected in securities prices, i.e. even insider information is of no use.
The strong form can be disposed of fairly quickly. If insider trading is no use, why are they likely to prosecute Martha Stewart? Nobody involved in the market actually believes that insider information is useless.
The semi-strong form is the most interesting case. American Enterprise Institute resident scholar Kevin Hassett on Friday presented his new work "Bubbleology," studying the phenomenon of stock market bubbles.
Hassett, although holding an economics doctorate from the University of Pennsylvania, not Chicago, is an ardent proponent of modern finance theory, complete with the obligatory sneering references to market practitioners who are incapable of comprehending the finer products of the economic mind.
Notoriously, he was also author, with James K. Glassman in 1999, of a tome called "Dow 36,000," which suggested that the appropriate valuation for the Dow Jones Industrial Index is at that level.
Hey, Kevin, if the market's so efficient, why isn't it capable of pricing equities within a reasonable margin -- say, 75 percent or so -- of the level you believe to be appropriate?
According to Hassett, the market is indeed efficient, except that at the "frontier" of human knowledge, bubbles may form. These bubbles are benign, since they result in huge resources being devoted to high-tech and human progress, far more than would naturally be so devoted.
He gave the example of the California gold rush of 1849, which filled up California much more quickly than it would otherwise have done.
Two can play at that game. I give the counterexample of the Yukon gold rush of 1897-98, which filled up the Yukon far more rapidly than would otherwise have happened. We're still waiting for that one to pay off -- it's only been 105 years.
Alternatively, the South African gold and diamond rush of 1870-1895, which resulted in South Africa being filled with speculators more quickly than would otherwise have happened, thus caused a fairly major war in the short term (Boer War, 1899-1902) and the world's most intractable racial problem in the long term.
However, the most important implication of Hassett's conclusion is that the efficient market hypothesis is tosh.
The existence of bubbles of the size of the Internet bubble of 1996-2000 is not just a minor anomaly in the theory. Because of its size, the bubble invalidates the theory altogether.
Market efficiency, of whatever form, is like virginity: the market cannot be efficient in some ways and not others -- if the market is non-trivially inefficient, the theory falls down and the conclusions that have been drawn from it are invalid.
There are other indications that the EMH is not what it might be. As is well known, the hypothesis has been used to justify stock index investing, since money managers cannot outperform the market.
However, according to a Morningstar study discussed Monday in TheStreet.com, over the past two years, fund managers with more than 20 years tenure in their position significantly outperformed both the stock market average and the average fund manager.
Since the tenure in question was accrued before the period over which returns were studied, there can be no "survivorship bias" in this study.
A further problem with the semi-strong form of the EMH is the quality of the information itself. Standard and Poor's Thursday reported that "core" earnings per share of the companies in the S&P 500 index, in the year to June 2002, totaled $18.48, not $26.74, as derived in earnings according to generally accepted accounting principles.
The principal sources of the difference were accounting for stock options, in which GAAP reported earnings did not provide any charge for this expense, and pension scheme accounting, in which companies had reported as earnings notional gains on their pension funds, when in fact those funds had shown a loss.
In S&P's view (with which I concur) neither factor is repeatable in the long term.
Thus, the "core" earnings of the S&P 500 Index, those earnings that could be expected to be sustained, were more than 30 percent lower than those reported -- and the current "trailing" price-earnings ratio for the index, even at today's allegedly depressed market levels, is about 48, not 35 as originally estimated.
This is important. It is now very clear that in strong markets, the information provided to investors becomes substantially degraded in quality, with accounting tricks used to cover up lapses in earnings growth. In such cases, the whole basis of the EMH -- that intelligent and rational investors make decisions based on all available information -- is degraded.
If the information provided is faulty, the decisions become sub-optimal, and the market ceases to be efficient (nor is it a random walk, because the great bulk of the errors are in the same direction.)
Even the weak form of the hypothesis is questionable. Studies have shown that the stock market is significantly leptokurtic, in other words, trends persist for longer than would be expected from a purely random distribution. This is information that can be and is used to trade profitably.
Studies carried out at Oxford University and the London School of Economics have provided further examples of potentially profitable trading anomalies. Thus, while the weak form of the EMH appears to be generally true, it too is not true in all circumstances and hence not an axiom that can be relied upon.
If, as I have suggested, the EMH is faulty, then huge swathes of current practice in the financial markets and the U.S. economy as a whole need to be re-examined.
Investors have been told for the past 30 years to invest in index funds, because money managers cannot beat the index.
Of course, most money managers cannot beat the index: since they represent the majority of invested money, it would be almost mathematically impossible for them to do so.
Most baseball players can't bat .350, either: the major league average is about .275. But just as there is the occasional baseball player, a Ted Williams or Barry Bonds, who at the peak of his career can bat .350, so it is clear that there are a few money managers who can, if given the right circumstances (freedom from second guessing and -- very important -- from corporate reorganizations) beat the market on a consistent basis.
Peter Lynch, at the Fidelity Magellan Fund, was one such manager; his successors were not so successful.
Like hitting .350, beating the market is not something that can be taught because, as a new technique is found, the opponent (the pitcher or other money managers representing the market as a whole) adapts to it, and the new technique ceases to work better than average. But the ability is real, not random.
Further implications, of great importance for the U.S. economy, are that dividends do matter, leverage does not necessarily increase returns, and options pricing models are inadequate.
Under modern financial theory, the investor should be indifferent between receiving his return in the form of dividends or capital gains. Since dividends are tax-inefficient (and do not benefit managers holding stock options), companies have moved increasingly to de-emphasize dividends and reinvest company earnings.
However, if the theory is false, then so is the conclusion. In practice, reinvesting dividends removes from investors a very useful control on management and encourages management to take risks with their money in order to increase the stock price.
Since the agency between stockholders and management is highly imperfect, the money subjected to that agency, and the period for which it is so subjected, should be limited as far as possible.
Not only should dividends be maximized, consistent with preserving the existence of the company, so too should returns of capital, in which stockholders in a slow growth company should themselves have the opportunity to re-deploy their money.
Certainly acquisitions and mergers, in which the agency problem is particularly acute, should be undertaken only after a supermajority (maybe 75 percent) vote in favor by the stockholders of the acquiring or merging company or companies.
There would be a lot of richer Time Inc. stockholders today, if that company hadn't successively merged with Warner Communications, Turner Communications and America Online. However, in each case the company's management made out like bandits.
As for leverage, conventional analysis in the '50s and '60s looked at the possible downturn in a company's operations in the worst imaginable bear market, and borrowed money only to the extent that it could be repaid even in that event.
Modern financial theory has thrown that technique out of the window: it is supposed to be more efficient for companies than individuals to borrow, while individuals can balance the risk on their portfolios by holding part of their wealth in Treasury bills.
Thus, since debt interest is tax-deductible and debt finance thus initially cheaper than equity finance, companies borrow money until it is almost impossible for them to borrow any more, even in a bull market for their services. This enables them to raise earnings per share artificially, by deploying the borrowed funds in their business (possibly repurchasing shares if no other profitable outlet can be found.)
Of course, management has stock options and makes a great deal of money during the borrowing phase, as earnings -- and thereby the stock price -- are artificially increased.
Once the downturn comes, and the company gets into difficulty, management expresses surprise -- but can quite often keep its jobs. The financial difficulties, which are easily manageable under the lax Chapter 11 bankruptcy provisions, generally result in a substantial partial reimbursement for the lenders, while the stockholders are wiped out.
The EMH, and more particularly the Capital Asset Pricing Model with which it is associated, also underpin the Black-Scholes options pricing model, variants on which have been used to value and hedge options positions in all markets since its invention in 1973.
Here, the problem is simple: the model rests on assumptions that are untrue.
In particular, for the model to be valid, prices must move continuously, without sudden jumps, and in a random walk pattern, without leptokurtosis -- the model has as a fundamental assumption Bayes' Theorem, formulated by Rev. Thomas Bayes (1703-61) under which the probability of several unlikely events all happening is the product of their probabilities.
In Bayes' analysis, based on simple mechanical experiments, events are truly independent and the theorem is valid.
In the real 21st century world, events are interdependent, and the chance of several of them happening at once is closer to the minimum of their chances (as it would be under fuzzy logic analysis) than their product -- in the case of four 10 percent probabilities, 10 percent not 0.01 percent.
Thus as Long Term Capital Management, which famously collapsed in 1998, and others have repeatedly found, extreme outcomes invalidating Black-Scholes hedging techniques happen far more frequently than is suggested by the mechanistic model. However options and the Black-Scholes pricing models remain popular, as they offer opportunities for nice profits by traders -- at the cost of the occasional catastrophic loss to stockholders.
More than any other single cause (except perhaps lax monetary policy), the EMH and modern financial theory in general was responsible for the late '90s bubble.
Far from being beneficial or confined to dot-coms, that bubble extended over the entire economy. Its unwinding, which has yet much further to go, is causing a deep recession with untold long-term damage to confidence, jobs and the well-being of the American people.
The EMH was built on shaky theoretical underpinnings. It is now tottering. Best push it over the edge and move on -- though I suppose nobody can take back the Nobel prizes!
(The Bear's Lair is a weekly column that is intended to appear each Monday, an appropriately gloomy day of the week. Its rationale is that, in the long '90s boom, the proportion of "sell" recommendations put out by Wall Street houses declined from 9 percent of all research reports to 1 percent and has only modestly rebounded since. Accordingly, investors have an excess of positive information and very little negative information. The column thus takes the ursine view of life and the market, in the hope that it may be usefully different from what investors see elsewhere.)
