by Editorial Staff
Updated: March 18, 2015
In this excerpt from The Wave Principle of Human Social Behavior and the New Science of Socionomics, author Robert Prechter dismantles the popular -- yet problematic – random walk theory, Wall Street's preeminent view of market behavior. Prechter also proposes his own, carefully thought-out solution -- see if you agree.
For years, some theoreticians have argued that stock price movements are random. Their assertion under the Efficient Market Hypothesis is that all investors make informed and rational decisions, weighing more or less identically the meaning of various events and conditions that affect markets and immediately adjusting investment values accordingly. Since no one can predict random outside forces, markets fluctuate randomly. Statisticians have run tests on financial market prices to demonstrate that they follow a "random walk" and are therefore unpredictable. "For two decades," said Fortune magazine in 1988, "finance professors have taught EMH as if it were as indisputable as the laws of gravity."
Go back to Chapter 4 and review the simple computer-generated model that Elliott Wave International designed to create a simulated market based upon the Wave Principle. Clearly every single movement in the model is utterly determined; after all, a simple formula produces it. Did you know, however, that the standard statistical methods of assessing the presence of determinism would find that the results of our model were random? How is this error possible? Stephen R. Cunningham, Assistant Professor in the Department of Economics at the University of Connecticut provides an answer. His research paper in the Review of Financial Economics reaches this stunning conclusion:
Neither the Samuelson-Fama tests for efficient markets nor the popularly-used augmented Dickey-Fuller (1979) test for unit roots can successfully discriminate between a fully deterministic time series, generated from a nonlinear (chaotic) process, and a random walk. A researcher applying these methods to a simple nonlinear price process would be misled into believing that such a series is a random walk.
This fact proves that the champions of EMH and random walk do not know, despite their claims, that markets are random. The failure of EMH researchers' statistical applications successfully to determine randomness in utterly determined chaotic data series invalidates the empirical basis of their work. This book, I hope, invalidates the theoretical basis of it, the idea that aggregate financial market behavior is based upon the efficient, rational processing (error #1) of extramarket information (error #2).
Standard-bearers for random walk smugly joke of producing random data series that look so much like stock charts that they fool people into thinking they are stock charts, as if this proves something. The definition of the word random is that the data can look like anything, so why shouldn't some of them look like stock charts? Similar charts of random data and stock prices no more prove that stock prices are random than they prove random data are stock prices. If randomly generated dots on a scale happened to produce music, it would hardly prove that music is random. In technical terms, as Paul Montgomery points out, such assertions commit the logical error of "affirming the consequent."
A champion of random walk, still plying his axioms in the very latest issue of Bloomberg's Personal Finance magazine, uses this common argument in favor of market randomness:
"More than 90 percent of professional mutual fund managers were outperformed by the S&P 500. If it were that easy to earn excess returns by exploiting the predictable patterns in the market, we should be able to find a substantial number of professionals who are able to do so."
This is nonsense, for several reasons.
(1) In the past decade, most of the money in the stock market has been handled by professionals. To demand that professionals outperform themselves is to demand the negation of a tautology.
(2) Using the S&P, one of the best-performing investments among all stock indexes, bonds and commodities in the entire world over the past 16 years, as a benchmark for money managers to exceed is an inverted straw-man argument.
(3) To demand that professional investors beat a bull market is to create a negative-sum game and then insist that the majority win at it. Every manager has to have some cash, and every manager pays commissions. By definition, it is impossible for the majority to beat a bull market. Even random walkers must reject this cute ruse. It is quite certain that if we were to isolate a bear market period in which many money managers beat the market simply because many of them held some cash, random walk proponents would not then declare such performance as evidence against their model.
(4) Beating the market is a false standard. To be consistent, random walkers should also insist that portfolio managers beat the market on the short side in bear markets. After all, market prices are simply ratios, making direction irrelevant. Carried to its extreme, their benchmark demands no less than constant outperformance in every market fluctuation, which is absurd. The only valid question is whether a manager makes enough money to make investing with him a good idea relative to what you would do as an individual.
(5) Nonrandomness hardly means that earning excess returns should be "that easy." This is a blanket substitution of one idea for another. Chapter 8 explains why it is anything but easy.
If professionals are operating under false premises, the patterns they perceive would be inconstant and therefore inadequate for reliable forecasting. In this case, their failures would not prove randomness, but epistemological error.
The argument presents a false dichotomy, an "either-or" that is not necessary. Random walk is a possible answer to the more general question of why few people do well at investing but hardly a necessary one. Impulsive herding behavior would explain why most people do not perform brilliantly in financial markets. It also happens to be the actual reason.
As A.J. Frost and I said in our 1978 book, "the Elliott Wave Principle challenges the Random Walk Theory at every turn."
As the Wave Principle reveals, the overall pattern of stock price movement is nonrandom and indeed so formally constructed that price fluctuation cannot (as argued in Chapter 18) be the result of reasoned decisions by well-informed individuals dealing continually with new information, as is commonly assumed.
It is my contention after closely observing the market and its participants for twenty years that few investment and trading decisions are based on reason, logic and knowledge gained from comprehensive research.