The Random Walk Theory
While EMT suggests that stock is always efficiently priced (and therefore you cannot outperform the market except as a consequence of luck), another theory suggests that price behavior is never based on anything predictable, but is completely random.
The random walk theory is the belief that price behavior cannot be predicted because it does not act on any predictive fundamental or technical indicators.
The random walk theory contradicts the widely accepted beliefs by both fundamental and technical analysts. Under the theory of fundamental analysis, the long-term value of a company is based on competitive strength, profitability, revenue and market expansion, working capital controls, and capitalization. Under the theory of technical analysis, price is predictable based on chart patterns, momentum indicators, and reversal signals such as those found in candlesticks.
Key PointThe random walk theory is based on a belief that stock prices cannot be predicted, and that all price behavior is the equivalent of a coin flip.
The random walk denies the reliability of any fundamental or technical analysis. It is at least as cynical as the efficient market theory and just as misleading. While short-term markets are clearly chaotic and inefficient, both short-term and long-term pricing can be tied to both fundamental and technical trends. In fact, far from being random, a more supportable belief about price behavior is that fundamental and technical indicators confirm one another, even though the timing is different. In other words, short-term price strength is derived from solid long-term financial strength, and long-term fundamental trends are reflected in the stock's long-term pricing trends as well.
The random walk theory is supported for the most part by academics and some economists, mostly those who rely on theoretical application and mathematical models and less on a study of how invested positions gain or lose value in the real world of the stock market. Those who subscribe to the random walk theory rationalize that prices have to be random because the market is efficient and prices in all currently known facts and influences at each moment. This is illogical, and you can prove this by simply observing the correlation between several forces: earnings news and prices; economic news and economic effects of companies (unemployment, inflation, interest rates, etc.); and changes in sector positions or cyclical movement based on economic change, to name a few.
The theory persists based on academic studies in spite of many developments in analysis, notably free online charting and instant price analysis as a result of the free charting and mathematical tools. The theory has been around for a long time, first introduced in an academic paper by British statistician Maurice Kendall (1953, The Analytics of Economic Time Series, Part I: Prices). Two important books expanded on the idea. The first was The Random Character of Stock Market Prices (1964) by Professor Paul Cootner. The theory was then expanded upon by Professor Burton Malkiel in A Random Walk Down Wall Street (1973).
Key PointThe flaw in the 'proof' of the random walk theory is that price behavior in the market is not the same as a statistical concept. There is a true cause and effect in the market.
Some efforts have been undertaken to 'prove' the random walk theory by statistical methods. A hypothetical stock's price is predicted based on coin flips, for example, when you know that there is an equal chance of the result being heads or tails. When a series of random coin flips is charted and described as 'past performance,' proponents of the random walk claim that the pattern can be used to predict future outcomes of coin flipping.
While the random events like coin flips will produce patterns, and those patterns can be interpreted as predictive tools, there is an important distinction between a coin and stock prices. We know that the coin flip is random and that all future coin flips are going to be random as well. However, in studying stock price behavior of a company with a known history of a specific volatility level, existing within a specific market sector and with a known competitive position, these known factors certainly affect price movement.
For example, a company with a strong dividend yield, low debt ratio, and consistent record of profits is likely to continue to experience rising prices over time. (If you doubt this, check the 10-year history of Wal-Mart or any other successful and well-managed company that dominates its sector). The reality of the factors influencing stock prices makes the coin flip exercise a statistically inapplicable comparison.
A problem in applying statistical modeling to stock prices, especially to prove the random walk theory, is that coins are not going to behave like stocks. Just because coin flips or any other nonâstock price events appear to be random does not mean that stock prices must also be random. Stock price behavior is not random, but is directly influenced by market volatility, financial news, investor and trader perceptions, and dozens of supply-and-demand realities. Just because a stock's price is difficult to predict does not mean it is also random.
The random walk theory conveniently ignores both price trends and momentum. The basic tenet of technical analysis relies on the observed fact that observation of past trends (price direction, momentum, volatility, duration of short-term trends, and more) can be used to predict future price movement. In fact, studies by many market experts provide a strong case to prove that price behavior is not random. Among these tests, a 10-year study by Martin Weber, a behavioral finance researcher, tracked market price changes. Weber found specific patterns that made long-term pricing trends predictable. For example, he found that stocks with upward price patterns and increased earnings tended to outperform other stocks for the following six months.
Key PointThe problem with the random walk theory is that it ignores the easily observed trends and momentum factors that do directly affect price movement.
This is an intriguing concept, because all of these evolutionary concepts also affect a company's ability to survive. Failure to compete (due to obsolete products or better-capitalized competitors), failure to adapt (developing new products as new technology or markets emerge), or the forces of natural selection (the market's preference for one company over another due to any number of reasons) are all interesting. These describe what happens to species in nature as well as to companies in the market.
One of the authors of this paper, Lo, explained this: 'Prices reflect as much information as dictated by the combination of environmental conditions and the number of 'species' in the economy.'
Within any cyclical period, a stock's price movement can appear very random; this does not prove that all prices act randomly at all times. The environmental model makes this point. If a number of animals in one isolated region starve to death and die out because there is no food, it does not prove that a species goes extinct by random selection. Other animals of the same species are likely to thrive elsewhere and even when one species does go extinct, there are usually logical reasons. Some may be hunted into extinction, while others fail to adapt to their environment or run up against the introduction of new competing species. Some may die out for no apparent logical reason, which also fails to prove that evolution is random.