Stock Market Trading Rules

Despite the widespread agreement, some participants in the debate remained skeptical. Indeed, prescient commentators of that era ocProzac Market ocasionally expressed the fear that the interrelationships of stockprice changes are so complex that standard tools like these cannot reveal them. That fear led to efforts to dispute the model by designing trading rules that could achieve above-normal returns by uncovering and exploiting these greater complexities. Among the most primitive though most illustrative trading rules was Sidney Alexander’s “filter technique.” This is a strategy designed to discern and exploit assumed trends in stockprices that, in Alexander’s piquant phrase, may be “masked by the jiggling of the market.” For instance, a “5% filter rule” for a stockwould say to buy it when the price goes up 5% (and watch it rise to a higher peak); then sell it when the price goes down 5% from that peak(and watch it fall to a lower trough); then short the stock(i.e., borrow it and sell it at the prevailing price, promising to repay with the same stock, to be purchased for the price prevailing at the time of repayment); then, when the price rises 5% from that trough, cover the short position. If this works, you get a gain on the initial sale plus a gain on the short position. More important, if it works, prices are following a peak-trough pattern. That means they are not random and the random walkmodel is contradicted. Alexander’s initial results indicated that such a technique could produce above-normal returns. Subsequent refinements of Alexander’s workby himself and others, including Fama, however, demonstrated that relaxing or changing certain assumptions eliminated the abnormal returns, particularly the original filter technique’s failure to note that dividends are a cost rather than a benefit when stocks are sold short. Alexander’s filter technique epitomizes the chartist or technical approach to stockanalysis and trading, under which a study of past prices (or other data) is used as a basis for predicting future prices. Indeed, Alexander’s filter technique is a conceptual cousin of limit orders and similar techniques prevalent in securities trading today. These techniques include conventional technical methods that rely on anomaly effects (the insider, month, weekend, and analyst effects) as well as the more unconventional methods (the hemline indicator, the Super Bowl indicator, and so on). These and related philosophies such as “momentum investing” and “sector rotation” remain staples of Wall Street futurology. They are widely and increasingly used by traders and recommended by investment advisers and brokers. They are nonsense, as many students of the random walkmodel (and EMT) recognize based on the foregoing analysis. They are nonsense not because of EMT but because they fly in the face of business analysis. “We shall dismiss these with the observation that their workdoes not concern ‘investors’, the trouble with all these tests of the random walk is that they are linear. They do not investigate the presence of nonlinear price dependence, something that in the early 1960s researchers simply lacked the computer horsepower to do. The trading rule test, for example, is linear in that it operates in chronological time (or real time). Neither it nor the other old tests consider the possibility that market time may be better understood from a perspective that is nonlinear. Einstein demonstrated that time is not absolute but works in dozens of different ways depending on the context, including forward (or linear), backward, circular, slow, and erratic (nonlinear), and can even stand still.