So when does a bear market end? Along with everyone else out there, I first started asking myself this question last year. Turning to historical bears for my inspiration, I started to make charts of old Dow Jones bears for an idea of how they looked. In hindsight, I can tell you that the trick to finding the October bottom in markets last year was not to use old American bear markets, but to use the Japanese bear market in 1990-1992 as your model. In the following chart I have aligned the Nasdaq, starting from its top in April 2000, with the Nikkei, starting with its top in January 1990.
The thing that struck me when I first saw this chart last December was the precise number of days both markets spend falling from top to bottom - in both cases the bear lasted exactly 647 trading days. Remarkable? Or coincidence? If you had the guts to buy the markets last October using this simple analog model as your guide, you would have done fairly well.
Before analyzing the analog more, let's turn to the greater question of analogs. When one compares any market to another, generalizations will occur. These generalizations allow for comparison, but they also hide the peculiarities underlying a market that naturally inhibit comparison. The Nasdaq in 2000-2002 and the Nikkei in 1990 were different markets in distant countries, driven by varying arguments, conflicting emotions, and entirely different underlying fundamentals conditions - thus their particulars prevent them from being compared. Yet both were manias governed by universal human emotions like fear and greed that have always existed - these universals allow for comparison. In the end, I think a balance is necessary. Analogs such as the one above are very useful for getting a general feeling about markets, but to follow them fanatically is a recipe for disaster.
How might this lead to disaster? My paper on the 1987 stock market crash posits that many isolated individuals started using an analog model for trading markets in 1987. This analog model was based on historical price action prior to the 1929 crash, and in the same simple style as the Nasdaq-Nikkei analog, purported some sort of basic linkage between 1929 and 1987. Because it attracted so much attention, the analog actually became a self fulfilling prophecy. The 1929 crash reoccurred in 1987. This is a very simplistic explanation, read the paper for something with more depth.
Having said all this, what can one do with the Nasdaq-Nikkei analog? Well, note that the rally in Japan petered out after about 10 months. It then commenced the lethargic up-and down bounce that would characterize Japanese trading through most of the 1990s. The Nasdaq, also 10 months into its rally, is reaching the same temporal point that the Nikkei reached before halting. The conclusion? Given the past usefulness of the analog in picking out a bottom, and given the tendency of people to behave in similar patterns irrespective of time, and given the chance that enough isolated traders might also use the Nikkei-Nasdaq analog such that it self-fulfills itself, it is now a better time to sell than buy.
Note: To see the data in raw form, go here.
© JPK and Lope
2002