Investment Advisors
When they're hot, they're hot, and when they're not they're not
Jim Dines, editor of the Dines Letter, is fond of saying that a trend in motion tends to remain in motion until it ends. Truer words are rarely spoken. Yet despite its being a tautology, Dines's statement contains a kernel of investment wisdom. It tells us that when in fact we have identified bona fide market trends, following them, on average, will be profitable.
The challenge is, and always has been, to discover and correctly identify such trends. The task is enormous not only because of the number of factors affecting the markets but also because of their complex interdependencies. The task is so Herculean that it is beyond the abilities of any one advisor to comprehend more than even a small fraction of the market's determinants.
A novel approach to meeting this challenge can be built on the deceptively simple premise that many analysts are better than one. A large number of market analysts will be able to comprehend more market determinants than any one financial advisor alone. The theory we have constructed on this premise has been termed the "hot hand theory."
The reasoning behind the theory is simple. Every financial advisor has a particular area of expertise. Furthermore, the indicators and determinants over which an advisor has his expertise will at some times bear more relevance to a market's trend than at others. It thus follows that an advisor will sometimes play the lead of a much "hotter" hand than at others.
This insight would be of little value to the investor if it were completely fortuitous when an advisor's analysis is accurate and when it is not. If being hot one month bore no relationship to being hot the next, past performance would provide no guidance whatsoever to future performance. Though, as the philosopher David Hume pointed out years ago, the future might not be like the past, in fact the markets do exhibit certain regularities. Thus, one would expect the advisors' track records to exhibit similar consistent patterns, as well.
The "hot hand theory" is thus based on the conviction that the advisors' performances do not follow a "random walk." Research has indeed revealed such patterns. It is not merely luck when a Charles Stahl or a Martin Zweig correctly analyzes the market. At such times, the particular market factors they analyze are very good indicators of the market's trend—and because the underlying market conditions that they have accurately perceived do not change quickly, their analysis will, once hot, remain so for some time.
To test this theory I created a hypothetical $10,000 portfolio for each of the major investment newsletters and invested it exactly as those newsletters directed, making the recommended changes along the way. At the end of each month, I computed the value of each portfolio. (My research began on July 1, 1980.)
The resultant performance ratings for each newsletter revealed certain fascinating patterns. Not only did they allow me to ascertain which advisor made the most money in any given month; they allowed me to determine whether an advisor's record of a given month was part of a consistent pattern and, if so, whether his portfolio's profitability was waxing or waning.
The accompanying graphs illustrate the value of locating a "hot hand" in the stock market. The first graph tracks the performance of Value Line's OTC Special Situations Survey (for the $10,000 portfolio created July 1, 1980).
The most striking characteristic of Value Line's performance is the consistency of its trend. Its portfolio increased in value for five consecutive months, from July to November 1980. In December it turned downward, and as of March 1 (as this article is written) it has been declining for three straight months.
Knowledge of such a trend in progress can prove very lucrative. By September 1980, for example, after following Value Line for only two months, it was clear that Value Line was playing the lead of a very hot hand. Over the two months of July and August 1980, its portfolio had appreciated some 19.2 percent, which is equivalent to an annual increase in excess of 100 percent! Acting on this information and purchasing Value Line's portfolio of recommended stocks would have resulted in profits of 43 percent over the next three months.
The performance ratings for Value Line also gave a good "sell" signal in December. Value Line did not lose its "hot hand" all at once, instead cooling gradually in December before steep declines in January and February. Those who held Value Line's portfolio through December lost a modest 4 percent and avoided the decline over the next two months of 15 percent.
Value Line's track record is to be contrasted with the performance of Martin Zweig's portfolio (Zweig edits the Zweig Forecast). Zweig's performance shows clearly the phenomenon of holding a hot hand, losing it, and then regaining it. By September 1980 it was clear that his portfolio was doing very well, appreciating over the previous two months by some 11 percent. And in the month of September, it continued to perform well, gaining another 7.1 percent in value.
But then Zweig began to lose his hot hand. The market continued to surge ahead, yet his portfolio became stagnant. In October it gained only 1 percent in value and then lost that much in November. (The S&P 500, by means of comparison, appreciated some 7 percent in November alone.) Predictably, Zweig's portfolio lost value at an even greater rate in December, losing an additional 2 percent.
In January, however, in spite of a Granville sell signal that sent havoc through the market, Zweig's portfolio turned up again. The up trend was confirmed with another increase in February, when Zweig's recommended stocks gained an additional 2.5 percent. This run up came in spite of a falling DJIA (notice the contrast between Zweig's performance and Value Line's in the previous graph; Value Line lost 15 percent in January and February of this year). As this piece is written, March 1, it looks like Zweig's portfolio is once again following a hot hand and thus is a good bet.
The value to investors of the Performance Ratings is also illustrated by comparing the records of the various "gold bugs." This last year witnessed extreme volatility in gold and silver prices. Some newsletters were able to make money in spite of that volatility, while other newsletters were at the mercy of price fluctuations.
The accompanying graph compares the London price of gold (end of month) with the value of Jim Dines's model portfolio of gold and silver mining stocks. As you can see, Dines's portfolio has fluctuated in value in close proportion to the fluctuations in the price of gold.
To be fair, we must point out that Dines's model portfolio is not specifically created for short-term trading (Dines has held this actual portfolio from January 1975). Yet we should also point out that the greatest losses in his portfolio's value came in January and February of this year, months for which he had been forecasting sizeable increases. In any case, it is clear that his portfolio's value is fairly closely tied to fluctuations in the prices of the metals themselves and thus is not a good hedge against volatility.
This record is to be contrasted with the performance of Charles Stahl's portfolio of gold and silver (Stahl edits Green's Commodity Market Comments). As is clear from the graph, Stahl's portfolio has thrived both when gold and silver have been wildly fluctuating and when they have been traded in a relatively narrow range. Stahl's impressive record—the best of any of the newsletters I followed which traded gold and silver—would lead one to follow his advice when wishing to trade gold and silver.
A final illustration of the worth of the performance ratings came in evaluating the record of Joe Granville's portfolio of recommended stocks. Granville has frequently asserted that market timing is of predominant importance; "you don't need to worry too much about your individual stocks," Granville wrote in the middle of a rather buoyant stretch in the market last fall. "As long as the market is okay, the majority of stocks will be okay."
The graph of Granville's portfolio over the latter part of 1980 and early part of 1981 testifies to the fallacy of his assertion—at least in the bull market which existed then. Over the six months from July to December 1980, when Granville advised investors to be 100 percent long on the market, Granville's portfolio appreciated negligibly. By December 24, 1980, his original $10,000 stood at $10,103, or an increase of just over 1 percent. During the same period of time, the DJIA's original $10,000 on July 1 had increased to $11,041, up 10.4 percent (and the DJIA itself lagged the broader market averages). Clearly, Granville was incorrect in asserting that stock selection was not as important as market timing.
What this teaches us to do is to follow Granville's timing but not his stock selections. For example, if you had followed Granville's market calls by investing in a mutual fund indexed to the DJIA, you would have made money over the last six months of 1980 at a rate some 10 times faster than you would have by buying the stocks Granville recommended.
The performance records of the various investment newsletters can be put to profitable use. They can be used to ascertain who has a hot hand and who ought to be avoided; they can be used to determine which advisor thrives best in periods of price volatility and which do best when a clear trend has emerged; and finally, they can be used to adjudicate disputes such as the one between the relative importance of market timing and stock selection.
The potential applications of this approach are many, and this article just scratches the surface. But I am convinced that the information on the relative performance of the newsletters can be an extremely valuable tool in picking the winners and avoiding the losers, with a minimum of risk.
Mark Hulbert is the editor of the Hulbert Financial Digest.