Strategies for Trend Following Markets

Rising strength-following markets are the best of times. If contrarian markets are a time of tilling and sowing, then strength-following markets are a time of harvest.

When both prices and the Spread are rising, the most profitable strategy is to buy and hold the strongest stocks. During these periods the trend is strong, and the best performing stocks attract a large and enthusiastic following. Avoid weak issues, even those which have built impressive bases. These stocks are likely to underperform stronger issues.

The signal feature of strength-following markets is that, on average, the higher the RS rank, the better the forward performance. The chart below shows a period during which the Spread rose. The black (lowest) line indicates the average performance of all stocks in the subject universe. The solid blue line shows the cumulative forward performance of the set of all stocks with an RS rank of one (1) or more. The dashed lines show the performance of increasingly highly ranked subsets of strong stocks. The higher the RS rank, the higher the forward performance.

By contrast, the performance of strong stocks during contrarian markets turns the above chart on its head. The chart below shows the performance of stocks with an RS rank of one or better during a contrarian period, as the Spread declines. Again, the benchmark is shown in black. The solid blue line is the forward performance of all stocks with an RS rank greater than one (1). Note that the average strong stock underperforms the benchmark. As RS increases, performance deteriorates further. The dashed lines show the performance of increasingly highly ranked subsets of strong stocks. The higher the RS rank, the lower the forward performance.

Ranking Stocks

When traders are in a strength-following mode, the most profitable information is the relative-strength ranking of stocks and groups. Focus on the strongest stocks within the strongest groups

Once RS is determined for each target in the universe of interest, ranking by RS is easy. Shown below is the rank, from strongest to weakest, of 84 groups as of late August, 1999.

Because the benchmark is always equal to 100, groups above 100 are relatively strong, while those below 100 are relatively weak.

While not neccessary, it is nevertheless useful to normalize RS rank as a percent of maximum and minimum extremes. Entries equal to or greater than 100 (the benchmark) are compared to the maximum entry. The top group in the above list (Semiconductors) is, of course, 100 percent of the maximum entry. The next, Oil Drilling, has an RS which is 80 percent of the maximum, and so on.

Entries below 100 are compared to the minimum entry. To get this calculation right requires that 100 be subtracted from each entry before a comparison is made. So, for instance, the bottom group, Mfg. Housing, converts to a reading of -33.75 (66.25 minus 100). The next group up from the bottom, Drugs, converts to a reading of -25.38, and so on. Then divide each converted entry by the absolute value of the lowest in the list, in this case Mfg. Housing, to arrive at a percent of the minimum. For Mfg. Housing, divide -33.75 by 33.75 and then multiply by 100. The result is 100 (-33.75 / 33.75 * 100 = -100).

A similar calculation is carried out for all other groups with an RS rank of less than 100. The calculation for Drugs is: (74.52-100) / 33.75 * 100 = -75. The Drug group is only 75 percent as weak as Mfg. Housing.

Below is a truncated version of the RS ranking, converted to indicate percent of maximum and minimum RS (allow for rounding errors).

The Janus Strategy

The chart below shows the results of strength-following and contrarian strategies over a period of twelve years, from early 1987 to mid-1999. The target universe is composed of a diversified list of group indexes.

The protocol used is simple. The relative strength of each group index is calculated using price data from the previous twenty-week period. Groups are then sorted by relative strength. Those with an RS > or = 1.00 are assigned to a strength-following portfolio, and those with an RS < 1.00 are assigned to a contrarian portfolio. Performance over the following week is recorded for each portfolio. Weekly results for each portfolio as well as for the benchmark (average group) are cumulated over the twelve-year test period.

Over the whole twelve-year period, strength-following proved to be the more productive strategy, beating both the benchmark as well as a contrarian strategy. This is due to several factors. First, the Spread rose for seven of the twelve years cumulatively, generating more favorable results for strength-followers about 58% of the time. Second, when the Spread rose, prices also rose nearly 70% of the time. In contrast, a falling Spread gave contrarians the advantage for only five years out of twelve, and then prices rose only 50% of the time.

Even though a strength-following strategy outpaced a contrarian strategy over the full period, traders nevertheless favored laggards over leaders better than 40% of the time.

In the chart below, periods during which a contrarian approach yielded better results than strength-following are shown in yellow.

THE JANUS STRATEGY

While a strategy of staying with relative-strength leaders has outperformed the average, shifting strategies in accord with the direction of the Spread, from strength-following to contrarian bargain-hunting and back again, has yielded even better results.

The Janus Strategy takes advantage of the two-sided nature of the market, holding only relatively strong groups as the Spread rises, then switching to RS laggards as the Spread falls.

In the following example, a simple ten-week moving average of the Spread is used to trigger the switch back and forth between contrarian and strength-following strategies. When the Spread is above the moving average, RS leaders are selected; when the Spread is below the moving average, RS laggards are selected. The next chart shows the twelve-year Spread against its ten-week moving average.

The next chart shows the cumulative results of strategic switching from RS leaders to RS laggards in accord with the direction of the Spread:

The fact that a crude switching mechanism can lead to such an improvement in performance is evidence that the Janus Strategy is robust.

Caution: The methods discussed in this chapter should not be used mechanically to make trading decisions. A ten-week moving average was employed for purposes of back-testing only. In a test of this sort, it is best to adopt some simple, fixed switching device in order to escape the criticism of having rigged results with perfect hindsight. The point of these demonstrations is only to impress on the student the strategic importance of the market forces captured by the Spread.

Caution

I have included the present chapter reluctantly. There is a risk to learning the use of mechanical indicators before the student has mastered the principles taught and demonstrated in earlier lessons. Beginning traders who rely too heavily on indicators, oscillators and other such devices before they have learned to read the tape directly with skill and judgement do themselves a disservice.

I recommend that you give this chapter a cursory review, then return for a more careful study only when you have gained some proficiency in trading without these tools.