In the next three lessons, we discuss another approach to relative strength. This method looks separately at 1) relative performance in rising markets and 2) relative performance in falling markets, then combines these two analyses in a unique presentation.
A trader is like a coach. It's the coach's job to put his/her strongest assets on the field, whether the challenge is offensive or defensive.
Risk involves loss, or, more precisely, the probability of loss. Because the probability of loss is greatest when markets are falling, the defensive qualities of a stock are most severely tested during periods of market decline, just when the risk of loss is most acute.
To do well on defense, a stock must out-perform the benchmark during periods when the benchmark declines. In the graphic below, the benchmark separates strong defensive performance (to the left of the benchmark) and weak defensive performance (to the right). The benchmark is always equal to 100.
|
|
Rising markets are risky, too. If a stock cannot score as the market rises, it subjects the trader to another risk, lost opportunity. This risk is highest as the market rises. Therefore, offensive strength is best measured as the market advances.
To beat the benchmark offensively, a stock must out-perform the benchmark as the benchmark rises. A weak offense, on the other hand, under-performs the benchmark as the market rises. The benchmark separates strong offensive performance (above the benchmark) and weak offensive performance (below the benchmark). The benchmark is always equal to 100.
|
A two-dimensional matrix may be used to display the combined offensive and defensive qualities of a target.
|
Here's how to locate a target within the Offensive/Defensive Matrix:
1. Compute the daily percentage changes for the benchmark for 100 days* (column B in the spreadsheet displayed below);
2. In separate columns (C and D), list changes greater than or equal to zero (Plus Days) and changes less than zero (Minus Days);
3. Compute the daily percentage changes for the target over the same 100 days (column E);
4. In separate colums (F and G), first list the target's daily change for all days when the benchmark's change is greater than or equal to zero (F), then list the target's changes for all days when the benchmark's change is less than zero;
5. Compute separate totals of the benchmark's Plus Days and Minus Days over the 100 day period, then compute separate totals of the target's Plus Days and Minus Days over the same period.
|
6. To compute Offense, divide the sum of the target's Plus Days by the sum of the benchmark's Plus Days, then multiply the result by 100:
Sum of Target Plus Days/Sum of Benchmark Plus Days x 100
7. To compute Defense, divide the sum of the target's Minus Days by the sum of the benchmark's Minus Days, then multiply the result by 100:
Sum of Target Minus Days/Sum of Benchmark Minus Days x 100
*Any number of days may be used. 100 days is a good choice for intermediate-term indications.
In the spreadsheet example above, the results are:
OFFENSE = .422/.484 x 100 = 87
DEFENSE = -.297/-.235 x 100 = 126
The target has put up a weak offense over the last 100 days. An offensive score of 87 means that the target did only 87 percent as well as the benchmark as the benchmark rose, and places the target below the offensive benchmark.
The target is also weak defensively. A defensive score of 126 means that the target lost 126% of the total given up by the benchmark during those periods in which the benchmark declined. This places the target to the left of the defensive benchmark.
The position of the target is shown in the matrix below. A target in this position is weak in both rising and falling markets.
|
If the benchmark is considered as a target which is measured against itself, then the benchmark's defensive and offensive scores are always equal to 100. For that reason, the benchmark is always located at the center of the matrix, at the intersection of the offensive and defensive benchmark indications.
|
Now suppose we alter the volatility of the benchmark-as-target. If volatility is increased by, say, ten percent, then offensive and defensive scores will each increase by ten percent to 110 and 110, respectively. That new target will be located to the northeast of the original benchmark. If this computation is carried out repeatedly, randomly increasing or decreasing the volatility of the benchmark-as-target, these targets will all fall on a diagonal line extending southwest and northeast from the center of the matrix.
|
Theoretically at least, an infinite number of such targets could be calculated, forming a solid diagonal line. This line is The Benchmark Equivalence Line, or BEL.
|
Along the BEL, offensive and defensive performance offset each other. Targets on the BEL to the northeast of the benchmark (matrix center) are stronger offensively but weaker defensively than the benchmark. As a result, the relative performance of those targets is quantitatively equivalent to the benchmark's. There is a behavioral difference, however, which is qualitative; targets to the northeast of the benchmark which lie on the BEL are more volatile than the benchmark. These targets rise more than the benchmark as the benchmark rises, but fall more than the benchmark during those periods when the benchmark falls.
|
Similarly, targets which lie on the BEL to the southwest of the benchmark are less volatile than the benchmark. These targets drop less than the benchmark as the benchmark falls and move up less than the benchmark during periods when the benchmark rises.
In the matrix below, identify each target as 1) offensively stronger or weaker than, or equal to, the benchmark, 2) defensively stronger or weaker than, or equal to, the benchmark, 3) overall stronger or weaker than, or equal to, the benchmark, and 4) more or less volatile than, or equally volatile to, the benchmark. The correct answers are given below the next graphic.
|
|
A number of years ago, shortly after I had worked out the calculations covered in the last two lessons, I showed my business partner the results. He took one look and asked, "Why don't you show the way the target moves over time?" Eureka! This was a simple yet brilliant idea, and I rushed back to the computer. Later, I showed Jack the chart. "It looks like a worm," he said, and the name stuck.
In the lesson, "Offense and Defense", you were introduced to the method for calculating the position of a target within the offensive-defensive matrix. To create a worm chart simply repeat that calculation sequentially over some period. When the results are plotted on the matrix, the target will trace out a line from the beginning of the period to the end.
In the matrix below, a target moves from an initially strong position in the northwest quadrant, across the BEL, to a relatively weak position in the southeast quadrant.
|
From the above chart we see not only the relative strength of the target, but the trend of relative strength against the backdrop of our matrix.
The next series of charts shows actual worm charts as of September, 1999. Each worm covers a period of twenty weeks from start to end. Data points are weekly.
The strongest sector during the late summer was Technology. Note that the worm gathers near the BEL before breaking through in a direct northwest direction. Such action is typical of targets which are emerging as strong leaders. The northwesterly trend of the worm indicates that both offensive and defensive strength are increasing at once, a very bullish sign.
|
The worm chart of the Financial sector presents an entirely different picture. The worm is weak and meanders toward the southwest (weakness) over the period.
|
The next chart shows an index made up of closed-end European country funds. The group has been basing for nearly one year.
|
The worm chart of the Europe group (below) shows that a recent history of relative weakness has placed the group southwest of the BEL. Nevertheless, the worm is trending very strongly toward the northwest. Note, too, that the worm is situated to the left of the defensive benchmark line. This is evidence that the group is now relatively resistant to decline and may be sold-out.
|
Superficially, the bar chart of the Internet group (below) is similar to that of the Europe group. Both groups show basing activity after a decline. In both, volume tends to dry up as basing proceeds, a generally positive development.
|
However, the worm of the Internet group shows a relative-strength trend quite different from that of the Europe group:
|
Beginning from a strong position NW of the BEL, the Internet group charted a direct course to the SE quadrant, giving up both offensive and defensive strength in equal measure. Until the worm turns, this group will not be a good candidate for purchase.
The next chart of the Retail Building Supply group shows a worm which tracks the BEL to the northeast.
|
Quantitatively, relative strength of the Retail Building Supply group roughly approximates that of the benchmark, indicated by a worm which remains near the BEL. Qualitatively, this group is becoming more volatile, relative to the benchmark, as the worm tracks toward the NE.
The worm chart of Retail Office Supply (below) demonstrates another way in which the worm respects the presence of the BEL by its behavior. Note how the worm first moves north toward the BEL before veering away to the east just as the BEL is approached.
|
After a quick rally from support in the mid-70s, MCK drops again on large volume to D (see below). This time, support is broken, establishing a pattern of lower highs and lower lows. The new low at D and a huge increase in volume on the decline together constitute a clear sign of weakness. After the initial low is made at D, the stock rallies weakly to supply in the high 70's. This may be an opportunity to sell, but the stock is still very oversold. We decide to wait for a better opportunity, should one come.
|
After a shakeout on volume to a minor new low, MCK rallies to E. The rally begins well, but immediately runs into trouble as the previous zone of supply is reached. Distribution above 80 is broad. The small base which has formed around 70 is certainly not yet enough to mount a serious challenge of supply evident in the neighborhood of 80. We are now alert for an opportunity to short close to protective supply.
The stock rallies, and manages to eke out a recovery high just above the peak of the earlier rally from D. Volume remains high and the high-low spread narrows. This is a sign that buyers are again encountering sellers at this level. The new recovery high has doubtless stopped out or scared out weak shorts, a bearish development. This shakeout has provided the best moment to date to establish an initial short position. We place an order to sell short at the market.
|
Our short begins to work almost immediately, a sign that we are on the right side. A new low is followed by a quick rally, indicating the presence of support in the high 60's. Only a solid move below support will convince us that we should add to our position.
|
So far support has held. Trading becomes dull around 70 and volume declines in a way that suggests that sellers are temporarily exhausted. By now, a fairly broad base has been built up between the high 60s and mid-70s, and volume has steadily declined. This is evidence of reaccumulation, and the rapid drop-off in volume and quiet price action may set up a rally. We have to pay close attention to our trade.
The stock begins to rally (below) on moderately improved volume. At F the price betters a daily high made four sessions earlier, and the price appears to be moving easily. We elect to exit our short position until the situation turns more favorable.
|
The rally carries above supply at E to 90, the point at which MCK first encountered distribution. At G, price reverses a strong intra-day upthrust on extremely heavy volume, action which we recognize as climactic. The rally has taken price into territory controlled by heavy-weight sellers. The high-volume reversal is an invitation to sell short. All conditions required to short are now fully realized. The follow-up rally back to supply at 90 provides ample opportunity to short at the market. The price is extended and near important supply. The weight of the evidence now allows us to take on a larger short position than we did in our initial probe of the short side at E.
The next chart tells the rest of the story.
|
In the following examples, identify each of the four key features of good short candidates.
Below: Is the clearest opportunity to short at the high in this chart? If not, where?
|
Below: Once the price breaks down from the top, is the opportunity to short gone? What is the best point to establish an initial short position? Where should the short position be increased?
|
Below: This stock exhibits an axis about which distribution takes place. How does the axis help to define levels at which shorts should be placed?
|
Below: Where is the sign of weakness? Where should the initial short be taken? Are there other well-defined opportunities to sell short? How many?
|
Below: What should you watch for now?
|
Below: How does the reverse use of Trend lines help to identify a sign of weakness?
|