Current
Observations
March 4, 2011
--THE CYCLES--
In searching for a potential top, we decided to look more closely into the theory of a friend who used to write a market letter but who now concentrates on money management. His name is Terry Laundry and his theory of identifying possible market tops is based on his so-called T Theory.
Terry's basic premise is that there are ways of measuring cash buildups that generally occur during market declines as people are selling and those cash buildups comprise the left-hand side of the letter T. Once the cash buildup ends, a rally begins and the distance from the center post of the T to the right side of the T where market tops occur matches the distance from the beginning of the cash buildup to the center of the T. In other words, the length of the left-hand side of the T determines the length of the right-hand side of the T where market tops are formed. Although the theory is far more sophisticated than this simplistic explanation, the simplistic interpretation would simply say that a market rally lasts as long as the preceding decline. Terry has experimented by using different advance decline oscillators and volume oscillators over the years to determine the length of the cash buildup, the left-hand side of the T. But one of the easiest ways to measure Ts is to use the cumulative advance decline line of the New York Stock Exchange. The large chart on the front page of today's newsletter will attempt to show you how eerily accurate these advance decline Ts can be and also how we need to be prepared for the possibility of a market top of great importance scheduled to occur over the next 2-3 months. That is not necessarily Mr. Laundry's own interpretation of his theory, but I have had some long discussions with Terry over the past few months and pointed out to him how his theory pinpointed to within four trading day the top of the advance decline line in 2007, a top which ultimately led to the second worst bear market of the past 100 years.
The chart above shows the daily advance decline line of the New York Stock Exchange going back to 1926. First we will show you how Terry's Ts led to the most important tops in the advance decline line of the last several decades. One of the things that Terry discovered over the years is that the center post of some Ts is best located at the midpoint of two market bottoms rather than at the first or second bottom individually. As we started to experiment with time spans using Terry's T Theory, we also found that some Ts are best measured by beginning the left side of the T midway between a double top rather than at one individual top or the other. We also noticed that sometimes locating at a midpoint between two tops or two bottoms worked best but at other times, more accurate results were obtained from using an individual top or bottom rather than the midpoint of those tops or bottoms. Let's give you one more general rule before we get to the application of the theory. In determining where the center post of the T should be placed, the general rule is that a center post is placed at the first bottom which leads to a higher high than the high seen between the current low and the low preceding it. So, for example, the 1974 low which is the lowest point on the chart is probably not a good candidate for the center post of a T because the rally following the 1974 bottom failed to move above the high between the 1970 and 1974 bottom.
One could argue that the 1982 bottom was a candidate for the center post of a long-term T because the rally after the 1982 bottom appeared to at least match if not slightly surpass the high between the 1974 and 1982 bottoms. In reality, one could argue that either the 1982 bottom or the midpoint between the 1974 and 1982 bottoms would be good candidates for the center post of a T and we will show you momentarily how well each of those choices worked in pinpointing a future top in the advance decline line. But we believe the prime candidate for a center post of the very long T formed from the tops in the A-D line seen in the 1950s could well be the low seen on July 23, 1984. Even though that low failed to move as low as the 1974 and 1982 bottoms, it was the first low that led to a significantly higher high than the high registered between the two lows preceding that high. It is our contention that the T measured by a time span from the 1956 and 1959 highs to the 1984 center post could well identify a very major top, at least in the advance decline line if not in the overall market.
That's enough for theory. Let's get to the results provided by some long-range Ts and what they could be telling us about the stock market's future.
The chart above, as noted earlier, is a daily chart of the cumulative advance decline line of the New York Stock Exchange going back to 1926. Notice the long-term basing pattern which formed between the lowest point on the chart on December 23, 1974 and the low on July 24, 1984. There were three distinct lows formed during that basing pattern. The dates of those lows were December 23, 1974, August 12, 1982, and July 24, 1984. Although the 1974 low turned out to be the most important one, it was not a prime candidate for the center post of a T because the advance which followed that bottom did not move above the high established between the preceding two lows, namely the 1974 low and the 1970 low.
The first real candidate for the center post of a large advance decline line T according to my understanding of Terry's theory (and please note that it is indeed my understanding and interpretation and not necessarily the view of Terry Laundry himself), occurred at the August 12, 1982 low in the advance decline line. That low was followed by a high on September 11, 1978, that slightly surpassed the highest point achieved between the lows of 1974 and 1982, namely the high of June 16, 1983.
Having located a potential center post for a very long-term T that could, in turn, determine the location of a potentially very important market top, the question becomes what date should be used for the left-hand side of the T. There are three candidates. There was almost an exact double top on the daily advance decline line registered on March 15, 1956 and March 13, 1959. Each of those is an obvious potential candidate for the left side of the T. Because those tops formed an almost exact double top, however, one could argue that the exact midpoint between the two highs could be used for the left-hand side of the T.
That exact midpoint is the date of September 12, 1957. Those of you who have the newsletter with colored charts can see a double arrowed green horizontal line near the upper part of the chart that delineates the time span between September 12, 1957 and August 12, 1982. T theory would argue that if the left side of the T has been properly placed, an equidistant right side of the T as measured from the center post on August 12, 1982, could determine the date for a potentially very important market top. In our recent research into T theory, we have noticed that trading day counts have led to more nearly accurate resolutions than calendar day counts although a legitimate argument can be made for either technique. There were 6256 trading days on the left side of the T from September 12, 1957 to August 12, 1982. If you measure forward another 6256 trading days from the August 12, 1982 center post to determine the resolution for the right side of the T, the calculated date turns out to be May 30, 2007.
The exact all-time high on the daily advance decline line of the New York Stock Exchange prior to what was arguably the second worst bear market in modern United States stock market history occurred on June 4, 2007, just three trading days away from the calculated date. Mind you, this calculation covered a time span of almost half a century from the left side of the T in 1957 to the right side of the T in 2007. Coincidence? That is always a possibility, of course, but let's continue to look at the data associated with the chart and see if we can argue that the other important top on the advance decline line that occurred on April 3, 1998, and followed the basing period between 1974 and 1984 could also have been determined by a long-term T.
When the rally in the advance decline line accelerated to the upside after the 1994 bottom, an examination of the chart would show three potential candidates for the left side of a long-term T and another three potential candidates for the center post of a T. The shortest T of the potential candidates would be a T constructed with its left side at the peak on March 13, 1959 and its center post at the midpoint between the 1974 and the 1982 bottoms (remember that the 1974 bottom did not qualify as a center post because of the explanations given earlier). The exact center of those two bottoms was October 17, 1978 and that is the date we will use for the center post of the proposed T. There were 4914 trading days between March 13, 1959 and October 17, 1978. That time span represents the left side of the T. If we measure another 4914 trading days from the October 17, 1978 center post, the calculated right side of the T would be March 26, 1998. Six trading days later, the advance decline line reached an exact peak which preceded its sharpest decline of the prior 25 years. Another coincidence? Of course, it remains a possibility, but the theory has now been shown to identify the two most important tops in the advance decline line over the past 30 years within just a few trading days.
Now look at the chart again. The advance decline line of the New York Stock Exchange has once again moved to new all-time highs. Is there a way to determine what could well be a final high on that indicator before the worst decline of all occurs? Let's examine the possibilities. According to our interpretation, there are only two possibilities remaining for the right side of the T which extends out into the current time zone and beyond from the three potential centerposts. Moving backwards from the potential centerpost of July 24, 1984, the center post which would give the furthest possible topping zones because it is the furthest to the right in time, one potential left side of a T beginning on March 13, 1959 has already expired (October 27, 2009). There are two remaining candidates for the left side of the T, and we believe the more logical of the candidates is the midpoint between the double tops registered in 1956 and 1959, namely September 12, 1957. Remember, that was used as the left side of the T whose center post on August 12, 1982, projected the 2007 top on the advance decline line to within three trading days.
There were 6,749 trading days between the left side of that T and the last possible logical candidate as a center post, namely July 24, 1984. If we extend the right side of the T out an equivalent 6,749 trading days from the 1984 center post, it points to April 27, 2011. We give that date because trading day counts have been more accurate in the past than calendar day counts, but for the sake of flexibility we will tell you that the same time spans measured on a calendar day basis would carry out to June 5, 2011. The last possible resolution, according to our interpretation would be the furthest top to the left of the long-term advance decline line chart in association with the furthest centerpost candidate to the right. That would entail using the date of March 15, 1956, in association with the date of July 24, 1984, a time span of 7125 trading days. If we count forward another 7125 trading days from July 24, 1984, it takes us to October 22, 2012. It is interesting to note, however, that the prior two Ts that virtually pinpointed the important tops in the advance decline line since the base building years in the 1970s and 1980s each used a midpoint of either a double top or a double bottom (but not both. In either case). In constructing what turned out to be the Ts that pinpointed the tops. This last possible resolution that points to October 2012 as a possibility uses neither a midpoint of tops or a midpoint of bottoms in its calculation. It is one of the reasons that we favor the calculation that points to a resolution between April 27 and June 5 of this year.
How does this calculation of a potentially very important top in the advance decline line scheduled to occur within the next 8-14 weeks affect the scenarios we wrote about in our last newsletter? It is impossible to know but we can attempt some intelligent speculation. Remember, the Ts we have been discussing here theoretically help to identify tops in the advance decline line, not the overall market indexes. As an example, the top in the advance decline line that was predicted almost to the day in 2007 was not a top in the overall market. The major market top in the indexes did not occur until four months later. There are almost always divergences between the major indexes and the daily advance decline line at important market tops. That means that even if a very major high is reached on the daily advance decline line between April 27 and June 5, there would be a very real chance that the final highs in the market indexes and averages would not be registered until sometime later.
Here is a tentative scenario that takes into consideration the rather bullish newsletter that was written in April and the rather bearish one that is being written now. The daily advance decline line could reach a major high between April 27 and June 5 and that top just might fit in with the resolution of the 784 calendar day cycle mentioned in the February newsletter and/or the Armstrong 8.6 year Economic Confidence Model cycle due to resolve in mid June, from which point that indicator and the overall market will suffer a relatively sharp decline. That sharp decline might end around early July in association with the 212 week cycle. From that point, we might begin to see the breadth numbers as determined by advances and declines be far less impressive, and begin to lead to the type of major divergence that generally accompanies important market tops.
That is one of many possible scenarios. The important point to keep in mind is that if the last top projected by T Theory resolves in the next few months, it could be a long time indeed before what could be a major top would be again exceeded. Because the very long term cash buildup phases will have completely expired (except for the outside possibility of October 2012 noted above), we can only surmise that it would take a significant period of time for another cash buildup phase to be generated. This conclusion agrees in general with our comments over the past several months regarding market valuations. The dividend yield chart on page 3 of our December 2010 newsletter suggests the possibility that we are facing another subpar decade in terms of market performance, a decade that could even see no gain based on the current dividend yields. That possibility fits in neatly with the possibility that the advance decline line could be approaching a very major top. Let's remember there is still a strong suggestion that 2011 will be a positive year overall based on the fact it is a pre-election year within the presidential election cycle. Let's also remember that we continue to have outstanding nominal four-year upside projections for both the Dow and the S&P calling for 14,043.48-15,922.51 and 1500.15-1736.56 respectively. Those projections along with the suggestion from T Theory that the earliest a top will be seen in the advance decline line is in late April suggests to us that we should not be too anxious to take a bearish stance as yet.
This page updated on
4/23/11
