Prepared by Peter Eliades' Stockmarket Cycles

The chart argues loudly that the stock market could well be at a major crossroads. If the chart looks familiar to our longer-term subscribers, it is with good reason. Over the past two plus years until around November 1996 we were watching the parallel channel that was formed by the lows of 1942, 1974, and 1982 on the lower channel and the highs of 1956, 1961, and 1966 on the upper channel. The S&P 500 stalled at that line in 1996, but when it blew above the upper channel line in November 1996, it was clear that the upper channel represented by the dashed line in the chart was not going to end the bull market. To show the continuing importance of that initial upper channel, however, the March-April 1997 decline was halted almost exactly as it returned to the line from above.

You should all know by now that we have an unswerving belief that the stock market is far from a random animal. We continue to believe that when this greatest of all bull markets does end, it will end at a point that will astound future market historians because it will prove once again there is “form and function” to the market's apparent random movements. Our next candidate for a stopping point is the new upper channel on the chart. This channel is also an exact parallel to the lower line formed by the 1942, 1974, and 1982 lows, but this one starts at the March 1937 high. Longer-term subscribers know we like to work with exact numbers when we construct these channels, but we should explain a few points relating to the channels. The prices used are closing prices on a daily basis for the S&P 500 Composite Index. On the chart, it appears as if the line drawn through the low of April 28, 1942 (7.47) and October 3, 1974 (62.28) goes exactly through the closing low of August 12, 1982 (102.41). On a calendar day chart, the line is actually at 104.11 on August 12, 1982. On a market day chart the line is at 102.15 on August 12. We use both a calendar day and market day chart because until 1952 there was Saturday trading on the New York Stock Exchange and it is difficult to adjust properly for that fact. Cycle theorists differ on whether time periods should be measured in calendar days or actual trading days. We use them both. We also used both the exact close on October 3, 1974 and the exact close on August 12, 1982 separately as the second points from the April 28, 1942 bottom. The differences were minor, but they give us a range of prices to look for. On August 6, the S&P 500 closed at 960.32. Depending on which of the above variations we use to construct the parallel channels, the upper parallel channel line fell somewhere between 947.21 and 970.88. For all practical purposes, the S&P reached its upper parallel channel line on that day. There is, of course, room over the next several days to move around 1% higher, but we would say that a move more than 2% higher than the 971 level over the next few weeks would be breaking above this second parallel channel. There are many technical arguments that say it will not happen.

Before we get into the technical indicators, we want to discuss what we believe are some important cyclical considerations for the next year or so.

The “8” year within each decade is associated with some fascinating statistics. Over the past 100 years, the low for the year in every “8” year has occurred in the first quarter of the year. There are no exceptions. On average, the low for the “8” year was seen by February 23. The dates for the lows are as follows: 1898 (March 25), 1908 (February 13), 1918 (January 15), 1928 (February 20), 1938 (March 31), 1948 (February 11), 1958 (February 25), 1968 (March 22), 1978 (March 1), 1988 (January 21). It is an extraordinary pattern and it suggests that whatever happens over the next 5-8 months, the low for the year in 1998 should occur before March 31. There is another pattern that is due to resolve next year, and it has also been very consistent and very impressive. There is a virtually perfect eight-year cycle that goes all the way back to at least 1910. Since that year, every eighth year has seen excellent buying opportunities that have usually lasted at least well into the following year. The last three resolutions 1974, 1982, and 1990, are the three landmark lows of this current bull market. Going back further, 1942 was one of the most important lows of the century, the 1950 lows have never been returned to, the 1958 lows have never been violated. The next year in the sequence is interesting because 1966 saw both an important top and an important bottom. The top was far more important because it was not convincingly surpassed until over sixteen years later, but the October 1966 low was indeed important. It led to a 35% advance over the next two years and it was not penetrated decisively until 3½ years later.

How in the world do we interpret those precedents? One argument would be that we are scheduled to see an important low before March 31, 1998, a low that could mark a bottom that might arguably last several years. The least we should expect to see is a rally of around 50% within 16 months of the 1998 low. The average gain from the eight-year cycle low to the next significant high of the following year and a half was 53.8% over the next 69 weeks (69 weeks was the average time period taken to reach the high of the next year and a half).

We believe such a rally could again be considered reasonable from a low sometime in the first quarter of 1998. That's the good news. The bad news is that we believe the decline preceding that rally will be so great that the 50% or so rally due off the early 1998 low will not even retrace 50% of the preceding decline. The implication of that statement is that we should see greater than a 50% decline prior to March 31, 1998. We base that judgment on several factors relating to cycles, technical indicators, and valuation models.

There is little question that the current stock market indices are at one of their most extended positions in stock market history. That is not a subjective statement and does not, by itself, imply there must be a large decline. There are objective ways of measuring how extended a particular market really is. One of those ways is a concept we mentioned in our letter dated June 27, 1997. It deals with a 316 period RSI (RSI stands for Relative Strength Index, a normalized oscillator attributed to Welles Wilder). We have pointed out that there have been only three general time periods this century prior to 1996 when RSI 316 on the DJIA reached 60+. Two of those prior three occasions ended with the two greatest market crashes of this century, 1929 and 1987. We believe one of the criteria for expecting a market crash is historical overextension of prices. On July 30, the DJIA reached 60+ for the third time in the past 18 months. We believe it was a sign that the market top was no more than two weeks away from that date.

Are there other ways of objectively measuring the overextension of a market based on long- term history? Frank Barbera, in his latest newsletter (The Gold Stock Technican), used a five-year moving average on the Dow and then constructed a detrended oscillator to determine how far above and below the five-year moving average the Dow was over the past 82 years. Frank notes there have been only three other times this century when the Dow moved over 70% above its own five-year moving average, 1929, 1937, and 1987. Of those three years, the smallest subsequent decline was the 1987 crash when the Dow fell over 41% intra-day in less than two months. On July 31, the DJIA was 77.1% above its five-year moving average. That fact alone does not mean the market has topped, but it puts this advance into historical perspective. No advance of this magnitude this century has ended well. The great distance above the five-year moving average is a good indication of potential vulnerability, and at current levels, the vulnerability is great.

The chart in this section shows the New York Stock Exchange daily advance-decline line accompanied by the McClellan Oscillator. It is important to remember that the only data used to determine the value of the McClellan Oscillator are the daily advance-decline line numbers. Technically speaking, when we search for divergences using the McClellan Oscillator, we should use the daily advance-decline line as the comparison. Notice the remarkable negative divergences that have built up since the high on the oscillator in early May. There is a series of at least four lower highs on the oscillator as the A-D line continued to make higher highs. Most of the new highs on the daily A-D line over the past week or so were made with the oscillator in negative territory. As this section is being written on Friday at 2:00 p.m. Eastern time, there are 2,100 more declining issues than advancing issues on the New York Exchange, so we have entered a preliminary reading of the McClellan Oscillator on the chart around the —137 level. It makes the chart look even more negative because it takes the oscillator to its lowest point since the April rally began. If the oscillator remains that low at the close, it could be looked at as a “kick-off” move to the downside, a high momentum move that often marks the beginning of a new trend, in this case a decline.

If we were to summarize the more important reasons behind looking for a major top in this time period, we would point to the Coppock Curve “Killer Wave,” the RSI 316 pattern, and the number of months up per 40 and number of months up per 100.

Perhaps we should review each of those concepts and explain why they are not just pointing to a potential top, but a decline of potentially historic proportions to follow.

The Coppock Curve “Killer Wave” has been described in detail over the past several months. Let's just say that when the Coppock Curve turns up from below zero, it gives a major buy signal. If it turns up from well above the zero level as it did earlier this year, it signals a market that is vulnerable to a significant decline. That configuration has occurred on only three prior occasions in the past 40 years, preceding the December 1968 top, preceding the January 1973 top, and preceding the August 1987 top. The 1968 top led to a 37% decline intra-day in a year and a half. The 1973 decline led to a 46.6% decline intra-day in 23 months. The 1987 decline led to a 41% decline intra-day in eight weeks! That's the full history of “Killer Waves.” We are scheduled to see a final top no later than September 12, 1997.

RSI 316 was fully explained in our June 27 newsletter. Readings higher than 60 on the RSI 316 for the Dow Industrials have occurred prior to 1996 in only three general time periods this century, 1928-29, 1954-55, and 1986-87. In each of those time periods, a final market top followed an initial 60+ reading by no longer than 523 calendar days. The 1929 top led to a decline of 89.5%, the 1956 top to a 20.6% decline, and the 1987 top to a 41 % decline. Yesterday's all time intra-day high on the Dow marked the 545th calendar day since the first 60+ reading in this time period (February 8, 1996) so a top is now overdue based on this indicator.

The third technical factor that points to a potentially historic decline is the measurement of months up per 40 and months up per 100. At the end of July, the Dow had been up 31 out of the past 40 months and up 69 of the past 100 months. At no time this century had the market been more extended. The 31 out of 40 months up established a new record for the century. The 69 out of 100 had occurred only once before, December 1961. It should be clear from these three technical indicators that the market is ripe for a decline of historical significance.

At the close today, the Dow Industrials gave a nominal 10-day projection down to 7896.20 ñ 37 points on a closing basis. The projection has not yet been confirmed but it would be confirmed with a sideways to down close on Monday, August 11.

Be aware that, due to the nature of price projections, sharp market advances lead to the possibility of even lower price projections. That possibility now exists and we will report the projections both here in the newsletter and on our daily telephone updates.

August 8, 1997 Peter Eliades

Stockmarket Cycles

P.O. Box 6873, Santa Rosa, California

Added to the WWW 08-15-97
Last updated on 08-15-97

Hosted by:
One Crossroads Place
610 West Maple Ave, Suite WWW
Independence, MO 64050
(816) 252-4080
sysop@kcmo.com

wmeubank@ocp.kcmo.com