--THE CYCLES--
Martin Armstrong is a controversial character to say the least. He is currently in prison for contempt of court charges. The amazing fact is he has been in prison for over seven years for contempt of court charges. We don’t know if records are kept on the subject, but this must certainly be some sort of record for prison time on contempt of court charges. We are not going to go into the details of the charges against Armstrong because that is not the purpose of this newsletter. We have no idea whether he is guilty of the spectacular charges under which he was incarcerated in January 2000.
He was a market colleague of ours and we met Armstrong several times. He personally invited us to speak at a symposium in Princeton New Jersey where the featured speaker was former British Prime Minister Margaret Thatcher. As we recall, it was a Cycles symposium and our recollection is that it was 15-20 years ago. Like so many people who spend their time analyzing markets, Martin was very opinionated. He was not an easy person to like. Nevertheless, he had developed what he called an Economic Confidence Model that related to cycles and we have always been attracted to anything related to cycles.
Those preliminary comments lead us to this chart. In December 1979, Armstrong published a chart of his Economic Confidence Model. The overall structure of the model is based on an 8.6 year business cycle. According to Armstrong, the cycle builds in intensity to form long waves of economic activity measuring 51.6 years. He was careful to say that his 51.6 year cycle should not be confused with the Kondratieff wave which was generally accepted to be an average length of around 54 years.
In an article written on September 26, 1999, Martin wrote, “I didn’t know how to go about such a quest to find if the business cycle was definable. Admittedly, I began with the very basic naive approach of simply adding up all the financial panics between 1683 and 1907 and dividing 224 years by the number of panics being 26 yielding 8.6 years. Well, this didn’t seem to be very valid at first, but it did allow for a greater amount of data to be tested compared to merely three waves described by Kondratieff."
Armstrong went on to discuss both a 51.6 year cycle (comprised of six 8.6 year cycles) and a 309.6 year cycle (comprised of six 51.6 year cycles). We will concentrate on the basic 8.6 year cycle in this discussion. Armstrong writes his cycle dates in fractions of a year. In order to determine exact dates, the fractional part of the year must be multiplied by 365 then translated to an actual date within that year. He used 1929.75 as a reference point for testing his 8.6 year cycle because the crash that year was obviously one of the great economic turning points of the 20th century in the United States. He began applying these 8.6 year cycles in real time in 1976.
It is important to point out that the cycle dates are not necessarily intended to be turning points for the U.S. securities markets. Some of them have turned out to be just that with uncanny accuracy while others seem to apply more closely to other financial markets. One of the important 8.6 year economic confidence tops was scheduled for 1989.95. Glance at the chart again to see where the Nikkei average reached its important top. Coincidence? At that point, it could well have been. But something that happened just over two years earlier made the possibility of a coincidence less likely. Armstrong began to notice that, despite the fact the 8.6 year periodicity was determined by taking the average number of years between financial panics, some of these predetermined dates were significant either exactly on the appointed date or within a very few days of that date. For example, in 1987 the date of an expected cycle resolution was 1987.8. That fraction of a year calculates exactly to October 19th. A cycle resolution chart published 10 years earlier had pinpointed the exact day of the greatest single day crash in the history of the modern securities markets in the U.S. Coincidence? Perhaps. The next 8.6 year cycle bottom after the 1987 crash was scheduled for 1994.25. That calculates to be between April 1st and 2nd of that year. The first trading day in April turned out to be April 4th because of the Easter weekend which preceded it and on that exact resolution day, both the Dow Jones Industrial Average and the S&P 500 registered their lows of the year and never looked back. That low would prove to be the springboard for one of the great bull market moves in history. Coincidence? Perhaps.
Arguably, the most popular market index over the past decade or more has been the Nasdaq Composite. We decided to show you a group of weekly charts of the Nasdaq Composite with all of the Armstrong cycle dates. Although the cycles theoretically often resolve to an exact day, we're using weekly charts in order to save space. We have placed arrows above and below the weeks when the cycles were due to resolve. Before we do that, we will give you the dates of those cycles to the exact day according to our own calculations.
January 19th, 1977, March 14th, 1979, May 8th, 1981, June 4th, 1982, June 30th,1983, August 26th, 1985.
August 26th, 1985, October 19th, 1987, November 21st, 1988, December 13th, 1989, January 8th, 1991, February 6th, 1992, April 1st, 1994.
April 1st, 1994, May 24th, 1996, June 23rd, 1997, July 20th, 1998, August 16th, 1999, September 12th, 2000, November 7th, 2002.
November 7th, 2002, January 1st, 2005, January 27th, 2006, February 23rd, 2007, March 22nd, 2008, April 19th, 2009, June 13th, 2011.
Each paragraph of dates listed above starts with an 8.6 year cycle bottom. The last date in each paragraph is the end of the next 8.6 year cycle and that date is repeated at the beginning of the next paragraph as the start of the next 8.6 year cycle. Although the cycle dates should be viewed as turning points rather than specific tops or bottoms, we will let you judge from the Nasdaq Composite charts just how accurate the alternating tops and bottoms prove to be. The sequence of cycles from each 8.6 year cycle bottom proceeds in the following order: 2.15 years to a top, 1.075 years to a bottom, 1.075 years to the main peak of the 8.6 year cycle, 1.075 years to a bottom, 1.075 years to a top, and finally 2.15 years to the next 8.6 year cycle bottom.
The three charts presented here follow the 8.6 year cycle and all its sub cycles from the first resolution in January 1977 through the most recent resolution scheduled for 2007.15 which calculates to the 23.75th day in February. The cycle dates given in the paragraphs above are either the main date of the calculation or the following day in order to account for the fractional remainder. For example, February 23.75 would either be February 23rd or February 24th.
We repeat our admonition that Armstrong did not represent that all the cycle dates would correspond with the United States securities markets. The easiest example is the major top scheduled for 1989.95. The scheduled date was December 13th, 1989. On that date the Nasdaq Composite was at 451.14 and although it rallied just over 4% to a subsequent high in July 1990, that July top was followed by a very substantial closing price decline of greater than 30%. The Japanese Nikkei average, on the other hand, registered its all-time high within days of the December 1989 cycle resolution and began an agonizing bear market that ultimately took prices down just over 80 percent in 2003. Even now, over 17 years after that late 1989 top, the Nikkei is 55% below its all-time high. That might be a good bedtime story for those who attempt to preach the sanctity of long-term investing.
As we noted earlier, we will leave it to the reader to judge the effectiveness of these cycle resolutions. There are two or three resolutions on the chart that are worthy of further comment. The resolution due on March 14th, 1979 was ideally scheduled as a top but we have moved it below the price action where we placed bottoms because it turned out to be a launching point upward on the Nasdaq of very major significance which is clear from the chart. The next signal worthy of comment was scheduled for October 19th, 1987. Although it was ideally scheduled to be a top, there should have been little question when that day was over whether it turned out to be a top or bottom. The resolution date had pinpointed the largest single day percentage decline in the history of the New York Stock Exchange and it turned out to be an exact resolution as the low close on that date has never since been violated. The interesting additional point to be made is that the ideal October 19th, 1987 resolution date was just two weeks away from the preceding high close on the Nasdaq registered on October 5th, 1987. That close was important enough to mark a high that would not be breached for another two years. The 1994 resolution was scheduled for April 1.25, the first or second day in April of that year. It was an exact resolution as the first trading day in April after a long Easter weekend marked the low for the year on both the Dow and the S&P. The Nasdaq Composite, on the other hand went on to register a lower low a few months later. The 2006 resolution for January 27.375 was scheduled to be a bottom but ended up being closer to a top than a bottom so we placed that up arrow above the price action. Beyond that, there are no further explanations required. We leave it to the reader to judge the efficacy of these resolutions in relation to the Nasdaq Composite.
This whole preamble leads us to the most recently scheduled resolution. Interestingly, the exact resolution was scheduled for February 23.75, 2007. In some of Armstrong's prior writings, he mentioned a specific resolution date of February 27th but we are guessing that was simply a miscalculation because virtually all other specific dates mentioned by him agreed with our calculations in terms of the fractional parts of a year. The fascinating part of the equation is that whether you use the apparently accurate February 23rd-24th (a Friday-Saturday) or the apparently miscalculated February 27th, you must so far walk away very impressed. The February 23rd date was within 2-3 days of all-time highs in the Dow Jones Industrial Average and several of the global indexes that have been performing so spectacularly over the past few years. The February 27th date marked the exact day of global financial panics with the ultra hot Shanghai market seeing its largest single-day decline in a decade, a drop of almost 10%. The more important question relating to either of the dates is whether they will simply mark short to intermediate term market tops (that has already been established) or whether they will be of the same type of significance as the late December 1995 resolution on the Japanese Nikkei. It is, of course, impossible to know the final answer. But one should indeed see enough in the charts provided to very seriously ponder that possibility. For all too long now, the market has ridiculed those who maintained it was at bubble valuation levels on an absolute basis (there are those who maintain the market is greatly undervalued on a relative valuation basis and there is justification for that opinion), has ridiculed those who pointed to the amazing long-term complacency of investors based on indicators such as sentiment readings and volatility index readings, and has ridiculed those who deigned to suggest that there is no new paradigm and that the market is priced for perfection in the future. The one thing we can assure you of is there will be no perfection in the future.
