Annual cycle for Dow - back to old patterns
Moon cycle for Dow - mass psychology
In this article the algorithm of choosing astro cycles will be introduced. However, before discussing it, I need to inform you that this is not a completed theory. Obviously this technology will be modified from time to time while our knowledge regarding this issue becomes more fulfilled. The models based on astro cycles are kin to phenomenological models that are used a lot in the theoretical physics (physicists work with phenomenological models when they do not have a full knowledge regarding the analyzed subject; instead, they state some rules/observations regarding this subject with understanding that this is not a complete knowledge yet. That is why experiments are needed: to confirm the whole set of rules or parts of it.)
For our models, there are just two rules that have to be met for the chosen astro cycles.
Rule #1 stability (called before split): the analyzed cycle should be stable i.e. it should work now at the same manner as it has worked for some time before (for example, 5 years ago, 10 years ago, etc.). To understand the cycles stability, look at the composite diagram below. Here you see, besides the main bold diagram, that there are three thick diagrams - red, purple and blue. This is Annual chart calculated for Dow since 1947 till 2012 :
A blue curve shows how this Annual cycle worked since 1947 till 1969, purple is for the period since 1969 till 1990, while the red one is for the years 1990 through 2012.
ALL THESE COMPOSITE CURVES - BLUE, PURPLE AND RED - SHOULD SHOW APPROXIMATELY THE SAME DIRECTION, THE SAME TENDENCY OF THE PRICE MOVEMENT. Only when all three show the same direction, we can consider the cycle as a stable one.
As you see, the Annual cycle above is pretty stable; it reflects some tendency in the price movement while the Sun passes different signs of Zodiac.
Look at another stable cycle, Mars-Saturn heliocentric:
This diagram shows that the price reaches its top when the angle between Mars and Saturn in helio reaches 180 degrees.
Look at the next example. This cycles is not stable (Mars-Jupiter geo):
Rule #2 Walk Forward Efficiency is positive: the cycle may be pretty stable though it does not mean that this cycle is suitable for forecasting. Stability means just that some AVERAGE tendency related to some particular astro cycle is present. This condition alone does not mean that we can use this cycle to get a forecast.
We have to conduct a more detailed analysis for this cycle - Walk Forward Analysis (WFA).
In order to conduct WFA for the analyzed cycle, click "WFA" button:
In several seconds (this is how fast we do the calculations - just seconds!), you will get Walk Forward Efficiency for our cycle (Mars-Saturn helio). In this example, it is 8 plus versus 2 minus, i.e. 80% positive results. The amount of positive results should be significantly more than negative results (this is the rule #2)..
Walk Forward Analysis shows how the analyzed cycle forecasts the future. For example, when we perform WFA for the Annual cycle and use the price history for Dow till August 2012, this is what the program does:
- it calculates the Annual cycle using price history till August 2011 ("in sample" interval), then it creates a forecast based on that Annual cycle (forecast for one year) and watches how the model based on that Annual cycle actually forecasts the next year - since August 2011 till August 2012 ("out of sample" interval). If the forecast is good (positive correlation), we have +1; otherwise we have -1.
- it calculates the Annual cycle using a different price history interval - till August 2010 - and watches how this Annual cycle forecasts the next year (which would be a year since August 2010 till August 2011).
"By default" the program performs this procedure 10 times, this is "sample size" parameter. If you have long price history, you can increase the sample size parameter clicking this button:
Annual cycle for Dow - back to old patterns
I' have faced a problem regarding the Annual cycle for Dow. If we download the Dow data from the beginning (1885), the Walk Forward Analysis does not show a good result:
we have 30% positive results only.
However if we download the price history from 1950 the results are much better:
I do not understand why the most important cycle is so hard to research in terms of formal criteria. The possible explanation is: we have to consider two Annual cycles - Fast Annual - that is based on the last 3-15 years of the price history, - and Annual that is based on the whole available price history (it reveals old 120 years aged patterns). In other words, we can assume that the stock market "keeps in mind" the newest patterns together with the old ones. When the stock market is "ready for a big drop", I always watch Annual based on the whole price history (a blue curve on the chart below). As an example, look at this chart with two Annuals before a huge drop in 2008:
In critical moments the stock market remembers its "old" patterns. This drop is explained in terms of "September drop" which is a very strong pattern in the Annual cycle while for the last three years prior 2008 this September drop was practically ignored by the stock market. Red curve represents "Fast Annual" which is based on three last years. Please do not forget that this is a phenomenological theory only as I have mentioned in the beginning of this article.
Moon cycle for Dow - mass psychology
This is ore more enigma for me. In general, the Moon cycle works this way (this is a composite diagram: price changes while the Moon moves through Zodiac):
The Moon in Zodiac cycle 
works better than Moon phases cycle, I was
surprised by this fact. In general Dow follows this cycle; Dow is high when the Moon
passes Libra-Scorpio and low when it passes Aquarius-Pisces. This cycle is very
weak, so usually I ignore the Moon cycle. But one fact revealed by the new
module has given me a new sight on the Lunar cycle.
During 12 months before a huge drop in September-October 2008, the Moon cycle has worked much better than it usually works. Walk Forward Efficiency for the year 2008 shows +11 versus -1, usually this value varies between +7-5 and +4 -8.
I suspect this fact is related with mass psychology. Somehow instability of the stock market makes mass psychology more sensitive to the Moon effect. The similar effect took place in the year 1987.