The most universal proof of the 12 celestial cycles of light from 11.1 years of sunspot cycles, the annual cycles, the monthly spin of the sun, the daily (circadian) spin of the planet, and the intraday heleioseismologic cycles of the sun, is found in the economic activities and the financial markets.
Yes there is periodicity in the finacial markets relevant to the 12 celestial cyles mentioned in post 2 ! And no strict periodicity it is not the main observable in the prices! But there is literally (stochastic) periodicity in the volatility of the prices and the volumes. Using spectral analysis for the last 50 years on the volatility of the index SnP500, it has been found a clear strict periodicity of 11 years. It coincides with the Global Climate sunspot cycle. For a reference on a scientific publication here is the link
http://www.ckscientific.com/new_page_3.htm
Often periodicity in the prices appear for a time interval and then disappear. Happily the periods are quite fixed. I put two images of a table called the 12 rainbow celestial cycles with the most common such periods, that almost all of them have their origin in nature, or in Cosmic Economy as Buckminster Fuller would put it.
Although natural environment is strictly periodic as in the table above, and so also it may be with the volatility of the prices,the effect on the prices is not exact periodicity. The rule is: if a periodicity of nature, with period p, affects (modulates) economic prices, then what is observed in the prices, is a sequence of constant trends (drift) of duration equal to p/2, either up or down in a quite random way. What is invariant is the duration p/2 of the constant trend (drift) but they are not strictly alternating up and down as in a periodic phenomenon. We may have a run of 3 such up trends of duration p/2, then 2 down, then 1 up then 4 down etc. I call it Rainbow Walk of period p/2.
Of course the real final observable result is the superposition of all the effects of the 12 cycles, not of a single only cycle of them. But the previous mechanism was described when we isolate a single frequency.
According to my statistical analysis and trading systems backtests, the last 10 years, it seems that the dominating optimal frequency to trade is that of Blue, in other words the one or two days sessional periodicity. It is by far the most robust and consistent frequency with excellent trading results. It is somehow expected as the Day-Night periodicity is also the strongest from a physical point of view. I conceive it as the coupling of the planet spin with the sun's spin. This frequrency is somehow invisible in the charts of daily bars. It has to be discovered with appropriate statistical and backtest tools.
Simple statistics of the daily bars Open-High_Low_Close in Excel prove e.g. that in the period 2006-2011, the pairs of day that if the first day the market goes up (down correspondingly) , hen the second day the market goes down (up correspondingly), in other words the opposite direction are 54% of the cases while , the next day going in the same direction is only 46% of the cases. This "small" edge of 4% is an indication of the daily periodicity and may give the opportunity to design great systems! Although this property is not universal for all instruments, there is a universal system, at the daily periodicidy , that could be called planet-spin system. A. Dukas, had conducted a hybrid (half-manual) version of it during 2003, wih remarkable performance, with his own indicators, for all cfd's. .It is a waving-pattern trading system utilizing slow highpass and fast lowpass filters (see post 32).As the volumes of transactions are very strongly periodic from day to day ( see e.g. forex, due to the sessions), the rules of volumes are very significant for such intraday systems (see post 10 for the rules of volumes). The ForexGrowthBot spike-based system, (especially if focused only on terminal spikes) is a part of the Planet-spin system. Terminal spikes are more rare than initial spikes.
More than 80% of the money that I have ever lost in trading from 2003 to 2011 were lost in experimenting and trying to code a 100% automated system at the scale of 3 hours or 1 hour or less that would perform better than a corresponding 100% automated system at the scale of one or two days. The periodicities at the faster intra-day scales are quite more faint than the 1 or 2 days periodicity. So the present state of the art (2011) in econometry, trading system design and also in information systems and platforms, does not seem to permit the coding and run of an 100% automated trading system (without reinvestment of profits or money management and intraday optimal adjustments of position sizes and pyramiding, see posts 32,33 ) at the faster intra-day scales that would outperform a (half) manual set-and-forget daily system at the scale of 1 or 2 days. But even if the advance of science expertise and technology would one day produce such an automated system, then utilizing this very technology and shifting the application to the frequency of 1 or 2 days, would produce a half-automated set-n-forget system that most probably would perform again better.
The situation is different if reinvestment of profits and intraday optimal adjustments and pyramiding is utilized , (see posts 32,33) : Although periodicities faster than 1-2 days, e.g. 30 minutes or 180 minutes may be less clear and weaker , by applying fast intraday reinvestment of profits, optimal adjustments of position sizes and pyramiding ,( see posts 32,33) , thus an exponential growth on the lots sizes, it is possible to surpass in speed of capitalization of systems based on 24-48 hours periodicity where forecasting is easier. This speed can be increased by, concentrating in spike days and refraining from trading during calmer days. Still the combination of defining the almost constant trend at 24-48 hours, and 20 days background, and then making the optimal adjustments of positions sizes, and pyramiding at faster intraday periodicities, but over the constant trend over the 24-48 hours, seems to me to be the really optimal. Nevertheless, selecting strong spikes that last 1-10 days may lead to large intervals without trading, while for a few days during the spike, a very fast intraday tarding.The latter really gives God's speed in capitalization.
The second best frequency is that of the spin of the sun, in other words the monthly periodicity (Yellow rainbow frequency). The trades last about 10 trading days (2 weeks). Although it is the 2nd best frequency, it is some how easier with this (almost) periodicity, to design a universal system, in other words a system that applies with the same rules, to all instruments (stocks, commodities, forex etc). I have designed and tested such a system, which I call it Star-Spin.It is a waving-pattern trading system utilizing slow highpass and fast lowpass filters (see post 32). Example of 100% automated, though neural networks such system, is the TradeVantage system. Also the Bill Williams terminal spikes system at h4-timeframe, is essentially part of a spike-based Star-Spin system.Terminal spikes are more rare than initial spikes.
The hybrid conduction of the planet spin system, is more difficult than the hybrid conducation of the star-spin system, due to irregular hours that waves appear inside the day. It is better an 100% automated version of the planet spin system , which meets with more irregularity and noise though than a 100% automated version of the star-system.
An interesting question is if the above 12 rainbow frequencies have effects over the price movements that are mutually correlated or mutually independent. At a 1st simple simulation of the above 12 frequencies we may assume for the sake of simplicity mutual independence (As if a superposition of 12 independent stochastic processes). But my more than 10 year statistical analysis and backtests, show that neighbouring frequencies do have correlated effects in the prices. I have observed two types of correlations a) The rainbow walk at a particular frequency has upper and lower reflection barriers that their distance depends on the amplitude of the next lower frequency. I call them Rayleigh barriers and I discuss them at a special post (35). b) Trends (as runs of rainbow vectors) usually start and are correlated with spikes of the faster frequency (I discuss it also in a subsequent post (32) about spike-based trading systems).
These rainbow frequencies are key and characteristic frequencies of absolute, cosmic origin from 11 years to 5 minutes. These frequencies create corresponding periodicity in the volumes and the volatility of the prices of the markers. All indicators, and forecasting measurements on the prices must be tuned to these frequencies if we want to have optimally successful prediction. According to my statistical analysis and trading systems backtests, the last 10 years, it seems that the dominating optimal frequency to trade is that of Blue, in other words the one or two days sessional periodicity. It is by far the most robust and consistent frequency with excellent trading results. It is somehow expected as the Day-Night periodicity is also the strongest from a physical point of view. I conceive it as the coupling of the planet spin with the sun's spin. This frequrency is somehow invisible in the charts of daily bars. It has to be discovered with appropriate statistical and backtest tools.
Simple statistics of the daily bars Open-High_Low_Close in Excel prove e.g. that in the period 2006-2011, the pairs of day that if the first day the market goes up (down correspondingly) , hen the second day the market goes down (up correspondingly), in other words the opposite direction are 54% of the cases while , the next day going in the same direction is only 46% of the cases. This "small" edge of 4% is an indication of the daily periodicity and may give the opportunity to design great systems! Although this property is not universal for all instruments, there is a universal system, at the daily periodicidy , that could be called planet-spin system. A. Dukas, had conducted a hybrid (half-manual) version of it during 2003, wih remarkable performance, with his own indicators, for all cfd's. .It is a waving-pattern trading system utilizing slow highpass and fast lowpass filters (see post 32).As the volumes of transactions are very strongly periodic from day to day ( see e.g. forex, due to the sessions), the rules of volumes are very significant for such intraday systems (see post 10 for the rules of volumes). The ForexGrowthBot spike-based system, (especially if focused only on terminal spikes) is a part of the Planet-spin system. Terminal spikes are more rare than initial spikes.
More than 80% of the money that I have ever lost in trading from 2003 to 2011 were lost in experimenting and trying to code a 100% automated system at the scale of 3 hours or 1 hour or less that would perform better than a corresponding 100% automated system at the scale of one or two days. The periodicities at the faster intra-day scales are quite more faint than the 1 or 2 days periodicity. So the present state of the art (2011) in econometry, trading system design and also in information systems and platforms, does not seem to permit the coding and run of an 100% automated trading system (without reinvestment of profits or money management and intraday optimal adjustments of position sizes and pyramiding, see posts 32,33 ) at the faster intra-day scales that would outperform a (half) manual set-and-forget daily system at the scale of 1 or 2 days. But even if the advance of science expertise and technology would one day produce such an automated system, then utilizing this very technology and shifting the application to the frequency of 1 or 2 days, would produce a half-automated set-n-forget system that most probably would perform again better.
The situation is different if reinvestment of profits and intraday optimal adjustments and pyramiding is utilized , (see posts 32,33) : Although periodicities faster than 1-2 days, e.g. 30 minutes or 180 minutes may be less clear and weaker , by applying fast intraday reinvestment of profits, optimal adjustments of position sizes and pyramiding ,( see posts 32,33) , thus an exponential growth on the lots sizes, it is possible to surpass in speed of capitalization of systems based on 24-48 hours periodicity where forecasting is easier. This speed can be increased by, concentrating in spike days and refraining from trading during calmer days. Still the combination of defining the almost constant trend at 24-48 hours, and 20 days background, and then making the optimal adjustments of positions sizes, and pyramiding at faster intraday periodicities, but over the constant trend over the 24-48 hours, seems to me to be the really optimal. Nevertheless, selecting strong spikes that last 1-10 days may lead to large intervals without trading, while for a few days during the spike, a very fast intraday tarding.The latter really gives God's speed in capitalization.
The second best frequency is that of the spin of the sun, in other words the monthly periodicity (Yellow rainbow frequency). The trades last about 10 trading days (2 weeks). Although it is the 2nd best frequency, it is some how easier with this (almost) periodicity, to design a universal system, in other words a system that applies with the same rules, to all instruments (stocks, commodities, forex etc). I have designed and tested such a system, which I call it Star-Spin.It is a waving-pattern trading system utilizing slow highpass and fast lowpass filters (see post 32). Example of 100% automated, though neural networks such system, is the TradeVantage system. Also the Bill Williams terminal spikes system at h4-timeframe, is essentially part of a spike-based Star-Spin system.Terminal spikes are more rare than initial spikes.
The hybrid conduction of the planet spin system, is more difficult than the hybrid conducation of the star-spin system, due to irregular hours that waves appear inside the day. It is better an 100% automated version of the planet spin system , which meets with more irregularity and noise though than a 100% automated version of the star-system.
And the third best frequency is that related to the year (planets orbit around the sun, Red in the above rainbow table). I call it seasonal system. An example of a 100% automated universal system, based on seasonal periodicities on all commodities, stock indexes, currencies etc is that of Keith Fitshen, called Aberration which is running succesfully from 1980 till today (2012).(see e.g.
It is very important therefore to realize, that if we are seeking to maximize the performance of say a seasonal trading, we cannot do it by looking in frequencies faster that the daily period. (Even very fine scalpers of 3-5 pips per trade , are based on daily sessional periodicities, and a stationary ranging channel that has no small scale periodicity at all.). We should rather resort to two different factors:
a) Apply it to a multitude of stocks , thus increasing the time we are with open positions in the market
b) Utilize options (e.g.as Chuck Hughes is doing)
If we want nevertheless to trade at an intraday fast time frame, then the trading at the fast time frame, like e.g. m5 (5 minutes bars) , cannot be the same and should be entirely different from trading at a slow time frame, like e.g. d1. The reason is that at a fast intra-day time frame the periodicities are weak, the micro-trends unstable, and the noise of large amplitude. (see post 44 about neutral grid trading)
The phase of the 1st daily spin or periodicity is the obvious day-night hourly phase, as defined by the volumes of transactions. The annual spin or hidden periodicity is again the obvious seasonal again as defined by the volumes of transactions (especially in the agriculture commodities). And the monthly spin or hidden periodicity is strangely enough defined by the full moon and new moon. In the history of technical analysis it was formed a secrete "Delta society" when it was discovered how clearly statistically full moon and new moon was correlated with the tops and bottoms of the prices of gold (see e.g. http://www.gold-eagle.com/editorials_05/rosen112006.html)
Therefore by combining the a) Planet spin (day) b) Sun's spin (week or month) c) Planet orbit (year and its first, or second harmonic, e.g. quarter) we get the best and simplest combination of periodicities of the markets for consistent and more successful forecasting.It is very important therefore to realize, that if we are seeking to maximize the performance of say a seasonal trading, we cannot do it by looking in frequencies faster that the daily period. (Even very fine scalpers of 3-5 pips per trade , are based on daily sessional periodicities, and a stationary ranging channel that has no small scale periodicity at all.). We should rather resort to two different factors:
a) Apply it to a multitude of stocks , thus increasing the time we are with open positions in the market
b) Utilize options (e.g.as Chuck Hughes is doing)
If we want nevertheless to trade at an intraday fast time frame, then the trading at the fast time frame, like e.g. m5 (5 minutes bars) , cannot be the same and should be entirely different from trading at a slow time frame, like e.g. d1. The reason is that at a fast intra-day time frame the periodicities are weak, the micro-trends unstable, and the noise of large amplitude. (see post 44 about neutral grid trading)
The phase of the 1st daily spin or periodicity is the obvious day-night hourly phase, as defined by the volumes of transactions. The annual spin or hidden periodicity is again the obvious seasonal again as defined by the volumes of transactions (especially in the agriculture commodities). And the monthly spin or hidden periodicity is strangely enough defined by the full moon and new moon. In the history of technical analysis it was formed a secrete "Delta society" when it was discovered how clearly statistically full moon and new moon was correlated with the tops and bottoms of the prices of gold (see e.g. http://www.gold-eagle.com/editorials_05/rosen112006.html)
As the most direct effect of the underlying cycles is the cycles of volumes and volatility of the prices of the markets, the easiest method to exploit them is the volatility-short/volatility-long techniques. Although this terminology comes from the trading of options it is possible to create somehow artificial options with ordinary positions (even at forex). Examples of volatility-long techniques are the break-out methods, and examples of volatility-short techniques are the B. Williams angulation-counter-trend method or other counter-trend methods that focus on retracements of spikes. If to the 2 states volatility-increase (spikes or trending), volatility-decrease (detrending ) we add the state of intermittency or noise (ranging) we get the 3 elementary qualitative dynamics of the market in respect to its cycles.
Finanly I must remark that in Economics the most celebrated business cycles, are the Kitchin cycle of 5.5 years, the Kuznets (Nobel prize winner) cycle of 22.2 years ( a full solar sunspot cycle of 11.1+11.1 years) and the Kondradieff cycle of about 66 years. Schumpeter tried to combine all the tree of them, but at best their models are of deterministic dynamic systems (Chaos theory) while I believe that the right formulation is of random coefficient dynamic systems, that their cycles are actually the input influence on the economic phenomena not the output observable behavior.
A very interesting video by Mike Maloney, where among other things . we watch Kondradieff cycles of Gold and Stock prices is at
http://www.youtube.com/watch?v=tj2s6vzErqY
A nice book on business cycles is that by Lars Tvede Business Cycles , Routledge editions 2001.
See also this video about the 6-months cycles in various sectors
https://w3.absoluteprofits.com/Finance/ABS/LP/ABS-Secret-Calendar?dkt_nbr=v9fxenc1
For measuring with indicators and utilizing some of these cycles in forecasting , as they are modulating the non-marshalian demand-supply waving pattern see posts 20, 21, 32.
Finanly I must remark that in Economics the most celebrated business cycles, are the Kitchin cycle of 5.5 years, the Kuznets (Nobel prize winner) cycle of 22.2 years ( a full solar sunspot cycle of 11.1+11.1 years) and the Kondradieff cycle of about 66 years. Schumpeter tried to combine all the tree of them, but at best their models are of deterministic dynamic systems (Chaos theory) while I believe that the right formulation is of random coefficient dynamic systems, that their cycles are actually the input influence on the economic phenomena not the output observable behavior.
A very interesting video by Mike Maloney, where among other things . we watch Kondradieff cycles of Gold and Stock prices is at
http://www.youtube.com/watch?v=tj2s6vzErqY
A nice book on business cycles is that by Lars Tvede Business Cycles , Routledge editions 2001.
See also this video about the 6-months cycles in various sectors
https://w3.absoluteprofits.com/Finance/ABS/LP/ABS-Secret-Calendar?dkt_nbr=v9fxenc1
For measuring with indicators and utilizing some of these cycles in forecasting , as they are modulating the non-marshalian demand-supply waving pattern see posts 20, 21, 32.
Probably the best instantaneous rewarding "why?", of manual trading is the joy and satisfaction in playing, among the situations of higher or lower uncertainty of what will happen in the global economy and markets, so as to plan and conduct a strategy that lets you know, on occasions, through data information and mathematical, and economic principles, what will happen with acceptable low uncertainty.
(And from a macroscopic point of view, probably the stronger reason for studying scientifically the capital and money markets, is that among all civilizations in the galaxy, this earthly civilization, is the only one with extremely complicated , massive alienating and evil economics. The more advanced civilizations do not even use money inside them, but only for simple commerce with different civilizations. For a mind therefore perceiving all reality, it is challenging to go through the underlying mathematical and scientific laws of such a rare economic phenomenon.)
