I love cycles.  I can’t help it.  I see them in everything around me.  Everything has a rhythm.  In fact, if things did not have inherent change, the human perceptual system is designed to ignore them.  If someone puts their hand on your arm, at first, you notice it. But then, after a few minutes, that initial sensation fades and as long as it remains, without any movement, your brain will more or less ignore it.  If it moves again, the perceptual system will pick it up and start over again.  So it is with cycles, and especially with stock market cycles!  This is because when it changes, it changes your pocketbook; a particularly sensitive perceptual system to say the least (at least for me anyway).  So I spend many hours scheming about how to quantify these things and convert them into cash.

One of the things I love the most about cycles is, they are anticipatory.  By nature, a cycle extends into the future.  This is different than any other form of market analysis, because a cycle is a projection in time.  Most people are not even capable of imagining the stock market in terms of time alone.  They are too busy dealing with the perceptual and sensory issues of price.  For me the time component is where it is at then; it is the component of analysis others ignore and it is the single biggest edge you can have in trading.

For this reason, I have correlated hundreds of cycles to the stock market that are all around us from the pulsing of electrical charges on our ionosphere (a non-linear cycle) to fixed cycles, such as a ten day (half trading month) cycle (a linear cycle).

Stock market cycle analysis becomes very complex due to other factors most people never think about.  For example, the issue having to do with missing data; the stock market is open 6.5 hours a day and is closed on weekends and holidays.  Other factors relate to market participation.  There are substantial changes in volume at various times of day.  Markets that trade over night have thin participation at night, while other times of day have certain seasonal or cyclical patterns themselves.  Volume is higher in the mornings around the open and in the afternoon around the close.  Graphically, the pulsating volume pattern looks like a suspension bridge as these cyclical volume patterns form.

A market may have multiple component cycles running at any given time.  For example, by doing cycle analysis, there may be waves that are 10, 14, 19, and 23 etc. periods in length occurring in the data (see the image below depicting these values in a Fast Fourier Transform of the S&P 500 — FFT).  When you add all these component sine waves together, it gives you a wave very similar to the market.  From there, you can use these cyclical components to project the next wave.

These time cycles may or may not give you information about the component amplitude (or variation in price).  In other words, it may give you good projections in time, but without regard to highs and lows.  Because of this, various waves may project a high or low, but it may be higher or lower than another wave. This is one of the toughest areas of cycle analysis; projecting price levels or relative amplitudes; a topic which is beyond the scope of this article.

When doing FFT analysis over time, some component waves appear over and over again when doing analysis, and others are less significant and don’t appear frequently.  The idea is to find the ones that are more or less constant and ignore the others.  There are also statistical tests you can do to try to determine if the component waves have a “significant” impact on the target wave (the traded market).   One example of this is the Chi Square test for statistical significance.

When many relevant waves line up together, you will typically find significant market turning points.  Again, this gets tricky because you are missing significant portions of real time (for example, what happens when the real wave peaks at two o’clock in the morning on a Sunday?).  This begs the question:  if there are cycles in the data (and clearly there are), then are they occurring due to external factors?  If so, does the fact that I am missing more than 80% of the data (the market is open 32.5 hours out of 168 hours in a week – 1- (32.5/168)= 81%), impact my ability to accurately discern the presence of the cycle?  Is the FFT valid at all using only 19% of the time?

This is another thing I really love about cycle analysis.  The very nature of our world and the structure of reality as we believe it to be is called into question through this science of cycles.  It actually spills over into metaphysics and even calls into question such issues as determinism and the question of free will.  I am, of course,  not intending to get too esoteric here, but when you see some of these cycles in action, their accuracy is so repetitively stunning that at times you will begin to wonder if you are staring down the face of creation itself;  that spills, of course, over into religion and all kinds of wonderful considerations about the nature of reality—Ok, enough on that!!  How’d we get from the subject of trading to the nature of reality anyway ?  :-)

There are other methods of determining the presence of cycles-  Maximum Entropy Method (MEM), Trigonometric regression, and any number of proprietary methods I have seen that are amazing.  Much work on this was done by J.M Hurst and another lesser known mathematician named Claud Cleeton, who took Hurst’s work and expanded it considerably.  I have coded all these works extensively, in addition to the works of others and have developed my own proprietary methods for cycle work.  Some of these cycle methods are so powerful, I can actually project stock market pivots for extended periods into the future.  My wife asked me, why don’t you tell anyone?   I told her, I don’t tell anyone this, simply because I know nobody would believe me. They’d say I was a quack even before they verified it because they would choose to not believe it in advance (maybe because they wouldn’t want to call their entire belief system and construct of reality into question or some other trivial issue like that).

Imagine what a world we would have if everyone was open to any idea-  would it be a better world?  Oops- there comes that esoteric stuff again…

A couple nights ago I was reviewing one of the cycles I actually computed in December of 2007 for the year of 2008. For fun, I will post the remaining projection for 2008 (today is November 21st, 2008).
Make a note of it and send me your comments in January.  Here are the anticipated turns:

Long - 2008/11/25
Short - 2008/12/02
Long - 2008/12/12
Short - 2008/12/19
Long - 2008/12/25
Short - 2008/12/30

The cycles for this projection run close to 60% accurate (meaning they are profitable) with an average trade of 12 S&P points; not bad for a year in advance!  The short side in recent months has been more reliable for obvious reasons averaging closer to 25 S&P points. It made over 600 S&P points so far in 2008- a grand whopping 76% of the current level of the index!  Either way, the statistical probability of the above historical result on the projected cycle is inconceivably low and makes a strong case for the existence of significant order in the markets that most people choose to deny without any intent of honest investigation.

Perhaps that is why cycle trading gives such an edge in trading – simply because even if you told people about the cycles they would ignore it-  The funny thing is, if the stock market is the collective financial behavior of market participants, then they choose to ignore themselves ;-)
Have some fun with it!