Un-necessarily negative to be eschewed as a matter of course.
Berk Janestein wrote a book on cycles in the market. Much discussion has been discussed on cycles. Many predictions have been tried.
I see a rough sharp bottoming cycle that goes 45 52 20 38 32 28 10 50 25 20 (days)
It may have a few curves overlapping, if you go back far enough with fourier. (foor - ee - yare) Maybe it will yield ("i" before "e", except after "c") to a probabilistic predictor with a (dare we hope?) 5% error. Maybe the cycles are longer, but overlapping this way to make a chop, as you sometimes see in the Bay of Biscay, reflecting off the shore, when the fr. fleet is riding out a storm.
If one experiments with a pascal's ( pass - kalz) triangle
back to no variance, and takes it forward with probabilistics, pos or negative, with the first row all positive (no meaning or sure method otherwise...) Then, the choices are roughly 10, 30 or 50 :)
I favour 30 as it is closer the mean. That is, at 10, 30 or 50 days from the last dip, we have another dip. Which is more likely? The most safe for your money.
What cycles have we in Universe and nature, any fool can see? Well for one thing there is winter, summer, fall, and spring. There are also, years, months, days and weeks. These go without argument. Moons wax and wane. (64 day cycle with gravity..). Comets come and go. (Bill Haley, sadly is no longer playing.) There is also a crop cycle that grows and dies. Sunspots spot and despot. Dictators come forth and de-dictate (With alarming regularity in Africa. You can set your watch by rebellions in Liberia) There is a 10 month, and 40 month (or 39) in the market. There is also the trader's cycle of 12 to 16 weeks. (get out of it, if it has not made a profit. It's a comin' down, down, down..)
In 1974 a firm in New Zealand put together a bunch of cycles and ran the bejeepers out of their fourier analysing, linear regression prediction program. It said, that contrary to popular opinion, the market would fall off a ledge in that year. "How odd!" was all they could say. "I wonder if there is a mistake in the program?" The market fell. The vengeance of its fall surpassed all proverb. Was the program right, or just lucky? Unfortunately it was right again, and in the past as well a few times. But luck being such a powerful part of probability this is not sufficient to predict future success. Besides the SEC said you must put in that disclaimer. It kind of spoils the advertising.
It pays to ignore cycles if you cannot figure them out. If you can, you had best call yourself Nostradamus. Leave yourself about 15% time and level "rubber", wear a long green robe, a coned hat with the moon and stars and the odd planetary body on it. (On the hat put a single eye in the centre. In the center of the eye, a $ sign.) Affect a long white beard. Talk in riddles. Make a lot of money in the market. When people ask you, "Oh great seer, where will it all be tomorrow?", shake your head, look of out at the clouds from the open window of the turret of your castle laboratory, and with a furrowed brow, say "The future alone knows about itself, and poor, poor man does not. How do you feel about a cup of coffee?"
EC<:-}
Market Data 
Markets 
AI
