Most active traders (and casual market followers) have, at times, noticed periodic behavior in stocks they are following. Assuming you are one of these people you have probably thought "how can I use these types of fluctuations as a method to make money in the market?" I'm sure you can think of multiple methods to accomplish this task - here are mine.

Start with the big question: How difficult would it be to find the stocks where this periodic behavior occurs?

Answer: As with many things, the actions required are simple in concept. The implementation, however, is a tad more difficult. Assuming you have access to market data (in other words internet access) there are 2 steps, if you are interested in creating this type of report for yourself - or at least how it can be done - read on...

1st step: You need an algorithm to manipulate the data - Most people think this is best left to math majors, but this is not entirely true. There are hundreds of ways to fit curves to functions. There are also tons of technical books (mostly text books) that have very in depth and hardly readable explanations. Fortunately, most of these methods lend themselves to numerical methods of calculation - translation: someone has written a program to do just about any fit that's out there (your average spreadsheet has several of these methods built-in).

This leaves you with 2 options:

1) Learn the math and program the data manipulation to your liking (probably the best method - it's the only way you can be sure you are getting the results you want).

2) Research and find a program that's out there that will do the fit the way you want (you may have to pay).

The best advice I can give you on this step is to pick a method with the most variables for you to manipulate. This will allow you to "tweak" the fit to your liking. Remember, your goal is to tweak the fit to find actual stocks that are fluctuating predictably - and therefore provide opportunities to make money off them.

2nd step: Create an efficient/readable report - I use HTML and make charts with Perl, but then again I design and create for the internet, and this may be more complicated than necessary. Probably the easiest way for the average person would be to create a template in a spreadsheet. Once this is made all they need to do is enter the dates and closing values of stocks in their respective fields - and let the spreadsheet do the rest.

I have found the critical part is to be able to make and view charts of the data and the fit. Numbers and calculations can't tell you how well the fit shows change in a stock 1/100th as well as a single glance at a good chart.

That's it. My experimentations with this subject have found the magic is in the fit. As stated earlier fits are a dime a dozen (I paraphrase of course), the hard part is finding one that actually provides useful information. The beauty, I am happy to report, is that once stocks following this behavior are found, they are often remarkably predictable. In case you are wondering, yes I have done this type of calculation and, although it did take a few iterations, it did produce accurate, useful, and useable results.