How does your system make a forecast or prediction?
We have developed a set of proprietary computer algorithms that detect patterns in stock data, then finds similar stocks in the past that had the same pattern. From the pattern and other similar stocks (we call them reference stocks), we can measure the probably of movement, both up and down, from the next day's opening price.
Unlike traditional technical analysis, our system looks at price and volume over time in order to make a forecast. We have found that volume is extremely important as it is a true measure of the commitment of other investors' decisions. Since most trading volume today is made by mutual funds, and these funds trade huge amounts of money, they require many days or weeks to move into or out of a stock. We have found that many of these actions frequently form predictable patterns.