System Overview

updated 3/7/2007

Website History
The STSS neural network resulted from a one person full time effort started early in 1995. I intended to build a personal trading system which would outperform Value Lines average annual 18% Return On Investment (ROI). At the time, Value Line was generally regarded as the best service available. I figured if I could get to 20% it would be a very good achievement and I'd make more money trading. I didn't start out to build a short term trading system. The short term system simply worked out as the best of many systems which I designed and tested. If you think about it, it makes perfect sense. It's easier to predict what will happen one day in the future then it is to predict what will happen six months in the future. What are you doing tomorrow? What will you be doing six months from now?
After a full year of design work the system was complete. Happily, the appropriate back tests far exceeded my original goal of 20%. It then occurred to me that others would be interested in trading this system. I learned how to build a website and The Short Term Stock Selector was born on 4/11/96. In order to establish a track record each trade was carefully recorded and logged as the position closed. We now have an extremely large Trading History which continues to grow.

What to Expect
Each evening after market close the STSS neural network searches a database of stocks for potential trades. It finds between 0 and 200 issues which appear ready to turn upward. It then analyzes these potential trades and predicts a gain or loss for each. Each pick is rated "A", "B", "C", "D" or "E". Those with the highest scores are rated A and those with the lowest are rated E. Over a long period of time A picks should outperform all others. Over a short period of time anything can happen, the market is risky and not perfectly predictable. Those viewing the free page are not shown this rating information until the trade is closed. Although the neural network screens approximately 800 stocks nightly, it is very picky. It looks not only at the attributes of the individual stock but also at the market in general. Buying and selling tends to run in short cycles. When the market heads down the neural network tends to open many positions. When it heads up it tends to close them. The neural network can't create buying opportunities, it can just identify them. You have to be a little patient. There will be times when no "A" picks are available for a week or more. Then again, I've seen as many as 150 opened in a single day. The average position is closed out in 7 market days. Winners will average less than 7 market days, losers will average more.
Some astute viewers have asked, "How often do you retrain the neural network?" Neural networks can be retrained to emphasize the most recent factors influencing market behavior. The STSS neural network has not and will not be retrained. It is specifically geared to evaluate "Fear" and "Greed". When the neural network issues a lot of buy recommendations it will do so during periods when bad news is prevalent. You will be afraid to buy because of the bad news. Conversely, when it is selling the news will be good and you will feel like it is doing the wrong thing. As long as these emotions influence human thinking this neural network should do its job just fine.
If you have questions be sure to read our Frequently Asked Questions section as it provides more specific information.

Disclaimer
© Neural Net Trading Systems 1996 - 2008