NeuroShell Trader Professional 5.0 was built with a stock scanning facility that allows you to scan a large number of stocks looking for high or low values of some particular indicator, or stocks which meet a particular criteria as expressed in an indicator (which can be built with the Indicator Wizard and might be quite complex). That scanning process can also be used with our neural net addons Adaptive Turboprop2 and Adaptive Net Indicators, since they are essentially neural nets embedded in indicators. So you can pick stocks that are reading high on a neural net, for example.

The stock scanning facility won’t scan for backtested trading strategies or nets trained with the Prediction Wizard. However, what follows is a method to scan with such models. It is essentially using our “batch processing” ability to make models using all of the stocks (or other issues) present in the multiple chart pages of a chart.

Lets suppose you want to find issues that are currently doing well with the RSI indicator. That is we want to find stocks that have made money with an RSI strategy, not just stocks that have a high or low RSI reading. Load a chart with several dozen of the issues in which you are most interested. Then insert a trading strategy using RSI for buying (you may want to omit short selling) but do not select optimization (at least not at first). Backtest over the period you are interested in, say the previous 4 weeks. It won’t take more than a minute to go through the dozens of stock issues you have loaded.

Next, insert the “System Equity: All Trades” indicator which you will find in the category called “Trading Strategy: System Information”. This will accumulate the equity over your backtest period.

Now go to the View menu and turn on the Portfolio view. You’ll notice that one of the columns is labeled “System Equity”. If you click on the top of this column you will see your stocks get sorted in the portfolio view by their performance over the backtest period (alternate clicking changes the sort from ascending to descending and back).

That’s it. Now you can buy the stocks at the top of the list, because they have responded best to RSI recently.

Here are some modifications:

1. You can include parameter selection optimization if you desire in the trading strategy. The main disadvantage is that your “scanning” can take a while, depending on your computer speed.

2. You can add many more dozens of stocks, but only if you have the RAM for them. Keep the amount of data loaded low, and your model simple to get more issues loaded.

3. You can include short selling in your trading strategy.

4. Do the same thing with the Prediction Wizard.