This page gives a general description on how AID generates its signals. Details are highly technical, are trade secrets and beyond the scope of this page.
AID uses two strategies to generate signals, strategy B and strategy G. This is because some investments work better with strategy B than with strategy G and vice versa. There are differences in both strategies but there are similarities. The common elements will be discussed in the text after this paragraph.
Both strategies consist of three steps. Step 1 is filtering, step 2 is determining tops and bottoms and step 3 is generating signals.
Step 1 is filtering (smoothing) the price data. This done to reduce the effects of random noise and short-term trading. Prudent investing looks at long term trends while ignoring random noise and short-term trading. Filtering the price does this because it is what prudent investors do.
Step 2 is classifying the today’s price as a top, a bottom or neither. This is important in determining buy and sell signals. Classification is done on filtered (smoothed) data.
If the classifier finds a top, it generates a sell signal. If it finds a bottom, it generates a buy signal. If it finds neither, no signal is generated. All classifier findings are reviewed by a human looking at a chart of the price data. This is because classifiers can make errors or trigger on small variations. Having a human review the classifier findings give you the benefit of both human and machine decision making.
Step 3 is the final step in generating signals. Simply determining top, bottom or neither status is not enough. You need a price for your limit order.
Limit prices are determining by taking past data and making a forecast of the next day’s price with a first order autoregressive model. The price forecast is the midpoint ((High Price + Low Price) / 2). The midpoint is used for reliable execution of limit orders.
Here are links if you have more questions: Click here to browse investments. or Click here to see how AID works.