For each time range there is a table with symbols:

• At the top of the table are symbols with positive forecasts, a “buy” signal.
• At the bottom of the table are symbols with negative forecast, “sell” or “short” signal.
• For trading decisions use the most recent forecast.

It is recommended to consider both the signal strength and predictability.

• The signal strength is the absolute value of the current prediction of the system. The signal can be positive (predicted increase), or negative (predicted decline).
• The predictability is the historical correlation between the prediction and the actual market move for that particular market. For trading one should prefer stocks with the strongest signal and largest predictability. Please note that the longer term forecasts have usually higher predictability.
• The table colors are indicative of the signal. Green corresponds to the positive signal, red to the negative signal. Deeper color means stronger signal, lighter color: weaker signal.

More about the predictability:

• The "strength" of the prediction is the Predictability P, which ranges between minus 1 to plus 1.
• This metric is an adaptation of the Pearson correlation coefficient.
• P=-1 means the actual market movement was against the prediction.
• P=0 means no correlation between the prediction and the actual market movement.
• P=1 means perfect correlation between actual market movement and its predicted change. Any value of P above zero means positive predictability, the higher the better. . For stocks we monitor and predict the Predictability P ranges generally between P=0.2 and P=0.7.

This is how P is calculated: for each prediction point we calculate the past correlation between the prediction and the actual value, then average the results for all prediction points, while giving more weight to the recent performance.

The P values may change with time for two reasons:

• The predictability P has a component emphasizing the recent predictability, thus it can change with the market conditions.
• As the machine keeps learning, the values of P generally increase.