The predictive machine learning algorithm is finely tuned to forecast positive stock performance 1 year from today. Predictive scores above 0.5 is statistically validated to have an accuracy of 80%, while scores between 0 - 0.5 is validated to have an accuracy of 60%.
We look at analyst over-perform/buy/hold/under-perform/sell reccomendation data, 5 year company growth estimates and, current shorts interest as a percentage of shares floated.
Total positions held by institutions and the trend of increased positions are analyzed. New and sold out institutional holders are also considered in our predictive analysis.