- Time Frame: Short Term
- Update Schedule: Every Morning at 7am.
- Source: OPTiON-i Cybernetics
The XLB 2SD models are built using a machine learning technique called ensembling. These models are built from 8 different machine learning algorithms to produce a more robust probability. Each algorithm is taught to classify time periods that are showing characteristics that lead to 2 standard deviations higher or lower over given time period. Each algorithm uses a pool of 189 hand picked price, sentiment, and breadth indicators that are updated daily. From these features, the 2SD finds a edge in regards to large movements over the coming days. The oscillator we present is the difference between the probability of a 2 standard deviation move higher and a 2 standard deviation lower. The most interesting and actionable time periods are when there is a large discrepancy one way or the other in the 2 sd oscillator. The higher the oscillator the higher probability of a 2sd move higher relative to lower and vice versa.