Dominant Factor™ methodology screens each candidate asset against a set of about 500 DataCore proprietary factors that are based on market observables (these risk factors have been carefully pre-selected, using Artificial Intelligence techniques). It tests for statistically significant non-linear models linking asset total returns and those of the factors within the fitting window (typically, the last 3 years) and retains for the asset  only those asset-specific factors that show significant explanatory power.

It then constructs asset-specific risk-reward profile that incorporates asset’s projected losses in the event of severe crises based on established links and observed long-term behavior of the proprietary factors retained for the asset. Asset selection decision is then based on its risk-reward profile and risk tolerance parameters set for a particular index.

As a result, a sector (industry or geographic) is considered attractive when capital starts flowing into the sector, producing a rally in its components. But, unlike common methods (such as Black-Litterman), which tend to be fooled by speculative bubbles and suffer when these burst out, DataCore portfolios flee away from a sector when the rally becomes nervous, announcing instability and a potential severe downturn.

In a way, by its nonlinear approach, the Dominant Factor methodology detects the behavioral effects of traders’ sentiment on prices and is able to interpret them, so as to avoid catastrophes. Instead of following the crowd, the system reads its sentiment from its impact on price behavior and acts accordingly. This results in a portfolio that is cautious during the rallies that are fragile, and thus cuts down losses by a significant amount in the downturns. Clearly, the Beta adjusts appropriately according to the market conditions, which is for us the very meaning of "Smart Beta" and "Smart Gamma".