The model invests in stocks from the Russell 1000 index based on Benzinga’s real-time financial news feed.
Natural Language Processing (NLP) on media content reveals sentiment and time decay which indicates the strength and the longevity of an article’s position towards tradable assets. Positions are evaluated and entered daily and are held conditionally until either the sentiment score drops below a predefined threshold or an exit criteria is met (i.e. Stop Loss or Target Gain).
This model relies on features extracted from news sentiment set to measure the “buzz gauge” which identifies and extracts both sentiment and volume in conjunction with additional fundamental and technical factors to identify long opportunities from the Russell 1000 universe. The model retrains dynamically on a roll forward basis to account for changes in market regime.
