Using Kpler Refineries to predict earnings metrics

Kpler's daily refinery data tracking 900+ global refineries enables investors to predict earnings for major oil companies like Phillips 66 weeks before official reports, with machine learning models achieving 96.9% accuracy in forecasting quarterly refining revenue. This groundbreaking analysis demonstrates how real-time refinery production data, combined with ensemble modelling techniques, can forecast critical earnings metrics like refining throughput and revenue with remarkable precision, offering traders and analysts a significant information advantage in energy markets.