VaR Change Analysis

Automated commentary framework for VaR change analysis

  • Goldman Sachs
  • Designed a machine learning framework and an automated commentary framework that attributes movement in a percentile of two regressed PnL distributions to its corresponding risk factor regressor variables
  • Used time series filtering and reduced computation time by 90%, correlation clustering for joint movement reporting
  • Outperformed the results of a state-of-the-art through higher accuracy, faster computation, and cluster outputs