Ascertaining the Uncertain: methodology and implementation in Model Risk Quant
Given the number of models now used in finance and their role in decision making, model risk quantification is of increasing interest to the board of financial institutions as well as regulators. The goal of model risk quantification is to ascertain if the uncertainty in the output of a model or/and a network of models can have direct consequences for model governance and wider risk management. The first half of this webinar will outline a methodology (jointly developed by Prometeia and Basinghall Analytics) for: a) estimating the uncertainty in a variety of models (due to inherent model limitations); b) adjusting this uncertainty over time in light of new information (collected as part of model monitoring); c) assessing circumstances causing a model exceeding its risk appetite tolerance limits. This methodology also produces a margin of conservatism estimate, which can be seen as a point of comparison for the regulatory MoC. Experienced practitioners and renowned academics will facilitate a round table discussion in the second half of the webinar. Panelists: Dr. Nasir M. Ahmad, Managing Partner at Basinghall Analytics; Prof. dr. Bart Baesens, Professor of Big Data & Analytics at KU Leuven and lecturer at the University of Southampton; Dr. Tiziano Bellini, Head of Risk Integration Competence Line Europe at Prometeia; Anurag Chavan, MD and Head of Model Risk Management at HSBC.