CreditManager: Risk Attribution Module
categories: RiskMetrics CreditManager, Product Documentation, general
Credit risk statistics in CreditManager are produced by a simulation-based model. Changes in these statistics between two dates are driven by various factors such as portfolio composition changes, the passage of time and changes in market data such as yields, credit spreads, correlation market factors and transition matrices (TM). Understanding and attributing these changes across time is an important part of the capital allocation process.
This paper highlights CreditManager’s Attribution Module that is designed to systematically address changes in simulated statistics such as VaR and Expected Shortfall (hereafter called capital), as well as non-simulated statistics such as current value and mean horizon value, across two points in time. The top level change in the statistics is broken down via a cumulative, sequential changing of the individual factors (Figure 1). This can be further attributed across portfolio drilldowns defined via tags. In the following sections, we present a practical overview of the methodology before moving to a case study showing the module in use. We conclude with examples of the CreditManager work flow.