CRO Quantitative Risk Modeling & Analytics Flashcards

Stress testing in risk management is BEST described as:

Answer: Evaluating the impact of extreme but plausible adverse scenarios on an organization's risk exposure

Stress testing evaluates how a portfolio or organization would perform under severe but plausible conditions that go beyond normal statistical assumptions.

In credit risk modeling, 'Loss Given Default' (LGD) measures:

Answer: The proportion of exposure that is lost if a borrower defaults, after recovery

LGD represents the fraction of the total credit exposure that is ultimately lost following default, net of any recoveries from collateral or legal proceedings.

Which risk model validation activity is considered MOST critical per U.S. regulatory guidance (OCC SR 11-7)?

Answer: Independent review of model conceptual soundness, data quality, and outcomes analysis

SR 11-7 guidance specifies that effective model validation must independently assess conceptual soundness, data integrity, and ongoing performance monitoring.

A CRO reviewing a credit scorecard notices the Gini coefficient has declined significantly over the past year. This MOST likely indicates:

Answer: The model's discriminatory power between good and bad credits has weakened

The Gini coefficient measures a credit model's ability to discriminate between defaulters and non-defaulters; a declining Gini signals the model has lost predictive power.

What distinguishes a 'fat-tailed' distribution from a normal distribution in the context of risk modeling?

Answer: Fat-tailed distributions assign higher probabilities to extreme outcomes than the normal distribution

Fat-tailed distributions have heavier tails than the normal distribution, meaning extreme losses occur more frequently than normal distribution assumptions would predict.

In operational risk modeling under Basel III, the Loss Distribution Approach (LDA) combines which two components?

Answer: Frequency distribution and severity distribution of loss events

The LDA models operational risk capital by separately estimating the frequency of loss events and the severity of those losses, then combining them via convolution into a total loss distribution.