Credit Risk Management Glossary

Adaptive Control Systems	An iterate process using a champion/challenger testing to determine the optimal policy. Adaptive control systems can be used for approval, administration and collection functions. They can make use of behavior and scoring tools.
Bankruptcy Modeling	Generally uses behavior-type scoring to predict likelihood of bankruptcy. It can be used in all three stages of the credit cycle (underwriting, administration, collections).
Bankruptcy Scoring	A scoring system that predicts the odds of an account going bankrupt using characteristics of bankrupts that are statistically significant.
Behavior Scoring	A scoring system using ongoing loan repayment and purchasing behavior to predict the odds of an existing account becoming or remaining bad. Accounts are generally scored monthly or quarterly. It can also be applied to loan renewals, line: increases, and collections.
Bureau Scoring	A scoring system based on credit bureau information only. It can be used separately or with an application scoring system.
Credit Score	The numerical sum of weights applied to various characteristics that have been shown to predict credit worthiness.
Fraud Screening	A process used to protect financial institutions from customer fraud. Examples would be comparisons of credit bureau data to application data or to data bases of known perpetrators.
Good/Bad Analysis	Provides a quick snapshot of the loan portfolio's stability. It is the ratio of good loans versus a bad loan over time.
Line Increase/Decrease Matrices	Used to decide changes in the credit limits based upon different criteria. For example, line increases could be granted based on a percentage of credit utilization and behavior score. As well, multiple matrices may be used to adjudicate the loan application.
Portfolio Stratification	Portfolio is segmented into significant characteristics which are then analyzed. An example would be to segment delinquent accounts by cycle and then segment the accounts in each delinquency cycle by credit score.
Post Campaign Reviews	Can consist of long-term and ongoing analysis. The purpose of these reviews is to determine if campaign goals have been met. Actual delinquency, sales volumes and rate of return are compared to the forecasted objectives.
Propensity Modeling	Using demographics and/or scoring, predicts the likelihood of product usage and attrition used with revenue scores to predict out profitability.
Revenue Scores	Risk/return analysis; using credit scores, account revenue/profitability is calculated to find the optimal credit strategy to maximize return.
Risk Rating	A classification method to benchmark a loan portfolio's total risk.
Roll Rates	A method to measure expected delinquency. The underlying assumption is that future accounts will continue to flow :through delinquent buckets as they have in the past. Dividing the current month's delinquency bucket by the prior delinquency bucket, calculates the month's roll rates in the previous month.
Scorecard Validation	Is an ongoing testing process to ensure that scorecards in use are performing as expected. Scorecards are benchmarked against original performance estimates of good/bad odds for each score range.
Static Pool Analysis	A pool of loans from a specific time period that has ongoing analysis conducted upon it. Analysis would examine such things as delinquency, prepayments and rate of return.
Stress Testing	Sensitivity analysis of a loan portfolio to learn how it would perform under different scenarios (i.e. high delinquency and high margins, high delinquency and rapid prepayment, etc.).
Vintage Analysis	Similar to static pool analysis except that only delinquency and arrears are analyzed from a given pool of loans. This analysis provides insights about how delinquency for a pool of loans from the same time evolves. As well, changes in credit policy can be monitored using this method.

Credit Risk Management Systems