CRM Credit Risk Migration Analysis

Understanding the risk associated with the various types of loans constituting a bank's loan portfolio plays an integral role in its ability to properly measure its risk adjusted return and estimate its formula based reserves. To better understand these risks throughout the business cycle, the stewardship and strategic charges of Management often suggest the use of data modeling tools and frameworks to estimate the Probability of Default ("PD") and Loss Given Default ("LGD") parameters which in turn enable the forecasting of credit loss. The use of such risk parameters replaces a number of the subjective elements inherent in many banks' credit loss estimates with the Bank's own historical risk migration and loss data.

Probability of Default/Transitional Risk Matrix

CRM's Credit Risk Migration Analysis model estimates the probability of default by reference to the average observed monthly migration of loan balances between a bank's risk grades, expressing such migration in a transitional probability matrix. Such matrix may be developed across product types and geographies as well as by collateral code and/or loan vintage depending on desired results and available data. Once formulated, the transition risk matrix:

• Enables automated monthly monitoring of risk grades, defaults, losses, and pay offs;
• Projects run-off and defaults as well as provides a comprehensive future-state projection of portfolio composition;
• Provides Probability of Default metrics by risk grade at valuation date;
• Provides a basis to test both the design and implementation of internal controls surrounding a Bank's internal risk rating system;

Expected Loss Factors

In addition to the control testing procedures enabled by the transitional risk grade matrix, the probability of default yielded by such matrix may be paired with the corresponding loss given default, also calculated using bank-specific losses, to estimate potential credit loss, which in turn

• Provides a defendable and repeatable means of evaluating the potential credit loss contained in a loan portfolio
• Supports an ALLL methodology that is comprehensive, well-documented, and consistently applied

Stress Testing/Loss Distribution Model

CRM's Loss Distribution Model utilizes the bank-specific PD and LGD metrics, combined with a bank-specific intra-portfolio correlation factor, to provide a probabilistic look at a spectrum of credit loss possibilities. CRM's Loss Distribution Model:

• Enables forecasting of cumulative credit losses over a desired time horizon.
• Allows a Bank to plan capital and allowance levels

Contact us to find out more.

Back to Services Index