Keeping Up With CECL

There have been continual conversations within the financial institution industry as how to best prepare for the upcoming regulatory changes required by CECL (Current Expected Credit Losses). While various methodologies are being considered, the Expected Loss or PD/LGD approach has been discussed in many of those evaluations.

To recap:

Expected Loss = Probability of Default X Loss Given Default X Exposure at Default

(EL=PD X LGD X EAD).

PD/LGD calculates the probability of loans experiencing default events and matches them to the percentage of the defaulted loan balance that is ultimately charged off. The formula can be applied by loan count, but most lenders take a balance approach which gives more weight to larger loans and shows the percent of total balance of the portfolio that has defaulted over the period. The loss rate is discovered by multiplying the PD by the LGD, and can then be applied to the loan portfolio balance to determine expected future losses. EAD is seen as an estimation of the financial institution’s exposure. For fixed loans, EAD is generally set to one which represents the current outstanding balance. For commitments, EAD is categorized in two ways, drawn and undrawn with the latter being subject to an estimation.

Dual rating systems typically assign a rating to the general creditworthiness of the obligor (PD) and a rating to each facility outstanding (LGD). The facility rating considers the loss protection afforded by assigned collateral and other elements of the loan structure in addition to the obligor’s creditworthiness. Dual rating systems have emerged because a single rating may not support all of the functions that require credit risk ratings. PD is the likelihood that a loan will not be repaid and fall into default. It must be calculated for each borrower. The credit history of the borrower and the nature of the investment must be taken into consideration when calculating. As a scale for comparison, PD can range from 0% to 100%. If a borrower has 60% PD, it is considered a less risky company versus a company with an 80% PD.

The challenge presented by CECL is that the PD/LGD models used today don’t include any elements of the reasonable and supportable forecasts required by the new standard. While there’s no specific regulatory guidance, this methodology could require quantitative adjustments to one or more of the factors, an additional factor layered into the calculation, or simply addressed qualitatively through narrative or through judgment as it relates to the data sets. Particular attention should be made to how PD/LGD modeling evolves. Whether qualitative adjustments are made as “top of the model” adjustments to the traditional models, or become embedded in the calculation. Either way, expect some level of documentation to be required in order to support the forecast adjustment.

An important step in applying CECL is to determine the appropriate portfolio segmentation. A portfolio segment is the level at which a financial institution develops and documents a systematic methodology (such as PD/LGD) to determine its allowance. There are broad implications for the appropriate selection of portfolio segmentation in the application of CECL. In determining segmentation, loans should be pooled by similar risk characteristics. The PD of a borrower not only depends on the risk characteristics of that particular borrower but also the economic environment and the degree to which it affects the borrower. Thus, the information available to estimate PD can be divided into two broad categories. Macroeconomic and borrower specific. Some examples of these two categories for commercial real estate include:

Macroeconomic information

• Changes in home ownership percentages which can impact the rental market
• Seasonally adjusted rate of apartments under construction
• Local, regional, and national job growth
• Local, regional, and national vacancy rates
• Rent rates

Borrower centric information

• Revenue growth
• Payment history
• Number of times delinquent in the past six months
• Loan-to-value ratios
• Internal risk rating with sufficient stratification
• Collateral adequacy

This information is specific to a single borrower and can be either static or dynamic in nature. By segmenting the portfolio accurately, one can narrow the required macroeconomic information per pool segment.

LGD is the fractional loss due to default. The fractional component is a result of the recovery rate defined as the portion of a bad debt that is recoverable. The value of the underlying collateral assigned to the loan will lessen the bad debt and subsequent loss.

Under the new current expected credit loss model, financial institutions will be required to use historical information, current conditions and reasonable forecasts to estimate the expected loss over the life of the loan.

The transition to the CECL model will bring with it significantly greater data requirements and changes to methodologies to accurately account for expected losses under the new parameters. Segmenting the portfolio with similar risk characteristics will refine a more accurate estimate in projecting the future expected losses required in the coming standard. The PD/LGD approach, with modifications, continues to be discussed as one methodology that is manageable.

Author Bio

Baker Hill empowers progressive financial institutions to increase revenue, reduce risk, and drive more profitable relationships.

Streamline business, consumer direct and indirect lending with our common origination platform. Understand profitability and risk at every level with our sophisticated business intelligence and analytics. Monitor and maintain a healthy financial portfolio with our statement analysis, exception, and risk management solutions.

Explore our solutions below for more information on how to generate sustainable growth, minimize risk and increase profitability.