Operationalizing CECL, if done right, can add value by enhancing risk-based pricing for products, boosting risk-adjusted profitability, optimizing portfolio/product mix, and even helping you make more informed acquisitions or divestitures. Credit risk models developed for CECL can help management gain greater insight into the nature and performance of their loans and asset portfolios and better understand what risks are worth taking. Effectively leveraging the CECL use test is one key step in moving beyond a check-the-box compliance approach to CECL and driving real value for your financial institution.
CECL Use Test Defined
CECL requires financial institutions to include CECL model outputs, internal reporting metrics and required disclosures in an institution’s day-to-day risk management activities and functioning in order to improve its business decision making. Specifically, CECL will push banks to refine their origination (including estimating cost of credit); underwriting; portfolio analysis; and risk measurement, monitoring and reporting processes as part of sustainable business-as-usual activities. To accomplish this,
Getting more from CECL
- Lines of business (LoBs) should develop advanced metrics to better understand and monitor the credit quality of their portfolios and, whenever possible, their individual loans. The LoBs should attribute their expected losses to specific macroeconomic, portfolio-specific and loan-level factors. This exercise should lead to improved risk-based pricing, higher risk-adjusted profitability and lower expected losses.
- The bank should determine the reasons for changes in the estimate of expected credit losses from one period to another.
- The bank should also strive to use CECL for improved credit risk appetite setting and capital allocation for granular level portfolios and large individual loans.
Required disclosures under CECL are not meant to be merely a check-the-box compliance exercise. Smarter firms should fully leverage these disclosures and other internal metrics to assess the credit risk inherent in their LoBs, portfolios, and individual loans with greater precision.
For example, suppose a specific retail LoB originates a large number of new loans that are mostly prime but that also include a much smaller percentage of near-prime loans. After assessing the overall initial credit quality of this new portfolio and existing underwriting criteria, the LoB buckets all the loans into the prime segment because, on an aggregate basis, the new portfolio exhibits prime characteristics under the bank’s credit ratings system. The modeling team then estimates the expected losses for this portfolio, but then determines that losses are higher than its peers with similar portfolios and wonders why.
With deeper analysis, the bank finds that a small percentage of customers indeed had near-prime risk characteristics but were labeled prime as part of the bigger portfolio. At origination, these near-prime customers were offered pricing that was low relative to their risk.
Could the bank have lowered its initial estimate of net expected losses? Yes. Here is how.
The bank should have split the larger portfolio into two risk pools: prime and near-prime. The loans should have been priced in accordance with individual customer’s respective risk profiles, with higher prices for the near-prime customers. Higher prices would have earned higher revenues and offset some of the expected losses in the near-prime segment, thereby lowering the net expected losses for the overall portfolio.
The bank could mitigate future losses by capturing data at a granular level. It could spend more time to scrutinize the following information, which is embedded within the required CECL disclosures:
- reasons for impairments (pricing was not risk-based)
- the number of positions that are in an unrealized loss position (loan counts and proportion of loans in prime vs. non-prime segments)
- the severity and duration of the impairments (for each segment)
Two key lessons from this exercise? First, create risk segments based on pre-specified business and risk criteria to be able to conduct proper risk attribution. Second, quality data is imperative—the bank could not have offered differentiated pricing for the two segments if it did not have granular loan level data.
CECL can also inform bank strategy so that it is able to choose those business segments that provide higher risk-adjusted returns, operate within its credit risk appetite and improve its capital allocation. The bank can also revamp its operating procedures so that all originations are examined with greater precision for credit risk. To achieve deeper insight into expected losses, risk managers should examine the following performance:
- Risk-based pricing
- Risk adjusted returns
- Default rates
- Delinquency rates
- Charge-off rates
- Percentage of nonperforming assets.
- Loan Debt-to-collateral-value ratios
- Third-party guarantees
- Current levels of subordination
- Geographic concentration
- Industry analyst reports
- Sector credit ratings
- Volatility of the security’s fair value
- Interest rate changes since purchase
- Macroeconomic factors such as unemployment or HPI
- Demographic factors, etc.
The bottom line? Don’t look at CECL as simply a compliance exercise. Leverage risk and performance insights to inform strategy, lending and pricing decisions to improve your results. Grant Thornton’s integrated CECL approach can help address your full range of CECL concerns.
Strategic Risk & Operations Advisory
+1 312 754 7145