The credit loss standard (CECL) issued by the Financial Accounting Standards Board (FASB) overhauls the current impairment models for loans, leases, and debt securities and also impacts commitments. It removes the “probable” threshold under the “incurred loss model” for recognizing credit losses. We will explore the new CECL Standards in
terms of what they do and how they impact the financial Industry, and what it will take to implement CECL in summary.
Then we’ll take a look at preliminary disclosure format examples: scope, sample sizes, and data requirement. We’ll close with a review of how Mortgage Industry Advisory Company (MIAC) is helping our clients prepare.
Overview of the Rules
Firms will be required to report the current estimate of lifetime loan losses, incorporated into the Allowance for Loan and Lease Losses (ALLL): thusly, CECL brings fair value into the ALLL. While a DCF approach was considered by FASB in exposure drafts, the final standard allows any approach, as long as it is reasonable. Institutions, auditors, and regulators will decide, so early discussions are
encouraged.
Both quantitative and qualitative methods are to be utilized jointly in a process that is generally described. Although there is a strong a bias to the use of cash flow models with assumptions powered based on quantitative data and “reasonable” scenarios, a justified historical loss-based result, which is quantitative only, could suffice.
The basis of Allowance Estimates is that CECL requires that estimates be based on relevant information about past events, including both qualitative and quantitative factors such as Historical loss experience with similar assets, then-current conditions, including evaluations of the borrower’s creditworthiness, and reasonable and supportable forecasts that affect the expected collectability of the financial assets’ remaining contractual cash flows.
Historical experience is quantitative information, so many models are based on it. Reasonable forecasts and conditions assessments are qualitative in nature as they must provide a forecast direction and estimate. Any factors not otherwise addressed in the final process are
likely to be qualitative in nature.
Other qualitative factors include:
- Changes in lending policies and procedures, collections, etc.
- Changes in the experience of management and staff
- Changes in the quality of the loan review system
Financial assets carried at amortized cost less a loss allowance will reflect the current expected cash flows to be collected and the income statement will reflect the credit deterioration or improvement.
If financial assets are carried at fair value with changes in fair value recognized through other comprehensive income (OCI), the balance sheet would reflect the fair value, but the income statement would reflect credit deterioration or improvement.
Under some circumstances, the institution can elect not to recognize expected credit losses on assets held at fair value. The conditions are that both the FV exceeds or equals the amortized cost, and if
expected, credit losses are insignificant.
What Will it take to Implement CECL?
An array of new processes will be required, including but not limited to management, governance, risk reporting, controls, and functional integration.
Program management will be the start, as will be newfound coordination among functional areas such as finance, originations, credit, operations and technology, and a revised governance and risk
management framework.
Segmentation of loan, lease, and debt portfolios into clearly identifiable portions with similar and discrete characteristics is the next step. This means specific identification and description of the
characteristics of the assets will be used to model probability of default (POD) and loss given default (LGD) to make the ALLL processes Basel compliant, plus development of specific credit risk modeling and forecasting models and tools.
FASB will require institutions to develop well-documented data management and data quality processes. Validated Extract-Transform-&-Load (ETL) systems and procedures and data quality reporting must be created.
Next, firms will identify control points in the origination and operations areas to ensure adequate support and data capture, and integration of new tools within Financial and Regulatory Reporting
Procedures.
Importantly, dual (jurisdiction?) reporting institutions—CECL and IFRS 9—firms will need to:
Handle 12-month (IFRS 9) versus lifetime (CECL) credit loss projections and reporting in models and report generation; and Estimate credit losses for future draws on commitments for IFRS and
for commitments that cannot be unconditionally canceled for CECL.
Planning for CECL
There are definite steps that organizations need to take to prepare for integrating these standards. This will start with defining a revised governance standard for CECL, establishing a steering
committee/task force with members currently in high-level positions in
finance, originations, credit, and operations who have management backing. This team needs to be given budget and action authority to implement prescribed procedures.
Then, as required, firms will need to identify appropriate external consultants and partners who can participate and contribute to the process, and should seek ways to integrate them into the process
early. These firms should be able to create and/or evaluate models for conformance with the new standards.
Next, these teams will determine the resource needs involved in each area and inventory what exists today, beginning with performing initial portfolio segmentation of all loan, lease, and debt assets
held, determine what data is needed, and what models and the technology needs by portfolio.
The committee will need to examine existing models and methods used in the ALLL process to determine which have the potential to meet the new requirements. Generally, cash flow forecasting models offer potential, and static models do not. Firms will perform pilot evaluations of the potential impact of CECL on the financial statements and the organization over the next two years, and then develop revisions to policies and procedures, to audit policies, and to model risk
management practices.
These findings will be used to build the “how-to” document or “roadmap”. Key objectives and milestones should be set along these lines:
Segmentation
Warehousing
Metrics
Modeling
Scenarios
Portfolio Segmentation: Segment the loan, lease, and debt portfolio into meaningful segments, so that then, firms can appropriately define data elements needed for each. This leads to identification of the critical statistical drivers of performance, which will be used in the
modeling and reporting, and defining data models.
Establish Data Warehousing: The CECL plan should include the data warehousing and capture infrastructure, tools and data models required, which will require implementation of powerful data capture methods, in monthly snapshots, including credit data and performance
data. Firms will need to obtain and reconstruct as much historical data as possible, plus review existing and alternative models, research credit modeling, loss modeling, and voluntary prepayments
modeling.
Firms must identify models that can be used or repurposed for different products, internal and external, and identify leverage opportunities and efficiency gains from current models, processes,
workgroups, and modeling approaches (ALLL, DFAST, or internal vs.
vendor models).
Develop Metrics and Assumptions: MIAC helps firms use these data models to develop key reporting metrics and descriptions of the drivers used in the credit, loss, and prepayment models; determine
assumptions; and drive the building of a descriptive narrative of all the models selected and the processes to be used.
Model Design and Development: Institutions will utilize the warehouses and the tools to develop historical analysis and internal models, and integrate results with existing systems. The task will be to expand these systems to handle scenario inputs, develop financial and managerial reporting format, as FASB will require lenders to define and explain the use of any credit-grading systems and other
qualitative inputs to the ALLL process.
Design Reasonable Scenarios: CECL urges banks to develop a narrative to explain the basis of the scenarios and why they are deemed to be reasonable. Regulators will review the scenario rationales with senior management, so institutions are preparing currently with external
consultants and auditors.
Firms will begin parallel test runs of the CECL ALLL process and the existing process, and test the process and the models and back-testing models. The challenge is to identify differences (size), volatility in results and capital impacts, and evaluate strategic considerations
across product lines. Naturally, product pricing, origination standards, processing, collections, and operations will be of the essence.
The CECL Committee is encouraged to write narratives, and work closely with auditors and external partners to finalize the narratives and reporting.