The largest cohort of CECL adopters by volume is now beginning to prepare for CECL which is only 6 quarters away. Whatever your plan, execute it. As a leader in CECL advisory and management, Valuant has already seen a YOY increase in 2023 CECL partnerships of 75%. As volume continues to grow, resources will begin to thin, and fees will rise.
Much like tackling that home renovation, when it comes to CECL you can DIY or outsource it to a 3rd party such as Valuant. In either case, we have prepared an instructional series to guide you along your preferred path.
Methodology and assumptions are important to a CECl calculation but documenting the conversion process may be equally important.
Detailing each step in the conversion process will make investors, examiners, and external auditors comfortable.
Key CECL Considerations
- Accurate data
- Communication with the Board and management
- Understanding of how your peers are implementing CECL
- Proficiency of FASB official guidance
- Documentation of methodology selection process
- Reasonable and Supportable forecasts
Governance & Controls
Given the complexity and subjective nature of a CECl calculation, governance of both the process and the model will be critical to defending the calculation and its assumptions.
The Board of Directors and its audit committee must define the governance and controls necessary to support management’s judgements made prior to implementation.
Before the CECL process can begin, management needs to assess the risks associated with the development of a CECL model as well as the impact of any material misstatements. At the direction of the Board, management should define the control environment by preparing a CECL model procedures document that identifies:
- CECL Committee
- Purpose and risk considerations
- Strategies and assumptions
- Communication of progress and results
- Model risk management and validation
This committee should not be a committee of one or two people. It should represent the key stakeholders of the allowance calculation. This committee should be Responsible, Accountable, Consulted and Informed (RACI Matrix).
What Will Your Data Be Used For?
- Allowance for Credit Losses
- PD/LGD or Other Loss Rate Calculations
- Financial Reporting – Footnotes and Call Report Disclosures
- Prepayment Rates
- Unfunded Commitments
- Income Recognition
Critical Data Elements
A Data Governance Program can help identify the Critical Data Elements for the CECL Model.
CDEs represent data elements that are critical to success for the business outcomes.
CDEs originate from specific tables or fields within Loan Systems or Data Warehouses.
Data Lineage describes how a CDE is derived.
What data quality metrics are in a place to verify CDEs?
Data Quality Dimensions
- Accuracy – Accurately represents real-world values
- Timeliness – Represents reality from the required time period
- Uniqueness – Recorded only once, not duplicated
- Validity – Conforms to the format, type or range intended
- Completeness – Complete in terms of required potential data
- Consistency – Represented the same way across the data set
Segments and Classes
In evaluating financial assets on a collective (pool) basis, an entity should aggregate financial assets on the basis of similar risk characteristics, which may include any one or a combination of the following:
- Internal or external (third-party) credit score or credit ratings
- Risk ratings or classification
- Financial asset type
- Collateral type
- Effective interest rate
- Geographical location
- Industry of the borrower
- Historical or expected credit loss patterns
- Reasonable and supportable forecast periods.
FASB ASC 326-20-30-3
“The allowance for credit losses may be determined using various methods. For example, an entity may use discounted cash flow methods, loss-rate methods, roll-rate methods, probability-of-default methods, or methods that utilize an aging schedule. An entity is not required to utilize a discounted cash flow method to estimate expected credit losses. Similarly, an entity is not required to reconcile the estimation technique it uses with a discounted cash flow method.”
What Does it Mean?
ASC 326 does not prescribe a specific loss methodology to calculate the ACL.
Each methodology has its own pros, cons, and applications. In addition to the technical aspects of each calculation, other operational considerations and constraints may impact model selection. These include but are not limited to data availability, data integrity, loss history,
expertise, understandability, efficiency in use, internal control, and other financial reporting considerations.
Loss Methodology Selection
The most simplistic CECL methodology; calculated by applying the observed historical loss rate to a portfolio balance amortized over its remaining life. Maybe the most familiar and most easily understood as compared to current ALLL methodologies.
NCO-based survival model that uses loan level attributes to summarize total incurred losses over an observation period
Loan-level model that takes the calculation one step further than Open Pool by calculating both the defaulted dollars and expected losses over an observation period, thereby bifurcating the default event and the loss event
The vintage analysis leverages a closed pool concept by capturing the losses by origination year, or vintage, of a loan’s life.
Lifetime Loss Rate Methods
- Ease/familiarity of the calculation
- Inherently considers some assumptions such as prepayment speeds
- Can capture migration, life cycle curves
- Extensive data requirements! Generally, need to capture a full or multiple economic cycles
- No R&S Forecast or explicit economic variables – may increase need for Q Factors
- Requires judgment when applying vintage loss history to newly originated loan portfolios; assumes the historical portfolio represents current
- Hard to implement for long term loans (e.g.,15- or 30-year real estate loans)
Cash Flow Model with PD/LGD
- Projects future cash flows over the life of the loan portfolio
- Discounted or NonDiscounted
- Generally accepted as best practice for modeling credit losses
- Probability of default (PD) and loss given default (LGD) are two key components
- The calculation can be challenging for smaller institutions
Advantages to using a PD/LGD loss methodology in combination with a DCF
- Insights into the portfolio that support improvements in the bank’s underwriting and collection procedures
- Management can incorporate changes in the bank’s business strategy by using inputs entered directly into the DCF model
Discounted Cash Flow
- Ability to run multiple scenarios
- Provides insights into trends that exist over a historical lookback period
- May not need a full economic cycle worth of data
- Flexibility and customization of inputs
- Does not explicitly capture migration
- Still has intensive data requirements
- Could require considerable amount of personnel, IT, data and modeling resources
- Complexity – can get into analysis paralysis
To Discount or Not to Discount?
The discount rate on PCD Assets is not the same as other assets – how will this be managed?
If an entity estimates expected credit losses using methods that project future principal and interest cash flows (that is, a discounted cash flow method), the entity shall discount expected cash flows at the financial asset’s effective interest rate.
If an entity estimates expected credit losses using a discounted cash flow method, the entity shall discount expected credit losses at the rate that equates the present value of the purchaser’s estimate of the asset’s future cash flows with the purchase price of the asset.
Factors to Consider
Forecasting and Q Factors
The measurement of expected credit losses is based on relevant information about past events, including historical experience, current conditions, and reasonable and supportable forecasts that affect the collectibility of the reported amount.
Reserve = Past events + Current conditions + Supportable Forecast
The forecast component is possibly the biggest change from the ALLL. What does Reasonable and Supportable mean? How long is the forecast period?
CECL is a life of loan calculation but the forecast should only be as long as it can be properly documented and supported. This might only be two years for some Segments. For any timeframe beyond the forecast period, consider reverting historical loss information for the balance of the contractual life not covered by the forecast period.
➡️ TIP: A CECL calculation needs to be as quantitative as possible. This includes your forecast. The more quantitative your calculation, the more defensible the examiners will find it.
- A mechanical forecast or forecasts based on correlation of bank historic performance with economic variables or performance indices
- Qualitative factors
- A combination of both mechanical forecasts and qualitative factors. For banks looking to develop a mechanical forecast, an option to consider is the scenario approach. This approach looks at one or more stressed scenarios based on forecasted economic scenarios and their expected impact on default (PD) and loss.
- The 9 standard factors identified in the 2006 Interagency Policy Agreement should still be considered along with new factors that reflect the portfolio or the market
- Be careful: some factors may “doublecount”your quantitative calculation
- Can be used in conjunction with a mechanical forecast or by themselves CECL guidance suggests that any calculation be as quantitative as possible.
- Make the qualitative factors quantitative and less subjective
- Document and support
Use Qualitative Factors to account for one-time or un-modeled events.
- Start fresh. This is not the ALLL: Q factors are now the forecast
- Factor adjustments must change from your ALLL
- Anchoring your Q factors
- Carry over all but the economic adjustment for your baseline
- Benchmark CECL coverage ratios to inform adjustments
- Factors should be less subjective and based on facts
- More economic and less portfolio performance
- Develop a matrix approach to lessen subjectivity
- Document logic and assumptions
Disclosures & Reporting
Public Business Entities
Firms that are considered Public Business Entities (PBE) are required to file 5 years of vintage performance disclosures with their GAAP financial statements PBEs that are not SEC filers must file 3 years and add an additional vintage each year until 5 years have been presented (Optional for non-PBEs). These reports are great for internal reporting.
The purpose of vintage disclosures is to allow the reader to understand the credit risk within the portfolio as well as the methodology considered for the CECL loss estimates.
CECL does not prescribe specific types of vintage disclosures, leaving it to the firm to determine the correct amount of disclosure. The guidance does suggest that vintage data include data by segment and/or product type with specific credit quality indicators that might include:
- Loan to Value
- Performance Metrics