R&C: What are the most impactful challenges you have encountered during implementation of the current expected credit loss (CECL) accounting standard for financial institutions (FIs)?

Bharodia: Each organisation has its own unique challenges. Depends on the size of the institution, the challenges are coming from data sourcing, model implementation, adjustment criteria setup, mean reversion logic determination, results validation and compressed timelines. The data attributes needed for model results analysis and special or individual asset treatment provide significant sourcing challenges which may impede the achievement of desirable CECL implementation. The availability of the models on time and attribution analysis may create significant implementation challenges. Finalising the mean reversion logic and adjustment criteria requires detailed analysis on model results. Validation of results requires transparent calculations within the solution. Compressed timelines with concurrent results validation and operational setup force a team to deal with competing priorities.

Hopping: While the challenges vary based on the size and sophistication of the institution, organisational challenges are among the most common and impactful. The risk and finance teams do not speak the same language, they are typically siloed in the organisation, and bringing them into one cohesive project team is a challenge on its own. Another challenge is with data and models. They are often in the critical path to understanding the impact of CECL on reserves. Most institutions are creating new models and updating some existing models to repurpose them. This is typically done while modellers keep up with their ‘day job’. Dirty or incomplete data and new data requirements slow this process down. Further, these risk teams are not used to the process and controls that finance will impose. Having these model outputs go into financial disclosures is a game changer for risk teams.

Jan-Mar 2019 Issue

SAS Institute