CECL is often perceived as a new regulatory and reporting burden. But it shouldn't be if you plan out how to structure your data to make reporting easy and with tools to drill down into your customer data to understand performance and improve future results.
There are many 'rules' of good reporting. CECLNow tries to apply as many as possible in our set of CECL forecast documentation and analytic tools. These include:
Build the Data Foundation
You need to pull together your customer data. There will be specific data that you need for CECL. But you also may have other data that can help profile your customers to build a better understanding of their performance and opportunity. Think broadly about what data to capture and store.
Comprehensive and Focused
Reporting should answer as many questions as possible. This can mean standard reports or the ability to generate reports as needed. Ideally the non-standard reports could be saved to make it easy to update in the future.
Understandable
Cryptic data has no value. The key metrics used in reporting should be clearly explained. Too many abbreviations distract rather than inform.
Actionable
The reporting should lead to recommendations. What pockets of risk should be avoid? How should we grow our business? Providing supporting detail from the reports makes it valuable for decision-making and performance tracking.
Consistent Calcuations
Shared definitions should be used across all reporting so that all readers understand the reports in the same way over time. This can also reduce effort and speed reporting because time is not wasted on recreating the wheel.
Supports Vintage Analysis
Vintage analysis measures key metrics based on the age of the accounts. Looking at differences between different origination dates tells you whether your performance is getting better or worse.
Context
Along with vintage analysis it's important to compare like groups of accounts. Summary level data can mask both pockets of risk and areas of opportunity so drilling down along key dimensions is required. It's also important to filter data to fine tune recommendations.
Easily Distributed
The reports and recommendations should be handed to people who can make decisions and take action. It's also important to build a shared understanding of results and trust in the data.
Relates to Objectives
The reports should detail more than numbers/metrics. How does the report relate to your objectives, your KPI’s? How do the numbers relate to my bottom line?
Compare Forecasts to Actuals
Forecasts should be tracked to see how accurate they were. This can lead to important learning to improve future forecasts. The time series structure of your data should include point in time forecasts.
Once you build out your data assets you've found the gold mine that awaits. Next it's time to start digging.
