Employee turnover at fast food restaurants can be as high as 150% each year. That's right, some of these business owners have to hire their staff each year and then replace half of those new hires again before the year is over.
In banking of course jobs are much more stable, even when industry pressures mount and competition runs high. One area where pressures have been bubbling is how best to leverage the data you might need for CECL to support compliance and hopefully even drive in more profits.
Banks must assess risk over the life of a loan or security, so getting enough data to achieve CECL compliance can be a substantial challenge for many banks. You should estimate expected credit losses in light of historical information, current information, and supportable forecasts of future events and circumstances too. Also, be sure to add an educated guess for prepayment rates.
Of course, not every bank has the historical data it needs to solve this problem. Even when it does, it might find that the history is too static. History should include economic changes over the time frame in question, or an unusual event (like a crisis) that isn't likely to repeat.
Some bankers may also feel unsure of their ability to use modeling assumptions. To avoid garbage in garbage out, you have to think about specifically tying either internally generated economic forecasts or those acquired to effects on your loan portfolio. While most bankers might be comfortable applying small qualitative judgments to reflect short-term assumptions, most prefer to use historical loss as the primary driver.
These are all good reasons to use external economic data to augment CECL calculations. To do that, you will need reliable data from public and/or private outside sources.
To begin, we suggest that your team reviews the baseline domestic macroeconomic variables from the OCC for Comprehensive Capital Analysis and Review (CCAR) and Dodd-Frank stress testing (DFAST) purposes. Regulators use these big bank variables to develop economic scenarios, so these are good starting points for relevant data. Another source for decent peer data or state average data is the FFIEC.
Also, for national metrics such as disposable income growth, unemployment, and housing prices, look to local and regional indices. Fed data is a good potential source for those perhaps.
Before you incorporate external historical information from any source into your CECL numbers though, you should also answer some questions about that data.
Regulators and auditors want you to understand things like: how clean the external historical data is, whether the source is reliable, do trends in national or state-wide data reflect your local conditions, and are your external data choices consistent with other forecasts or history.
Then, be sure to compare/contrast all of this with whatever historical data your bank already has, as well as with peer experience and industry benchmarks.
All of this may feel like a tall order. However, we can help. Contact us to discuss CECL and your data today.
