We surfaced data by United Van Lines that looked at customer migration patterns in 2018. Given so many boomer customers of community banks are moving around, we thought you might like to know the top states for retiree destinations (possible deposits) were: NM (43%), FL (39%), AZ (37%), SC (37%) and ID (34%).
As boomers continue to retire and move, bankers continue to try to understand the ramifications of CECL. After all, the CECL model represents a massive change in loan loss reserve methodology, and as such FASB has allowed a long window for implementation. Yet, as we are all well aware, time passes quickly and the implementation date will be here before you know it.
Despite it being out for some time, many banks still have questions around CECL. One of the most prevalent we hear is: "How do I choose a method and what will auditors and regulators expect as documentation for the one chosen?" To help you, here are some guidelines you can use:
First, determine if the data you have will be adequate under each of the methods to calculate a statistically sound reserve. Probably the biggest driving factors in choosing an appropriate method are what type and periods of data are available, and the number of losses the bank has actually experienced within each homogeneous group.
Next, get to know the data. Once you understand the data you are working with and the method(s) that won't work (due to a lack of data or lack of loss observations), you will ultimately surface the methods that will work. This is especially true for community bankers, who generally have a handful of loss observations vs. the large amounts of data that mega banks may have.
Third, to get a good handle on the data, take an inventory of the loan level data you have available. Be sure to list the years of core data available; prepare a list of loss observations and determine how many loss observations you have for each homogeneous group; identify the time periods of call data you have access to for your bank; and identify the time periods of other industry data you may have available.
Fourth, have a general understanding of how the different methods work and how much data is needed to calculate a statistically sound reserve.
Fifth, know the limits. In general, if your universe of loss observations is limited, most of the methods will not work unless you make some assumptions. These can be hard to substantiate to auditors and regulators, so care must be taken. In addition, limited data and loss experience may lead you to augment loss experience with industry data or use industry data only. In these cases, issues can arise, so be careful and thorough.
Sixth, if your bank has adequate data and a sufficient amount of loss observations for a variety of methods, further discussion will be needed.
Seventh, when you discuss the methods, some logic that may be used includes: determining any obvious methods that won't work and why; narrowing down a list of the methods which will work; documenting the attributes of the short list under each of the methods; making a decision and stating the reason why (should focus of the type of data you have and the risk profile).
If you have questions about this, we are here to help because we know CECL can be tricky. Contact us today to learn more.
