Using Peer Data to Estimate Loss Experience
The use of peer data to calculate loss experience in the ALLL is typically reserved for institutions that have little or no loss history. De novo institutions, for instance, do not have the loss experience necessary to perform the ALLL calculation; rather, they must rely on peer analysis to determine the corresponding loss experience to assign to the various pools in their portfolio.
While peer data is an accepted way for institutions to calculate loss, it is not as comprehensive as historical loss or other more granular forms of calculating loss, such as migration analysis. Because peer analysis uses data from other institutions, it does not capture the inherent loss experience of the institution performing the calculation.
As such, it is a regulatory expectation to graduate from peer analysis to other loss methodologies once an institution has a substantive pool of data. Typically, institutions will need to collect at least four quarters of data in order to graduate to more robust loss methodologies.