Demystifying static pool analysis

As I work with and educate credit unions across the country on loan portfolio analytics, Static Pool Analysis seems to be one of the most confusing and difficult to understand topics.   There are a number of ways that static pool analysis can be used to manage risk in the loan portfolio, and prevent and predict future losses.

A static pool of loans is a grouping of loans with similar credit risk characteristics that were originated during a specific time period.  At a high level, this may be all indirect loans originated in the year 2010.  At a more granular level, this may be all loans originated in the “D” credit tier during the first quarter of 2010.  What’s important to static pool analysis is that it includes loans originated during a period of time where internal and external risk factors were basically the same.  Once the pool has been established, no loans are removed from the pool and no loans are added to the pool; the pool remains static.  The reason that the static nature of the pool is important is that if new loans are added, then the new loans will dilute the performance of the loan pool.  Let’s say, for example, in April 2014 you are analyzing the delinquency of loans originated in February of 2014.  If you have $1 million in loans originated in February and $10,000 of these loans are now more than 30 days past due, your 30-day delinquency would be 1%.  If you included March originations of another $1 million, your 30-day delinquency would fall to .5%, even though the loans originated in March have not even matured 30 days.  This is a simplistic example, but it serves to demonstrate how adding loans to a static pool can dilute the results of the analysis and provide inaccurate data.

To some degree, this demonstrates how current methods used by credit unions to evaluate loan risk are insufficient, and why regulators insist on static pool analysis.  A common metric used by credit unions today is the annualized loss ratio.  The formula for this metric is the total charge-offs for the last 12 months divided by the average portfolio balance for the last 12 months.  If the credit union has a growing loan portfolio (meaning the balance of the portfolio today is higher than one year ago), this calculation could yield a loss metric that hides the poor performance of older loans in the portfolio, as the higher number of new loans has not matured to the point of loss.  Static pool metrics provide a more accurate picture of a loan pool’s performance.

Therefore, it is critical that a loan pool is fully formed prior to conducting an analysis.  It would be inaccurate to include loans originated in 2014 in a static pool analysis of delinquency of loans by origination year as the 2014 pool has not fully formed.  It would not be complete until the end of December of 2014.  It would be appropriate to include January and February in an analysis of loans pooled by month of origination as those months have been fully formed.

Another important point is that when calculating loss on a static pool, the calculation should be a comparison of the loss amount and the origination amount of the loan, not the current balance of the portfolio.  This will enable your credit union to use this data to predict portfolio performance.  If your credit union originated $20 million in loans in 2011, and since origination your credit union charged-off $100,000 on loans originated in 2012, then the loss ratio would be $100,000 divided by $20 million to get to a cumulative loss ratio of .5% on loans originated in 2011.  Assuming nothing has changed, using this information one could predict that you would also lose .5% on loans originated in 2013 after they had matured for two years.  If you originated $40 million in 2013, then you could estimate that you will have lost $200,000 from that pool of loans by the beginning of 2016.  Your estimates become more precise as the precision of your pools increases.  If I did this analysis by credit tier, for example, I could be more precise with my estimates in the case that the percentage of my portfolio in “A” tier was different in 2013 than it was in 2011.  Using static pool analysis can increase the accuracy of predicting loss, repayment speeds, and rates of return to name a few examples.

Michael Cochrum

Michael Cochrum

Michael has worked in the consumer lending industry since 1989. In 1999, he joined the credit union industry, working for the Texas Credit Union League’s credit union. Mr. Cochrum ... Web: www.cudirect.com Details