HealthScore Trends: Changing Score Distributions

What did the credit union industry HealthScore distribution look like ten years ago as compared to today? In this post we explore the score mixture from 2010 through 2019, highlighting impressive changes in credit union scores, and a recent plateau that could indicate a pending downturn.

Setting Up The Analysis

For this analysis we extracted year-end scores for each year 2010-2019. We then segmented each year’s scores into score buckets as follows:

  • 0-2.99
  • 3-4.99
  • 5-6.99
  • 7-8.99
  • 9-10

We also calculated the percentage of credit unions in each bucket, and charted the data, which we share below.

Note that when we established our score ranges we looked back at 20 years worth of credit union data. A score of five is generally “average” performance based on that history.

Breaking Down the Score Distribution

The first set of data to share is the count of credit unions in the various score buckets, which is shown in the table below. To aid in your interpretation of the data… for the cycle ending 12/31/2010 there were 139 credit unions that scored 2.99 or less, 3,811 with scores between 3 and 4.99, 3,278 credit unions with scores between 5 and 6.99, and so on.

Cycle Date0-2.993-4.995-6.997-8.999-10Total CUs
12/31/20101393,8113,27826307,491
12/31/20111233,3983,43428507,240
12/31/2012842,7703,70839716,960
12/31/2013672,7923,48234606,687
12/31/2014502,3083,57946506,402
12/31/2015331,9163,66253606,147
12/31/2016351,7223,54460805,909
12/31/2017241,5273,45568305,689
12/31/2018191,0963,58579205,492
12/31/2019211,0383,47181905,349

And here is that same data in chart form:

The amazing run of improved performance we have written about every quarter for the last 24 quarters, or six years, has everything to do with the clear-out of credit unions in the sub-five score buckets. In the chart you get a very good sense of just how many credit unions scored below average back in 2010 – nearly 4,000 – and how few do now relatively speaking.

And here is another way to look at that same data – by the percentage of total credit unions contained in each score bucket.

Cycle Date0-2.993-4.995-6.997-8.999-10
12/31/20101.86%50.87%43.76%3.51%0.00%
12/31/20111.70%46.93%47.43%3.94%0.00%
12/31/20121.21%39.80%53.28%5.70%0.01%
12/31/20131.00%41.75%52.07%5.17%0.00%
12/31/20140.78%36.05%55.90%7.26%0.00%
12/31/20150.54%31.17%59.57%8.72%0.00%
12/31/20160.59%29.14%59.98%10.29%0.00%
12/31/20170.42%26.84%60.73%12.01%0.00%
12/31/20180.35%19.96%65.28%14.42%0.00%
12/31/20190.39%19.41%64.89%15.31%0.00%

And here is the associated chart:

As noted in the table and chart focused on the count of credit unions, we had a large number of credit unions performing below average in 2010. The count showed more than 4,000, which at the time was more than 50% of all credit unions. We now show less than 20% in that same range as illustrated in the chart above.

Key Findings and Interpretations

Reports on credit union industry performance often focus on the aggregate credit union balance sheet and income statement. For example, if total outstanding loans for all credit unions increases then such reports indicate credit unions are growing loans. Of course this is not factually incorrect, but aggregate reporting can mask how credit unions are really performing both individually and collectively.

Here is what we mean. A relative handful of very large credit unions can be growing loans while at the same time a much larger percentage of credit unions are shrinking in loans. The aggregate may be increasing thanks to those large institutions, but to report that credit unions are growing loans hides that fact that “credit unions” really aren’t growing loans at all.

We believe our HealthScore solves this particular challenge in that the score is not calculated on aggregate data, but on individual data. The HealthScore, then, is a summary of individual credit union data that identifies the performance movements of credit unions overall. When credit unions are looked at this way, real individual/industry challenges and issues reveal themselves.

To that point, here’s what we’re picking up on as individual/industry issues – both positive and negative:

  • Mergers: In the chart highlighting the count of credit unions you can see that 2,142 credit unions ceased to exist over the 2010 to 2019 time period. Most of those credit unions were merged into others, and most, but not all of those merged credit unions were in the below average score buckets. Their removal aided the improvement in credit union scores overall. In other words, the industry improved, in part, because of the demise of hundreds of credit unions.
  • Score Plateau: For most of the last decade, especially starting with 2012, credit union performance improved substantially quarter after quarter. The percentage breakdown chart shows this very clearly. However, the last year shows a very distinct “leveling off” of scores. In 2020 it is very likely we will see growth in the sub-five score bucket as larger numbers of credit unions in the 5-6.99 range see performance deteriorate due to slowing loan growth and sustained margin pressure.
  • Merger Pace: In 2019 the year-over-year decline in the total number of credit unions dropped below 3% for the first time in a long time – we think driven by the improvements illustrated in the tables and charts above. However, depending on whether economic pressures persist, we could see an acceleration of merger activity in 2021. We think this will come from two segments. The first will be from credit unions already living in that sub-five bucket. The second will be from those credit unions noted above that experience performance issues that drop them into the sub-five bucket.

Preparation is Key

The best thing for credit unions to do presently is to ensure clear understanding of the real root causes of current performance, and then to identify the risks likely to impact those causes over the coming year. If, for example, you are a credit union that lives by indirect lending, you are likely already seeing some growth pressure. High car costs, the durability of new cars purchased over the last couple of years, coronavirus(!!) all may further impact car purchase volumes in 2020. You should be actively assessing your cost structures and flexibilities now to ensure you can aggressively trim costs while maintaining good member service if car loan demand continues to slow throughout the year.

If you need some help with this kind of inspired discussion, let us know. You can either complete an online proposal request form, or schedule a phone consultation.

Inviting Participants

We’re looking into the creation of a monthly HealthScore tracking report. This effort would involve a representative sample of credit unions agreeing to send us key ratios every month that we then turn into mid-cycle scores. We believe such a report would allow for more rapid identification of emerging industry performance trends over our typical quarterly report cycle.

If we decide to execute this idea, we’ll likely reach out directly to credit unions we consider “representative” – but will also include credit unions that have a desire to be involved. If you want to put your credit union on that list, let us know by sending a note to healthscore@glattconsulting.com.

1 comment

Leave a Reply

%d bloggers like this: