Make better lending decisions through data analysis education.
Scorecard Series: Call Report TrendsIn the last blog, we went over the Visible Equity Database scorecard. For this entry, we will be going over call report trends. For any questions as to how certain statistics are calculated, see the Call Report Statistics Descriptions document (Page 1, Page 2, Page 3).
Federally insured credit unions report data to the NCUA on a quarterly basis. This collection of data is known as the call report. The main advantage of call report data is that the data are collected on all federally insured credit unions. Because of this, we don’t have to do any clever sampling techniques like on the previous database scorecard. The biggest disadvantage, however, is that the data are only collected every quarter. So the data are collected every three months, and they typically are not made available for public use until a couple months after the end of the most recent quarter. So for these reports, we are using data up to Q4 of 2016.
Here is our call report trends scorecard. This is quite a bit busier than the previous scorecard, but we’ll break it down carefully to make it more digestible. With this scorecard, we’re reporting a bunch of statistics for various categories over the course of four years: Q4 2013, Q4 2014, Q4 2015, and Q4 2016. We tried to keep the statistics only to things that are useful, and frankly, we only have four years because that’s what fits on a single page. The different categories we’re looking at include general information, loans, shares, assets, delinquencies, and charge-offs. Let’s go ahead and jump into the top section.
This first section reports general information: number of active credit unions, number of members, the member to credit union ratio, number of borrowers, and the borrower to member percentage. Let’s look at a single year to simplify things.
Here we are looking at the statistics for Q4 of 2016. Below the Q4 header we see sub-headers that say “Agg.,” “Grow.,” and “Med.” The “Agg.” sub-header stands for aggregate. This column simply aggregates the data for all credit unions. For example, the total aggregate number of members for all credit unions as of Q4 in 2016 was about 108 million.
The next sub-header, “Grow.,” stands for growth. This column is also on the aggregate level, and it shows the year over year growth. For example, the aggregate number of members grew by 4.10% from Q4 of 2015 to Q4 of 2016.
The last sub-header, “Med.,” stands for median. This column is NOT on the aggregate level. Instead, with this column, we found the statistic of interest for each credit union and then took the median. This gives an idea of what to expect for a “typical” credit union, whereas the aggregate level shows more of the credit union environment as a whole. Using the member example again, we see that the median number of members for a credit union is 3,434 as of Q4 in 2016. Notice that the median for credit unions and the member to credit union ratio report NA. This is because it doesn’t make sense to find the median number of credit unions for each credit union, so we chose to just block out those cells. Each year of each category in this report is broken down in this aggregate, growth, median manner.
Going back to the General trend section...
...we can see these same statistics year over year. So for the number of credit unions, we can see how many credit unions are active under the Agg. column for each year as well as the year over year growth. It looks like the number of credit unions has decreased by about 4% every year. However, it appears that the total number of members has increased every year. Thus, the member to credit union ratio has also increased every year. Borrowers have also increased by about 6% every year. An interesting indicator for activity is the borrower to member ratio, which we see has increased year over year. One final note to point out is that some statistics in this report, like the borrower to member ratio, are reported as percentages, and in such cases, we put the growth column in terms of basis points for clarity.
With loans, we are only looking at the total loans, average loan balance, and yield on average loan. Total loans is the total amount of loans (in $) for all credit unions. Average loan balance is the total amount of loans divided by the total number of loans. And yield on average loan is the total interest on loans minus the total interest refunded divided by the average of the total amount of loans for the current quarter and the previous year end.
We see consistent growth from year to year for total loans and average loan balance, and we see very small decreases for yield on average loan. The medians follow the same growth pattern as the aggregate level. Total loans and average loan balance consistently increase, while yield on average loan slightly decreases. Note that yield on average loans is annualized. It’s not obvious from this report because we are looking exclusively at Q4 data, but future iterations of this report not looking at Q4 will have certain statistics annualized.
With shares, we’ll first focus on the top section. The statistics we’re reporting include total shares, average shares per member, total loans divided by total shares, total regular shares, total share drafts, total money market shares, and total share certificates. As far as these last four share types go, we chose to report on these four because they make up the vast majority of all shares.
Looking at just 2016, we see that all of these statistics are on the rise. Most notable are total shares, total regular shares, money market shares, and share certificates. These all seem to have increased quite significantly in the past year when compared to previous years. For example, with share certificates, we see that in previous years there were decreases or only slight increases. The most recent call report shows that share certificates as a whole have increased by almost 5% over 2015. Another interesting finding is that the median for share certificates has barely changed over the last four years. This implies that the increase we’re seeing on the aggregate level is probably due to a smaller number of credit unions as opposed to all credit unions as a whole. If we take a look at average share per member, we see that this has increased every year. Since we already know membership is on the rise, this implies that shares are growing at a faster rate than members, so members are putting more money into share accounts. It’s interesting that all of these statistics are increasing while the number of credit unions is decreasing.
Now we’ll take a look at the bottom portion of the table. In this section, we’re going into a little more detail about the four share types. The categories this time are percent of total, growth, and rate. The percent of total category shows how much of all shares is made up of the share type in which we are interested. In 2016, regular shares made up about 36% of all shares, and under the growth column, we see that there was a 137 basis point increase over 2015. The rate column shows the median dividend rate for that share type, so the median regular share dividend rate in Q4 of 2016 was 0.12%. An interesting takeaway is that share drafts, money market shares, and share certificates have recently declined as a total percentage of all shares whereas regular shares have increased. Perhaps members are taking money from these accounts and putting it into regular shares.
This next category is on assets. We have total assets, return on average assets, and net long-term assets divided by total assets. The big takeaway here is that assets as a whole have been increasing over the past four years, return on average assets have not changed significantly, and net long-term assets divided by total assets went from an increase between 2012 and 2013 to a decrease ever since.
With delinquencies, we are looking at total amount of delinquent loans, the amount of delinquent loans divided by the amount of all loans, and the amount of delinquent loans divided by the amount of assets (we are considering delinquent to to be at least two months past due). Total delinquencies decreased between 2012, 2013, and 2014, but since then have increased. It’s actually quite surprising that the median amount of delinquencies increased quite a lot between 2015 and 2016, going from about 110,000 to about 125,000. But not surprisingly, total delinquencies to total loans and total delinquencies to total assets seem to be slowly changing from decreasing to increasing. However, the magnitude of these changes appear to be quite small.
Finally, we have charge-offs, and we’re looking at total net charge-offs, bankruptcy charge-offs divided by total charge-offs (which are both year to date), and net charge-offs divided by average loans. Just like delinquencies, we see that charge-offs went from decreasing to increasing. However, the most recent increase in charge-offs is quite high at about 22%. With bankruptcy charge-offs compared to total charge-offs, we see that this stat is consistently decreasing. So charge-offs as a whole are becoming less due to bankruptcies. And net charge-offs to average loans is essentially not changing much at all.
This scorecard gives a good look into how the credit union industry as a whole has been performing over the past few years. The next scorecard we’ll review breaks these statistics down by peer groups, so stay tuned for that blog.
Make better lending decisions through data analysis education.