The credit union loan portfolio has expanded 45.8% in the past 5 years. To serve members more efficiently, credit unions are trying new strategies. Anecdotally, Callahan & Associates has observed an uptick in the use of automated decisioning for the consumer loan portfolio. Additionally, Callahan clients have requested more information on the automated decisioning activities of other credit unions in this area of business.
To learn more, Callahan & Associates survey 333 credit unions in the fall of 2016 to learn about automated decisioning practices in the consumer lending portfolio. The survey consisted of 7 questions, and respondents ranges in asset size from $1M to $15B.
Read on for insights from the survey. Then, dig deeper into different asset-based ranges using the interactive features below.
Insight 1: Does asset size affect whether a credit union has a consumer loan origination software?
Answer: Yes
Of the 333 respondents to the survey, 63.66% used automated decisioning; 36.34% did not.
When examining asset bands, there is a positive correlation between size of the credit union and Loan Origination System (LOS) usage. Approximately one-quarter, 26.67%, of credit union respondents with less than $100 million in assets use automated decisioning, compared with 94.74% of credit unions with more than $1 billion in assets.
There is also a loose correlation between asset size and the intent to implement an LOS. For example, 16% of respondents with less than $100 million in assets said they were planning to implement automated decisioning. For credit unions with $250-$500 million in assets, that number jumps to 41%.
The impersonal nature of an LOS was the leading reason for why a credit union does not use automated decisioning in its consumer loan portfolio.
Select an asset range in the feature below to examine the differences among five groups.
CONSUMER LENDING AUTOMATED DECISIONING BY ASSETS*
FOR 333 CREDIT UNION RESPONDENTS | DATA AS OF 06.30.16
Callahan Associates |
*Click on a peer group below and the graphs will adjust to display data for that asset band.
var divElement = document.getElementById(‘viz1526486032155’); var vizElement = divElement.getElementsByTagName(‘object’)[0]; vizElement.style.width=’1000px’;vizElement.style.height=’1027px’; var scriptElement = document.createElement(‘script’); scriptElement.src = ‘https://public.tableau.com/javascripts/api/viz_v1.js’; vizElement.parentNode.insertBefore(scriptElement, vizElement);
var divElement = document.getElementById(‘viz1515617160423’); var vizElement = divElement.getElementsByTagName(‘object’)[0]; vizElement.style.width=’1000px’;vizElement.style.height=’1027px’; var scriptElement = document.createElement(‘script’); scriptElement.src = ‘https://public.tableau.com/javascripts/api/viz_v1.js’; vizElement.parentNode.insertBefore(scriptElement, vizElement);
Insight 2: Does automated decisioning negatively affect consumer loan delinquency?
Answer: Unlikely
The median consumer loan delinquency for the credit union respondents that reported using automated decisioning is 49 basis points, compared with 55 basis points for those credit unions that did not report using automated decisioning.
The ranges for the middle 50% distributions are similar for both groups: 0.31%-0.91% for those that do and 0.25%-91% for those that do not.
Insight 3: Are there correlations between automated decisioning and:
-
Efficiency?
-
Consumer loan origination growth?
-
Loan-to-share ratio?
-
Average member relationship?
(1) Answer: Yes, slightly
When examining the entire group of credit union respondents, those that use automated decisioning have a median efficiency ratio of 79%; thats 5 percentage points lower than those that do not.
However, the median efficiency ratio for credit unions with less than $100 million in assets is 4 percentage points higher for institutions that use consumer automated decisioning. Perhaps executives at smaller credit unions are accepting the upfront costs of implementation with the expectation for future growth and efficiency.
(2) Answer: Yes.
Credit unions respondents that use auto decisioning for their consumer loans have higher rates of five-year consumer loan origination compounded annual growth rates (CAGR). The median for those that auto decision is 13%. Thats 5 percentage points higher than the 8% CAGR for those that do not.
(3) Answer: Yes.
The median loan-to-share ratio for credit unions that reported using automated decisioning is 81%. Thats higher than the median loan-to-share ratio for credit union respondents that do not use automated decisioning.
Credit unions that are more loaned out could have higher loan demand and a greater need for an LOS. Conversely, credit unions that have an LOS could have the ability to generate more loans.
Either way, the correlation is evident.
(4) Answer: Yes.
Average member relationship correlates positively with consumer loan origination systems. The median average member relationship for credit union respondents that auto decision is $16,971, with half of the results between $14,329 and $20,474. Comparatively, credit union respondents that do not use automated decisioning have a median average member relationship of $13,318, with half of the results between $10,861 and $15,937.
CONSUMER LENDING AUTOMATED DECISIONING PERFORMANCE DATA*
FOR 333 CREDIT UNION RESPONDENTS | DATA AS OF 06.30.16
Callahan Associates |
*Click on a peer group below and the graphs will adjust to display data for that asset band.
var divElement = document.getElementById(‘viz1515617160423’); var vizElement = divElement.getElementsByTagName(‘object’)[0]; vizElement.style.width=’1000px’;vizElement.style.height=’1027px’; var scriptElement = document.createElement(‘script’); scriptElement.src = ‘https://public.tableau.com/javascripts/api/viz_v1.js’; vizElement.parentNode.insertBefore(scriptElement, vizElement);
Insight 4: How often to credit unions evaluate their consumer decisioning criteria?
Answer: Yearly.
The most frequent response to whensurvey respondents evaluate their criteria yearly (34%). Also, many respondents specified looking at the criteria when changes occurred in the lending program.
Some respondents said they evaluated their criteria weekly when they first implemented the LOS and then less frequently when their programs were up and running. One respondent reported,I will review if the underwriters see a loan that shouldnt have gotten through the engine or one that should have gotten through.
The responses indicate credit unions still rely on human oversight to ensure quality control within the automated decisioning process.
Insight 5: What percent of loans are auto approved or denied, compared with manually approved and denied?
Answer: 30.38% of loans are auto approved or denied.
Although credit unions are manually reviewing the majority of loans, 30.38% of loans are auto decisioned. The composition of those automatically decisioned loans is 22.81% auto approved versus 7.57% auto denied (which equates to a 75.1% approve and 24.9% deny breakdown within automated decisioning).
Given the average loans outstanding at credit unions that use automated decisioning is $765.0 million, 30.38% automatically decisioned loans can make a big impact on the portfolio.
Additionally, based on the scatterplot below of the credit union respondents, an increase in auto approved loans does not correlate with an increased consumer delinquency rate.
Insight 6: Who has the biggest market share in the loan origination system market?
Answer: Meridian Link.
Meridian Link is the most popular LOS provider among survey respondents. It has not only the most clients overall but is also the most used provider across all asset ranges. Fiserv and Symitar come in at second and third, respectively. Fiserv offers four platforms. From least to most popular, they are: Velocity, XP2, LoanCierge, and Loan Desk. CRIF is the fourth most popular LOS provider among survey respondents. In addition to the SAIL and ACTion platforms represented in the survey, CRIF offers three other platforms: Mark IV, BizMark, and LoanCenter.
CONSUMER LENDING AUTOMATED DECISIONING USER TRENDS*
FOR 333 CREDIT UNION RESPONDENTS | DATA AS OF 06.30.16
Callahan Associates |
*Click on a peer group below and the graphs will adjust to display data for that asset band.
var divElement = document.getElementById(‘viz1515680167852’); var vizElement = divElement.getElementsByTagName(‘object’)[0]; vizElement.style.width=’1000px’;vizElement.style.height=’1027px’; var scriptElement = document.createElement(‘script’); scriptElement.src = ‘https://public.tableau.com/javascripts/api/viz_v1.js’; vizElement.parentNode.insertBefore(scriptElement, vizElement);
