Tuesday, July 28, 2020

NCUA Overstates the Interest Rate Differentials Between Credit Unions and Banks

The National Credit Union Administration (NCUA) is overstating the interest rate differential between banks and credit unions for various loan and deposit products.

NCUA uses data from S&P Global Market Intelligence, which compares the national average rates for 23 common loan and deposit products at banks and credit unions, as well as the average rates for these same products at banks that converted from credit unions.

However, a 2020 paper by a Credit Union National Association economist -- Jordan van Rijn -- and others are critical of previous studies that used naive estimates from institution- and branch-level interest rate data.

The authors write that studies relying on this data are subject to selection bias. The authors note that studies using institution- and branch-level data do not reflect the actual rates paid by households, but the best advertised rate.

In other words, advertised or average rates do not control for household and loan characteristics. Therefore, the authors write that this best advertised rate may significantly differ from the actual rates paid by consumers.

The study, which examines new and used car loan rates, found that credit unions offer lower rates than banks on auto loans, but found that the interest rate differential is smaller than the interest differentials implied by institution- and branch-level data.

The paper compared its results with the data reported by NCUA. It concluded that this selection bias can explain about half of interest rate differential on new auto loans and approximately a quarter of the interest rate differential on used car loans.

Given this selection bias associated with institution- and branch-level data, which leads to an overstating of the interest rate differential between banks and credit unions, NCUA should remove this information from its website.

However, if the agency continues to publish this information, it needs to include a disclaimer that its interest rate differential data do not reflect the actual interest rates paid or received by consumers and the differences are overstated. This disclaimer should be in bold, large type at the very top of the website.

The paper, Financial Institution Objectives & Auto Loan Pricing: Evidence from the Survey of Consumer Finances, can be found on the Social Science Research Network (www.ssrn.com).

No comments:

Post a Comment


The content is provided for educational purposes only, with the understanding that neither the authors, contributors, nor the publishers of this site are engaged in rendering legal, accounting or other expert or professional services. If legal or other expert assistance is required, the services of a competent professional should be sought.

Comments appearing in response to articles appearing on this site do not necessarily reflect the views of the ABA. ABA makes no representations regarding the truth or accuracy of commentary or opinions that may be posted in response to the articles that appear on this website.

The inclusion herein of any link to a website, either in the text of an article or in a comment, does not denote any approval, sponsorship, or endorsement by the ABA, and ABA is not responsible for the content or opinions expressed on those linked websites or related commentary. This content is not licensed to third parties sites and is not affiliated with any third party site. Any reference to the author or this content on any third party site on the Internet is not authorized by the ABA.

It is the policy of the American Bankers Association to comply fully with all antitrust laws. Certain discussions should be considered off-limits, including those that contain competitively sensitive data such as price and cost information, or statements that could be construed as reflecting an attempt or desire to control or influence a particular market or markets. Future pricing or other prospective competitive information should never be shared.