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).
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).
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment