💾 Archived View for gmi.noulin.net › mobileNews › 5605.gmi captured on 2023-06-16 at 18:00:46. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2023-01-29)
-=-=-=-=-=-=-
2015-09-28 08:18:45
Cheaper funding is not enough to induce banks to lend more
Sep 26th 2015
AFTER credit dried up in America in 2008, the Federal Reserve scrabbled for
ways to perk up spending. One trick it tried was to offer banks concessionary
funding, hoping they would lend more to consumers and so induce Americans to
open their wallets. An NBER working paper* published this week by Sumit Agarwal
of the National University of Singapore, Souphala Chomsisengphet of the
Treasury Department, Neale Mahoney of the University of Chicago and Johannes
Stroebel of New York University looks at data from hundreds of millions of
credit cards from 2008 to 2014 to work out why the results were so
disappointing.
First, the researchers looked for evidence of pent-up demand for consumer
credit. That involves comparing the credit-card balances of people who have
very similar credit scores but end up on different sides of the various
spending-limit thresholds that lenders impose on cardholders. The scale of
excess borrowing by those with the higher limit, the theory runs, gives a sense
of how much more those with the lower limit would spend if they could suddenly
borrow more. This is important, because banks tend to boost their consumer
lending not by lowering interest rates on credit-card balances but by
increasing credit limits. (Indeed, the paper assumes that the interest rates
banks charge do not change even as their own funding becomes cheaper.)
The most creditworthy customers, it turns out, are the least keen to splurge
when extra credit is offered. For every dollar their credit limits increase,
they boost their borrowing by $0.23. Even that is an exaggeration: by further
digging through the data, the researchers establish that the borrowers with the
best credit records are only shifting their borrowing from card to card to take
advantage of improved terms not borrowing any more in aggregate. At the other
end of the scale, those with the muckiest credit histories borrow an extra
$0.58 for every $1 hike in their credit limit.
But that is not the whole story. The researchers then take a bank s
perspective, and ask to whom it makes most sense to lend. Boosting credit
limits draws in extra interest payments and charges, but there are costs too.
If it is mainly the highest-risk borrowers who take advantage of higher limits,
or if the higher limits encourage more reckless borrowing in general, then
default rates will climb, eating away at profit margins.
The researchers use the same trick as before to work out how much more likely
people are to default when offered a higher credit limit. Combined with
information on revenue from interest and fees, they can then work out the extra
profits banks earn from boosting the credit limits of customers of different
degrees of creditworthiness. They find that for the riskiest customers, income
from fees and interest does not increase quickly enough to compensate for
rising default rates among these newly unleashed borrowers.
The researchers calculate that the rational response to a reduction of a
percentage point in the rate at which banks themselves can raise funds is to
boost the credit limits of the 37% of cards issued to those with the highest
credit ratings by $2,203 each. For the 28% of cards with the lowest ratings, in
contrast, they should only loosen limits by a measly $127. In other words,
there is a big mismatch between those who are willing to borrow more, and those
to whom it makes sense for a bank to lend. That explains why the Fed s efforts
to boost the economy had a much smaller effect than expected. All the same, it
would be odd to criticise the banks for trying to steer clear of risky
borrowers. It was a failure to do that, after all, that got America into the
mess from which the Fed was trying to extract it.
consumer lending during the great recession , NBER working paper no. 21567.