-
New proposal would adjust definitions in the agency's premiums formula that banks had said was too burdensome. An FDIC board member said the plan may still be too complicated.
March 20 -
WASHINGTON — Months after the Federal Deposit Insurance Corp. revamped its large-bank assessment system, banks are urging the agency to revisit new reporting requirements stemming from the plan.
May 23
Banks are worried that the Federal Deposit Insurance Corp. is placing too much emphasis on the stressful years of the financial crisis in determining lenders' concentration of high-risk loans.
The FDIC is trying to ensure that the risk of holding large amounts of high-risk assets, including subprime loans, is appropriately reflected in large banks' deposit insurance assessments. To do so, the agency is using the years from 2007 to 2011 — smack dab in the middle of housing and financial crises — as a baseline for determining how consumers with comparable credit risk perform during periods of major stress.
Banks are not too thrilled about basing the time period for determining the probability of defaults on arguably the riskiest period for troubled loans in recent history.
In a comment letter sent at the end of May, the American Bankers Association and five trade groups listed the time period for determining whether a consumer loan is higher risk as one of the main issues they want changed in the FDIC's proposed rule. Banks would prefer that the final rule specify that the time period be updated every two years and not be set to an arbitrary time period.
"It is standard industry practice to
Last year, the FDIC broadened its pricing formula for large banks to make deposit insurance premiums more sensitive to high-risk loans. The FDIC required financial institutions with over $10 billion of assets to report the amount of "subprime" consumer and "leveraged" business loans, which are combined with other metrics to calculate a bank's pricing assessment.
"One of the main goals of the new pricing system was to better price banks by risk," said Brenda Bruno, a senior financial analyst at the FDIC. "We're interested in how consumers are behaving during a stress period."
Bruno, speaking during a webinar last month titled "The New Subprime Definition," said the FDIC wanted more accuracy and consistency in reporting, and to reduce the reporting burden for financial institutions.
The FDIC issued proposed rule-making on
Subprime consumer loans are defined as loans with a 20% probability of default within two years, or securitizations in which over half of the pool is backed by such loans. Non-traditional mortgages are excluded from the definition.
Under the FDIC proposal, banks would determine the probability of default using the average of two 24-month default rates — from July 2007 to June 2009, and from July 2009 to June 2011. They would report the outstanding balance of consumer loans in their retail portfolios stratified by 10 product types. Banks typically need a minimum sample size of 1,200 loans to meet the FDIC's proposed requirements.
Tyler Davis, a senior financial analyst at the FDIC, said in an interview that the agency is looking at periods of stress for a reason.
"What we are interested in is whether banks are originating a lot of credit to consumers that look risky relative to other consumers," Davis says. "It's difficult to differentiate performance among institutions when everyone is performing well. So even if the credits are performing well today, we know these consumers are weaker than others and in a stress scenario they're more likely to have a higher default rate."
"This is the latest experience of stress that we've had, it's fairly recent and there's no plan to update that," he added.
The FDIC has generally received high praise from banks for using a probability of default definition, rather than relying on three-digit credit scores, to assess the risk from subprime loans across comparable institutions, says Sarah Davies, a senior vice president of analytics at VantageScore Solutions.
It would have been more difficult to get an apples-to-apples comparison among banks based on credit scores because there are so many different credit scoring systems and many banks used their own proprietary scores, she says.
"One of the challenges with the way a lot of risk regulation is written is this notion that if I just reference this particular three-digit credit score, I'm referencing a particular risk," says Davies. "But default rates change over time, as we've seen with the way the economy and the recession have played out."
For example, a credit score between 691 and 710 had a 10% probability of default from June 2008 to June 2010, up from a 6% propensity to default in the period from June 2003 to 2005, according to research from VantageScore.
Banks have expressed some concerns because the FDIC's proposals did not consider how large banks should treat consumers who have no credit histories or ratings.
Davies says banks may be gratified to know that fewer than 1% of consumers who did not have a credit score defaulted on loans in qualifying accounts during the height of the mortgage crisis from 2007 to 2009.
"When we looked at the default rates of unscorable consumers, they were much lower than 20%, so they aren't high risk," she says. "For the large bank population, it is a comfort that there is not some great unknown risk sitting out there."
Still, Davies expects additional commentary from the FDIC going forward on how banks treat consumers who do not have credit scores.
