NEW YORK, April 4 (Reuters Breakingviews) - Bank watchdogs don’t have a crystal ball when it comes to spotting bank runs. They nearly had something close, though. The Federal Deposit Insurance Corp, which insures Americans’ savings, launched a project in 2020 that would have used technology from outside companies to flag risks just like the ones that felled Silicon Valley Bank. The initiative fizzled because of skepticism from the regulator’s new leadership, and a culture wary of relying on private firms. Reviving it in some form could in future save the FDIC money, stature and time.
The FDIC is one of several agencies that watches over American banks, but it’s the one that picks up the tab when a lender fails. It manages the $128 billion fund that pays back depositors with savings below $250,000 if their bank collapses, and must also find new owners when a bank goes into receivership. Given the FDIC’s overarching mission of protecting the sanctity of its fund, the failures of SVB and Signature Bank last month did not cover the agency in glory: Fearing a systemic crisis, it had to take the unusual step of reimbursing all their depositors – even those with millions in the bank. Like other regulators, it was blindsided by the speed of the banks’ collapse.
The FDIC at least knew what it didn’t know. Under former Chair Jelena McWilliams, an appointee of then-President Donald Trump, it invited companies to create prototype data tools to help bank supervisors monitor credit risks, depositor behavior and the effect of interest-rate moves. Thirty-three firms applied for this “rapid phased prototyping,” and four were shortlisted in mid-2021, including Palantir Technologies (PLTR.N) and S&P Global Market Intelligence. Their presentations convinced even some skeptical FDIC officials, say people familiar with the process.
Then the plan hit a wall. McWilliams quit in 2022 after her Vice Chair Marty Gruenberg, a Democrat-affiliated appointee, challenged her over the rules surrounding bank mergers and ultimately won her role. Gruenberg, on the FDIC board since 2005, did not support the rapid phased prototyping data project, fretting that it amounted to outsourcing supervision, according to people familiar with the situation. As a result, no bid was declared a winner. For all but the biggest banks, the FDIC continues to rely on quarterly snapshots known as “call reports,” and the findings of its on-the-ground inspectors.
SVB would probably have failed regardless. Even had the new data panopticon been built, banks would have had to opt in, unless Congress expanded legal limits on what information the FDIC can demand. Even so, recent bank failures show why such a project makes sense, not just in foreseeing problems but in engineering a smoother cleanup. When finding a new buyer for a bank like SVB the FDIC must set up a “data room,” and sourcing current information and models can take several days. The challenge of luring a buyer contributed to a $20 billion loss to the FDIC’s depositor-rescue fund.
The death of the 2020 project – and the fact it didn’t start years sooner – reflect deeper challenges at the FDIC. While banks have roundly embraced technology, spending billions on blockchain-based ledgers and cutting-edge analytics, the watchdog is slow to embrace technological innovations, say sources with personal knowledge, with multiple overlapping systems, many built internally. Risk aversion is sensible for a regulator, but can come at the cost of less efficiency. A lack of technical sophistication at smaller banks compounds the problem: Highly trained examiners use precious time to collect and transfer data using old technology that could otherwise be used to analyze the safety of the banking system.
Change is also hard won at an agency where employees tend to stay put for decades, thanks partly to a generous compensation and benefit scheme that pays an average $230,000 per year, based on figures in its annual report. The FDIC’s independent watchdog warned in February that 64% of its advanced IT experts will be retirement-eligible by 2027. Meanwhile, 36% of staff in the division that winds up failed banks are already eligible, more than twice the government-wide average. At the executive level, the median length of employment is 26 years, based on the FDIC’s website and other publicly available sources. Insiders joke that the agency’s abbreviated name stands for “Found Dead In Chair.”
While some staff don’t leave, others trained in up-to-date methods won’t stay. The resignation rate among examiners-in-training doubled between 2020 and 2022, the independent inspector found. A cap on potential rewards, standard at government agencies, means the FDIC struggles to compete for tech-native talent with Wall Street firms and financial technology startups. Past independent reviews have publicly called out the agency for poor storage of passwords, failing to update tech guides, and inadequate adherence to tech protocols, some of which the FDIC has contested. Such vulnerabilities at a bank might elicit a slap from regulators.
The collapse of SVB and Signature Bank might be a wake-up call. The $22.5 billion combined hit to the rescue fund is the biggest since 2009, a year when 140 banks failed rather than 2023’s two. Refilling the pot by charging banks higher levies is already proving unpopular with small lenders, who have considerable political support and point out that they are not to blame for recent troubles. The FDIC is reviewing its own failings, but members of the Senate are already preparing for further investigations, according to people familiar with the situation.
Change may come whether Gruenberg champions it or not. The FDIC’s new vice chair, Travis Hill, worked at the agency under McWilliams – including the period in which the ill-fated data project was still live. Hill was reappointed by the Senate in December as a bipartisan pick. He also personally helped draft a 2018 rollback in bank regulation that Gruenberg has called a “bad idea,” setting the scene for potential disagreements. The long-serving chair has seen off rivals before, but in the FDIC’s weakened state, now might be a good time for a newcomer to propose a more modern approach.
Follow @johnsfoley on Twitter
