The Financial Stability Oversight Council recently released its annual report. Not surprisingly, this year’s report reviewed the banking failures in the early part of the year. In discussing the “vulnerabilities in the banking sector” that resulted in the bank failures, the Report cited poor risk management practices and heavy reliance on uninsured deposits as among the chief causes. Among lessons learned, the Report noted the need to improve data and monitoring of uninsured deposits and depositor composition. Pointing to the current regulatory proposals relating to requirements for additional long-term debt, enhanced resolution planning measures and stronger governance and oversight and supervision practices, the Report notes these as deterrents to future crises and concludes that, overall, the US banking system remains resilient. The Report devotes substantial attention to a discussion of the refinancing risks of CRE loans, noting these are elevated due to, among other things, the sizeable amount of upcoming maturities in 2024, as well as weakening in CRE credit quality. The Report observes with concern the increased participation of nonbank financial institutions in the provision of financial services. For example, the Report cites potential risks related to role of nonbank mortgage servicers (nonbanks service over half of all mortgages, and the share of nonbank originations and servicing now stand at record highs). The Report also points to the rapid increase in private credit as a trend to monitor. For the first time, the Report devotes substantial attention to the risks associated with the use of AI in financial services—these risks can include safety-and-soundness risks, including cybersecurity and model risks, as well as consumer compliance risks. Securities and Exchange Commission Chair Gensler focused his remarks relating to the FSOC report on AI and the risks associated with AI in financial services—linking these back to the SEC’s proposals relating to the use of predictive data.
Chair Gensler refers to AI as “the most transformative technology of our time.” However, he cautions that the “possibility of one or even a small number of AI platforms or data aggregators dominating raises issues with regard to financial stability.” In addition, AI may compound risk as a result of reliance on models, leading to “herding” behavior and increasing the interconnectedness of the financial system. Chair Gensler notes the need to update model risk management tools. As well, Chair Gensler points to the need to consider the need to rethink system-wide or macro-prudential policy interventions. See the full text of his remarks.