Ever since the failure of Silicon Valley Bank in early 2023, research has been probing whether US banks adequately protected themselves against the risk of rising interest rates. (For more, read “Are US banks hiding their losses?”)
The most vulnerable banks were the least protected, according to a paper from Chicago Booth’s João Granja and a team of researchers, whose work adds to a growing chorus of findings that banks failed to insure themselves against a well-telegraphed set of rate hikes. Worse, many such banks used highly opportunistic risk-management strategies, assuming that any downside risk they made with their portfolios would be borne by uninsured debtors and the Federal Deposit Insurance Corporation.
The researchers analyzed federal call-report data and US Securities and Exchange Commission filings from 2021 and 2022. Only 6 percent of the aggregate assets in the US banking system were hedged by interest rate swaps, they find. The most vulnerable banks actually spent 2022 reducing their already limited hedges, even as the Federal Reserve began aggressively hiking rates to rein in the worst inflation in decades.
In such an environment, why would any bank decrease its hedges? Because it can use an accounting gimmick to decrease the sensitivity of its book asset values to interest rate risk. When banks classify their securities as held-to-maturity, they do not need to mark them to market, even though they commit to holding those securities until they mature.
Just before monetary tightening began in 2022, most assets in the US banking system were exposed to (rather than hedged against) interest rate risk.
Altogether, the paper finds banks reclassified $1 trillion of their securities to held-to-maturity during 2021 and 2022, keeping their declining values hidden. (In contrast, securities classified as available for sale are marked to market, and their values may fluctuate dramatically in a rising-rate environment.)
The researchers also find that more vulnerable banks—those with lower capital ratios, a higher share of uninsured depositors, and a larger proportion of assets exposed to interest rate risk—were more likely to reclassify securities as held-to-maturity, “focusing on short-term gains but risking further losses if rates rose,” the researchers write.
Equity holders had little incentive to hedge the bank’s assets against interest rate risk, opting instead for what the banking industry calls “gambling for resurrection.” That is, the accounting gimmicks allowed the banks to avoid scrutiny while they bided their time, waiting for their gambles to pay off. In the meantime, they raised insufficient amounts of capital, betting that no run on the bank would happen—and that if it did, the government would bail them out.
The question of how banks account for their portfolios isn’t new: It was one of the biggest issues in the 2008–09 financial crisis. This time around, auditors should have recognized assets that were inappropriately marked as held-to-maturity, but even banks audited by the Big Four accounting firms including KPMG and PwC were just as likely to allow the aggressive classification, the researchers find.
On the other hand, US banks that were regulated federally rather than by state banking departments were less likely to reclassify securities, all else being equal. And “more robust enforcement” of existing held-to-maturity accounting rules by regulators, as well as auditors, might help, the researchers say.
However, it’s doubtful that auditors and supervisors can ever be certain that banks truly have the ability to hold to maturity securities that are marked as such, the researchers acknowledge. Even robust enforcement will leave some residual risk.
João Granja, Erica Xuewei Jiang, Gregor Matvos, Tomasz Piskorski, and Amit Seru, “Book Value Risk Management of Banks: Limited Hedging, HTM Accounting, and Rising Interest Rates,” Working paper, March 2024.
Ten ways investors are, or should be, using large language models
What AI Sees in the Market (That You Might Not)
Combining two transit policies generates far better outcomes than either policy on its own.
Charge Drivers to Improve Public Transit
Analyzing both numbers and narrative context in company reports led to better price forecasts.
Large Language Models Can Improve Stock Market ForecastsYour Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.