
A New York Fed study has found evidence that stock investors attached a stigma to banks that turned to the Federal Reserve’s Discount Window during the March 2023 banking turmoil, a result that could help explain why some institutions are reluctant to use the central bank’s traditional emergency lending channel.
The researchers said investors became more sensitive to balance-sheet weaknesses after a bank’s first Discount Window borrowing. They did not find the same statistically significant shift after a bank first borrowed from the Bank Term Funding Program, or BTFP, the temporary facility the Fed introduced as the crisis unfolded.
That distinction matters because the Federal Reserve’s lender-of-last-resort tools work best when banks are willing to use them early, not only after funding stress has become acute. If managers believe a trip to the Discount Window will be read by investors as a warning sign, they may delay borrowing even when the tool is designed to stabilize a bank and the broader financial system.
Researchers tracked how investors reacted after first loans
In a Liberty Street Economics post published Tuesday, Natalia Fischl-Lanzoni, Martin Hiti and Asani Sarkar examined how stock investors responded when publicly traded banks first borrowed from Fed facilities during the first month of the 2023 bank run. The study was based on a recent paper and focused on whether emergency borrowing changed the way investors priced bank risk.
To measure that reaction, the authors built stock portfolios that isolated four vulnerabilities that were central to the 2023 turmoil: reliance on uninsured deposits, unrealized securities losses, the size of cash holdings and capital strength. They then estimated how closely an individual bank’s daily share return moved with those risk portfolios. A higher beta meant investors had become more attentive to that bank’s exposure to a specific weakness.
Before the run, those risk sensitivities were close to zero, suggesting stock investors were not especially focused on those balance-sheet issues. Around March 7, just before the crisis intensified, the betas turned positive and rose sharply. The authors then looked at what happened when a bank took its first loan from the Fed, reasoning that the first draw was the moment most likely to deliver fresh information to the market.
Between March 8 and March 31, 2023, 34 banks in the sample took a first BTFP loan and 33 took a first Discount Window loan. Another 207 banks used neither facility, giving the researchers a control group. Ten institutions borrowed from both programs, a reminder that banks facing stress often drew on more than one source of liquidity as conditions deteriorated.
Looking at the seven trading days before and after each bank’s first borrowing, the researchers found that banks using only the Discount Window saw a significant rise in investor sensitivity to uninsured-deposit risk and capital weakness, especially four to seven days after the first loan. By contrast, banks that used only the BTFP did not experience a statistically significant change in investor risk perceptions.
Facility design and history may explain the gap
The contrast is notable because the two Fed programs served related but not identical purposes. According to the Federal Reserve’s official BTFP page, the program offered loans of up to one year against high-quality securities valued at par, helping banks avoid forced sales during stress. The BTFP was introduced on March 12, 2023 and stopped making new loans on March 11, 2024.
The Discount Window is older, broader and more established in crisis management, but it also carries a longer history of market wariness. In the New York Fed post, the authors argued that history may have shaped investor interpretation in 2023. A bank willing to accept the less favorable terms of the traditional window could be seen as having exhausted cheaper or cleaner alternatives, or as needing to pledge assets such as loans that the BTFP would not take.
The terms mattered too. The blog post notes that a bank bringing underwater securities to the Discount Window could borrow against their market value, while the BTFP allowed eligible securities to be valued at face value. Borrowing costs were also typically lower at the BTFP. Those differences may have made the emergency program look less like a distress signal and more like a policy response tailored to the specific strains that erupted after the collapse of Silicon Valley Bank and the wider regional banking panic.
Still, the authors were careful not to overstate the comparison. They said the head-to-head difference between the two facilities was not statistically decisive. That means the evidence supports caution rather than a blanket conclusion that one form of Fed borrowing is always stigmatized and the other is never viewed negatively.
Why the findings matter for crisis policy
For policymakers, the study speaks to a long-running problem. The Fed can make liquidity available, but the tool is less effective if banks fear the market consequences of tapping it. In a fast-moving run, that hesitation can make funding stress harder to contain, increasing the chance that banks cut lending, sell assets into weak markets or seek more expensive funding elsewhere.
The findings also fit with a broader literature on Discount Window stigma. Earlier New York Fed research has argued that the problem persisted even after post-financial-crisis efforts to normalize borrowing and improve operational readiness. The new study adds a market-based angle by focusing on how equity investors responded in real time when banks first used Fed lending facilities during a live crisis episode.
For investors, the paper suggests that Fed borrowing is not interpreted mechanically. Markets appeared to separate a long-standing backstop that has often been associated with distress from a newly created facility aimed directly at the unrealized-loss problem exposed in 2023. In other words, the signal depended not just on the fact of borrowing, but on which facility a bank used and what that choice implied about its condition.
There are limits to the evidence. The study covered relatively small groups of banks, looked at a narrow event window and examined one unusual period in March 2023. The authors also noted that it is difficult to fully separate the market’s reaction to borrowing itself from the underlying reasons a bank needed to borrow. Even so, the results offer fresh evidence that the design and reputation of emergency lending tools can shape how markets receive them.
That leaves a practical policy question. If the Federal Reserve wants banks to use emergency liquidity facilities quickly when a panic begins, it may need not only generous terms and broad collateral rules, but also structures that reduce the perception that borrowing marks an institution as unusually weak.
Latest News
View all news- Walmart Raises Full-Year Outlook After Revenue Climbs 5.2% in Third Quarter
- Treasury Sanctions Alleged Crypto Scam Operator, Firms and Myanmar Commanders
- Fed Vice Chair Jefferson Warns AI Investment Could Raise Inflation Before Productivity Benefits Emerge
- U.S. Natural Gas Storage Rises by 20 Billion Cubic Feet, EIA Reports
- GAO Questions Cost Controls on $45 Billion U.S. Immigration Detention Funding