By Jeff Huther and Alison Touhey
ABA Data Bank
A contributing factor to the bank failures in spring 2023 was that the failing banks had inadequately prepared for using the Fed’s discount window. This led some commentators to recommend that banks be required to pledge, or pre-position, specific amounts of collateral at the Fed. Discount window data show that most domestic banks have access and periodically test their access but rarely use the discount window for business-as-usual purposes or out of need. There are likely many reasons for this, including preference for other liquidity providers, operational inefficiencies of discount window usage and, especially, the stigma of using the window. Very few of the domestic banks that regularly use the discount window have a presence in public stock and bond markets, a strong indication that publicly traded banks see a risk that discount window usage would adversely affect investor perceptions.
ABA analysis shows that banks generally pledge collateral to the discount window that cannot be used for funding from repo markets or from the Federal Home Loan Banks — banks have optimized their short-term funding sources based on considerations of cost and efficiency. Therefore, requirements to pledge specific amounts at the discount window will alter this structure, thus undermining existing funding arrangements and underlying markets. This would increase financial and operational costs for banks without creating offsetting gains elsewhere, making the system as a whole more fragile.
Most banks have ready access to the discount window
Today, most banks are operationally ready to tap the discount window. The Fed reported that 3,641 banks had access to the discount window in 2023, with pledged collateral of $1.8 trillion. Separately, the Fed publishes discount window transactions with a two-year lag. As Figure 1 illustrates, domestic banks that have used the discount window pledged around 5 percent of their total assets before the pandemic and around 12 percent of their total assets since then.
While over 2,500 domestic banks borrowed from the discount window at least once between the beginning of 2010 and the end of the second quarter of 2022 (Figure 2), only 102 borrowed an average of more than five times a year. Less frequent borrowers were mostly testing their ability to readily access the discount window — they borrowed a small amount overnight to ensure funding availability if needed.
Most regular users of the discount window are not publicly traded
On the other hand, frequent users of the discount window are unlikely to be just testing borrowing capacity. Of the 102 banks averaging more than 5 transactions a year, 73 had not raised funds from the capital markets (for example, by issuing stock or bonds; see Figure 3). It is very likely that, because most of these banks do not use traded securities for funding, they are less concerned with the stigma that is often associated with borrowing from the discount window. Other researchers have shown that discount window borrowers generally have lower account balances (reserves) at the Fed, also suggesting that these few regular borrowers are using the discount window out of need. That is, privately held banks can operate with lower levels of immediately usable cash reserves because they are comfortable turning to the window to meet unexpected outflows. Banks that have borrowed in equity or bond markets hold higher cash balances to avoid using the discount window.
In normal times, banks have deep alternatives to the discount window
In addition to discount window access, most commercial banks in the U.S. can borrow on a collateralized basis from the FHLBs and many larger banks participate in repo markets. Only some collateralized funding sources, however, are interchangeable. Treasury securities and mortgage-backed securities issued by government sponsored entities can be readily used as collateral in repo markets. These securities plus mortgage-related loans can be used as collateral for loans from the FHLBs. And almost all bank assets can be used as collateral for discount window borrowing, offering the broadest ability to tap liquidity and usually reserved for times of stress (Figure 4).
Banks prioritize collateral usage for day-to-day borrowing from private markets
The increase in collateral pledged to the discount window shown in Figure 1 has coincided with a decline in the share of pledged collateral that could also be used in the repo market or for borrowing from the FHLBs. The orange and blue bars in Figure 5 show the share of domestic banks’ pledged collateral that could have been used for borrowing in the repo market or from the FHLBs but instead was held at the discount window. The black hash marks in Figure 5 represent the share of domestic bank assets that could have been used as collateral in the repo market or for advances from the FHLBs.
The difference between the bars and the hash marks can be interpreted as banks’ preference for prioritizing lending from the private sector over the discount window. Using 2022 as an example, 28 percent of the collateral pledged by domestic banks that borrowed from the discount window was also eligible for the repo markets or with FHLBs. The hash mark for 2022 shows that 46 percent of domestic banks’ assets were repo market or FHLB pledgeable. This 18-point difference suggests that banks are selectively pledging collateral to the discount window that cannot be used in other markets. The difference between pledgeable shares and actual pledges widened markedly in 2020, presumably a consequence of the fiscal and monetary responses to the pandemic (which led to large increases in Treasury securities). Prior to the pandemic, the shares of collateral pledged to the window were similar to the composition of banks’ assets, so it is possible that the current bias of pledging non-marketable collateral to the discount window will not persist.
Conclusion
The under-representation of repo and FHLB collateral at the discount window is not surprising given the stigma of turning to the Fed, particularly with respect to the time needed to place collateral at the Fed and its slow return once the borrowing is repaid. A requirement to pledge specific amounts of collateral to the discount window could make banks more dependent on the Fed, crowding out well-functioning market options like the repo market and the FHLBs. More broadly, pre-positioning collateral at the discount window should be based on individual bank needs. Banks appropriately turn to private markets first and these markets meet a wider range of funding needs than the discount window (in particular, discount window loans are required by statute to have very short maturities while FHLBs provide advances that reduce bank duration risk). The current hierarchy enhances financial market stability — banks can rely on private markets to absorb day-to-day fluctuations while only turning to the Fed in its role as the lender of last resort.
Jeff Huther is a VP and senior economist in ABA’s Office of the Chief Economist. Alison Touhey is SVP for bank funding policy in ABA’s Office of Regulatory Policy. For additional research and analysis from the ABA’s Office of the Chief Economist, please see ABA Economic Research and Insights.
