Is Discount Window “Stigma” a Serious Problem?

Last year’s failure of Silicon Valley Bank prompted efforts to improve liquidity risk management within the banking system. Much of the attention has focused on improving access to the Federal Reserve’s discount window (DW), including attempts to reduce discount window hesitancy due to its perceived “stigma.” While the DW stigma is treated as an article of faith in some quarters, how strong is the supporting evidence? Do efforts to reduce the perceived stigma introduce risks of their own?

Policy debates around the use of the discount window rest on two essential premises. The first is that a DW stigma exists. The second is that such a stigma represents a serious problem. How do proposed legislative and regulatory fixes address these problems? More fundamentally, how well do these premises hold up?

Encouraging Use of the Discount Window

Most policy proposals consider ways to encourage banks to better prepare for accessing the discount window and to improve the Fed’s operational processes to make the DW more user-friendly. Some proposals also address discount window stigma more directly by more effectively concealing the users of the discount window and by making borrowing at the DW more attractive. In other words, regulators want more banks to use the lending facility, whether they need to or not.

Senator Mark Warner has introduced the Discount Window Enhancement Act, which seeks to promote better DW preparedness by banks and to make the discount window easier to use. The Warner bill would also require the Federal Reserve Board to “comprehensively review the weekly reporting of its balance sheet activities and consider changes to avoid market distortions that could inadvertently place individual financial institutions at a disadvantage.” The bill would also require regulators to give “positive consideration” to successful testing and pre-pledged collateral when making liquidity evaluations.

The Federal Reserve has also suggested requiring larger banks to pre-position some minimum amount of collateral at the discount window. Regulators would tie that amount to a bank’s level of uninsured deposits. This could make banks better prepared to address liquidity contingencies. It could also reduce DW stigma. Requiring all large banks to pre-position their collateral may make actions of specific banks less likely to stand out. TARP operated largely on this principle, which you could call the “I am Spartacus!” approach.

What’s the Downside?

These efforts to reduce DW stigma are not without risks. Making the discount window too convenient and attractive can discourage banks’ efforts to improve their liquidity profiles. Why maintain large liquidity buffers if you know the Federal Reserve will provide a lifeline at little cost? Changes to balance sheet reporting to make DW activities easier to hide may appeal to the Fed’s penchant for secrecy. However, these efforts may also conflict with the public’s right to know. After all, the Fed has a $7 trillion balance sheet and posted a $114.3 billion loss last year.

Establishing minimum pre-positioning levels can lead to some unintended side effects. As of December 2023, nearly 2,000 banks had pledged collateral to the discount window with an aggregate lendable value of $2.6 trillion. Pre-positioning requirements would presumably cause those totals to go up. But banks already have $7.66 trillion in pledged assets on their balance sheets. Pledged loans represent 31.70% of total loans and pledged securities represent 40.46% of total securities. Pledging more assets to the Fed may mean less collateral available to borrow from the Federal Home Loan Banks or from the private sector.

A Look at the Evidence

Discount window stigma is often treated as an unassailable fact. I wouldn’t claim to have done a comprehensive review of the available literature on DW stigma. But I don’t find the evidence I have seen especially convincing. First, let’s start with some data. As the chart below shows, there have been wild swings in the size of the Fed’s various credit programs. There was a huge spike during the Global Financial Crisis (GFC), with smaller but still meaningful spikes during COVID-19 and following the SVB failure. The discount window wasn’t the only or even the primary source of credit, but the data shows the Fed’s ability to inject massive amounts of liquidity into the system when needed. With $2.6 trillion in pledged collateral, banks also have a lot of dry powder.

Armantier et. al. provided evidence that some DW stigma exists. The authors compared rates paid at the discount window during the GFC with loans with similar terms under the Term Auction Facility (TAF), another Fed program with presumably less of a stigma. Banks preferred the TAF and would pay a premium to access that program rather than use the DW. However, preferring one Fed program over another is not the true test. The policy concern arises if banks risk insolvency to avoid the DW stigma. Favoring TAF to the DW is like preferring cake to ice cream. Preferring insolvency to DW stigma is more akin to preferring rat poison to ice cream.

Sam Schulhofer-Wohl of the Dallas Fed has described two types of discount window stigmas. “In-the-moment stigma” occurs when a bank faces a trade-off between immediate funding needs and concerns over adverse market reaction to borrowing from the discount window. While not exactly an enviable choice, it’s hard to see banks preferring insolvency to a perceived stigma. “Anticipatory stigma” occurs when banks don’t prioritize DW readiness because they find the actual use of the discount window so unattractive.

Anticipatory stigma appears more plausible. But regulators can also address this issue through the supervisory process. Examiners review a bank’s contingency funding plan with two primary objectives in mind. First, they determine whether management has identified a wide range of current and contingent sources of liquidity to meet unanticipated outflows. Second, they assess whether management has taken steps to ensure those sources are available when needed. If bank managers take the necessary actions, even if merely to please the regulators, that’s often good enough. Regulators don’t need to win hearts and minds.

A Tale of Three Banks

Some cite the Signature and SVB failures as examples of anticipatory stigma. Both banks were ill-prepared to access the discount window. Signature lacked pre-approved collateral. SVB had posted collateral but had not conducted test transactions and was not able to move securities collateral quickly from its custody bank or the FHLB to the discount window. But it’s unclear whether DW access would have helped much in these instances.

SVB was still profitable at the time of its failure due to the low rate (0.57%) it paid on its liabilities. But SVB earned only 2.73% on its assets, which consisted largely of long-term, deeply underwater MBS. Taking out loans at the discount window at 5.00% would ensure a huge negative spread. If you borrow at 5% and then invest them at 3%, don’t expect to make it up on volume. Accessing the discount window may have allowed the bank to survive a few more days (more likely hours) but not reduced costs to the FDIC.

Signature’s asset yields were a little higher, at 3.04%, but still well below the costs to borrow at the DW. Contrast Signature with First Republic, which took out $109 billion  in primary and secondary credit from the Fed to offset deposit outflows. First Republic survived for a couple of more months. But its eventual resolution also cost the FDIC $16.34 billion. That’s nearly six times the resolution cost of Signature ($2.82 billion). Not a great discount window success story.

Conclusions

None of this is to suggest that the Fed shouldn’t try to bring discount window operations into the 21^st Century. Supervisors should also continue to emphasize that banks need to consider a wide range of liquidity channels and have taken the appropriate steps to ensure the channels will be available if needed. However, policy proposals also need to take the downsides of encouraging discount window use into account and to consider that the DW stigma may be less of a problem than many assume.