Speaking to NYU’s “Money Marketeers” last week, Randy Quarles, the Fed’s Vice Chair for Supervision, shared his views on Fed policy, and particularly on steps he thinks the Fed should take to reduce the size of its balance sheet.
Perhaps better than anyone else at the Fed, Mr. Quarles understands the role that liquidity requirements play in propping‐up banks’ demand for excess reserves, and how those requirements foiled the Fed’s attempt to get the quantity of such reserves substantially below its crisis‐era peak.
Rather than accept that defeat, Mr. Quarles wants the Fed to try again, after first doing something to reduce banks’ demand for reserves. Reducing the Fed’s balance sheet, he says, is worth the effort because it will ultimately make its policies more credible:
Although I fully support the FOMC’s current plan to purchase Treasury bills and increase the size of the balance sheet in the very short term, over the longer‐term, I believe that the viability of balance sheet policies is enhanced if we can show that we can meaningfully shrink the size of the balance sheet relative to gross domestic product following a recession‐induced balance sheet expansion. In effect, I believe that balance sheet policies are more credible if we can show that there is not a persistent ratcheting‐up effect in the size of the Fed’s asset holdings.
Abundant, but Finite, Reserves
The danger to which Mr. Quarles refers, if only vaguely, is one that BPI’s Bill Nelson has repeatedly emphasized, namely, that unless something is done to reduce banks’ appetite for excess reserves, they might gobble‐up as many reserves as the Fed tosses their way, and still end up crying for more. “Such a dynamic,” Bill told a packed house at Brookings last December, “led the Norges Bank (the central bank of Norway) in 2010 to switch from a system with abundant reserves to a system with more scarce reserves.” As Norges Bank officials explained at the time,
When Norges Bank keeps reserves relatively high for a period, it appears that banks gradually adjust to this level…With ever increasing reserves in the banking system, there is a risk that Norges Bank assumes functions that should be left to the market. It is not Norges Bank’s role to provide funding for banks…If a bank has a deficit of reserves towards the end of the day, banks must be able to deal with this by trading in the interbank market.
Where Mr. Quarles differs from his Norges Bank counterparts is in holding that the Fed can shrink its balance sheet without abandoning its abundant reserves (or “floor”) operating framework. For him, as for most Fed officials, the Fed’s announcement early last year that it planned to stick to a floor system might as well have been chiseled on a stone tablet. “As the FOMC announced in January of 2019,” Quarles says, “the Committee intends to implement monetary policy in an ample‐reserves regime.” In light of events since last January, this digging‐in of official heels seems imprudent, to say the least. Yet it’s an all too common Fed practice, to which Fed officials are all the more inclined to resort when they have good reason to doubt the merits of some past decision.
Though Mr. Quarles rules out any alternatives to a floor system, he at least wants a floor system that keeps banks from becoming excess reserve junkies. In explaining how to limit banks’ appetite for excess reserves, he harkens back to something he said at a Hoover Institution conference back in May 2018. Asked by the very same Bill Nelson quoted earlier whether “despite how the LCR [Basel’s Liquidity Coverage Ratio requirement] is written, banks are told that they have to meet a material part of their high‐quality liquid assets requirements with reserves,” Mr. Quarles answered that “I do know that that message has been communicated at least in some supervisory circumstances in the past. I would say that that’s in the process of being rethought.”
Evidently, despite last year’s repo‐market turmoil, the Fed’s rethink has still not gone far. Hence Mr. Quarles’ New York remarks urging it forward.
I think it is worth considering whether financial system efficiency may be improved if reserves and Treasury securities’ liquidity characteristics were regarded as more similar than they are today—that is to say, that reserves and Treasury securities were more easily substitutable in the context of liquidity buffers. … I want to explore options that would maintain at least the level of resilience today while also facilitating the use of HQLA beyond reserves to meet the immediate liquidity needs projected in banks’ stress scenarios.
Assume a Can‐Opener Reliable Lender of Last Resort
The option Mr. Quarles prefers is that of having banks and their supervisors “assume” that the Fed stands ready to “provide liquidity to bridge the monetization characteristics of HQLA [High Quality Liquid Asset] securities versus reserves”:
If firms could assume that this traditional form of liquidity provision from the Fed was available in their stress‐planning scenarios, the liquidity characteristics of Treasury securities could be the same as reserves, and both assets would be available to meet same‐day needs.
Which brings us to the main shortcoming of Mr. Quarles’ remarks. For he proposes no actual changes to the Fed’s long‐standing last‐resort lending arrangements. Instead, he appears to consider those arrangements, and the Fed’s discount window in particular, adequate. The problem, he suggests, isn’t institutional: it’s psychological. The change that’s needed is a change in the beliefs of Fed supervisors and the bankers they supervise. Instead of treating Treasurys as less‐than‐perfect substitutes for reserves, they should assume henceforth that they can and will “rely on the Fed’s discount window” to make those assets practically identical whenever markets become stressed.
But this solution begs the question: assuming that bankers and their regulators can be coaxed into taking such a leap of faith, would their doing so be prudent? If experience counts for anything, the answer to that question is a resounding “no.”
The Stigma Problem
The rub is the notorious “stigma problem”—bankers’ reluctance to take advantage of the Fed’s discount window loans for fear that doing so will cause people to suspect that they’re in hot water. Though the problem has been around much longer, it became notorious during the last financial crisis. During that crisis, Ben Bernanke explains in his memoir, “banks were reluctant to rely on discount window credit to address their funding needs” because they worried that the word would get out, and that that would “lead market participants to infer weakness.” If that happened a bank’s normal funding sources could dry. Worse than that, it could suffer a run.
Proving that the stigma problem is real can be difficult, for there are all sorts of reasons why a bank might resist borrowing from the Fed. But in December 2007 the Fed set up an emergency facility—the Term Auction Facility—specifically designed to be a stigma‐free alternative to the discount window. So, it’s possible to gauge the severity of the discount window stigma by comparing its popularity with that of the newer facility, and especially by noting the extent to which banks turned to the TAF even when that meant paying more interest than the Fed’s discount windows would have charged.
Several years ago, Olivier Armentier and his coauthors did just that. And they found that the stigma problem was indeed severe. First of all (as the figure below, reproduced from an article by Atlanta Fed economist Larry Wall, clearly shows), while the TAF became an instant hit, banks remained reluctant to borrow from the discount window throughout the crisis. Just as importantly, “more than half of the TAF participants submitted bids above the DW [discount window] rate.”
On the basis of these findings, Armentier and his coauthors conclude that
Although the DW may still have a role to play as an emergency lending facility when a bank cannot find financing in the market for occasional and idiosyncratic reasons, one may question the ability of the DW as a channel to supply liquidity to a broad set of banks after a systemic funding shock.
Nor is there any reason to suppose that the stigma problem has gone away since the last crisis. On the contrary: thanks to a provision in the 2010 Dodd‐Frank Act compelling the Fed to publish bank‐specific data on its discount‐window operations after a two‐year lag, it seems to have gotten worse. Discount window borrowings have been considerably lower during the last decade than they were in the years leading to the financial crisis (see here), with recent figures the lowest on record. Most significantly of all, as David Benoit reported last November in the Wall Street Journal, discount window borrowings didn’t increase much when repo rates spiked last September. Yet that spike was an example of just “the kind of stress the discount window is designed to ease.”
Mr. Quarles instead thinks the stigma problem exaggerated. “While there has long been discussion about how the discount window is ‘broken’ because of stigma about using it,” he says,
we know it is still an important part of firms’ contingency planning and preparations. Banks currently pledge over $1.6 trillion in collateral to the discount window, which means that banks have gone to the trouble of working with their local Reserve Bank to make sure they have access to the window, if needed, and they have set aside a portion of their balance sheets as collateral to do so.
This is far from compelling. Experience shows rather that, while many (though hardly all) banks pledge collateral with their reserve banks, they may do so only in case they must borrow from the discount window for those “occasional and idiosyncratic reasons” to which Armentier et al. refer. They may have no intention of using the discount window otherwise. Since roughly half of the collateral banks pledge for primary credit borrowings consists of low opportunity cost commercial and consumer loans, such discount‐window insurance comes cheap.
A Stigma‐Free Alternative
If the discount window’s stigma problem can’t simply be waved aside, the Fed needs to take seriously Armentier et al.‘s conclusion that it
complement the DW by designing new “stigma proof” facilities specifically aimed at supplying liquidity to the entire banking sector. This is precisely what the Fed did when it created the TAF. Similarly, the BoE recently adopted a twofold approach: a DW for meeting idiosyncratic liquidity shocks, and a contingency liquidity facility activated in response to exceptional market‐wide stress.
While it’s tempting to suppose that a revived TAF might solve the problem, things aren’t so simple. Instead of standing ready to lend to any bank at any time, the TAF supplied funds only during auctions held once every two weeks. It also took three days to actually credit the funds to winning banks’ accounts. For these reasons alone, it might not be of any use to a bank facing a genuine stress scenario. The TAFs auction‐determined rates based on fixed credit allotments would also render it useless as a means for policing the upper limit of the Fed’s overnight target range.
What the Fed could really use is a full (that is, unlimited) allotment, fixed rate standing facility that’s also stigma‐free, for unless it’s so banks might still be willing to borrow on the private market for more than the Fed’s upper rate limit, thereby causing the Fed to miss its target, much as it did last September.
Is it possible to design such a facility? In a note written several years ago, the Atlanta Fed’s Larry Wall expressed his doubts.
The original TAF was not designed to set an upper bound on overnight rates. If the Fed wanted to establish a new TAF that would be effective for monetary control, several changes would have to be made in the new facility. First, the new TAF would need to provide funds at a fixed rate if it were to set an effective upper bound on short‐term rates. Additionally, the new TAF would have to provide for full allotment, that is, every bank would have to get the full amount of its request—otherwise, it would need to buy funds at the elevated market rate. Finally, if the new TAF were going to set a limit on overnight rates, it would need to be available every day and would have to settle the same day as the auction. In other words, the new TAF would have to operate just like primary credit if it is to set an effective ceiling on short‐term rates. But if a new TAF were set up with the same terms as primary credit, it would seem to be just as vulnerable to stigma as primary credit.
The SRF Alternative
If Wall’s “new TAF” won’t answer, could the Standing Repo Facility (SRF) proposed by Jane Ihrig and David Andolfatto, which resembles that idea in many respects, do so? Ihrig and Andolfatto believe it could. In particular, they say it “would not suffer from stigma problems that make the discount window an ineffective tool in these circumstances.”
Why not? Although the proposed SRF wouldn’t possess the same stigma‐reducing features as the TAF, it has other features that might serve just as well. For starters, instead of lending at a “penalty” rate, the SRF would supply funds at a rate only slightly above the Fed’s target‐range upper limit. Discount window loans, in contrast, are made at rates substantially above that limit—at present the surcharges are 50bps and 100 bps for primary and secondary credit, respectively. The lack of a penalty rate would prevent anyone from inferring that firms taking advantage of the SRF are doing so because they’re desperate. The SRF would, in other words, avoid the “adverse selection” problem that gives people a reason to doubt the soundness of any firm that borrows at above‐market rates. The SRF’s modest rate premium would also make it more tempting for banks to plan on using it in drafting their living wills.
But there’s perhaps a more important reason why the SRF would not stigmatize banks that take advantage of it. Because it exists in part to enforce the Fed’s rate target upper limit, while charging a rate slightly above that limit, it would become active only at times when reserve‐market stress causes private‐market overnight rates to reach the Fed’s target upper limit: at other times banks needing cash in a hurry, including large banks that have filed resolution plans, would be better‐off swapping Treasurys for cash in the private repo market. But once private market rates do reach the SRF rate, any bank might find it economical to borrow from the SRF. The presence of a substantial number of borrowers, borrowing at rates that actually carry no real penalty at all, would further help to avoid a stigma problem, if not to eliminate it altogether.
Nor are these the only reasons why an SRF might be stigma‐free. According to Bill Nelson (him again!), the fact that the SRF would probably accept only Treasurys and other first‐tier HQLA’s as collateral would also help to render it stigma‐free, since any “bank that has Treasury securities available to secure a loan from the Fed is unlikely to be a bank experiencing liquidity pressures.” Also, although SRF “transactions would be subject to the same two‐year disclosure requirement as discount window loans,” they “would be provided under Section 14 of the Federal Reserve Act, which authorizes monetary policy operations, rather than section 10B, which authorizes discount window advances” and to that extent would resemble the temporary repo operations the Fed once routinely engaged in with primary dealers, which bore no stigma on account of them. (Nor, one can now add, do any of the banks that have participated in the Fed’s recent, ad‐hoc repo operations appear to have feared being stigmatized by doing so.)
If It Ain’t Broken…
Why doesn’t Mr. Quarles consider the SRF alternative? Actually, he does: instead of having banks rely on the discount window, he says,
Another approach could be to set up a new program or facility: For example, there has been much discussion among market participants, as well as policymakers, about the potential benefits of setting up a standing repurchase agreement, or repo, facility for banks and how such a facility could improve the substitutability of reserves and Treasury securities for these firms.
However, Quarles adds that “While this option is still of interest, there may be benefits to working first with the tools we already have at our immediate disposal.”
Considering all the new tools the Fed has toyed with since the onset of the last crisis, one might find Mr. Quarles’ stance refreshing. I fear, however, that he has chosen the wrong occasion on which to put a kind word in for old‐fashioned bureaucratic inertia. For in this case any potential gains from sticking to the Fed’s old toolkit are far outweighed by the potential costs of having to rely upon a tool which, if it isn’t quite “broken,” is too close to it for comfort.