- Key insight: The biggest hurdle to shrinking the Fed's balance sheet is the amount of reserves in the banking system necessary to ensure smooth market function. Fears of 2019-like reserve scarcity have shaped balance sheet management in the post-COVID era.
- Expert quote: "They are haunted by this. They feel that was an own-goal and there's been a lot of work in the Federal Reserve System to find out when those risks are in danger of repeating." — Anil Kashyap, professor of economics and finance at the University of Chicago's Booth School of Business
- Forward Look: New Fed Chair Kevin Warsh wants the central bank's balance sheet to be smaller. This week he will give his first public remarks.
New Federal Reserve Chair Kevin Warsh wants to turn back the clock on the central bank's balance sheet to 2010. To get there, he will have to go through 2019.
The Fed's first attempt to shrink its balance sheet resulted in a
Reserves, or funds held by commercial banks at the Fed, make up about 45% of the liabilities on Fed's $6.7 trillion balance sheet. The other two major components are currency in circulation and the Treasury's general account, neither of which the Fed has direct control over.
When conversations about shedding assets from the Fed's balance sheet turn to the question of which liabilities will offset those reductions, reserves are the clear answer. But, inevitably, those discussions come back to those weeks of financial stress in September 2019 and the institution's imperative to avoid a repeat performance.
"They are haunted by this," said Anil Kashyap, professor of economics and finance at the University of Chicago's Booth School of Business. "They feel that was an own-goal and there's been a lot of work in the Federal Reserve System to find out when those risks are in danger of repeating."
Warsh, who will give his first speech as chair after this week's Federal Open Market Committee meeting, has his own history with the Fed's balance sheet to contend with.
Warsh was a Fed governor and member of the FOMC when it undertook its second round of quantitative easing — a move he was highly skeptical of at the time. According to a transcript of the committee's meeting in November 2010, he expressed a litany of concerns, including that large scale asset purchases would have little impact on lending to the real economy, disproportionately benefit large financial institutions and increase the Fed's footprint in financial markets.
Warsh ultimately voted for the action grudgingly, noting that he did so only out of respect for then-Chair Ben Bernanke and with a hope that the Fed would swiftly reverse course if the policy did not deliver on the objective of stimulating economic growth.
"If I were in your chair, I would not be leading the committee in this direction, and frankly, if I were in the chair of most people around this room, I would dissent," Warsh said to Bernanke at the time, adding, "I think this is called the Bernanke Fed for a reason. I've got a lot of confidence that if the risks that I talk about materialize, you will not hesitate and you will change your view, you will change this experiment."
Warsh called for inserting caveats into the Fed's policy statement that would have made it easier to end the expansion, but he gained little traction with the rest of the committee. He left the board four months later.
Ample vs. abundant
One of the key lessons from the episode was that "ample" reserves — meaning slightly more than banks need to comfortably settle payments and meet liquidity needs — are not always enough to keep the financial system moving smoothly.
Shortly before the crisis, banks believed the amount of reserves to be well above their needs. A survey of senior financial officers conducted by the Fed
That perceived excess disappeared quickly on September 17, 2019 following a confluence of events related to the Treasury's general account — another liability on the Fed's balance sheet that competes for space with reserves. An influx of corporate tax payments and a settlement of a large Treasury auction increased the general account, thus shrinking the supply of reserves to about $1.4 trillion. While still well above the aggregate lowest comfortable level, reserves were not evenly distributed throughout the banking system, with a handful of large banks hoarding large quantities and smaller banks bidding up borrowing costs to get more or retain what they had.
The Federal Open Market Committee — the Fed's monetary policy arm — held an emergency conference call on October 4, 2019 to discuss how to respond to the stresses in overnight funding markets. Patricia Zobel, then vice president of the Federal Reserve Bank of New York, told the group that a "higher level of reserves may be needed to accommodate distributional frictions in reserve markets."
In the months that followed, the Fed steadily added about $400 billion assets, which corresponded with $200 billion in additional reserves held by banks. How this would have played out of the longer term is unclear, as the COVID-19 pandemic in March 2020 instigated the Fed's biggest balance sheet expansion to date — more than doubling in a matter of 18 months.
Still, the post-2019 view toward reserves in non-crisis moments remains prevalent. In a speech
"I start from the view that problems emerged when reserves fell below 8% of GDP," Waller said. "One might argue that banks are now larger relative to GDP, so they may desire a bit more reserves. Furthermore, there is also a genuine concern that it is not only the total amount of reserves that matters but also the distribution of reserves across the banking system. So, I would add a buffer to the 8% of GDP that I cited earlier and assume 9% is the threshold below which reserves would not be ample."
Based on this calculation, Waller argued that the minimum amount of reserves the system needed was $2.7 trillion — around $600 billion less than the amount of reserves at the time. He said the Fed could get to that optimal reserve level by continuing the balance sheet reduction schedule it was pursuing at the time. However, the FOMC voted to end that program during its December meeting and the amount of reserves in the system has climbed back to more than $3 trillion.
This leaves the banking system in a position it has been in since 2020: one in which reserves are not only ample but "abundant," or a level that is well in excess of what banks need.
Darrell Duffie, professor of finance and management at Stanford University's Graduate School of Business, said the Fed is making the logical choice by erring on the side of too many reserves.
"The Fed's being conservative, but if you think about the asymmetry on the costs and benefits, the cost of going too small is very high, the cost of going too big is not that high, if you're talking within the range of a few $100 billion," Duffie said. "So, I think the Fed's current policy is appropriate until they figure out a way to reduce the overall demand for reserves by banks."
Policy remedies
Duffie, who has studied the Fed's balance sheet and the 2019 liquidity crisis extensively, said the Fed has several options for curbing reserve demand among banks. He said a liquidity saving mechanism could be incorporated into Fedwire — the central bank payment rail used for bank-to-bank transfers — to reduce the volume of reserves needed to settle payments on a day-to-day basis. Similarly, the Fed could convince banks that they will not be penalized if they overdraft their accounts to meet payment obligations, so long as they are sufficiently collateralized.
Yet, banks have shown a reluctance to trust that they will not be cited for taking advantage of such leniencies. Fears of supervisory rebuke have also made banks unwilling to use the Standing Repo Facility, the liquidity provision system established by the Fed in response to the 2019 liquidity crisis.
"Until recently, banks have said they are stigmatized by the idea of using the Standing Repo Facility, because it would indicate that they aren't meeting the spirit, at least, of their regulatory requirement to be self-sufficient," Duffie said. "Because, if you go to the Fed for more reserves, that means you didn't have enough on your own, and [an executive] might worry … that would signal to the Fed that you weren't meeting this liquidity requirement."
Duffie noted that the Fed has made some changes to the facility that have resulted in a modest uptick in use by banks, but more work should be done.
Others say the key to reduced demand should come through regulatory reform, namely changes to the liquidity requirements and changes to certain capital charges — like the supplemental leverage ratio — to make it easier for banks to hold other highly liquid assets to manage their liquidity needs.
Andrew Levin, an economics professor at Dartmouth University and a former staffer at the Fed Board of Governors, said many of these reforms would require coordination across multiple agencies.
"The right way to shrink the Fed's balance sheet is in close coordination with the [Financial Stability Oversight Council] and Treasury's Office of Debt Management," Levin said. "Part of the issue in 2019 was that the Fed tried to do it in isolation."
Others say more drastic reforms are needed. Norbert Michel, vice president and director of the Cato Institute's Center for Monetary and Financial Alternatives, said the most effective way to reduce interest in reserves is to stop paying interest on them.
In the broad sweep of Fed history, interest on reserves is a relatively
"What you would have is more like what you had pre-2008 for almost a century, and … for the entire post-war period," Michel said. "You would have banks in the market with each other and financial firms in the market with each other, and with banks funding themselves in what looks a lot more like a market-based, free enterprise-type system, one that works — one that did work very well, by the way."
