Board of Governors of the Federal Reserve System

When I was asked to participate on this panel in the middle of last year, the prevailing metaphor regarding Federal Reserve balance sheet policy was "as boring as watching paint dry." Well, times have changed, and I commend the conference organizers for their foresight. Today I would like to discuss some of the recent decisions that the Federal Open Market Committee (FOMC) has made regarding the balance sheet and lay out a rough framework for some further issues that are on the horizon.¹

In January, after much discussion, including in previous meetings, the FOMC announced its intent to continue operating in a framework of ample reserves.² In this regime, active management of the reserve supply is not needed. The Federal Reserve controls the level of the federal funds rate and other short-term interest rates primarily through the use of administered rates, including the rate paid on reserve balances and the offered rate on overnight reverse repurchase agreements. This regime is sometimes referred to as a floor system, because the administered rates place a floor under the rate at which banks and others will lend in the federal funds market. In adopting this framework, the Committee stated its intention to continue operating as it has for the past decade.

The announcement was an important step in our normalization process. And we are now set up to make further decisions on the eventual size and composition of our balance sheet. Before providing more context on those decisions, let me first provide a little more detail around our decision to remain in the current framework of ample reserves.

The most important factor in the decision was that the current system has worked very well. It has supported the achievement of our dual-mandate objectives of maximum employment and price stability. And it has shown itself to be flexible and well suited to maintaining interest rate control through various changes in money markets, bank regulation, and the Federal Reserve's balance sheet. Since the FOMC began lifting interest rates in December 2015, money market rates have generally moved closely with the federal funds rate, which in turn has followed changes in administered rates.

Now that the decision on the operating framework has been made, a natural next step is to contemplate the appropriate size of the Fed's balance sheet and reserves and the process for getting there. In line with the requirements of operating with ample reserves--and boosted by the growth in nonreserve liabilities--the Fed will maintain a larger balance sheet and reserve supply relative to the pre-crisis period, with the goal of remaining on the flat portion of the reserve demand curve. I would note that reserves have already declined appreciably from their peak, falling by $1.2 trillion to a current level of around $1.6 trillion. At the same time, we have seen a substantial increase in our nonreserve liabilities, such as currency in circulation and the Treasury General Account balance. In our statement on Policy Normalization Principles and Plans, we outlined an intention to hold no more securities than necessary to implement monetary policy efficiently and effectively.³ As the balance sheet continues to shrink, we are now in the process of determining that necessary size.

Ultimately, the size of the balance sheet will be determined by a number of factors, including demand for nonreserve liabilities, such as currency (which has been rising), and, importantly, the quantity of reserves necessary to remain reliably on the flat portion of the reserve demand curve. Survey results suggest that banks have greatly increased their demand for reserves in the post-crisis period. Responses to the September 2018 Senior Financial Officer Survey report that banks would be comfortable with a level of reserves in the system in the neighborhood of $800 billion, taking into consideration the level of interest rates at the time.⁴ In part, this increased demand reflects a response to regulatory changes introduced after the crisis. These changes include, importantly, the Liquidity Coverage Ratio (LCR), which has improved banks' liquidity resilience by requiring firms to hold sufficient high-quality liquid assets to cover potential outflows during times of stress. Reserves, along with Treasury securities, are favored under the LCR, and, consequently, firms currently meet a sizable fraction of their LCR requirements by holding reserves.

Notwithstanding survey results, the level of reserve demand remains quite uncertain. It is possible that, over time, the preferences of banks will shift, or that demand will prove more price elastic than banks are currently expecting. As I have discussed previously, bank holdings of reserves to meet LCR requirements could shift toward Treasury securities, as aggregate reserves decline, without much upward pressure on the federal funds rate.⁵ That said, even if uncertain, it is probably safe to say that reserve demand is much higher than before the crisis.

As we work to calibrate ample reserves, there are some tradeoffs that are worth noting. For example, we could operate with a level of reserve balances at the lower end of what might be considered ample. In that case, there would likely be occasions when unexpected declines in the supply of reserves or increases in the demand for reserves would require an open market operation to offset temporary upward pressures on the federal funds rate. Alternatively, we could operate with an average supply of reserves large enough to keep the federal funds rate determined along the flat portion of the reserve demand curve even with an unexpected shift in the supply of or demand for reserves. This approach would be operationally convenient but would also leave the size of the balance sheet and reserves larger than necessary most of the time. In my view, it might be appropriate for us to operate somewhere in between these two extremes, with a sizable quantity of reserves large enough to buffer against most shocks to reserve supply. On those few days when that buffer is likely to be exhausted, we could conduct open market operations to temporarily boost the supply of reserves.

With so much uncertainty over the level and slope of the reserve demand curve, a degree of caution is warranted. As outlined in the minutes of the January FOMC meeting, the Committee has discussed ending the reduction in the Fed's aggregate asset holdings sometime in the latter half of this year, with still-ample reserves in the system.⁶ At that point, one option discussed, without any decision being made at this point, is to hold the level of total assets roughly fixed for a time. Even as the total size of the balance sheet remains fixed, the composition of the liabilities would gradually change, in part as demand for currency grows in line with the economy. Over time, the gradual increase in nonreserve liabilities would displace reserves as the overall balance sheet remains fixed. This plan would substantially reduce the pace of the decline in reserves, allowing us to gradually approach our assessment of the appropriate amount of reserves for the efficient and effective implementation of monetary policy. Of course, in the longer run, once we reach our preferred level of reserves, the balance sheet would have to resume growth to match a continued increase in demand for nonreserve liabilities.

I would like to wrap up with a brief discussion of some of the other decision points we will encounter as we continue the process of normalizing our balance sheet. In particular, what does the Committee judge to be normal in regard to the type and duration of assets that we will hold? On composition, in line with our previously announced normalization principles, I favor a return to a balance sheet with all Treasury securities, allowing our mortgage-backed securities (MBS) holdings to run to zero. In those principles, we also state that while we do not expect sales of MBS as part of the normalization process, later we would be open to limited sales to reduce or eliminate residual holdings of MBS. In regard to duration, moving to shorten the duration of our holdings could increase the Fed's ability to affect long-term interest rates if the need arose. However, it might be preferable to have the composition of our Treasury holdings roughly match the maturity composition of outstanding Treasury securities, minimizing any market distortions that could arise from our holdings. Over the course of our upcoming meetings, I look forward to what promises to be an interesting discussion on these issues with my colleagues.

Finally, in assessing our balance sheet policy, it is important to point out that the Fed remains entirely focused on meeting its statutory dual-mandate objectives of maximum employment and price stability. The normalization of the balance sheet is not a competing goal. If ever it appears that our plans for the balance sheet are running counter to the achievement of our dual-mandate objectives, we would quickly reassess our approach to the balance sheet.

1. These remarks represent my own views, which do not necessarily represent those of the Federal Reserve Board or the Federal Open Market Committee. Return to text

2. For more information, see the Committee's Statement Regarding Monetary Policy Implementation and Balance Sheet Normalization, which is available on the Board's website at https://www.federalreserve.gov/newsevents/pressreleases/monetary20190130c.htm. Return to text

3. The Committee's Policy Normalization Principles and Plans were adopted on September 16, 2014, and can be found on the Board's website at https://www.federalreserve.gov/monetarypolicy/files/FOMC_policynormalization.pdf. The FOMC adopted addenda to the Policy Normalization Principles and Plans on March 18, 2015 (PDF), and on June 13, 2017 (PDF). Return to text

4. The survey is available on the Board's website at https://www.federalreserve.gov/data/sfos/files/senior-financial-officer-survey-201809.pdf. Return to text

5. See Randal K. Quarles (2018), "Liquidity Regulation and the Size of the Fed's Balance Sheet," speech delivered at "Currencies, Capital, and Central Bank Balances: A Policy Conference," a Hoover Institution Monetary Policy Conference, Stanford University, Stanford, Calif., May 4. Return to text

6. See Board of Governors of the Federal Reserve System (2019), "Minutes of the Federal Open Market Committee, January 29-30, 2019," press release, February 20. Return to text