The Fed has offered some idea of how it will start reducing the size of its balance sheet, but there’s no way to anticipate how or when it will end.
What will be the impact on the real economy? If it results in more tightening of financial conditions than the Fed desires, how will it be offset? How big will the balance sheet still be when the Fed deems it fully normalised? At what balance-sheet size would the competition for reserves from financial institutions start overriding the effect of the interest paid by them? If the economy once again tips into a downturn, will the Fed consider once again growing its portfolio holdings?
Although these questions don’t yet have answers, the many attendant complexities are explored in a deeply researched, comprehensive new note from Nomura’s Lew Alexander. You can read the whole thing by clicking here, but we highlight and discuss a few excerpts below.
First, on the expected pace of the decline in the Fed’s holdings:
The FOMC has laid out how balance sheet roll-off will go forward once it decides to proceed. Figure 1 shows the projected monthly pace of roll-off. The Federal Reserve currently holds $1.8trn in agency MB securities, which tend to amortize at a rate of about 1% per month, with some seasonal variation. The $20bn maximum cap for MBS roll-off should be enough to allow essentially unrestricted roll-off – unless interest rates fall and prepayments spike – by the end of next year.
The pace of roll-off of the Federal Reserve’s Treasury holdings depends on the particular securities in the SOMA portfolio. Treasury amortizations are uneven month to month. This reflects the pattern of Treasury issuance, which, historically, has been higher in the middle month of each quarter. The $30bn maximum monthly cap on roll-off will be binding, and therefore the Federal Reserve will be purchasing new Treasury securities, in some months for the foreseeable future. This will facilitate the Federal Reserve’s ongoing securities lending operations, which promote the efficient functioning of the Treasury market. Nonetheless, under the FOMC’s plan for Treasury roll-off, the Federal Reserve’s holdings of Treasury securities should decline at a rate of about 11% per year over the next five years.
Second, an extended passage setting out the tradeoffs between shrinking the balance sheet and retaining a large portfolio that uses interest on reserves and reverse repos to better control rates:
The scale of reserves and RRP transactions that the FOMC will ultimately choose to maintain depends on trade-offs between competing concerns. A large Federal Reserve balance sheet means that the Fed has a “big footprint” in money markets. This has pros and cons. A “big footprint” may enhance the efficiency and effectiveness of the Federal Reserve’s control over short-term interest rates.
It may also enhance financial stability by reducing complexity and reducing the incentive for private maturity transformation.
However, a “big footprint” also means that the Federal Reserve is effectively displacing private financial intermediation in short-term funding markets. It also means that the Federal Reserve interacts with a wider range of counterparties and has a direct impact on a wider range of short-term interest rates. A direct, obvious, and controversial aspect of this displacement of private financial activity is the interest that the Federal Reserve pays on reserves and its RRP transactions.
At the moment, the Federal Reserve appears more inclined to maintain a relatively large balance sheet and the “big footprint” in money markets that comes with it. However, a new set of policymakers will have to revisit these decisions. In the scenarios outlined above, the earliest that the Federal Reserve would have to start expanding its balance sheet again is July 2020. By that time, the Federal Reserve is likely to have a new Chair, a new Vice Chair, its first Vice Chair for Regulation, and a new President of the Federal Reserve Bank of New York.
In this context, it is important to note a key connection between the debate over financial regulation and the policy issues related to the size of the Federal Reserve’s balance sheet. This connection relates to short-term funding of long-term assets. Such “maturity transformation” was a major accelerant in the GFC, and important aspects of the regulatory response – notably the expanded use of simple leverage ratios and money market reform – were designed to limit that activity.
The Trump Administration said it makes sense to ease restrictions that have made it costly for banks to raise short-term liabilities to fund high-quality long-term assets. The administration seems much more supportive of facilitating such activity in the banking system. Given these attitudes, the Trump Administration and their potential appointees to the Federal Reserve Board may be less inclined to support the Federal Reserve maintaining a large balance sheet and a “big footprint” in money markets.
In this context, the positions put forward by The Clearing House (TCH), whose members are large banks, are noteworthy. Research from TCH has been very critical of the leverage ratio and other regulations that made it expensive for banks to expand the size of their balance sheets. In addition, other research from TCH advocated that the Federal Reserve should return to a small balance sheet, arguing that a small Federal Reserve “footprint” in money markets is less distortionary and that a lower level of interest on reserves would be less politically controversial.
On the key regulatory issues, the Trump Administration seems closer to the positions of TCH than those of current Federal Reserve officials, notably those expressed by Chair Yellen in her recent speech at Jackson Hole. It is not clear whether the incoming leadership of the Federal Reserve is going to be as supportive of maintaining a large balance sheet, and a “big footprint” in money markets as current Federal Reserve officials. Key decisions on the Federal Reserve’s balance sheet are years away, but attitudes of Trump’s Federal Reserve appointments on regulatory issues are worth monitoring.
My own position on this has been clear for a while: A large balance sheet can bring tangential benefits in addition to the enhanced control over short rates, and a residually big imprint in money markets for the official sector (the government and the Fed collectively) is a stabilising force. It’s reasonable to explore just how big the balance sheet needs to be, or how its composition might be remixed to give the Fed better control of the yield curve in the future, but size on its own isn’t a problem.
Still, it’s good to revisit these tradeoffs from time to time, and especially useful is the discussion of how the Fed’s larger balance sheet is consistent with the trajectory of new regulations since the crisis. For all their complexity, one of the clearest effects of these regulations has been to crimp the financial sector’s ability to create debt that trades in money markets with the illusion of near-perfect, money-like safety — either by puncturing the illusion or restricting the debt creation in the first place. Similarly, a large Fed balance sheet has crowded out competition from the private sector by giving a wide range of counterparties access to money-like holdings when economic conditions make such holdings appropriate.
Third, a passage on how the Fed has attempted to preclude a re-enactment of Bernanke’s taper tantrum:
An important aspect of balance sheet normalization is that the two drivers of long-term interest rates – term premia and the expected path of short-term interest rates – are likely to move in opposite directions. Higher term premia will tighten financial conditions. The first order effect on the economy would be a reduction of aggregate demand. With inflation still notably below the FOMC’s target, the FOMC will want to offset the economic impact of higher term premia. In other words, the FOMC will probably moderate the trajectory of its short-term interest rate hikes in order to counteract the economic impact of higher term premia. That means declines in the expected path of short-term interest rates will offset some portion of the rise in term premia on long-term interest rates.
We can use the Federal Reserve Staff’s FRB/US model to assess how rising term premia will affect the US economy, the trajectory of short-term interest rates and other financial assets. Even with the Federal Reserve’s well-communicated intention to move very gradually, there may be a discrete reaction in term premia and mortgage spreads. Figure 15 shows the impact of a sustained 50bp increase in term premia, combined with a sustained 25bp increase in the spread between retail mortgage rates and comparable Treasury yields (Figure 15 shows deviations from the baseline).
The direct effect of the rise in term premia and mortgage spreads is a reduction of aggregate demand, increasing the unemployment rate and lowering inflation. The FOMC is assumed to follow an inertial Taylor rule, so it can lower the path of short-term interest rates. Asset markets are assumed to be forward looking. Consequently, the expected short rate component of ten-year yields falls when term premia and mortgage spreads rise. This immediate response offsets a significant part of the rise in term premia on longterm interest rates.
Of course, a gradual adjustment of the Federal Reserve’s balance sheet should (may) generate a gradual adjustment of term premia and mortgage spreads. Figure 16 shows the impact of a gradual adjustment of term premia and mortgage spreads that is comparable to the base case shown in Figures 10 and 11. The total adjustment of term premia is about 80bp, but it takes place over many years. The rise in term premia eventually depresses the path of short-term interest rates. Investors anticipate the lower path of short-term rates, which holds down the expected rate component of long-term interest rates now. Thus, the prospect of balance sheet adjustment in the future holds long-term rates below where they would otherwise be right now because short-term rates will be lower in the future. Long-term interest rates would still rise from current levels, but by only about half of the adjustment in term premia.
The adjustment in term premia does imply a steeper yield curve over time. The logic of these simulations is that adjustments of the Federal Reserve’s trajectory for short-term interest rates keep the overall stance of monetary policy near the baseline path. However, with the adjustment in term premia, the mix of short- and long-term interest rate is quite different. That is, with balance sheet adjustment, the effective tightening through higher long-term rates is offset by a more accommodative path of short-term rates.
The attempt to offset any reduction in demand from the shrinking of the balance sheet by lowering the anticipated path of rates would be consistent with recent speeches by Bill Dudley and Lael Brainard. Whether the rest of the Fed will go along is anyone’s guess given the forthcoming, and still highly uncertain, personnel changes.