Fed Balance Sheet Normalization: When Boring Is Good

Introduction

At the start of 2017, Federal Reserve (Fed) members began sharing their views on the Fed’s balance sheet with increasing frequency. This renewed attention to balance sheet normalization brought back memories of the 2013 “taper tantrum,” with some investors warning that the Fed’s exit from the Treasury (UST) and mortgage-backed securities (MBS) markets would result in a sharp rise in term premium, steepening of the yield curve, and widening in mortgage spreads. At Western Asset we had a more sanguine outlook, as we expected the Fed to take a cautious, gradual and well-communicated approach to reducing its purchases, and we believed that MBS valuations already reflected the downside risks of a Fed exit (see As Fed Reinvestments End, New Opportunities Arise for Bond Investors, March 28, 2017).

The Fed’s discussions culminated at its June meeting, when it reiterated its intent to start balance sheet normalization this year and further detailed the mechanics of its plan for reducing the balance sheet in an Addendum to the Policy Normalization Principles and Plans. Under this plan, the Fed will initially reduce its monthly MBS and Treasury reinvestments by no more than $4 billion and $6 billion, respectively. If runoff falls short of the caps in any month, the Fed will not sell the difference; if, however, runoff exceeds the caps in any month, the Fed will reinvest the difference. Every three months, it will increase the monthly MBS and UST caps by $4 billion and $6 billion until they reach $20 billion and $30 billion, 12 months later (Exhibit 1). Assuming that economic and financial conditions hold, we believe the tapering program will begin in September or December, in line with market expectations.

Market Impact

To be sure, as the Fed goes through the process of normalizing its balance sheet, private investors will need to absorb a significant amount of supply (Exhibit 2). In our view, however, the Fed has learned from the 2013 experience and designed a program that can indeed “run quietly in the background” as the Fed intends². This will leave the federal funds rate as the Fed’s main tool for adjusting monetary policy, while maintaining some balance sheet flexibility to deal with more exigent conditions in which the Fed’s policy rate is at or near zero.

Why do we believe this amount of supply will be manageable?

• Pace and Magnitude of Runoff:
  The runoff schedule laid out is very gradual, offering the market ample time to adjust to the new program. Once the runoff is fully ramped up, which we expect to occur toward the end of 2018, the maximum amount per year that can run off is $600 billion. If rates remain close to current levels, however, supply from runoff will not come close to this amount (Exhibit 2). In fact, when runoff peaks in 2019, the Treasury and MBS supply created would add only about 2% to the balances currently outstanding. If rates rise to follow the forward curve, supply will be even lower due to slower prepayments. On the other hand, if in any given month there is a rally in rates that results in the Fed’s MBS portfolio experiencing runoff greater than the cap due to higher prepayments, the Fed will reinvest the excess amount, providing a countercyclical backstop of sorts to the MBS sector.
• Size of Long-Run Balance Sheet:
  As the slow balance sheet reduction program emphasizes, the Fed’s balance sheet does not need to shrink all that much. A number of the initial concerns about having a large balance sheet have proven unfounded. Contrary to what many feared, the impact on inflation has been muted at best. Another concern was that the large amount of excess reserves would frustrate any attempt to raise rates. The relatively new interest on excess reserves (IOER) and term repo programs allow the Fed to influence market rates even with a large balance sheet. Most estimates for the terminal size of the balance sheet range from $2.5 trillion to $3 trillion, around 15% of GDP. By the Fed’s own projections, the balance sheet would reach that range sometime between 2020 and 2021.³
• Supply Versus Duration:
  The Treasury has flexibility over how it refunds the issuance maturing on the Fed’s balance sheet. For example, the Treasury can issue bills, which have low duration, or it can issue longer-maturity, higher-duration bonds. The Treasury could initially increase auction sizes in shorter-maturity USTs in order to facilitate a smooth transition to a smaller Fed balance sheet. Given the growth in the government money market funds resulting from the introduction of new SEC rules in 2016, any increase in Treasury bill issuance would likely be easily digested (Exhibit 3).

• Fundamental Factors:
  While technical factors can affect valuations, many fundamental considerations ultimately determine the level of UST yields and MBS spreads. To state the obvious, UST yields will be lower in a slower-growth, lower-inflation environment than in a higher-growth, higher-inflation environment. Changes in inflation expectations, as well as global developments, can also increase or decrease risk premium. Our outlook is for moderate growth and low inflation to continue, giving value to USTs as diversifiers against assets with credit risk in our portfolios.

  MBS valuations are influenced by the level of mortgage yields, changes in volatility, MBS hedge-adjusted carry (i.e., carry net of duration and volatility costs), and value relative to other asset classes. The current backdrop of improved spreads versus USTs, low and declining volatility, and cheapness relative to corporates is positive for the MBS sector.

• Global Central Bank Policies:
  While we expect the Fed to initiate the balance sheet normalization process later this year, we do not expect material changes to the large scale asset purchase programs currently being conducted by the European Central Bank (ECB) and Bank of Japan (BoJ). We believe that both the ECB and BoJ will proceed cautiously when removing accommodation due to persistently low levels of inflation across the eurozone and Japan. The ongoing involvement of these central banks will continue to provide support for fixed-income valuations globally.

Conclusion and Investment Implications

We believe that the Fed’s monthly cap structure will minimize the impact of balance sheet reduction on the rates and MBS markets. We would also characterize the Fed’s goal with respect to not only the balance sheet but also the fed funds rate as gradual normalization rather than aggressive tightening given the prolonged absence of realized inflation. The Fed’s careful approach should keep UST rate volatility subdued and support MBS spreads.

In our broad portfolios at Western Asset, we have been underweight agency MBS given our preference for other asset classes and the risks posed by the Fed’s exit from the sector. MBS spreads have indeed widened over the past year in anticipation of Fed balance sheet reduction, and are now at levels reflecting a worse scenario than what the Fed ultimately delivered in June. In addition, MBS have underperformed USTs despite the downward trend in volatility and tightening in credit spreads this year. We have accordingly increased our allocation to agency MBS, bringing portfolios close to neutral relative to their benchmarks. MBS spreads may leak wider over time as the Fed approaches full withdrawal from the market, but the carry will serve to offset that. Given the slow pace of Fed balance sheet reduction and widening of MBS spreads that has already occurred, we believe that the magnitude of further cheapening will be limited, and that MBS are poised to outperform USTs over the next nine months to a year.