- After ending its QE bond purchasing program in 2014, the Fed pledged to reinvest the principal from its maturing bond holdings in MBS and USTs “until normalization of the level of the federal funds rate is well under way.”
- As expectations for further monetary tightening have risen, investors are now wondering when the Fed will end reinvestments.
- We expect the Fed to introduce its taper plans sometime around the end of this year, resulting in wider MBS spreads. The magnitude of widening will depend on the level of primary mortgage rates, Treasury issuance and Fed SOMA strategies, fundamental factors, and investor positioning.
- As the Fed’s dominance in the MBS market recedes, the importance of active management will increase.
Through its first quantitative easing (QE) program, the Federal Reserve (Fed) began purchasing agency mortgage-backed securities (MBS) in November 2008 and US Treasuries (USTs) in March 2009. Since ending its third asset purchase program (QE3) in 2014, the Fed has continued to participate actively in both the MBS and UST markets by reinvesting the principal received from its maturing bond holdings. As a result, the Fed’s holdings have hovered around $1.75 trillion in MBS and $2.6 trillion in USTs for almost 2.5 years. Anticipation for change is now building as expectations for further monetary tightening and Fed member comments on balance sheet normalization have brought the topic of Fed “taper” (i.e., a gradual end to reinvestments) front and center.
We believe the Fed will introduce its plans to taper reinvestments sometime around the end of this year, with the likely result of some MBS spread widening. Yet several factors, including improved current MBS valuations and the low volatility environment, indicate that the MBS sector will remain attractive over the near-term, particularly as a replacement for other high-quality investments that may have weaker fundamental valuations.
We further detail our views on Fed tapering and relative value in the agency MBS sector in the notes that follow.
When will the Fed stop reinvesting in MBS and USTs?
When the Fed completed QE3, it pledged to reinvest maturing principal in MBS and USTs “until normalization of the level of the federal funds rate is well under way.” Prior to the March Federal Open Market Committee (FOMC) meeting, Fed members had indicated that this level would be reached when the federal funds rate reached 1% to 2%. In the press conference following the March meeting, however, Fed Chair Janet Yellen stated that the threshold was not wholly quantitative in nature and added the “balance of risks and confidence in economic outlook” as another condition to tapering reinvestments. The Fed’s current projection for the terminal funds rate is 3.0%. Assuming that this projection does not change materially, then 1.5% is a reasonable estimate for rate normalization that is “well under way,” although in light of Yellen’s recent qualification, tapering may not occur until later in the hiking cycle.
We further surmise that housing reform, which Treasury Secretary Steven Mnuchin has stressed as a top priority for the Trump administration (see our previous analyses: Laying the Foundation for Housing Reform and Forcing the Hand of Housing Reform), could lead the Fed to proceed cautiously around ending MBS reinvestments. We expect the administration to reveal details around its plans for Fannie Mae and Freddie Mac over the next 12 months, which is likely the same window of time required before the economy reaches the Fed’s quantitative and qualitative thresholds for taper.
Weighing all these considerations, we expect the Fed to begin communicating its taper program sometime around the end of this year. While this may clash with the administration’s plans for housing reform, we believe the Fed’s desire to initiate the process before the end of the current Chair’s term will prevail.
Will the Fed taper MBS or Treasuries first?
Once the Fed determines the economy has met the Fed’s conditions for taper, we believe that reductions to reinvestments in MBS and USTs will commence simultaneously and be phased in over a nine- to 12-month period. This schedule would be similar to the Fed’s reduction of new Treasury and MBS purchases from December 2013 to October 2014. At the current level of mortgage rates (4.25% to 4.375%), the net supply of MBS would increase every month by $2 to $3 billion until the end of the taper period. It stands to reason that this pace would increase in rate rallies and decrease in rate selloffs, as the amount of maturing principal on MBS would be greater in lower-rate environments and smaller in higher-rate environments. This would mute the typical directionality of mortgage flows in rate moves due to convexity hedging, i.e., the need for MBS holders to add duration as their portfolios shorten duration in a rally and to sell duration as their portfolios extend duration in a selloff.
Will the Fed sell MBS?
As Chair Yellen affirmed in her most recent appearance before the Senate Banking Panel, the Fed has no plans to sell either USTs or MBS. Yellen’s term as Chair does expire at the end of January 2018, and a new Chair confronted with new circumstances would not necessarily adhere to this plan. Nonetheless, given the tenuous nature of the economic recovery, we do not anticipate a new Fed regime will sell assets in significant size in the near term. Further, we believe the Fed will remain sensitive to any disruption to the housing sector of the economy that might occur as a result of asset sales.
How could this impact MBS spreads?
The Fed owns 29% of the agency MBS market, and its reinvestments currently constitute about 25% of origination. Given this significant presence, we expect spreads to widen once the Fed begins communicating its plans for ceasing reinvestments. The magnitude of spread widening needed to accommodate a Fed exit will depend on several factors, including:
- Level of primary mortgage rates:
- Once the taper has been completed, the market will need to absorb an additional $200 to $225 billion of MBS supply per year, assuming that primary rates remain in the current range. At lower mortgage rates, this number will be larger, since there will be more organic supply and faster prepayments on the Fed’s MBS portfolio. At higher mortgage rates, this number will be smaller, since there will be less organic supply and slower prepayments on the Fed’s MBS portfolio.
- Treasury issuance and Fed SOMA (System Open Market Account) strategies:
- If the Treasury issues bills with low duration to replace the issuance maturing on the Fed’s balance sheet, then MBS yields are likely to rise more than Treasury yields, whereas if the Treasury issues longer maturities with higher duration, then MBS spreads may not widen as much.
- The Fed has communicated a desire to eventually return to a Treasury-only portfolio. If the Fed targets a larger long-run equilibrium size for its Treasury SOMA portfolio, this would result in wider MBS spreads, whereas if the Fed targets a smaller steady state Treasury portfolio size, this would mitigate MBS spread widening.
- Fundamental factors:
- Volatility, MBS yields, MBS hedged-adjusted carry (i.e., carry net of duration and volatility costs), and relative value between asset classes.
- Investor positioning:
- Bank deposit and loan growth versus UST and MBS holdings, investor (e.g., domestic banks, international accounts, money managers) overweights/underweights to MBS.
Our estimate of the magnitude of widening is in line with consensus at 15 to 20 basis points (bps), although there is margin for error due to variance in the factors described above. One potential upside for Ginnie Mae valuations is that banks may increase their demand for Ginnie Mae MBS, which receives the same regulatory treatment as USTs, in order to maintain their liquidity coverage ratios as the Fed withdraws reserves.
How are we positioning our MBS portfolios ahead of the Fed tapering?
While we believe that valuations are likely to cheapen once taper eventually approaches, we believe current valuations and hedge-adjusted carry provide adequate compensation over the next 9 to 12 months. We view MBS as appropriately priced for the near-term risk of tapering, as option-adjusted spreads have widened 20 to 25 bps from the tightest levels of 2016 (Exhibit 2) and now reside on the wider end of the three-year range. From a fundamental standpoint, we believe that the Fed’s intention to raise rates gradually will contain volatility in longer-dated rates, which should benefit mortgages. We also view the MBS sector’s convexity profile favorably. At today’s rate levels, only 27% of the 30-year mortgage universe has the rate incentive and equity available to refinance,1 limiting the risk of further duration extension. Additionally, the current technical landscape, in which the Fed is reinvesting and origination remains low, limits the potential for spread widening over the short-run horizon.
In our allocations to production coupons, we began to rotate from TBAs to specified pools in December 2015, based on our view that attributes protecting against faster prepayments were undervalued and that TBA carry would begin to deteriorate as reduction in the Fed’s activity level from outright purchases to reinvestments would adversely affect the composition of MBS issuance. Spreads on specified pools have improved as both of these views were subsequently validated. We currently hold a bias for TBAs over specified pools as we believe the latter sector is fully priced.
Almost a decade ago, the Fed launched QE in an unprecedented effort to combat damage from the financial crisis. The looming withdrawal of QE will likewise be a unique event. As the Fed’s dominance in the MBS market recedes, the importance of active management will increase. We expect events over the next 12 months to present major opportunities for relative value-oriented MBS managers to capture alpha.
- Morgan Stanley Truly Refinanceable Index, as of March 15, 2017.