Will Fed tapering lead to headwinds for the mortgage market?
As the US Federal Reserve starts to normalize its balance sheet, some anticipate that this monetary tightening could create risks for agency mortgage-backed securities (MBS). However, considering the history of quantitative easing (QE) since the global financial crisis, we believe that tapering of reinvestments alone is unlikely to lead to near-term widening MBS spreads. Longer term, as long as the Fed does not actively sell MBS holdings, a gradual tapering of reinvestments and a balance sheet in run-off could mean only marginal headwinds for the mortgage market.
A balance sheet larger than pre-financial crisis
Beginning in the fall of 2017, the Fed plans to start shrinking, or running off, the balance sheet reinvestment program for Treasury maturities and MBS principal paydowns by $6 and $4 billion per month, respectively. These amounts (caps) will be reviewed quarterly and increased in equal increments until they reach $30 billion for Treasury run-off and $20 billion for MBS. The ultimate size of the Fed’s balance sheet is unknown, but will likely be significantly larger than pre-2009 levels of $800–$900 billion.
The ebb and flow of reinvestment amounts of agency MBS and Treasurys could lead to relative performance differences, as the chart indicates. Over the short term, agency MBS can be expected to outperform, while in the first half of 2018, Treasury reinvestments surpass agency MBS as the amount of Treasury maturities increases. Reinvestment amounts were calculated as the difference between the Fed’s expected MBS paydowns, Treasury maturities, and their respective run-off caps.
A look at history may be telling
The concerns regarding agency MBS from a shrinking balance sheet stem from the amount of securities owned by the Fed, more than a quarter of the nearly $7 trillion mortgage market. While a rapid balance sheet reduction through active selling would certainly impact MBS, a gradual fading of reinvesment purchases and a balance sheet that runs off through principal repayments should be palatable. As an example, from 2010–2011, the Fed’s MBS holdings similarly started shrinking as QE1 ended in March 2010, while their Treasury holdings increased as the Treasury-focused QE2 was resumed in Novermber 2010. Over that time frame, spreads were only 10 basis points wider in what on the surface looked liks a signficant headwind.
Given this backdrop, we think MBS spreads are unlikely to widen in the near term from the Fed announcement alone. A sharp move +/-25 basis points, however, can change prepayment fundamentals and affect spreads.
Ion Dan is portfolio manager, senior structured products analyst & trader at Macquarie Investment Management