In its most recent statement, the FOMC affirmed that it intended both to complete its announced purchase of $1.25 trillion of agency-guaranteed mortgage-backed securities (MBS) and to slow its purchases, and that it anticipated ending the program by the close of the first quarter of 2010. This FOMC is distinct from a Board of Governors program and can be extended or curtailed without Treasury or other governmental approval.
In an interview yesterday on CNBC, Professor Robert Shiller, one of the principals behind the Case-Shiller Index of housing prices, responded to a question about the impact that the Fed’s termination of its MBS purchase program by saying he was uncertain as to the impacts it would have. But he clearly felt that there were significant downside risks to both the housing market and the recovery and that these risks would likely increase.
From the very beginning of its MBS purchase program, the Federal Reserve has been the most significant player in the GSEs’ MBS market. Over the period for which data are readily available, from January through January 2010, Fed purchases averaged nearly 70% of GSE MBS issuance. In some early months the Fed bought virtually all of the GSEs’ issuance, particularly in some segments like securities backed by fixed-rate obligations, and recently the Fed has averaged about 60% of the newly issued GSE MBS.
It is important to understand what the impact of the program has been; but also, as the question to Professor Shiller suggests, understanding the effects that the Fed’s withdrawal will have on mortgage rates, the housing market, and ultimately the economy is perhaps even more critical. Commentators on the program have differed widely on the impact it has had on mortgage rates. Estimates have ranged from 50 to 150 basis points in reductions in mortgage interest rates. The manager of the Federal Reserve’s Open Market Desk, Brian Sack, estimates that mortgage rates are about 100 basis points lower than they would have been without the program. At the same time, while there is consensus that the program has narrowed the spread between MBS rates and Treasuries significantly – perhaps on the order of 100 basis points – Stroebel and Taylor (see Stroebel and Taylor, Jan. 27, 2010: http://www.voxeu.org/index.php?q=node/4531) suggest that after controlling for prepayments that it is hard to identify more than a 30-basis-point impact of the program on spreads.