Given your function as counterparties for monetary policy operations, I thought it might be useful to take this opportunity to speak about the implementation of the policy decisions that the Federal Open Market Committee (FOMC) made at its last meeting. As you know, the FOMC took two additional policy steps aimed at supporting a stronger economic recovery and helping to ensure that inflation, over time, is at levels consistent with the dual mandate. First, the FOMC decided to extend the average maturity of its holdings of securities by purchasing $400 billion of Treasury securities with remaining maturities of six years to 30 years and selling an equal amount of Treasury securities with remaining maturities of three years or less. Second, the Committee decided to shift the reinvestment of principal payments on its holdings of agency debt and agency mortgage-backed securities into agency mortgage-backed securities (MBS) in order to help support conditions in mortgage markets.
The broad parameters of these policy steps were communicated in the FOMC statement. However, the implications of these actions for financial markets will depend importantly on the details of their implementation. For this reason, the Trading Desk at the Federal Reserve Bank of New York (the Desk) issued a statement at the same time as the FOMC statement specifying the operational details of its plans to implement the FOMC’s decisions, as it has done after other balance sheet actions.2 Today I will highlight some of those operational details and explain some of our thinking for putting them in place.
On the Maturity Extension Program
Let me start with the Maturity Extension Program (MEP). The motivation for this program was the same as that for previous asset purchase programs by the Federal Reserve. As has been discussed on many occasions, the effects of the asset purchase programs are thought to arise from the amount of duration risk that they remove from the portfolios of private investors. By removing duration risk, the Federal Reserve puts downward pressure on longer-term real interest rates, which in turn pulls down private borrowing costs and makes broader financial conditions more supportive of growth. The MEP was intended to have the same effects, only under a program that did not involve an expansion of the Federal Reserve’s balance sheet.3
Duration risk can be measured in a variety of ways, but one common measure for a securities portfolio is ten-year equivalents, or the amount of 10-year Treasury notes that an investor would have to buy to be exposed to the same amount of duration risk contained in the portfolio. Some of the staff work that calibrates the economic impact of the Federal Reserve’s balance sheet policies assumes that the effects on yields and financial conditions are driven by the amount of ten-year equivalents that the Fed takes into its portfolio.
By that metric, the effect of the Maturity Extension Program is about equal in size to that of the large-scale asset purchase program that ended in June of this year (what we have referred to as LSAP2). This fact was highlighted in a recent post to the Liberty Street Economics blog.4 In both cases, the effect of the program was to remove about $400 billion of 10-year equivalents from the market.
However, while similar in their intent, there are differences between a maturity extension program and an asset purchase program. Let me highlight three of them that are related to the structure of the MEP.
First, the size of the MEP is limited by the amount of shorter-term Treasury securities that the Federal Reserve could sell. As announced, the program involves selling about 80 percent of our holdings of securities maturing within three years, leaving little room to add to the MEP. This constraint affected the choice of securities to be purchased, as the purchases had to be at long maturities in order for the program to remove a meaningful amount of duration risk.5 We anticipate that the average duration of our purchases will be just over 10 years, or nearly twice as long as that for the purchases conducted as part of LSAP2.
Second, unlike an asset purchase program, the MEP has implications for the amount of debt that the Treasury has to issue over the next several years. As you know, when the Federal Reserve has maturing holdings of Treasury securities, the proceeds are automatically rolled over into newly issued securities at Treasury auctions. However, under the MEP, we will have sold $400 billion of securities maturing through June 2015. With these securities no longer being rolled over to the Federal Reserve, the Treasury will have to issue a larger amount of debt to private investors, thereby reintroducing duration risk into the market. This consideration was another reason why the purchases in the MEP had to involve very long-term securities, in order to ensure that the program removed a meaningful amount of duration risk from the market even after the Treasury’s issuance.
Third, the program requires the Desk to actively sell Treasury securities maturing over the next three years. This element of the program was not intended to put upward pressure on shorter-term Treasury yields. Indeed, with the FOMC indicating that it anticipates that economic conditions are likely to warrant the current level of the federal funds rate through mid-2013, I would have expected the 2-year Treasury yield to remain well anchored. However, there has been some modest upward pressure on the 2-year Treasury yield since the FOMC meeting—a development that the Desk will continue to monitor.
In terms of operations, the program is being carried out over the FedTrade system, which allows participation by all primary dealers in a competitive bidding process for their own accounts or on behalf of other market participants. Operations to sell securities were designed to essentially have the same structure as purchase operations, in that we take bids across a range of securities and select those that appear most attractive.6 To date, we believe that the execution of our operations has gone well.
On the MBS Reinvestments
Let me now turn to the implementation of the other policy initiative from the last meeting, the decision to shift our reinvestments into agency MBS. As noted earlier, the intent of this action was to help support conditions in mortgage markets. One factor prompting this decision was the ongoing widening between secondary market rates on MBS and Treasury yields, which was keeping mortgage rates higher than they otherwise would have been. Previous communications had highlighted that changes in market conditions, including a widening of the MBS spread, could prompt the FOMC to shift its reinvestments into MBS rather than Treasury securities.7
This decision came as a surprise to the markets, in part because the FOMC had communicated that it seeks to return to a Treasury-only portfolio over time. However, the move to a Treasury-only portfolio has always been described as a longer-run objective, in contrast to near-term policy decisions that are instead driven by economic and financial developments that have implications for the FOMC’s ability to meet its objectives.
In any event, the MBS spread should now reflect that the Federal Reserve is active in this market. The Desk anticipates that the pace of MBS purchases will run at around $25 billion over the next several months. Beyond that horizon, the pace of MBS purchases will depend on the evolution of refinancing activity, as governed by the path of longer-term interest rates, housing activity, and other factors.
The MBS purchases are being conducted internally by the Desk over the TradeWeb platform. As with our Treasury operations, the process for transacting in MBS is structured to follow a competitive process across primary dealers to ensure that we are receiving fair market prices for the securities. To date, the purchases have gone smoothly, and market liquidity seems to be quite good.
Our purchases are focused on newly-issued securities in the To-Be-Announced (TBA) market, given that these are relatively more liquid and more closely tied to the primary mortgage rate.8 In particular, we have set the distribution of our purchases over agencies, coupons, and terms to be roughly proportional to an estimate of new market origination. However, we make some exceptions based on the objective of the program and concerns about market functioning, which to date have led us to reduce the allocation to securities issued by Ginnie Mae and to 15-year MBS relative to what would have been called for under a proportional approach.
Lastly, you may have noticed that the FOMC directive to the Desk provided the authority to conduct dollar roll transactions to facilitate the settlement of our MBS purchases. We will take an approach of engaging in dollar roll transactions if the pricing of the roll moves to levels indicating significant scarcity of the securities on which we are expecting delivery. This is a prudent approach for addressing market functioning concerns while maintaining the policy objective of the program.
In conclusion, let me say that all of us have a role in the operational success of the Federal Reserve’s balance sheet programs. On our end, we try to design programs that meet the FOMC’s objectives while not being detrimental to the functioning of financial markets. On your end, we rely on your ongoing and active participation in these operations, both for your own accounts and on behalf of other market participants, under the terms that we have specified. To date, I think that both sides have lived up to their responsibilities, resulting in the effective implementation of these programs. Going forward, we hope to continue this track record, regardless of the direction that policy takes.
1My comments do not necessarily reflect the views of the Federal Open Market Committee (FOMC) or any other members of the Federal Reserve System.
2The FOMC statement can be found at http://www.federalreserve.gov/newsevents/press/monetary/20110921a.htm, and the Desk statement can be found at /markets/opolicy/operating_policy_110921.html.
3In a speech in July 2010, I noted that the asset purchase programs had removed duration risk from the market not only through the size of the SOMA portfolio, but also through its composition (see /newsevents/speeches/2011/sac110720.html). The MEP is designed to operate exclusively through the composition of the balance sheet.
5In this regard, asset purchase programs contain a degree of freedom that policymakers do not have under a MEP. Under an asset purchase program, policymakers can choose the optimal range of maturities to be purchased and can scale the program up as needed to achieve the desired effects.
6The algorithm for determining which bids to accept in our sales operations depends only on a comparison to market prices. This approach does not involve an additional assessment of relative value across securities, in contrast to the algorithm for purchasing securities. One reason for this difference is that we are selling the vast majority of our securities holdings, and thus there would be little benefit to the relative value assessment if most of the differences in value are persistent.
7Chairman Bernanke’s Jackson Hole speech in August 2010 (see http://www.federalreserve.gov/newsevents/speech/bernanke20100827a.htm) and the minutes from the August 2010 FOMC meeting (see http://www.federalreserve.gov/monetarypolicy/files/fomcminutes20100810.pdf ) indicated that reinvesting in MBS rather than Treasury securities might become desirable if market conditions were to change. In a speech from October 2010 (see /newsevents/speeches/2010/sac101004.html), I noted, “A significant widening of MBS spreads to Treasuries … could affect policymakers’ decisions about which assets to purchase.”
8To be sure, the spread between the primary mortgage rate and the secondary rate on MBS is very wide, reflecting a variety of factors. However, those factors should not reduce the sensitivity of the primary rate to the secondary rate over time.