This question is on all market participants’ minds. Attention is now centered on whether the process will begin, as some FOMC participants have suggested, as early as the September FOMC meeting. Indeed, it seems highly unusual that, in the first week of August and in the wake of the controversial press conference held after the FOMC’s June meeting, four Federal Reserve Bank presidents, widely considered representative of the full spectrum of views among FOMC participants, have gone on record suggesting that tapering of the Fed’s $85 billion/month asset purchase program could begin as soon as the September. The four are President Fisher (hawk), Presidents Pianalto and Lockhart (centrists), and President Evans (dove). To be sure, we have heard ad nauseum from FOMC participants that their actions will be conditioned upon “incoming data.” Putting all that aside, there are both personal and political considerations as well as theoretical economic issues that complicate the tapering formulation process.
First, there are some interesting internal FOMC political considerations that might give current FOMC members pause as they consider starting the tapering process this year. To do so with only three FOMC meetings left before a new Fed chairman might be seated would effectively pre-commit both that new chairman and the reconstituted FOMC to a policy path that virtually no members would have had a say in formulating.
Consider first the situation on the Federal Reserve Board itself. President Obama has indicated that he will nominate a new chairman of the Fed sometime this fall, so Chairman Bernanke is already a lame duck. Additionally, Governor Duke has left the Board, and Governor Raskin has been nominated for the number-two position at the Treasury. Governor Powell’s term is up on January 31, 2014. These transitions mean that there are soon to be four vacancies on the Board of Governors. Finally, should Governor Yellen not be named to replace Chairman Bernanke, there would be little reason for her to stay on. That would leave only two current Board members – Governor Tarullo, a lawyer, and Governor Stein (the only economist), neither of whom experienced the events of the 2007-2008 financial crisis firsthand at the Fed.
Add to this unprecedented turnover at the Board the fact that the only Federal Reserve Bank president who will vote both this year and next is the New York Fed’s President Dudley, a permanent FOMC member. Finally, Cleveland President Pianalto, who is scheduled to have a vote next year, has announced her retirement; and President Fisher, who is also scheduled to vote next year, reaches the mandatory retirement age of 65 in 2014. This means that there potentially could be as many seven new voting members on the FOMC next year, none of whom are currently in place. The present FOMC has no idea what those people’s views are or what their policy preferences may be. To initiate a tapering policy this year under such circumstances could be highly disruptive should the new FOMC desire to pursue a different policy program next year. All of this argues for caution on the part of the current FOMC, especially given the turmoil that has roiled markets recently over policy concerns and the lack of evidence that the economy has suddenly picked up sufficient steam, such that policy actions would be required at this time.
Another key issue concerns the theory behind the Fed’s asset purchase programs and the growing evidence regarding their efficacy or lack thereof. Fed officials have been dogged in their attempts to distinguish among the FOMC’s zero interest rate policy (holding the Federal Funds rate between 0 and .25 percent), the interest rate paid on reserves, its asset purchase programs, and its communications and forward-guidance tools. Because nominal interest rates cannot fall below zero, there is a limit to how far accommodative policy can be extended by lowering interest rates.
One way to think about this is that in fixing the price of Federal Funds, which constitute tradable excess reserves at the Fed, the FOMC is controlling the quantity of reserves in the banking system and ultimately the money supply. Low rates are accommodative because the opportunity cost to banks of holding low-yielding assets in the form of deposits at the Fed is high relative to the returns that can be made by making loans (and in doing so, also increasing the money supply through the deposit expansion multiplier). The opposite applies when rates are high and policy is restrictive. But at the zero bound it is no long possible to lower rates and encourage expansion of bank reserves and the money supply indirectly, so the Fed provides extra accommodation by operating directly on the supply of bank reserves through its asset purchases. The Fed pays for the Treasuries and MBS by writing up banks’ deposits held at the Fed, and in that way it provides further accommodation by increasing the quantity of excess reserves, and hence Federal Funds, directly.
Asset purchases also do another thing. When Treasuries and MBS are taken out of the private sector market, their prices increase and their yields decline. This makes them less attractive to hold relative to other higher-yielding assets such as equities, corporate debt, and loans. This is termed the portfolio balance effect. Indeed, many have argued that because of the portfolio balance effect, yield-seeking funds have found their way into equities and are behind the increase in the stock market. Of course, this was one of the intents of the asset purchase program: the hope was that an increase in perceived wealth would stimulate consumer spending, encourage investment and promote economic growth.
There is another component of this policy, however, that is rooted in the theory of the term structure. That theory in its simplest form, assuming no inflation, holds that real longer-term interest rates can also be represented by a series of real short-term rates. That is, if real long-term rates are higher than real short-term rates, this implies that investors expect short-term rates to rise in the future. The logic is simply that an investor could, for example, invest in either a two-year obligation or two one-year obligations – a one-year spot rate and a one-year forward contract on the same instrument. Aside from a small fee for giving up liquidity by holding the two-year instrument, an investor who does not expect short-term rates to rise would be indifferent as to the choice of holding two one-year instruments versus the two-year instrument. However, if the investor expects short-term interest rates to rise, then he or she would always opt to hold the sequence of two one-year investments unless the rate on the two-year instrument was sufficiently high to make the investor indifferent as to which option was chosen. So by comparing the rate on a two-year investment with the rates on a sequence of two one-year investments, it is possible to determine, for example, whether short-term rates are expected to rise in the future, because the rate on the one-year forward instrument will be higher than the spot rate if short-term rates are expected to increase. The opposite would hold if short term rates are expected to fall.
In the case of the Fed’s $85 billion/month asset purchase program, which involves the purchase of long-term Treasuries and MBS, the Fed is taking these assets out of the private market, bidding up their prices, lowering their yields, and interfering with both the normal term structure and market expectations about future rates. By doing so, the FOMC is over riding market expectations and is de facto signaling that it intends to keep short-term interest rates lower than the might otherwise expected. In fact, this is exactly what the FOMC has said in the statements released after its meetings. This is the so-called signaling channel that has been investigated recently by economists at the San Francisco Fed. (See Bauer and Rudebusch, “The Signaling Channel for Federal Reserve Bond Purchases,” FRB San Francisco Working Paper Series, August 2012.) Another FRB San Francisco paper (see Cúrdia and Ferrero, “How Stimulatory Are Large-Scale Asset Purchases?” FRBSF Economic Letter, August 12, 2013) supports the importance of the signaling interpretation. The authors estimate that without the signaling effect, the second asset purchase program, known as QE 2 (when the Fed purchased an additional $600 billion in securities), would have added only 4 basis points to real GDP growth and 2 basis points to inflation. But when they also estimated forward guidance effect, they conclude that it dwarfed QE2 alone by adding another 9 basis points to GDP growth (but only another 1 basis point to inflation).
Now, what about the $85 billion asset purchase program? Using data from the Treasury on its net issuance of debt in 2013 to date, we find that the Federal Reserve has purchased 75% of those securities. So, if the Fed were to begin to scale back its purchases, more Treasury supply would be available to the private sector, putting downward pressure on prices and raising rates. According to the expectations theory the rise in rates, because of the FOMC’s tapering would logically be interpreted by the market as a signal by the Fed that short-term rates will rise sooner than previously expected. Thus, notwithstanding the FOMC’s view that it was not (according to the theories it was following) tightening policy, markets would be led to conclude the opposite. Indeed, the abrupt reaction to even the hint by FOMC participants that the FOMC might consider tapering its program, demonstrates that the market interpreted the talk as a signal that rates would rise sooner than expected.
So what does this mean for investors? There clearly is a disconnect between theory and evidence and it is currently impacting the FOMC’s intended policy. This, together with the personal/political considerations surrounding the composition of the Board of Governors, its leadership, and the makeup of the voting presidents, makes divining what is likely to happen even more difficult. One thing seems rather clear at this point and that is other factors besides just “incoming data” will be in play, which will only serve to increase volatility and place a premium on hedging by investors.