Goldman is not used to being snubbed when it comes to its policy recommendations. Which is what essentially happened during the Jackson Hole speech in which Bernanke left the QE3 door slightly open... but not enough to appease Bernanke's Goldman alum superior at the New York Fed. As such, since at least a few days have passed without Goldman reminding of who calls the shots, here is probably the most comprehensive summary of the Jackson Hole aftermath, and what the market will now expect to come out of the Fed on September 21. Whether Bernanke will go ahead and listen to Goldman, is unknown - the Fed risks incurring the wrath of Goldman at its own peril. What is perhaps most interesting about self-Q&A are the Goldman proposed "radical" measures that the Fed could consider to employ in addition to LSAP and Twist which so far have proven to be ineffective: "There are three main ways in which the Fed could be more radical: (1) an extension of the QE program into markets other than Treasuries and agency MBS, e.g., private sector securities, (2) a much bigger QE program, up to the extreme version of a promise to buy as many securities as needed to hit a specific yield target (i.e. a "rate cap" further out on the yield curve as then-Governor Bernanke suggested back in 2002), and (3) an explicit or implicit change in the Fed's policy targets." If these are truly hints as to what the Fed should do, then we hope readers have their gold $3000 calls firmly in place.
That said, to end once and for all all ridiculous debates over what is and what isn't QE3, here is Goldman with the most spot on definition of what further easing is and will be: "One final note on terminology: for simplicity, we have decided to label as "QE" any program of large-scale asset purchases that removes a significant amount of duration from the market, regardless of whether it is financed by creation of excess reserves or sales of short-duration securities. This is a little sloppy, but we believe that the economic distinction between the two modes of financing is so small--and the term "QE2" is already so well established--that we want to refer to them with the same basic term." Hence people can stop referencing momentum chasing FX traders in attempting to make their case of why Operation Twist is not QE3.
Q&A on Monetary Policy after Jackson Hole
Q: Will Fed officials ease monetary policy further?
A: Yes, we think so. At this point, we believe that the majority of the FOMC expects real GDP growth of around 2.5% in the second half of 2011, and perhaps 3% in 2012. As a result of growth that averages a bit above trend through the end of next year, the committee "…anticipates that the unemployment rate will decline only gradually toward levels that the Committee judges to be consistent with its dual mandate." Our own expectation is that real GDP growth will average nearly 1 percentage point below the committee's likely forecast and that the unemployment rate will go sideways to slightly higher. If we are right and the committee ultimately reduces its growth and employment forecast accordingly, we think that this shift in view will be accompanied by further monetary easing.
That said, additional easing is not a certainty. Following Monday's better-than-expected personal spending report for July, it looks like real GDP in Q3 may come in a bit stronger than our forecast of just 1% (annualized). Moreover, inflation has continued to come in on the higher side of expectations, and the "hard" economic data for July have been stronger than anticipated. If growth and inflation exceed our forecasts, additional easing would become less likely.
Q: Should Fed officials ease monetary policy further?
A: Yes, we think so. The US economy has a large--and probably increasing--amount of excess capacity, at a time when fiscal policy is already primed for restraint. Thus, the risks to the attainment of the Fed's dual mandate of maximum employment and low inflation are clearly tilted to the weak side at the current stance of monetary policy, which implies that further easing is a natural expectation. Indeed, our QE-adjusted Taylor rule--which is based on the Fed's past response to unemployment and inflation gaps as well as our estimates of the effectiveness of unconventional monetary policy--implies a need for additional easing equivalent to a cut in the funds rate by around 200bp.
There are two main arguments against further easing, but we don’t find them very compelling. The first is that "QE2 didn't work, so why should QE3 be any better?" We disagree with the view that QE2 didn't work. It is true that QE2 failed to ignite a more powerful recovery. However, we would attribute this to the combination of an even more adverse "baseline" pace of growth than we (and the Fed) had expected, and to the increase in oil prices. Moreover, our commodity strategists believe that most of the increase in oil prices was due to the tightening demand/supply situation in the oil market, exacerbated by the turmoil in the Middle East. Our belief that the moves in oil prices have been mainly driven by supply and demand rather than monetary policy is also consistent with the easing in oil prices over the past few weeks--a period in which the economic indicators have deteriorated and expectations of QE3 have grown.
The second argument against further easing is that most of the problems in the economy are not easily addressed by monetary policy but require either fiscal solutions or simply time. We agree that monetary policy is unlikely to be very powerful, but we do think that it would add a few tenths to GDP growth and would not have significant costs in terms of inflation given the large amount of slack in the economy (see "How Much Growth Boost to Expect from QE3?" US Daily, August 24, 2011). Ultimately, the question whether to ease is a cost-benefit calculation, and the benefits exceed the costs in our view.
Q: If it happens, what form will the easing take?
A: In his recent monetary policy testimony, Chairman Bernanke noted three ways of providing additional stimulus: (1) more explicit guidance about future policy, (2) changes in the size or composition of the Fed balance sheet, and (3) a cut in the interest rate on excess reserves (IOER) from its current 25 basis points (bp).
By stating their expectation that the funds rate will stay exceptionally low until at least mid-2013, Fed officials have already gone as far as they are likely to go with respect to (1). Moreover, while a cut in the IOER as per (3) is possible, it is unlikely to have a sizable effect given that the effective funds rate is only 8-9bp at present. This leaves (2) as the most straightforward option. Specifically, our current baseline expectation is a renewed asset purchase program that removes a similar amount of duration from the marketplace as the QE2 program announced last November. However, we think there will be some "tweaks" compared with QE2 that would be mostly designed to reduce the potential for "backlash" against the Fed at home and abroad. One possibility would be to finance the asset purchases via sales of shorter-duration securities as opposed to the creation of excess bank reserves (see "For More Easing, Will the Fed Go Big or Go Long?" US Daily, August 15, 2011). The advantage is that such a "twist" does not imply another controversial expansion in the Fed's balance sheet. Another possibility--not necessarily mutually exclusive--would be to announce not the ultimate amount of purchases but only a "run rate" that is reevaluated at each FOMC meeting and remains in place until further notice.
One final note on terminology: for simplicity, we have decided to label as "QE" any program of large-scale asset purchases that removes a significant amount of duration from the market, regardless of whether it is financed by creation of excess reserves or sales of short-duration securities. This is a little sloppy, but we believe that the economic distinction between the two modes of financing is so small--and the term "QE2" is already so well established--that we want to refer to them with the same basic term (see "QE2 as a Shortening of Treasury Debt Maturities," US Daily, October 25, 2010). (We draw the line at an increase in portfolio duration that is effected only through the reinvestment of MBS paydowns; we do not call this QE.)
Q: What assets would they buy?
A: Probably mostly longer-duration Treasuries. To be sure, the recent widening in agency MBS spreads has somewhat raised the probability that Fed officials might go back into that market, and a further widening could raise it further. It is also possible that MBS would be incorporated into a possible "twist" via sales of high-coupon securities (which have relatively short duration because of a higher probability of refinancing/prepayments) and purchases of low-coupon securities (which have relatively long duration). However, we think the hurdle against duration extension in the MBS market is relatively high because Fed officials remain determined to return to an all-Treasury balance sheet in the longer term. An increase in the stock of low-coupon securities that may not roll off the balance sheet for up to 30 years would make this process even more challenging.
Q: What is the timing?
A: The key question is what happens to the FOMC's forecasts. We suspect that the current forecast is not yet weak enough to lead to another big easing step, on top of the substantial change in forward-looking language at the August meeting. Based on our forecasts for the near-term economic data--including a 50,000 gain in nonfarm payrolls, a flat 9.1% unemployment rate, and a 48.5 ISM--we also do not expect a large downgrade in the forecast. As a result, our base case for September 20-21 is for the FOMC to state its intention to reinvest MBS paydowns at the longer-end of the Treasury yield curve, but only to signal the possibility of a larger program at a later date.
However, this is a close call. For one thing, our assessment of likely changes in the FOMC forecast could change in response to new information about the economy, including substantial disappointments in this week's economic data. Moreover, it is possible that the FOMC leadership has already decided to take another big step but is still building the internal support (we noted this possibility just after the last meeting; see "QE3 Now Our Base Case," US Daily, August 9, 2011). But our best guess is that a big step will only come a bit later.
Q: Does the Fed have more extreme options up its sleeve?
A: Yes, but the hurdles are very high, and a more radical approach is unlikely unless the economy and/or the financial markets perform substantially worse than we are forecasting. There are three main ways in which the Fed could be more radical: (1) an extension of the QE program into markets other than Treasuries and agency MBS, e.g., private sector securities, (2) a much bigger QE program, up to the extreme version of a promise to buy as many securities as needed to hit a specific yield target (i.e. a "rate cap" further out on the yield curve as then-Governor Bernanke suggested back in 2002), and (3) an explicit or implicit change in the Fed's policy targets.
Regarding (1), the general understanding of the Federal Reserve Act is that it does not allow purchases of assets which could result in an outright loss for the Fed. Therefore, a purchase of private sector assets such as nonconforming mortgages or corporate bonds (let alone equities or real estate) almost certainly requires funding from Congress, which is unlikely to materialize unless the level of distress in the economy or the markets rises sharply.
Regarding (2), while our base case is a QE program that removes a similar amount of duration as QE2, a bigger and bolder approach is certainly possible. The bigger the program, the more likely it is that it would ultimately need to involve a sizable expansion in the Fed's balance sheet because the scope for a "twist" reaches its natural limits. However, the limiting case of a "rate cap" would again require an extreme situation. Fed officials would undoubtedly worry about the tail risk that they might need to buy up the entire supply of securities in the targeted sector to make good on their promise.
Finally, regarding (3), we have again received some questions about the possibility that the FOMC might move to a nominal GDP target (see "The Fed Discusses Easing Options," US Daily, October 12, 2010, as well as this week's Economics Focus in The Economist). It's important to note that depending on the interpretation, the Fed's dual mandate (in which policy responds to both employment/real GDP and inflation) already has some similarities with a nominal GDP target (in which policy responds to the product of real GDP and the price level). The key differences are that (1) an announced nominal GDP target is much simpler and therefore more powerful than the hazier dual mandate, which is interpreted differently by different people; and more importantly, (2) the dual mandate is defined in terms of rate of change of prices, while a nominal GDP target depends on the level of prices. (There is not such a clear distinction with respect to the employment/real GDP component, which is typically understood to refer to levels in the dual mandate definition as well.) The implication is that a nominal GDP target, Fed officials attempt to "make up" for past undershooting of inflation via future overshooting. In other words, a move to a nominal GDP target is tantamount to a temporary increase in the inflation target. We believe that the skepticism expressed by Chairman Bernanke's in his 2010 Jackson Hole speech still applies, and do not expect a move to either a higher inflation target or a nominal GDP target.