Why Goldman Is Surprised By The Market's Reaction To The Twist, And What's Next For The Fed?
After spending the last few weeks 'helping' the Fed with its agenda, Goldman Sachs' Andrew Tilton seems a little disappointed by the market's reaction - reasoning that the FX and equity-investing plebeians will take longer to comprehend the less familiar 'twist' operation that has already been wholly discounted into the TSY curve. While he did not get all he wanted from this meeting (even though the 'twist' was larger than expected), Hilton wastes no time in looking to the future and the chance of further economic weakness leading to more dramatic Fed actions.
From Goldman Sachs US Daily : Fed Does the Twist, Though Not Everyone Dances
- This afternoon the Federal Open Market Committee (FOMC) announced plans to sell short-term securities on its balance sheet and buy longer-term securities, "doing the twist" in market parlance. We estimate that implementation--to be completed by mid-2012--will increase the total duration of securities on the Fed's balance sheet by nearly $400 billion ten-year equivalents. The Fed also indicated it will shift the reinvestment of maturing and prepaid agency debt and agency mortgage backed securities from Treasuries into agency MBS, providing incrementally more support for the housing sector. Overall, the easing action was more aggressive than our expectations or the market's.
- Still, not everyone danced along. Several Congressional Republican leaders wrote a letter to Chairman Bernanke at the outset of the meeting asking the FOMC to refrain from more stimulus. Three members of the FOMC--regional Fed presidents Fisher, Plosser, and Kocherlakota--dissented from the decision. And the market reaction was mixed, with the yield curve flattening as anticipated but equity prices down sharply, the dollar stronger, and overall financial conditions tighter on the day.
The justification for this action was, of course, the weak economic outlook. The statement emphasized the weak state of the economy, suggesting "continuing weakness in overall labor market conditions" and "only a modest pace" of growth in consumer spending. At the same time, inflation remained a secondary concern, with the statement noting the moderation in (headline) inflation in recent months and reiterating the expectation that inflation will "settle...at levels at or below those consistent with the Committee's dual mandate". While the FOMC still forecasts some improvement in the pace of growth over the coming year,there are significant downside risks to the economic outlook, including strains in global financial markets". The statement retained an easing bias, noting again that the FOMC "is prepared to employ its tools" to "promote a stronger economic recovery in a context of price stability".
Although the broad thrust of the statement and action were consistent with our expectations, the overall easing move was larger than we anticipated, for several reasons:
A bigger "twist" than most had expected. In previous commentary we had indicated an expectation that the Fed would sell perhaps $300bn in shorter-term securities maturing within 2-3 years (see yesterday's US Daily, "FOMC Preview"). The $400bn announcement was bigger than our expectations, and probably the market's as well.
A slightly higher share of purchases at the long end. Detail available on the New York Fed website (www.newyorkfed.org/markets/opolicy/operating_policy_110921.html) indicates that 29% of the purchases are expected to be nominal Treasuries of 20-30 years' maturity, probably a bit more than markets had been expecting. Together with the bigger nominal size of the twist, our calculations suggest the Fed will add nearly $400bn in ten-year equivalents to its balance sheet; our conversations with clients implied a market expectation consistent with our own view, in the range of $300-$350bn.
Reinvestment in MBS. The FOMC indicated that as agency debt and agency mortgage-backed securities mature, it will now reinvest these proceeds in agency MBS rather than Treasury debt. This implies flat rather than declining holdings of agency securities, and therefore incrementally more support for the housing market than previously. (Recall that some FOMC participants had objected to--and presumably continue to object to--purchase of agency rather than Treasury securities on the grounds that these interfered with credit allocation.)
The Fed made no change to its 25bp rate on excess reserve holdings; we had seen a slightly-better-than-even chance that this would be cut to around 10bp, in part as a signaling device and in part to help tamp down any effect of short-term Treasury sales on the front end of the yield curve. In the end, the FOMC apparently decided that the costs we identified--potential interference with the normal functioning of the federal funds market, unwanted interactions with deposit insurance fees, and effects on money market mutual funds--outweighed the modest benefits. (For more details, see “Revisiting the Rate on Reserves", US Daily, September 13, 2011.) The lack of a change in IOER also could be viewed as a sign that the FOMC is confident its conditional rate commitment is sufficiently credible to keep short-term rates low without other actions.
About a month ago, we explored the potential impact of a "twist" on interest rates and the broader economy (see "For More Easing, Will the Fed Go Big or Go Long?, US Daily, August 15). In our view, Fed asset purchases operate by reducing the supply of duration in the bond market, increasing the equilibrium price (and reducing the equilibrium yield) for medium- and longer-term securities. Thus, to provide stimulus via its balance sheet, the Fed can either "go big"--expand the balance sheet by purchasing more securities--or "go long"--increase the average duration of the securities it holds.
The tradeoff between these two options is illustrated in the chart below, reproduced from the August 15 US Daily. It shows the average duration of the Fed’s balance sheet on the horizontal axis and the total size on the vertical axis. Prior to the financial crisis the Fed's balance sheet had a total size of approximately $900bn and an average duration of between two and three years. Two rounds of asset purchases expanded the balance sheet to its current size of $2.6trn and also increased its average duration to about 4 1/2 years. Our past work on this topic suggests that in total, and holding growth and inflation expectations constant, this balance sheet expansion lowered ten-year Treasury yields by about 50 basis points. If the Fed wanted to lower ten-year yields another 25bp, it could choose a variety of options; two possibilities would be expanding the balance sheet while holding its average duration constant ("going big", point A) or holding the balance sheet constant while increasing its average duration ("going long", point B).
When fully implemented, the "twist" announced today will take the Fed almost all the way to point B. The Fed expects the average duration of the Treasury portfolio to increase to around 100 months, which would imply an average duration for the entire balance sheet of nearly six years (if the duration of the agency securities in the portfolio did not change on average). Given our previous work, this implies an impact on ten-year Treasury yields comparable to QE2, in the range of 15-30bp.
Although the Fed took a bold step with the twist, not everyone danced along. News media reported earlier today that several senior Congressional Republicans had written a letter to Chairman Bernanke at the outset of the meeting asking the FOMC to refrain from more stimulus. And once again, three FOMC members--Dallas Fed President Fisher, Minneapolis Fed President Kocherlakota, and Philadelphia Fed President Plosser--dissented, with the statement noting only that they "did not support additional policy accommodation at this time".
The market reaction was also mixed, as Exhibit 2 indicates. The yield curve did indeed twist--with the difference between ten-year and two-year Treasury yields flattening 11bp to 166bp. But other asset prices responded in ways atypical for monetary easing: the dollar rallied, equities slumped, and commodity prices fell. On net, our GS Financial Conditions Index actually tightened on the day, certainly not the reaction Fed officials would have been hoping for. Given the market most focused on the implications of the "twist" did react in the way we expected, we are surprised at the behavior of other markets; one possibility is that the unconventional move will take more time to digest in markets less familiar with its likely method of action. Indeed, we found in prior work that the equity and foreign exchange markets seemed to lag behind the fixed income market in pricing in asset purchases (see "QE2: How Much Has Been Priced In?", US Daily, October 7, 2010).
What's next for the Fed? Probably a continued easing bias, but without further actions in the near term unless the economy weakens further. After all, the Fed has just announced major easing steps focusing on the front end of the yield curve (its August 9 conditional commitment to keep the funds rate exceptionally low through mid-2013) and the back end (today's "twist"). The Fed plans to implement the "twist" through mid-2012, so it will likely take substantial further deterioration in the outlook to change course before then. While this clearly remains a risk, it is neither the Fed's base case nor ours.
If the FOMC chooses to do more, the most likely easing options would seem to be stronger versions of those just announced. In the case of the front end, the Fed could extend its rate commitment or tie it (per the proposal by Chicago Fed President Evans) to specific targets for unemployment and inflation. To ease longer-term rates further, the Fed would likely return to outright balance sheet expansion.
And for good measure here is the continuing saga of pricing in the TSY curve 'Twist' as 30Y is now -27bps from Tuesday's close!!
So it seems that we are all but fully discounting the impact on TSY markets and the slower growth outlook and non-expansionary balance sheet effects seem unlikely to cause significant portfolio rebalancing effects given no LSAP this time - or maybe as Andrew notes, we are slow on the uptake.