1. The headlines for last week’s first-tier US economic data releases were above expectations. And at least the employment report was genuinely better, with faster than expected private payrolls growth, significant upward revisions to past months, a 0.3% wage gain, and a firm household survey. The trends are still consistent with the gradual slowing in employment and wage income growth implied by our forecast, so the news is far from good in any absolute sense, but at least for now the deterioration in the labor market no longer looks as precipitous as it did after the last report. The somewhat better claims data of the past two weeks send a similar message.
2. The ISM picture was much less good, however. Although the manufacturing composite edged up from 55.5 to 56.3, the new orders/inventories gap fell again and taken by itself now points to a composite of clearly below 50 in a few months. Moreover, the nonmanufacturing composite dropped sharply, led by new orders, and the all-industry composite showed its largest month-to-month decline since November 2008.
3. The numbers over the next few weeks are likely to look decent. That’s partly because of the direct implications of the employment numbers for industrial production and personal income, partly because the housing indicators are likely to bounce from their extremely depressed current levels, and partly because the bottom-up indications for retail sales—the most important release in the next few weeks—are reasonably firm.
4. Overall the news is sufficiently mixed to make a big “QE2” announcement at the September 21 FOMC meeting unlikely. Although we suspect that the FOMC and the Fed staff will revise down their growth forecasts once more, the size of the revision is unlikely to be large enough to qualify as the “significant weakening of the outlook” identified by Chairman Bernanke as one key trigger for additional easing in his Jackson Hole speech. (The other was a meaningful drop in inflation and/or inflation expectations.)
5. Later this year or early next, however, we do expect a return to unconventional monetary easing. This is because we strongly disagree with the notion that the recent slowdown in activity is a temporary “soft patch” in an otherwise fairly decent recovery, which seems to underlie the Fed’s forecast of a reacceleration in 2011 after a modestly slower period in 2010H2. On the contrary, we believe that the stronger growth of late 2009/early 2010 was a temporary “firm patch” in an otherwise extremely anemic recovery, and there is a sizable (25%-30%) risk of a renewed recession. As this becomes clear, Fed officials are likely to act.
6. The most likely policy shift involves purchases of US Treasuries, although changes in the forward-looking language are also a possibility. Ultimately, any new purchases are likely to total at least $1 trillion, but today’s NYT interview with outgoing Vice Chairman Kohn suggests that Fed officials may only announce a smaller amount upfront and then adjust their plans in response to new information (seehttp://www.nytimes.com/2010/09/06/business/economy/06fed.html?_r=1&ref=business). The advantage of such a policy is that it may be an easier “sell” to skeptical officials, although the risk is that the markets will view it as half-hearted.
7. How effective is a return to QE likely to be? The uncertainties are enormous, butJari Stehn’s analysis of the first round of QE in late 2008/early 2009 concluded that it pushed down 10-year Treasury yields by 25bp and eased the GSFCI by 80bp per $1 trillion in purchases. We suspect the next round would be less effective in terms of easing financial conditions, not because of a smaller impact on riskless long rates but because there is much less room for spread compression in the credit markets. So a 50-60bp easing in the GSFCI (relative to what would happen without QE) may be a more realistic expectation. Based on historical linkages, this is worth about ½ percentage point on growth, or a bit less given that the mortgage refinancing channel of transmission is clogged by the large number of households in negative equity. If this is the right order of magnitude, a $1 trillion purchase would not have a dramatic effect on growth, but would not be insignificant either. Of course, Fed officials could buy more and/or supplement the purchases with changes in the Fed statement to reinforce the effect.
8. In the runup to QE2, communications will remain a challenge for the Fed. The problem is twofold. First, the FOMC is far from united, and participants (especially regional bank presidents) who are skeptical of the need for further action will continue to make their views known. This causes confusion in the markets, even if it ultimately has little bearing on the outcome. Second, the leadership wants to signal that more easing is on the table without talking too pessimistically, for fear of “scaring” those market participants who believe that the Fed has a privileged perspective on the fundamental outlook for the economy. The results are sometimes a bit odd—for example, the statement in the August 10 minutes that “…no member saw an appreciable risk of deflation…” which came just a few weeks after one member (President Bullard) had presented an analysis that strongly implied just such a risk. Given the range of different opinions and the conflicting objectives, the risk of further communication hiccups and resulting bouts of bond market volatility is high.
2. The ISM picture was much less good, however. Although the manufacturing composite edged up from 55.5 to 56.3, the new orders/inventories gap fell again and taken by itself now points to a composite of clearly below 50 in a few months. Moreover, the nonmanufacturing composite dropped sharply, led by new orders, and the all-industry composite showed its largest month-to-month decline since November 2008.
3. The numbers over the next few weeks are likely to look decent. That’s partly because of the direct implications of the employment numbers for industrial production and personal income, partly because the housing indicators are likely to bounce from their extremely depressed current levels, and partly because the bottom-up indications for retail sales—the most important release in the next few weeks—are reasonably firm.
4. Overall the news is sufficiently mixed to make a big “QE2” announcement at the September 21 FOMC meeting unlikely. Although we suspect that the FOMC and the Fed staff will revise down their growth forecasts once more, the size of the revision is unlikely to be large enough to qualify as the “significant weakening of the outlook” identified by Chairman Bernanke as one key trigger for additional easing in his Jackson Hole speech. (The other was a meaningful drop in inflation and/or inflation expectations.)
5. Later this year or early next, however, we do expect a return to unconventional monetary easing. This is because we strongly disagree with the notion that the recent slowdown in activity is a temporary “soft patch” in an otherwise fairly decent recovery, which seems to underlie the Fed’s forecast of a reacceleration in 2011 after a modestly slower period in 2010H2. On the contrary, we believe that the stronger growth of late 2009/early 2010 was a temporary “firm patch” in an otherwise extremely anemic recovery, and there is a sizable (25%-30%) risk of a renewed recession. As this becomes clear, Fed officials are likely to act.
6. The most likely policy shift involves purchases of US Treasuries, although changes in the forward-looking language are also a possibility. Ultimately, any new purchases are likely to total at least $1 trillion, but today’s NYT interview with outgoing Vice Chairman Kohn suggests that Fed officials may only announce a smaller amount upfront and then adjust their plans in response to new information (seehttp://www.nytimes.com/2010/09/06/business/economy/06fed.html?_r=1&ref=business). The advantage of such a policy is that it may be an easier “sell” to skeptical officials, although the risk is that the markets will view it as half-hearted.
7. How effective is a return to QE likely to be? The uncertainties are enormous, butJari Stehn’s analysis of the first round of QE in late 2008/early 2009 concluded that it pushed down 10-year Treasury yields by 25bp and eased the GSFCI by 80bp per $1 trillion in purchases. We suspect the next round would be less effective in terms of easing financial conditions, not because of a smaller impact on riskless long rates but because there is much less room for spread compression in the credit markets. So a 50-60bp easing in the GSFCI (relative to what would happen without QE) may be a more realistic expectation. Based on historical linkages, this is worth about ½ percentage point on growth, or a bit less given that the mortgage refinancing channel of transmission is clogged by the large number of households in negative equity. If this is the right order of magnitude, a $1 trillion purchase would not have a dramatic effect on growth, but would not be insignificant either. Of course, Fed officials could buy more and/or supplement the purchases with changes in the Fed statement to reinforce the effect.
8. In the runup to QE2, communications will remain a challenge for the Fed. The problem is twofold. First, the FOMC is far from united, and participants (especially regional bank presidents) who are skeptical of the need for further action will continue to make their views known. This causes confusion in the markets, even if it ultimately has little bearing on the outcome. Second, the leadership wants to signal that more easing is on the table without talking too pessimistically, for fear of “scaring” those market participants who believe that the Fed has a privileged perspective on the fundamental outlook for the economy. The results are sometimes a bit odd—for example, the statement in the August 10 minutes that “…no member saw an appreciable risk of deflation…” which came just a few weeks after one member (President Bullard) had presented an analysis that strongly implied just such a risk. Given the range of different opinions and the conflicting objectives, the risk of further communication hiccups and resulting bouts of bond market volatility is high.
No comments:
Post a Comment