Goldman: The Ball Is Now In The Fed's Court
What Zero Hedge has been saying for well over half a year has finally hit the mainstream, with pundit after pundit "suddenly" coming out of the closet and making the uber-bold proclamation that "QE3 is here." Yawn. That said, since Goldman's opinion is the only one that matters (see previous posts on this matter, especially those referencing the activities of one Bill Dudley at one "Pound and Pence"), here is Jan Hatzius explaining how the whole world now looks up to Bernanke to pick up the QE torch lit up in the past week by the SNB, the BOJ and the ECB, and take Central PlanningTM to escape velocity (which may well be needed if we hope to get Mars to bail out the Earth shortly). Specifically, when discussing what the Fed will announce on Tuesday, naturally follows Monday, or the day in which risk comes home to roost, Hatzius says the following: "First, we expect them to expand the scope of their “extended period” language to cover not just the exceptionally low funds rate but also the exceptionally large balance sheet. For example, they could rewrite the current forward-looking language in the statement to say that economic conditions “…are likely to warrant exceptionally low levels for the federal funds rate and exceptionally large asset holdings for an extended period” (our suggested change in italics). Indeed, our baseline expectation is that this change will occur at the August 9 FOMC meeting, although it is a relatively close call. Second, we expect the composition of the Fed’s balance sheet to shift toward longer maturities. This could happen via an increase in the average maturity of its reinvestment of MBS paydowns and/or a change in the reinvestment policy for its Treasury portfolio. However, we do not yet expect this for the August 9 meeting, although it is possible." Operation Twist 2 it is then, with unlimited purchases in the 2-7 year range to keep the yield at a sturdy 0%, and the 2s10s to surge record highs (alas, QE3 means inflation, inflation, inflation down the line) in a last ditch attempt to bailout America's financial system, which unfortunately has just entered wind-down mode.
From Goldman Sachs:
Subpar Growth Brings the Fed Back into Play
The headlines from the July employment report beat market expectations, but the US economy is nevertheless struggling. Earlier today, we downgraded our growth and employment forecasts and shifted to a view of renewed (limited) moves in the direction of monetary easing by the Federal Reserve—more likely than not at next week’s meeting. Today’s article discusses our forecast changes in greater detail.
Disappointment in the First Half of 2011…
Following last Friday’s GDP release, the current cycle now has the sorry distinction of featuring both the largest decline and the smallest recovery in real GDP of any postwar business cycle. Over the eight quarters since the 2009 Q2 trough, real GDP has grown just 2.5% (annualized), about ½ point below the pace in the first eight quarters of either the 1991-1993 or 2001-2003 recovery. All of the underperformance relative to the prior two recoveries has occurred in the first half of 2011, when real GDP grew just 0.8% (annualized).
A key question for gauging the growth outlook in the second half of 2011 and in 2012 is how much of this weakness was due to temporary shocks and how much of it reflects the economy’s underlying fragility. To answer this question, we have taken another stab at estimating the impact of three identifiable shocks that have hit the economy in early 2011, namely the supply chain disruptions associated with the Japanese earthquake, the sharp increase in energy prices, and the turn to fiscal tightening:
1. Japan-related supply chain disruptions.
In past research, we estimated that the supply chain disruptions associated with the Japanese earthquake would shave about ½ percentage point from real GDP growth in the second quarter, or ¼ point from the first half. This estimate was based on an assumption that the shortage of key components would keep production across the motor vehicle supply chain about 7% (not annualized) below its counterfactual baseline level in the second quarter. While the Q2 industrial production numbers are roughly consistent with this estimate, motor vehicle production in the Q2 GDP accounts—a series that is estimated more indirectly from data on final sales, net exports, and inventories—showed no meaningful decline in the second quarter. This does not mean that there was no effect; after all, vehicle production seemed to be on track for a healthy gain at the start of the second quarter. Nevertheless, we would now view the ¼-point figure as an upper bound for the true effect.
2. Higher energy prices.
We have written a lot about the effects of the sharp increase in energy prices since the end of 2010. As shown in Exhibit 1, our analysis implies that the 30% increase in crude oil prices from December 2010 to April 2011 should take off ¾ percentage point (annualized) from the average growth pace through early 2013, with a peak effect of about 1½ percentage point in early 2012.
However, we believe that the effect may have been more front-loaded than suggested by Exhibit 2. The reason is that our econometric estimates are based on the economy’s average behavior over the past 25 years. For much of that period, US households probably had a greater ability to spread the adjustment to an adverse shock to their real income over time than they do now. If they are now less able to spread the adjustment over time, this would imply that the impact of an oil shock on consumption and GDP growth should be more front-loaded now than implied by our econometric results.
A more basic look at the recent consumption and income data also supports the idea of a more front-loaded impact. In the first half of 2011, the increase in retail energy prices has subtracted 1½ percentage points from real disposable income growth. We can calculate how much of the real income hit passed through into real consumer spending by looking at the change in the personal saving rate. If saving had declined, this would imply that households “absorbed” some of the real income loss via lower saving. In fact, however, the saving rate was basically unchanged in the first half of 2011, suggesting that the 1½-point hit to real income growth fully translated into a 1½-point hit to real consumption growth. This suggests a hit to real GDP growth of about 1 percentage point in the first half of 2011, a little more than suggested by our econometric results in Exhibit 1. However, it suggests less drag going forward.
3. Tighter fiscal policy.
Finally, fiscal policy seems to have exerted a much sharper drag on GDP growth than we had expected. Following the passage of the December 2010 bipartisan fiscal package, we thought that fiscal policy at the federal, state, and local level would be roughly neutral in the first half of 2011. However, based on the actual spending and tax data now in hand, we estimate that fiscal policy subtracted nearly 1 percentage point from growth (see Exhibit 2).
Some of this was because spending in both the federal and the state and local sector contracted more sharply than we expected. But some of it is because there is surprisingly little evidence for any offsetting boost from lower taxes as the 2% payroll tax cut seems to have been fully offset by the loss of the “Making Work Pay” tax cut at the end of 2010, higher final tax settlements, and tax increases at the state and local level. The net result is that there was essentially no change in the average effective income and payroll tax rate on personal income in early 2011, as shown in Exhibit 3.
…And a Subdued Outlook in the Second Half
Exhibit 4 summarizes the preceding discussion. The line labeled “2011H1” starts from the Commerce Department’s current estimate that real GDP grew 0.8% (annualized) in the first half. It then subtracts the estimated impact of the Japanese earthquake (-0.25%), the energy shock (-1%), and fiscal policy (-0.9%) to arrive at an estimate of the “underlying” growth pace excluding these factors of 3%. This is clearly better than the very weak “headline” GDP growth in the first half.
But better does not mean good. Suppose that the underlying growth pace stays at 3% in the second half of 2011, and adjust for the three factors discussed above. We expect a ¼-point boost from the unwinding of the Japanese shock, but still see a drag of ½ point from higher oil prices and another ¾ point from tighter fiscal policy. This would point to a real GDP growth rate in the second half of 2011 of only about 2%, i.e. still clearly below the US economy’s longer-term trend.
A growth pace of 2% in the second half would probably put at least a bit of additional upward pressure on the unemployment rate. As shown in Exhibit 5, the link between the growth rate of real GDP and the change in the unemployment rate, known as Okun’s law, continues to be very tight. If growth is ½ percentage point below its 2½% trend rate for half a year, that implies an increase in the unemployment rate by a tenth or two. In such an environment, the risk of a renewed recession is likely to remain elevated.
Going Sideways in 2012…
If the economy “muddles through” the next few quarters without falling back into recession, our baseline expectation is a modest pickup in growth back to a trend rate of 2½% by the spring of 2012. This implies a flat unemployment rate of about 9¼% at the end of 2012—a terrible outcome 3½ years after the official end of the 2007-2009 recession when measured against any standard other than a renewed recession.
What will prompt even such a modest improvement in 2012? Apart from the forces of mean reversion—i.e., the further away the time period, the harder it is to disagree with a forecast of trend growth—the main reason for expecting some improvement is that we still see a decent amount of evidence of private-sector “healing” in several areas.
In the long-suffering housing market, the flow of new housing starts has been running far below underlying demographic demand for the past 2½ years. Although the stock of excess supply is still substantial, we have recently revised down our excess supply estimates in light of newly released data from the decennial US Census. We expect housing starts to rise gradually in coming years, although the recovery will be slower than in past cycles and it is still likely to take several years until the sector is back to normal levels of activity. On the price side, our house price model points to a stabilization in nominal home values in 2012 (although this prediction would change if we saw a renewed sharp deterioration in the labor market).
More broadly, household balance sheets have improved over the past few years. Household debt service as a share of disposable income is now back within the range seen prior to the debt boom of the 2000s; household credit quality has continued to improve; and at least according to the Fed’s Q2 senior loan officers survey banks have been easing their lending standards for consumer loans (the next survey is due after the upcoming FOMC meeting).
…And Near-Term Risks Tilted to Recession
The risks around our baseline forecast are tilted toward a renewed recession. At this point, we believe that the probability of a renewed downturn is about one in three. Most of the risk is concentrated in the next 6-9 months.
We are concerned about three specific issues. First, a worsening of the European financial crisis, and a failure of European policymakers to respond adequately, could lead to a further tightening of financial conditions and credit availability not just in Europe but also elsewhere including the United States. Specifically, we will probably need to see the European Central Bank step in to buy Italian and Spanish government bonds in substantial size to stabilize markets.
Second, fiscal policy might tighten more sharply than we presently expect. Going back to Exhibit 2, even our baseline forecast implies fiscal restraint of about 1% of GDP in 2012. The reason is that we expect expiration of the extended unemployment benefits, some discretionary spending cuts in the wake of this week’s debt deal, and some modest further restraint from state and local budgets. This restraint is one important reason why we expect growth to stay relatively slow in 2012. But this 1% of GDP would climb to 1½% of GDP if the payroll tax cut expired as well. Although we expect another extension, the risk to this assumption has increased given the fact that Congress did not include it in this week’s debt deal.
Third, the US economy has historically had a relatively low threshold before deterioration in the labor market has started to feed on itself and culminated in a full-blown recession. Exhibit 6 illustrates this fact. It shows that except in periods of early recovery—when the unemployment rate was still rising from the lagged effects of prior GDP declines—there has never been an increase in the three-month moving average of the unemployment rate by 35 basis points (bp) or more in which the economy was not in recession six months later. As of today’s employment report for July, the unemployment rate had risen 21bp from the most recent low point (although the spot rate was down on the month).
There are some reasons to believe that the 35bp rule might not be quite as iron-clad in the current situation as it appears to have been in prior cycles. For one thing, we still are in a relatively early phase of the business cycle. Firms are still operating with exceptionally lean cost structures and might therefore be slower to respond to soft demand with labor cuts than they have been in past slowdowns. Moreover, cyclical sectors such as housing and vehicle manufacturing are already depressed and unlikely to contract much further. Nevertheless, we are mindful of the historical pattern and would be quite worried by any further meaningful increases in the unemployment rate in coming months.
A Return to Disinflation
Our inflation views have not changed much. We expect core CPI and PCE inflation to peak around 2% by yearend 2011 but then to fall back to 1¼% by yearend 2012. While these figures are only very slightly below our previous forecasts, our conviction has increased that the large-scale underutilization of resources will result in substantial renewed disinflation.
In our view, much of the recent acceleration in core inflation reflects temporary factors. These include “mechanical” pass-through from higher commodity prices into goods and services with high commodity content, such as pet food and airline tickets, as well as the impact of the Japanese supply chain disruptions on auto prices. Meanwhile, wage growth remains only around 2% year-on-year, and unit labor costs have been essentially flat.
Indeed, more statistically based measures of core inflation do show some signs of deceleration in recent months. As shown in Exhibit 7, the Dallas Fed’s “trimmed-mean” personal consumption price index—which eliminates the largest price increases and declines—has slowed from 2.4% (annualized) in April to 1.3% in June. Likewise, related measures such as the trimmed-mean CPI, the median CPI, and the “sticky” CPI—which includes only categories whose frequency of price adjustment is below average—have also slowed in recent months. These series can be noisy, so this is far from conclusive evidence. But there are now at least some signs that underlying inflation is starting to slow again.
Small Further Easing Steps from the Fed
On the monetary policy side, we retain our long-standing call for no hikes until 2013, and it could well be even later. We now also believe that Fed officials will continue reinvesting maturing and prepaid securities until 2013.
A more immediate question is whether Fed officials will restart their quantitative easing program and announce another big increase in the size of their balance sheet. Our answer is “probably not” unless the economy falls back into recession. However, we now think that Fed officials will take two smaller steps in the remainder of 2011.
First, we expect them to expand the scope of their “extended period” language to cover not just the exceptionally low funds rate but also the exceptionally large balance sheet. For example, they could rewrite the current forward-looking language in the statement to say that economic conditions “…are likely to warrant exceptionally low levels for the federal funds rate and exceptionally large asset holdings for an extended period” (our suggested change in italics). Indeed, our baseline expectation is that this change will occur at the August 9 FOMC meeting, although it is a relatively close call.
Second, we expect the composition of the Fed’s balance sheet to shift toward longer maturities. This could happen via an increase in the average maturity of its reinvestment of MBS paydowns and/or a change in the reinvestment policy for its Treasury portfolio. However, we do not yet expect this for the August 9 meeting, although it is possible.
How powerful are these measures likely to be? Probably not very powerful by themselves. Based on New York Fed estimates, pushing back expectations for the start of the decline in the balance sheet by one year—a very generous assumption for the likely impact—would be the equivalent of removing expectations for one 25bp rate hike. The effect of a shift in the reinvestment policy depends on how far out the curve the Fed would move, but is probably also not huge. However, market participants would probably view either of these measures as “another step on the road to QE3” which might make them somewhat more powerful as a signaling device.