Goldman's Team Jan 2010/2011 Economic Roadmap, Key Risks, And Rate Projections
The latest from Team Jan - as always, a little blonde, a little dumped by Vince Vaughn, a little pessimistic, yet with enough overtones to satisfy those traders who have Sell axes. We expect the retort out of Team Jim (O'Neill) with bated breath.
The Federal Funds Rate—How Low for How Long?
*The November employment report was much better than expected, though still weak in absolute terms. However, other indicators released over the past few weeks, including this week’s ISM indexes have been a lot more mixed. We expect real GDP to grow 3% (annualized) in the fourth quarter, slow to a 1½% pace in late 2010, and then gradually reaccelerate in 2011.
*If our economic forecast is broadly correct, the federal funds rate is likely to stay at 0% in 2010 and, more likely than not, in 2011 as well. First, core inflation of 1% or less coupled with unemployment of 10% or more justify very expansionary policies. Second, “risk management” considerations argue for erring on the side of later rate hikes. Third, gradual tightening in fiscal policy and a cessation of Fed asset purchases are likely to substitute for—and hence delay—an increase in the funds rate to some degree.
*There are two main risks to our forecast. First, the economy could grow more quickly than we expect, although the upside surprise would need to be quite large to generate rate hikes in the short to medium term. Second, a very large runup in asset prices could trigger rate hikes before they seem warranted on real-economy grounds. However, asset prices would need to rise a great deal further to give credence to the notion of a renewed bubble, and credit growth would probably need to rebound sharply as well.
*In contrast, we are less concerned about the monetary policy implications of several other popular “bugbears,” including a dollar crisis, a big drop in potential output, or an outbreak of inflation.
The Federal Funds Rate—How Low for How Long?
On Wednesday, we rolled our outlook for the global economy into 2011. From a US perspective, our most controversial forecast is that the Federal Open Market Committee (FOMC) will keep its federal funds rate target in the current 0%-¼% range in 2010 and, more likely than not, in 2011 as well. This forecast is quite far out of consensus. Although a number of forecasters have recently moved toward the view that the Fed will be on hold in 2010, all but 1 of the 43 forecasters that submitted a forecast for the first quarter of 2011 to Blue Chip Financial Indicators showed a rate hike by that time. (At the time, we had not yet issued a 2011 forecast.)
The Rationale for Our Forecast
There are essentially three reasons for why we expect the funds rate to stay very low for very long:
1. Our economic outlook calls for very easy policy.
The most important reason for our call is that our economic outlook calls for very accommodative macroeconomic policies for a long time to come. It is very hard to see an economic justification for tightening policy when inflation is at 1% or less and the unemployment rate at 10% or more, as we expect for 2010-2011.
Indeed, our economic outlook—and even the Fed’s own forecasts—would justify even greater amounts of stimulus. We have long made this case using our estimated version of the Taylor rule. As shown in Exhibit 1, this rule currently points to a funds rate of below -5%. Under our inflation and unemployment forecasts, this number would fall to -8% in 2011; under the FOMC’s more upbeat forecasts, it would rise to -2% but still stay negative throughout the next two years.
Moreover, these results are not driven by our particular version of the Taylor rule, or for that matter the Taylor rule framework. Regarding the specific formulation, Glenn Rudebusch of the San Francisco Fed presented a Taylor rule in May that is virtually identical to ours. Regarding the overall framework, a recent study by John Williams, also of the San Francisco Fed, simulates the effects of various (hypothetical) sub-0% interest rate paths on unemployment and inflation using the Fed’s macroeconometric model. He finds that a policy of keeping the funds rate at -6% throughout 2009-2010 produces much more desirable outcomes, with both inflation and unemployment closer to the target, than the actual policy of keeping the funds rate near the zero lower bound. This is also very consistent with the Taylor rule estimates in Exhibit 2.
2. Risk management suggests erring on the late side.
If we look beyond the central forecast to the tails of the distribution, risk management considerations also argue in favor of tightening policy later rather than earlier. If Fed officials wait too long and inflation rises more than anticipated, they should be able to “catch up” relatively easily because they can tighten financial conditions to an essentially unlimited degree. Thus, the risk of an inflationary disaster scenario is remote unless Fed officials either commit massive errors or are somehow prevented from doing their job properly, neither of which we expect.
In contrast, if they tighten too early and the economy falls back into recession, Fed officials would have a much harder time correcting this error, as the funds rate is already at the zero lower bound. While they would undoubtedly mount a renewed all-out effort in easing credit via unconventional monetary policy measures, the effectiveness of these measures is unclear.
3. “Tightening” doesn’t need to mean rate hikes.
For much of the postwar period, a decision to tighten macroeconomic policy was tantamount to a decision to raise the federal funds rate. This was because the federal funds rate target (rather than the Fed‘s balance sheet) was the sole instrument of monetary policy and because monetary (rather than fiscal) policy was the dominant tool of overall stabilization policy.
But neither is true now. A tightening of the overall policy stance could come through fiscal restraint, an unwinding of the Fed’s balance sheet policies, a hike in the federal funds rate, or a combination of all three. Unless the entire burden is placed on the federal funds rate, this suggests that rate hikes could lag the cycle to a greater degree than in the past.
Indeed, one could argue (and some policymakers do argue) that the most natural sequencing would be to tighten the fiscal stance and the Fed’s balance sheet policies before raising the funds rate. Large-scale fiscal stimulus and Fed asset purchases were only adopted as emergency measures after the funds rate had hit the zero lower bound. Both policies can be thought of as attempts to get closer to the negative funds rate prescription of Exhibit 1. If and when this prescription changes—i.e. if and when the Taylor rate moves back toward zero—the logical first step might be to unwind these fiscal and balance sheet policies before raising the funds rate, especially given the structural imbalance of the federal budget and the markets’ hypersensitivity (misplaced, we believe) about the inflationary implications of a bloated central bank balance sheet.
It would be a mistake to take this “sequencing” argument too far. On the fiscal side, it will be difficult politically to actively tighten policy, i.e., raise taxes or cut spending at a time when the unemployment rate is around 10%. On the balance sheet side, only a minority of Fed officials seem to favor outright asset sales. Others worry, in our view correctly, that it may be difficult to calibrate the effects of such sales on financial conditions and would prefer hikes in the funds rate to precede outright asset sales.
But a softer version of the sequencing argument may still apply. Even if there is no turn to active restraint via tax hikes, spending cuts, or asset sales, it is highly likely that there will be some passive restraint. Exhibit 2 shows that we expect fiscal policy to turn from stimulus by late 2010 or early 2011. Moreover, the asset purchase program is likely to be completed in the first quarter of 2010, and this might lead to a tightening of financial conditions if the “flow” of purchases matters for either the level of long-term yields or the valuation of other assets. The larger these passive restraints, the further away is the point at which Fed officials will need to turn to federal funds rate hikes to in order to fend off inflationary pressures.
We see two main risks to our forecast:
1. A much stronger economic recovery.
It should go without saying that if our forecast for the economy is badly wrong, then we will likely be wrong about monetary policy as well. That said, we believe that the upside surprise would need to be large to trigger rate hikes, at least in the short to medium term. Even under the Fed’s view about growth in 2010, which is just over 1 percentage point higher than ours, we believe that rate hikes would probably still not come before 2011. Real GDP would need to grow substantially above trend—at least 4%, in our view—and this would need to start bringing the unemployment rate down quickly in order to get Fed officials to hike before the end of 2010. Even then, hikes might be delayed if core inflation continues to drift down, as we would expect even in such a buoyant growth environment.
2. Worries about a renewed asset bubble.
Even if the recovery does remain sluggish for a while, a big runup in asset prices could push the FOMC to raise rates before this seems justified from a real-economy perspective. Following their disastrous experience with the “Greenspan doctrine,” Fed officials are likely to become more willing to make a judgment that an asset price boom has gone too far, and use their policy instruments to restrain it.
But while this is a risk, the hurdles against a bubble-induced interest rate hike still seem high. First, an easy monetary policy should to some degree be expected to raise asset valuations, even if markets are completely rational. And since monetary policy works primarily via financial conditions, up to a point an asset market rally is “a feature, not a bug” of the Fed’s efforts to boost economic growth.
Second, at present it seems that asset valuations are in fact roughly in line with the long-term norm, rather than clearly above it. This is shown in Exhibits 3-5. Exhibit 3 plots the valuation of US equities, measured alternatively as Tobin’s q (the ratio of the market value of equity to its replacement cost) and the ratio of equity prices to the 10-year average of reported earnings per share in the S&P 500. Exhibit 4 shows the spread between the yields on Baa corporate bonds and long-term Treasury securities. And Exhibit 5 shows the ratio of home prices to rents, using the Loan Performance house price index. In all three cases, valuations are fairly close to the long-term average.
Third, it is important to remember that the most damaging types of bubbles involve strong credit growth. At present, however, credit growth remains sharply negative. Until that changes, it’s unlikely that Fed officials would get sufficiently worried about a renewed bubble to “clamp down.”
Fourth, even if Fed officials did see a renewed bubble, they have consistently maintained that higher interest rates are a very blunt tool for addressing it. The first choice would probably be “open mouth” policy and/or regulatory measures. Ultimately, however, a higher funds rate might follow if Fed officials did become truly concerned about a renewed bubble and other measures did not show the desired effect.
…and Overblown Worries
In addition to the genuine risks to our forecast, there are several other arguments for why rates might go up soon that seem much less compelling to us. While there is a grain of plausibility to some of them, they seem unlikely to become relevant anytime soon.
1. A dollar crisis.
The story that a sharp depreciation in the dollar will push the Fed into rate hikes is a popular one. This is especially true among observers based in countries outside the United States, which typically are smaller, more open to international trade, and therefore quite naturally more focused on the currency markets.
But we see two arguments against the idea that Fed officials will raise the funds rate to stave off a dollar collapse. First, the currency is less important in the US than elsewhere. While Fed officials certainly take account of its impact on financial conditions, the impact of currency changes on inflation, in particular, is quite minor. A common rule of thumb is that a 10% trade-weighted dollar depreciation raises the level of CPI by just ¼%.
Second, the dollar just isn’t that weak. It has certainly depreciated substantially over the past nine months, but we view this as the flip side of the normalization that has taken place in global financial markets. Indeed, according to the Fed’s broad trade-weighted index, the dollar currently is just 1% weaker than the average of the past two years. It would take a substantial further depreciation to start ringing alarm bells.
2. A big drop in potential output.
If potential output has dropped sharply, we might be overestimating the output gap and underestimating the risk of inflation. What do we make of this argument?
Not much. It is true that potential output is difficult to estimate in real time. Thus, it would be problematic to rely entirely on the ratio of real GDP to its pre-crisis “trend.” But this is not necessary, as there has also been a very large decline in observable measures of resource utilization. The unemployment rate has more than doubled, industrial capacity utilization remains near a postwar low, and home vacancy rates are at or near record levels.
Of course, it is possible that “sustainable” resource utilization has declined, and particularly that the non-accelerating inflation rate of unemployment (NAIRU) has increased, as it did in the 1970s. But even if we accept this and use the Congressional Budget Office’s peak NAIRU estimate for the postwar period—6.3% in 1978—the conclusion that we have a large amount of slack would not change. Moreover, productivity growth has so far been quite rapid coming out of the recession, which does not point to a big potential output hit, at least not yet.
3. Higher inflation or inflation expectations.
Partly because of the previous point, we do not worry much about higher inflation. The output gap is likely to remain massive under virtually any reasonable economic forecast, and this will continue to exert substantial downward pressure on core inflation. We are also not as concerned as many others about a “spontaneous” increase in inflation expectations, or one that is driven by a temporary headline inflation shock. If this happened, it could indeed call for a policy response. But at least for now, the available data on inflation expectations remain quite benign, and we do not expect this to change.
4. A decision to move from “crisis” to “low” rates.
Finally, it is sometimes argued that the FOMC should move away from the current “emergency” rate to a level that is still low (perhaps 1%-2%) but clearly above zero. But we do not understand the rationale for such a move. The Taylor rule and the Willliams study discussed above show quite clearly that the funds rate “should” be below 0%, not above 0%. In other words, the only sense in which policy is on an emergency setting is because it’s addressing an economy that is in desperate need of emergency stimulus.
Moreover, such a move could create significant practical problems. In our view, it is unrealistic to think that the FOMC could commit credibly to a move from near-0% to 1% or 2%. Therefore, the markets would likely assume that this move off “emergency rates” was the start to a much bigger tightening cycle, which carries the risk of a big tightening in financial conditions. This also makes us doubt that FOMC will seriously consider a shift toward higher rates before they are convinced that a significant policy tightening is in fact appropriate.