Goldman Keeps Its NFP/Unemployment Estimates Unchanged: -25,000 And 10.1%, Says This Is A U- Not A V-Shaped Recession
Goldman is known for changing its estimates within 24 hours of an NFP number. Today, there is no change, and it stays at -25,000, coupled with an estimation of the unemployment rate at 10.1%. Additional, some bearish observations on the US economy via Goldman uber economist Jan Hatzius, who is now convinced this is a U- and not a V-shaped recession, follow.
In the V-shaped cycles before 1990, the end of the recession was followed by monetary policy tightening within half a year or less. In the U-shaped cycles since then, it took 2½ to 3 years. Our expectation is that the current cycle will resemble the post-1990 cycles, as we expect Fed officials to keep short-term interest rates near zero in 2010 and, more likely than not, in 2011 as well.
Monetary Policy Implications of U vs. V
Last week’s US Economics Analyst argued that the recovery from the 2007-2009 recession has so far looked much more like the U-shaped recoveries following the 1990-1991 and 2001 recessions than the V-shaped ones following prior postwar downturns. This is particularly true as far as the labor market is concerned; indeed, if anything the current recovery has been even more “jobless” than the two prior ones, despite the fact that it followed a much deeper downturn. (We will get the January employment report on Friday at 8.30 Eastern Time; our estimate remains a 25,000 drop in nonfarm payrolls and a 10.1% unemployment rate, though the uncertainty is even larger than normal as we described in Tuesday’s daily comment.)
Today’s comment looks at monetary policy in U-shaped and V-shaped cycles. The bald historical facts are in the chart below. It shows the time lag between the end of the recession as defined by the Business Cycle Dating Committee of the National Bureau of Economic Research (NBER) and the first increase in the federal funds rate, broken down into separate lags between (1) the recession end and the peak in the (3-month moving average of) the unemployment rate and (2) the unemployment peak and the first rate hike.
The contrast between the two types of cycles could hardly be greater. In the pre-1990s cycles, each of the two lags averaged just a few months, for a total lag between the recession end and the first hike of less than half a year in every single case. In contrast, in the last two cycles, each lag has averaged more than a year, for a total lag of 2½ to 3 years.
Our labor market and monetary policy forecasts essentially envisage another 1991 or 2001-style cycle. We expect the unemployment rate to peak in the first half of 2011, 1½ to 2 years after the end of the recession. Moreover, we see no rate hikes until the end of 2011. If short-term rates rise in the first half of 2012, that would imply a total lag of 2½ to 3 years between the end of the recession and the first rate hike, very much in line with the experience of the past two cycles.
Of course, there are a number of risks to the view that this monetary policy cycle will resemble the last two. On the side of an earlier hike, we can see three main ones:
1. As discussed more fully in last Friday’s piece, financial conditions have eased more sharply (after a considerably bigger tightening) in this cycle than in the last two. This could translate into a more vigorous recovery in final demand and ultimately GDP than in prior cycles. There are some signs that this is happening in core areas of domestic demand such as personal consumption and capital spending, although they are quite tentative.
2. The level of the funds rate is lower. At present, the target funds rate is in a 0% to ¼% range, compared with a trough level of 3% in the early 1990s and 1% in the early 2000s. With core inflation at roughly similar levels as in the early 2000s and about 1 percentage point lower than in the early 1990s, this implies that the current funds rate is lower not only in nominal but also real terms.
3. A number of Fed officials worry greatly about the risks to inflation, inflation expectations, or another asset price “bubble” if the funds rate stays near zero for an extended period, despite the low current level of inflation and the large output gap. If this view gains greater currency within the committee, perhaps in the wake of an updrift in inflation expectations or a sharp recovery in risky asset prices, the FOMC might decide to hike sooner. We do not expect such an outcome, but we certainly cannot rule it out.
However, there are also risks on the side of an even later hike than in the prior two cycles:
1. The output gap is much larger than in either one of the previous cycles. A larger output gap implies that the Fed should give the economy far more “running room” than in previous cycles. This is the main reason why our version of the “Taylor rule” still points to a deeply negative funds rate.
2. So far, we can attribute most if not all of the growth in final demand since mid-2009 to the expansionary impact of fiscal policy. We therefore need a large acceleration in underlying final demand to offset the waning of the fiscal impulse. We expect such an acceleration, but it is not a foregone conclusion.
3. Unlike in previous cycles, a tightening of monetary policy need not involve a rise in short-term interest rates, at least not immediately. The Fed’s asset purchases are slated to end later this quarter, which could lead to a tightening of financial conditions to the extent that the Fed’s ongoing flow of purchases have held down mortgage rates and long-term interest rates more broadly. Moreover, there is a debate over whether Fed officials should ultimately sell assets outright to shrink the Fed balance sheet, perhaps even in advance of the first funds rate hike. We neither recommend nor expect this, but we also cannot rule it out. If it did happen, this would likely tighten financial conditions significantly and thereby postpone the first hike in short-term rates.