Global: The Recession of 2007
Joachim Fels (London)
Economists don’t like the “R” word…
Economists are usually reluctant to predict recessions before they have actually occurred. One reason is that recessions are rare events: they don’t happen very often and don’t last very long. The United States’ economy has been in recession only nine times in the last 60 years, or roughly once every seven years. Before the last recession in 2001, the economy even expanded for a full ten years. And the average recession has only lasted for about four quarters. Thus, at any point in time the chances are high that a doomsayer will be proved wrong.
Another reason is that there is no broad consensus amongst economists on what actually causes recessions. The most frequently cited candidates are monetary and fiscal policy mistakes, oil price shocks, wage cost pushes, sharp exchange rate adjustments, technology shocks and financial crises. But opinions remain divided on which of these factors are the most important and what combination of these factors will tip an economy from prosperity into recession. The safer bet to make for a risk-averse forecaster is thus to predict that the economy will usually grow at around its perceived trend rate, and sometimes a bit above or below that.
A telling illustration of forecasters’ reluctance to utter the “R” word is the post-mortem of professional forecasters’ performance in predicting the 2001 recession conducted by James Stock and Mark Watson, who used the forecasts compiled each quarter for the Philadelphia Fed’s Survey of Professional Forecasters (SPF). Stock and Watson find that as late as the fourth quarter of 2000, when US industrial production was already declining, the median SPF forecast was predicting strong economic growth throughout 2001. Moreover, even during 2001 the SPF failed to forecast the sharp declines in GDP as they were occurring. The authors conclude dryly: “Evidently, this recession was a challenging time for professional forecasters.” (see How Did Leading Indicator Forecasts Perform During the 2001 Recession? Federal Reserve Bank of Richmond Economic Quarterly, Summer 2003)
…but the bond market is toying with the idea
Financial markets are definitely less shy than economists about forecasting recessions, as illustrated by Nobel laureate Paul Samuelson’s famous quip: “The stock market has predicted nine out of the last five recessions”. A more reliable indicator for impending recessions than the stock market appears to be the bond market or, more precisely, the slope of the yield curve. A series of empirical studies suggest that the spread between long-term and short-term interest rates is a good predictor of future economic activity. Typically, an inversion of the yield curve has been followed by recession about a year later. Arturo Estrella from the New York Fed, who authored or co-authored many of these studies during the 1990s, notes that this empirical relationship appears to be robust over time and across different countries, with particularly strong results obtained using data for economies in Europe and North America. (A. Estrella, “Why Does the Yield Curve Predict Output and Inflation? The Economic Journal, July 2005, pp.722-755).
Against this backdrop, I take the dramatic flattening of the US yield curve this year seriously. Two-year-Treasury yields are currently trading only 8 basis point below 10-year yields, so it doesn’t take much from here for the yield spread to slip into negative territory. Based on past experience, there is now a significant risk that the US economy will slow sharply or even fall into recession in late 2006 or early 2007. In fact, this is precisely what the Eurodollar futures curve seems to be suggesting, too. While the futures markets suggest that investors are betting on another two or three Fed rate hikes between now and the middle of next year, the same market sees the implied Fed funds rate easing slightly (by 10 basis points or so) between June 2006 and June 2007. Thus, investors appear to be attaching a significant probability to an economic slowdown or a recession over the next 6-12 months, leading to a policy reversal at the Fed.
Taking the bond market seriously
Given professional forecasters’ apparent inability or reluctance to forecast recessions, and given the decent track record of the yield curve in forecasting recessions, the warning signals from the relentless flattening of the curve and the inversion in the June 2006 to June 2007 Eurodollar futures contracts should not be dismissed easily. A recession in 2007 is definitely very high on my radar screen.
Of course, there is no magic in the link between the yield curve and the business cycle. Fluctuations in the yield curve tend to be largely driven by the ups and downs in short rates, which are controlled by the central bank. Sharp cuts in interest rates usually lead to a steeper yield curve slope, while a monetary tightening would lead to a flatter curve. Thus, the established fact that inverted yield curves predict recessions is just another manifestation of the power that monetary policy wields over the business cycle. In fact, a recent high-calibre econometric study suggests that the short rate alone has actually been an even better predictor of GDP growth than the yield spread since the early 1990s (A. Ang, M. Piazzesi, M. Wei. “What Does the Yield Curve Tell us About GDP Growth?” Journal of Econometrics, 2005 forthcoming). Well, the Fed has now raised short rates by 300 basis points and is widely expected to add at least another 50 basis points to the bill over the next several months. In my view, it is first and foremost the combination of monetary tightening and a highly leveraged and asset-dependent economy that raises the spectre of a sharp economic slowdown or even a recession in 2007.
Of course, there are weighty counterarguments to this punchline. One is that despite 300 basis points of rate hikes, nominal and real short rates are still low in historical comparison. While that is certainly true, I would argue that the change in short rates matters as well, not only the level, and that the US economy must have become more sensitive to interest rate changes due to the massive expansion in household sector balance sheets. Easy monetary conditions have made the economy addicted to debt and rising asset prices. As for a heavily addicted junkie, a significant reduction of the dope should already be enough to cause cold turkey for the economy.
Another counterargument is that while short rates have gone up, asset prices in general, which are the main transmission mechanism from short rates to the economy, have held up well or have even gone up further. My response: watch out for the cracks to appear in high yield, equities and real estate next year for a sign that the transmission mechanism is working. Naturally, higher short rates and a flatter or inverted yield curve are only the first step in a series of events that will lead to a sharp slowdown or recession. The logical next step would be for high yield spreads to widen, house price increases to slow (early signs of which are starting to appear in the coastal regions) and, perhaps, equities to tank at some stage next year, which in turn would set in motion a financial accelerator leading to a significant slowing of economic growth.
Risk of 2007 recession also rising in Europe
As in the US, I believe the risk of a 2007 recession in Europe has risen recently, even though it is not as high (yet) as in the US. The recession risk is lower so far because the yield curve remains comfortably in positive territory. Note that an ECB working paper shows that the yield spread is a useful recession predictor in the euro area, too (F. Moneta, “Does the Yield Spread Predict Recessions in the Euro Area?” ECB Working Paper No. 294, December 2003). However, a potentially more aggressive monetary tightening by the ECB than the market currently foresees could flatten the yield curve further in the course of next year. Moreover, the new German government coalition is planning a sharp fiscal tightening consisting mainly of tax increases for 2007. (For details, see my colleague Elga Bartsch’s note Recovering on Borrowed Time, 15 November 2005). In Italy, where an election is scheduled for next spring, a new government might also slam of the fiscal brakes in 2007. Fiscal tightening in two large euro area countries accounting for about half of euro area GDP, combined with the lagged effects of monetary tightening during 2006, which could slow house price inflation, might well result in a sharp economic slowdown or even a European recession in 2007. Needless to say, this would be even more likely if a US recession came to pass. Thus, what I’ve described as type-II stagflation — a combination of relative economic stagnation in the euro area economy coupled with asset price inflation — could morph into an even more nasty combination of recession and asset price deflation — not too dissimilar from the Japanese experience during the 1990s.
Bottomline: New lows for bond yields?
The warning signs to watch out for on the road to possible a recession in 2007 are: (1) an inversion of the US yield curve and a further flattening of the euro area yield curve in the next several months; (2) a widening of high yield spreads; (3) a significant slowing of house price inflation; (4) a decline in equity prices. The road towards recession will be bumpy though. I wouldn’t be surprised to see temporarily stronger economic growth in both the US and the euro area as we go into the new year, which would lead to a further sell of in bond markets and more monetary tightening in response. But as night follows day, the latter would eventually bring on the slowdown or recession. I still see bond yields trending higher as we move into 2006. But if the recession of 2007 comes to pass, US and European bond yields could make new lows eventually. And Dr. Bernanke might actually have to climb on board of his proverbial helicopter in 2007 to flood the economy with paper money. In the meantime, all eyes on the yield curve!
morganstanley.com |
