A variety of fears have held the financial markets down in the past few months (see our note, “Four fears gripping markets”). Renewed concern about an upcoming US recession is one of them. The first activity data available for January – core retail sales up 0.6%, manufacturing up 0.5% and capital goods orders up sharply – suggest there is no short-term danger of a slump, but the preliminary PMI reading for February was disappointing enough to keep the issue very much on investors’ minds. Is the US economy on the brink of a new recession? What warning signs, if any, do we have?
1. A fall in the ISM manufacturing index to below 50 does not equate with recession
On the one hand, it would certainly be a mistake to dismiss the signals sent out by the manufacturing sector on the grounds that it has only limited weight in the US economy. Although manufacturing accounts for just 15% of output, it accounts for nearly 30% of the volatility in US output growth (see Graph 1). That makes it a key sector to observe and underscores the importance of tracking the ISM (Institute for Supply Management) manufacturing index.
On the other hand, history shows that the ISM manufacturing index – and its new orders component in particular – would have to fall to much lower levels to indicate that the odds of a recession are high. Those odds don’t start to become significant until the index breaches the 45 mark (see Graph 2).
In addition to the absolute readings, changes in those readings should be watched closely. But even so, sharp drops in the ISM manufacturing index are not always followed by recession. That’s because the manufacturing sector often has a shorter cycle than the economy as a whole. Several corrections in the ISM index can occur within a single phase of economic expansion (see Graph 3).
Today’s weak manufacturing data can be primarily attributed to two factors:
1.Extensive inventory destocking in a number of sectors.
2.The damping effect of lower oil prices on investment in extractive industries – largely in shale oil (see Graph 4).
In both areas, we are probably closer to the end of the correction than to the beginning.
2. What, if any, are the warning signs of recession?
Various arguments have been adduced in support of the claim that the US is heading for a recession:
1. The length of the current expansionary phase. This argument rests on the assumption that no economy can grow uninterruptedly and therefore concludes that – as the current expansion has far surpassed the usual duration of such phases – we should prepare for a recession. In a recent study that used the same methodology as that for calculating human mortality rates,(1) Glenn Rudebusch found that the historical record since World War II does not support the view that the probability of recession increases with the length of the recovery – whereas it did before the war. Reasons for that change include robust postwar growth in the less-volatile service sector and the government’s increasing role in stabilising the economy, often through countercyclical measures.
2. A slide in profits as a percentage of GDP. During relatively long expansionary phases (e.g., in the 1960s and the 1990s), profits start slipping midway through – that is, several years before the onset of recession (see Graph 6).
So if neither a prolonged expansion nor declining profits brings growth to a halt, what does cause recessions? Unfortunately, there are no open-and-shut explanations and the question continues to divide economists. Exogenous productivity shocks, financial shocks, rigidities in an economy, oil shocks, shifts in the expectations and preferences of economic agents due to extrinsic causes – there is no shortage of theories put forward to explain the cyclical nature of the economy, yet none of them has gained the upper hand so far.
That suggests it might make more sense to take an empirical approach to forecasting recessions. A look at the various factors causing business cycle volatility reveals that the following four components of demand have a particularly strong influence on fluctuations in GDP:
- Consumption of durable goods
- Residential fixed investment
- Non-residential fixed investment
- Change in inventories.
Although they generate only one fourth of total output, these cyclical components account for 80% of GDP volatility. And they are a leading indicator of all recessions. In fact, virtually all contractions in cyclical sectors pave the way to recession (see Graph 7).
Apart from the 1953-1954 recession – which was caused by a falloff in defence spending after the end of the Korean War (a budget that accounted for 15% of GDP at the time) – the contribution of cyclical components to aggregate demand has peaked in the runup to all other recessions. And whether that peak was reached well in advance or just shortly before the recession hits, it always occurred before the sharp contraction in GDP (see Graph 8).
Residential fixed investment stands out as a particularly influential factor. Research by Edward Leamer(2) shows that a slowdown in the housing sector has always preceded the onset of recession. Moreover, all housing-market downturns were accompanied by recession, except for those in 1951-1952 and 1966-1967; in both of those cases, a sharp rise in government spending (to finance the war efforts in Korea and Vietnam) was enough to absorb the shock. The other cyclical components tend to kick in at a later stage – once the recession is already under way.
What all these components have in common is their relative sensitivity to interest rates. Most recessions are in fact preceded by monetary policy tightening (see Graph 9), but as monetary policy works with “long and variable lags”, as central bankers are wont to say, there is no telling how long it will take for a given interest-rate hike to put the brakes on GDP growth.
The current round of tightening by the Fed is unlikely in any case to trigger a recession on its own. For one thing, interest rates are still well below the nominal GDP growth rate. For another, tighter financing conditions don’t consistently lead to recession in the US, as the data from 1987 and 1998 clearly show (see Graph 10). American banks began to impose tougher lending terms on businesses in 1998, but kept mortgage lending rates advantageously low. The upshot was that the recession didn’t hit until 2001, and it was an extremely mild one at that. Tighter lending terms for households seem to have had an impact faster (see Graph 11).
What are we to make of all this? History suggests it is extremely difficult to develop a theory that can predict with any degree of accuracy when the US economy will tip into recession. The housing sector would appear to play an important role in such shifts, particularly because it is so dependent on current financing conditions. However, the 2001 downturn shows that other segments of the economy can also trigger recessions; that time the segment was corporate investment. In any event, recessions almost invariably occur after monetary policy tightening is carried out when the weight of cyclical economic activities in the country’s GDP has already reached its peak. As those cyclical activities have relatively little weight at present, the Federal Reserve may be in a position to engineer a soft landing – that is, a slowdown significant enough to reduce inflationary pressure without causing a recession, as in 1994. That is undoubtedly what the Fed is aiming for, but it is too soon to tell whether it will succeed.
CONCLUSION
As we have seen, the ISM manufacturing index provides valuable warning signals for shifts in the US economy, but taken on its own isn’t enough to get a clear view of trends under way. That’s why we believe it must be coupled with other indicators. In our opinion, weekly initial jobless claims are another reliable indicator of where the US economy is heading – and the latest figures are quite reassuring. The unemployment rate tends to increase as an economy moves closer to recession (see Graph 12).
Given all of the above, we don’t currently believe that the US economy is in danger of falling into recession. The following changes could, however, call our outlook into question:
- A sharp decline in several cyclical components of output, reflected in indicators such as car sales and new home sales.
- A fall in the new orders component of the ISM manufacturing index to below 44.
- A rise in weekly initial jobless claims.
The US economy may experience temporary soft patches as a result of external shocks or endogenous factors, but most likely won’t go into recession. In our view, we are dealing with a mid-cycle slowdown rather than a prelude to a full-fledged recession. Our outlook is for the growth cycle in the US to continue through 2016.
(1) FRBSF ECONOMIC LETTER, 2016-03, February 8, 2016, will the Economic Recovery Die of Old Age?
(2) Edward E. Leamer, 2007. “Housing is the business cycle,” Proceedings – Economic Policy Symposium – Jackson Hole, Federal Reserve Bank of Kansas City, pages 149-233.
This document is not pre-contractual or contractual in nature. It is provided for information purposes. The analyses and descriptions contained in this document shall not be interpreted as being advice or recommendations on the part of Lazard Frères Gestion SAS. This document does not constitute an offer or invitation to purchase or sell, nor an encouragement to invest. This document is the intellectual property of Lazard Frères Gestion SAS.
