+1 Added to my documents.
Please be aware your selection is temporary depending on your cookies policy.
Remove this selection here

United States: towards a recession?


GDP stagnated in the first quarter of 2015. The weather does not explain the entire situation. The dollar’s appreciation has an impact on activity that is comparable to a tightening of monetary conditions. Should we fear the end of the cycle? In terms of profi ts, yes, but in terms of economic activity, it is still premature.

Rebalancing the added value sharing in favour of wages should allow consumer demand to drive growth and push businesses to continue investing (accelerator effect), starting in the second half. From a fundamental viewpoint, there is no reason the US economy should fall into a recession. There are no excesses to purge (neither on the side of investment nor on the side of private debt). Yet there is also no (or no longer) any reason for activity to grow much faster than the potential. The consensus among economists seems too optimistic, including for 2016.





Activity stagnated in the first quarter of 2015. The slowdown was partly expected, due to the harsh winter. Last year, growth fell into negative territory and then rebounded sharply in the second quarter. Still, the weather does not explain the entire situation – far from it. The dollar’s appreciation drags down growth directly (via its impact on exports) and indirectly, via its negative impact on corporate margins, which in turn weigh down business investment. In fact, then, the dollar’s appreciation has an impact on activity that is comparable to a tightening of monetary conditions.

Meanwhile, consumer spending remains solid, but the savings rate is rising in spite of the improvement seen on the labour market, the downturn in oil prices, and an asset base (finance and real estate) that is gradually returning to its 2007 peak. Behind the observed cyclical slowdown, it is the end-of-cycle question that is now being asked. For US activity has been rising steadily for the past six years, which makes for a very long expansion cycle with regard to the average post-war cycle. That said, any country can have quarterly setbacks, and they do not necessarily mark a downturn in the cycle. So what are we dealing with?

A cycle that is unique in post-war history

In practice, the length and scope of the cycle must be considered at the same time. Despite the length of the expansion cycle, this recovery is proving to be the softest in post-war history. Nearly seven years after the start of the “Great Recession,” economic activity is not yet back to its “potential” (i.e. the United States still has a negative output gap). This is unprecedented in the history of post-war cycles. No doubt one would have to consider the 1930s to find an episode of this kind (though that era’s statistical system provides an inadequate basis for measuring growth potential and making any comparison).

Most components of final demand have taken longer than in a traditional cycle to regain their pre-crisis level. This can be explained by the nature of the Great Recession. Indeed, it takes much longer to rise out of a recession when it is accompanied by a banking and financial crisis. The long-term work done by C. Reinhart and K. Rogoff on the impact of economic and financial crises does much to inform this point: “GDP growth and housing prices are significantly lower and unemployment higher in the ten-year window following the crisis when compared to the decade that preceded it”. In terms of cycle, we could probably interpret their remarks by saying that it takes more than one decade to close the output gap.

Based on the CBO estimates on potential GDP and its pace of growth, and upholding our own growth outlook, the output gap is not expected to close until 2019! In the 1930s, the Fed waited until 1937 to start tightening its monetary policy; history holds that this was an error because the tightening triggered another recession.

Comparisons with past post-war cycles should be made cautiously. Furthermore, that is what makes the current situation so hard to analyse. It is now impossible to tease apart what belongs to the economic cycle (direct and indirect impact of dollar trends or falling oil prices, for example) from more structural factors (lower potential growth due to the crisis).

Indeed, it is hard to explain in this cycle why business investment has not been more dynamic in spite of especially favourable conditions (high profits, low interest rates) and an aging capital stock. The slowdown in productivity gains (observed) and the weak investment dynamic bode ill for future growth potential, even before considering the impact of the aging population. Ultimately, cyclical and structural factors are working together and cannot be untangled, given the singularity of the current cycle.

What do the statistical approaches on growth in the second quarter of 2015 say?

The purchasing managers’ indices (ISM, PMI, etc.) used by most economists to assess very short-term growth did not predict the scope of the slowdown seen in the first quarter.

Therefore it is useful to look at other approaches. Abundant empirical literature has developed over the past 15 years to statistically infer the current quarter’s growth using economic data published over that same quarter. The work done by regional Fed gives a good illustration. Thus we can cite:

  • The index calculated by the Chicago Fed using a set of 85 existing monthly indicators. It is a matter of extracting the «common component» from this set. This index (built to be less than zero when growth falls below its potential) shows that Q2 started out with slower than potential growth (estimated at 1.7% by the CBO in 2015).
  • The Atlanta Fed’s work that developed a method to extract an estimate «in real time» of the current quarter’s growth using the highest-frequency data (here too with a purely statistical approach). The advantage over the previous method is that the data that are still missing do not prevent a very regular update of the estimate. The statistical information is continually incorporated into calculations for assessing GDP growth.

The method developed by the Atlanta Fed deserves all the more attention because it correctly estimated growth in Q1 (0.2% vs. 1.0% expected by the consensus). Today, this indicator (which has been available for Q2 since the beginning of May) shows growth of around 0.8% on an annualised basis for Q2! In both cases, the current consensus appears much too optimistic for the second quarter (still with 3% annualised growth anticipated in Q2).

These are not predictive models: they are not useful for ascertaining momentum in economic activity beyond the current quarter, and cannot be substituted for an analysis of the determining factors of demand (or supply). Yet their contribution is not insignificant. First, because, in periods of structural change, the traditional forecasting models (calibrated on past data) are more fragile. Second, because growth (yearly average) for the current year is more than 4/5ths known by the end of the first half. At this point in the year, projections for the second half weigh more heavily on the following year’s growth (yearly average) than on that of the current year. Finally, statistical approaches give valuable insight on the «surprise» in store (the spread in consensus on growth). This surprise can have a marked impact on the financial markets (via its impact on monetary policy anticipations).

Consumer spending and residential investment: the last leg of the cycle?

The drop in corporate profits, which is in large part tied to the dollar’s rise, does not necessarily signal the end of the business cycle. It may presage a rebalancing move in the share of wages in added value.

Insofar the investment recovery has been soft in this cycle, despite record profits, investment spending will stay flat with the turnaround in margins. That said, even if the investment rate is at its highest, there has been no overinvestment in the United States, except for the energy sectors (the correction of this over-investment is the reason for its decline in Q1). For productionfacilities are aging, and businesses must continue to invest, if only to keep he capital stock in shape. In the end, everything will depend on the trend in lobal demand. In an environment where external demand will be less dynamic (slowdown in global trade, appreciation of the dollar), the key clearly lies with household demand. Yet as far as that goes, the fundamentals remain solid, even though the second quarter is off to a disappointing start.

  • The labour market is still healthy: The improved situation on the labour front is probably the most convincing element for the US economy. The indicators built by the regional Feds, which summarise the trends in many, many variables, continued to improve over Q1.
  • Purchasing power: a supporting factor. The compensation component of the Employment Cost Index (ECI) accelerated (+0.7% in Q1 after +0.5% in Q4), despite the stagnation in activity and the downturn in profits. It is a much more reliable measure, on the macroeconomic scale, than hourly earnings from the monthly jobs report. This latter measurement, which shows that the wage increase is slowing, is in fact biased, since it mainly concerns unskilled jobs. Indirectly, this means that the labour market’s vigour is beginning to have an effect on certain job segments skilled jobs). In real terms, acceleration of compensation is also notable, even without factoring in the decline in oil prices (which is temporarily boosting purchasing power).

If we add job creation to the increased compensation, the rise in total payroll promises to be brisk. In other words, added value sharing is rebalancing in favour of consumers (and probably the middle class). Job creations, higher wages, and the drop in oil prices (to which consumers are very sensitive) are the reason for the high consumer confidence, which is at its highest level since 2007.

  • Property effects: still on the docket. The stock market has been in a stall since the year began. However, net household wealth (as a % of GDI) is close to its 2007 high. The rise in the savings rate in Q1 (from 4.5% to 5.5%) seems temporary to us (see graph). The increase in households assets is on a fairly solid base. Though equities look overvalued with regard to traditional metrics (PE or cyclically-corrected PE), the same cannot be said for real estate, which is playing a larger role in consumption.
  • Residential real estate: a long haul engine. Although the household accessibility index has fallen in the past two years, it is still 30% higher than its long-term average, despite the rise in housing prices. Yet the share of residential investment in GDP (3.3% in early 2015) remains decidedly below its long-term average (4.6%). We estimate it will take several more years to normalise the situation, given the creation of new households. Thus residential construction should contribute positively to growth, for at least as long as financial conditions are favourable.

Inflation and monetary policy: what is the consequence?

Core inflation is very low and has even been slowing down since the year began. That said, the start of acceleration in wages that we are seeing, paired with a slowdown in productivity, could soon drag down unit labour costs. On a 12 month horizon, then, it is likely that upward pressure will be seen on core inflation (which will nonetheless stay well below the Fed’s target). Such a movement, combined with the impact of the rise in oil prices, would be enough to encourage the Fed to opt for “less-accommodative monetary conditions” by raising its key interest rates. But before it does, the Fed should make sure that consumption is solid. It will be prudent rather than proactive. The spectre of the 1937 monetary policy error continues to loom over the Fed. The Fed will not take the risk of triggering a too-quick tightening of monetary conditions, because the economy is too fragile to bear too rapid a rise in real interest rates.

Ultimately, the first quarter is a real soft patch, which is probably not finished questioning economists about the forces at work. The consensus for 2015 growth (at 2.8%) does not factor in the persistent weakness in activity in Q2. However, it seems too soon to us to be counting on an end to the cycle. Rebalancing the added value sharing in favour of wages should allow consumer demand to drive growth and push businesses to continue investing (accelerator effect), starting in the second half. From a fundamental viewpoint, there is no reason the US economy should fall into a recession (unless bond yields soar or the equity market plunges). There are no excesses to purge (neither on the side of investment nor on the side of private debt). Yet there is also no (or no longer) any reason for activity to grow much faster than the potential. The consensus among economists seems too optimistic, for both 2015 and 2016.



Nearly seven years after the start of the “Great Recession,” economic activity is not yet back to its potential


















From a fundamental viewpoint, there is no reason the US economy should fall into a recession




Publication (Download)

Article in English

Article en Français

Cross Asset of May 2015 in English

Cross Asset de Mai 2015 en Français

Back up

Didier BOROWSKI, Head of Macro Economics
Send by e-mail
United States: towards a recession?
Was this article helpful?YES
Thank you for your participation.
0 user(s) have answered Yes.