Reflation?

Data released this week point to a rebound in activity and inflation. Yet it is too early to interpret them as the first signs of the reflation that some are expecting. Granted, inflation has made an impressive rebound, retail sales are reassuringly robust, and the turnaround in surveys, notably in the manufacturing sector, is encouraging. Yet we shouldn’t over-interpret these results nor forget the risks.

Last summer, consumer price inflation was limited to about 1% year-on-year, whereas in January 2016, it surged to 2.5%, more or less in line with the Fed’s target. This target is not based on the consumer price index (CPI) but rather on the less volatile personal consumption expenditures price index (PCE). In the past, PCE inflation of 2% – the Fed’s target – has corresponded to CPI inflation of about 2.5%. As in the eurozone1, the upturn in inflation must be seen above all as the result of the rebound in oil prices. The energy component of the CPI declined by about 10% year-on-year in summer 2016 but in January 2017, it rose by more than 11%. Last summer, energy was holding down inflation by roughly 0.7 percentage points, whereas today it is lifting inflation by about 0.8 pp. This perfectly explains the acceleration in prices. Core inflation, which excludes food and energy prices, has held at 2.3%, virtually the same level since last summer.

In February, the energy price base effect will be just as strong as in January, and it will not fade off until March, and then only gradually. If oil prices were to level off at current levels, the contribution of the energy component wouldn’t normalise until May, fluctuating between 0.2 and 0.4 pp through the end of the year. Inflation would continue to be driven upwards by energy, albeit to a lesser extent. As to core prices, they depend primarily on wages.

Despite its dynamic momentum, the job market has still failed to generate sustained wage growth. The US economy has been creating jobs since October 2010, more than 15 million overall in a little over six years. Yet wages have stuck to a soft trend. The average hourly wages of non-supervisory and production workers – our preferred measurement – only increased 2.4%. In 2006-2007, when the unemployment rate was as low as it is today (4.8%), wages were growing at a pace of nearly 4%. This is probably due to residual underemployment. Despite the job market’s impressive performance in recent years, there is still a large job gap. If from late 2007, job

creation had been strong enough to absorb new entrants on the job market (net of those quitting it), non-farm payroll employment would be about 6.5 million higher than it is currently. Although this helps calm fears about an inflationary surge, it also calls for strong demand in order to normalise the labour market.

In this respect, recent retail sales trends are very encouraging. Reported in value terms, this data are volatile. To isolate volume effects from price effects, we have to focus only on “core” retail sales. The control group excludes vehicle and fuel sales as well as food services and building materials, and thus follows more closely the consumption of goods as defined in the national accounts. Those core sales have accelerated in recent months, rising to an annualised quarterly rate of 3.5% in January, from 1.2% in September 2016.

For the moment, manufacturing production has picked up, but not at such a robust pace. In January, output rose 2.2% (annualised 3-month rate) after a year of virtual stagnation. But prospects are definitely encouraging. First, the decline in inventory is bound to trigger a rebound in production. The inventory to sales ratio slipped from 1.40 in May to 1.35 in December. Second, surveys have picked up strongly since the beginning of the year. In January, manufacturing ISM rose to 56, a cumulative gain of 6.6 points in five months, driven by solid components for production (+12.1) and new orders (+11.5).

Surveys by the New York and Philadelphia Feds point to another upsurge in February. Our NEM composite index – using weighted data reconstructed according to the ISM calculation method – gained 1.6 points between January and February. This is a spectacular turnaround from last summer’s trough, and in February, the NEM index hit 55.3, the highest level since year-end 2014, when US industrial output began to feel the strains of the drop-off in oil prices and the dollar’s rebound.

Seen in this light, we can see the expected rebound as a catching-up movement after the sluggish growth reported by the US economy between year-end 2015 and mid-2016. A welcome performance, but nothing miraculous.

Since Donald Trump’s election, mortgage rates have increased by about 65 basis points, enough to raise fears of a slowdown in construction. So yes, we are witnessing a cyclical rally. Yet the main explanations might be nothing more than temporary factors. Moreover, the strong dollar does not leave much room for a rebound in the manufacturing sector, whose external competitiveness is also undermined by the upturn in unit labour costs. It is too early to begin popping the corks.

by Alexandra ESTIOT