Economic data roundup – probability of US rate hike recedes after weak data

Post with image

Recent bad news for the US economy seems to have been positively interpreted by financial markets. The sudden drop in the ISM manufacturing index and the labour market report for August were disappointing. However, US equity markets responded by rising slightly. In Europe, Japan and emerging economies, equity markets have also risen over the past few days and emerging currencies have rallied too. The rationale of course is that weaker data reduces the probability of the US Federal Reserve (the Fed) raising US interest rates at its meeting on 20-21 September.

If rates were to be raised this month, it would pave the way for discussions about the chances of two increases this year and could possibly have triggered an increase of rate hike expectations for this year and next. That all looks less likely to us in the wake of the latest economic data.

Purchasing manager indices (PMIs) were released in the last week of August for most of those countries for which a PMI is calculated. There were positive and negative surprises in the data but overall the PMIs point to quite a lacklustre trend in the pace of global growth.

Although markets have kept a generally positive tone without any improvement in economic fundamentals, we have made no change to our current cautious stance in asset allocation.

Latest global PMI indices suggest no imminent change to environment of insipid growth

Our global GDP-weighted manufacturing PMI slipped to 50.9 in August, matching the average in the previous 12 months. The decline in the global manufacturing PMI was due to growth easing in developed economies; it was stable in emerging markets. While business sentiment improved strongly in emerging markets in July, this improvement stalled in August.

Exhibit 1: No change to the weak growth environment – changes in regional manufacturing PMI indices for the period between 2010 and 02/09/16

Manufacturing PMIs

The global services PMI also fell slightly. This was also due to a decline in developed economies. Overall, the services PMIs are also still pointing to modest growth. This left the global composite PMI essentially unchanged at around the average of the past 12 months. Remember that so far this year, global growth has averaged only 2.5% year-on-year with a slowdown in developed economies and stable, but relatively slow, growth in emerging economies.

The PMIs hardly signal any improvement in global growth momentum

Exhibit 2: Service PMI indices point to modest rates of growth – changes in the regional service PMI indices for the period between 2010 and 02/09/16


Source: Markit, BNPP IP as of 06/09/2016

However, apart from lacklustre global growth, the PMIs are signalling some improvement in emerging economies. The emerging market composite PMI had been below 50 between May 2015 and June 2016 and it has now been positive for two months. The problem is that the hard data is not providing much confirmation of such a trend. Rates of growth of industrial production have improved in emerging markets, but this is due more to a bottoming-out in Brazil and Russia than to genuine acceleration. Trade data has remained weak.

Exhibit 3: Composite PMI indices point to a lacklustre trend in global growth – changes in composite PMI indices during the period between 2010 and 02/09/16

comp pmis

Source: Markit, BNPP IP as of 5 September 2016

So where were the surprises in the PMI data?

Not so much in the eurozone, even though PMIs there softened. The decline of the manufacturing PMI further into negative territory in France and the sudden drop below the 50 level, which separates growth from contraction, in Italy were clearly disappointing and are reasons for concern. But the improvement on the services side in both countries was reassuring. In the US, the manufacturing and services PMIs fell, but the bigger surprise was the steep drop in the ISM manufacturing index (more on this below).

The main upside surprises came from the UK. The PMIs dipped firmly to below 50 in July after the Brexit vote, but recouped those losses in August. This confirmed improvements in other indicators in the UK as well as July’s strong retail sales, although most sentiment indicators have not yet recouped the Brexit losses. Although it now seems that the negative impact of the Brexit vote was initially overstated, we think it is still too early to tell.

In emerging economies, the halt in the improvement in Brazilian manufacturing was disappointing, as were the declines in China and South Korea. A weak number in Turkey had been more or less expected given the turmoil in the region. India stood out positively with an especially strong gain in the services sector. Japan disappointed with only a marginal improvement in the manufacturing PMI, which remained below 50, and a dip below this level in services. This contrasts with the tight labour market and the improvements in workers’ earnings, giving the Japanese economy very much a two-sided face.

Weak data for ISM manufacturing index rules out a rate hike from the Fed in September

While the Markit PMIs discussed above decelerated somewhat in August, the sudden drop to below 50 in the ISM manufacturing index was quite a shock to the markets. Given the abruptness and magnitude of the decline, it could easily be dismissed as a one-off. Indeed, without such a sharp drop in the Markit manufacturing PMI, a rebound next month looks reasonable. But four out of the six regional PMIs slowed and also four out of six indicated contraction.

The decline among the components in the ISM manufacturing index was broad-based. The new orders index dipped to below 50, while new orders minus inventories, which leads the overall index with a decent correlation, fell to a level which implies an even weaker index . The manufacturing sector also shed jobs in August. While this is far from exceptional, employment in the US manufacturing sector has essentially flat-lined since early 2015.

August’s labour market report, released on 02/09/16, was disappointing, although not outright weak. Employment growth slowed to 150 000 in August, down from an upwardly revised 275 000 in July. The more modest growth in employment by 97 000 in the separate household survey, together with a stronger gain in the labour force, led to an increase in the number of unemployed people, although this was not enough to be reflected in the unemployment rate, which held steady at 4.9%.

Wage growth slowed as well as the average workweek, pointing to slower growth in household income. In the first half of the year, consumption grew more strongly than income, adjusted for inflation. We expect the rate of consumption growth to slow and approach the rate of income growth. For it to hold up, improvement in wages, employment and hours worked are crucial.

Exhibit 5: The US non-farm payroll report for August was disappointing – changes in the monthly US non-farm payrolls report for the period from 2010 through 02/09/16


In our view, the main implication of this data is that it is now very unlikely that the Fed will raise rates in September. Prior to the data the Fed had successfully prepared the ground and could have taken another small step in its process of rate normalisation in September. This data has put a spanner in the works. As a result, the market-implied probability of a rate hike has now fallen by 10 percentage points to 32%.

With inflation and inflation expectations this low, we think this latest data confirms the view that the Fed will wait until December before undertaking any further rate increase. The prospect of a lower-for-longer environment means the rich valuations of many asset classes may persist but we remain cautious as we think this peace is fragile.

Exhibit 6: With the US Federal Reserve unlikely, in our view, to raise its key official rates before December 2016, equity markets may be able to hold their current overvalued levels of valuations – the graph shows changes in valuations for the principal regional equity markets during the period between January 2016 and 02/09/16

equitiesSource: Datastream, BNPP IP as of 6 September 2016

This article was written on 5 September 2016 in Amsterdam

Leave a reply

Your email adress will not be published. Required fields are marked*