Shocks & Surprises, Week Ending May 26th

The guiding principal of Shocks & Surprises is that it identifies that data which falls one standard deviation above or below the range of consensus expectations, or in the absence of a surveyed consensus, more than a standard deviation outside identifiable current trends.

One of the upshots of this is that if these Shocks & Surprises are distributed normally, we should expect roughly 16% of the data to Shock and 16% to Surprise, and the remaining 68% to conform to consensus or trend. That would represent a state of affairs we might call 'normality'. Over the last six weeks, we been a long way from normality: typically between 45% and on occasion 60% of the data coming across our screens has been either Shocking or Surprising. Statistically we have been living in distinctly non-normal times. I am sure you have noticed.

What is less obvious is that, news headlines notwithstanding, the last couple of weeks may be edging back to normality, in the sense that the economic data is beginning to conform once more to statistical normality. This is not wholly or solely a matter of the financial industry's stable of economists reconciling their short-term forecasts to the new reality; it also includes a reassertion of historic seasonal patterns, albeit perhaps at slightly lower levels than previously expected. Thus, two weeks ago, the proportion of Shocks & Surprises fell to 43.1%, and this last week it sank to 35.2% of the 70 pieces of data I tracked - which is within roughly three percentage points of statistical 'normality'.

More, the only part of the world which is still delivering a statistically abnormal proportion of Shocks & Surprises is, of course, Europe. In the US certainly, but also within Asia as a whole, we are no longer in abnormal territory.

So we must start with Europe, where this last week brought the advance readings of manufacturing PMIs for May. Dire reading they made, with Germany, France and the rest of the Eurozone showing the worst readings for around three years, and with new orders shrinking fast owing to faltering domestic demand more than falling export interest. In the UK, the CBI Trends Total Orders survey gave the same message, as it retreated to levels last seen in December, although export orders eased only marginally. Germany's Ifo Institute also published their monthly survey of Germany business sentiment, with readings for overall Business Climate, Current Conditions and Expectations all falling below the level of expectations – although even now German business expectations are running higher than the long-term average for the series. There's no such mitigation for Italian Consumer Confidence, which for the second successive month explored previously untouched depths of gloom: a reflection on the outlook for the national economy rather than current individual economic and financial experience.

Even in Europe, however, positive surprises remain. The Eurozone's construction output for March surprised, jumping 12.4% mom as the zone recovered (Germany particularly) from February's construction-halting freeze. In addition, there were two sharply positive surprises from the UK, when March output of services rose 0.5% mom, and April's budget cash surplus was roughly four times bigger than the previous year (and expectations). The regular run of UK data for 1Q becomes increasingly difficult to reconcile with the 0.3% qoq GDP contraction recorded in the national accounts.

The US gave us two surprises this week. The first was a 1.8% mom rise in the Federal Housing Finance Agency house price index: to put this into context, it was the highest monthly jump for at least the last 10 years, with prices jumping all across the US. There's no other real estate data reporting this sort of a jump, so for the time being it remains an unexplained anomaly, raising the suspicion that there's something surprising about the way the data has been produced, rather than it telling us something genuinely surprising about improving conditions in the US real estate market. The second surprise looks far more convincing: the final version of the Uni of Michigan Consumer Confidence survey for May contained a very sharp upward revision in the 'economic outlook' reading, which took the total index to its highest level since October 2007.

Set against that, orders for capital goods, excluding defence and aircraft, contracted 1.9% mom in April, extending a run that's really souring fast: January orders fell 3.1% mom, in February they rose 2.9%; in March they fell 2.2% and now in April they are down 1.9%.

Asia brought no positive surprises at all during the week, but few shocks either. The HSBC Manufacturing PMI for China made headlines as it retreated to 48.7 in May from 49.3 in April, but although this was towards the bottom end of current trends, it did not break them. Rather, the Shocks came elsewhere: Japan's 7.9% yoy rise in exports in April disappointed (exports to China fell 12.4% mom); Taiwan's April export orders shocked by falling 3.5% yoy (mainly thanks to a 8.2% mom fall in orders from US and 7.1% mom fall in orders from China and HK); and Singapore's industrial production also disappointed by falling 3.5% mom and 0.3% yoy in April.