Shocks & Surprises Worth Noticing from Last Week: US Labour, British Auto Sales, China's Leading Indicators

Although quiet, on balance last week produced the cheeriest balance between shocks and surprises since early July.  The three I've chosen to focus on are:
1. US Labour Markets - The data's volatility over the last couple of months is unsettling, but the best explanation for the fall in September's unemployment rate really is that more people are being hired.
2. British Auto Sales -  Virtually ignored by the market, September's auto-sales were 2.5SDs better than  (pronounced) seasonal trends, with more expensive marques doing best. This is another fairly sizeable data-point which contradicts the generally-accepted and universally-propagated  grim economic 'narrative'.
3. China's Leading Indicators.  Both sets of leading and coincident indicators concur that we have an upward inflection point in China's cycle.


1. US Labour Markets The best explanation for the fall in the US unemployment ratio in September to7.8% from a revised 8.1% in August is simply that more people really are being hired. The chart shows that during September the number in work rose by 0.6% mom, or by 873k mom.  It also suggests  that all that occurred was a reassertion of the jobs growth rate that has been visible since the beginning of 2011. That trend (R2 of 0.90 vs time) expects growth of 0.14% a month, whilst over the last three months the average gain has been 0.13% a month. 
In August the number employed fell 0.3% below that trend, September’s have it 0.1% over the trend. Non-farm payrolls rose by 114k, with the biggest net hirers being health and social assistance (+45k), transport & warehouse (17k), financial services and professional business services (both 13k). 

Suggestions that the numbers are being somehow fiddled look thin: first stop for this would be the seasonal adjustments, but there’s nothing visibly tricksy happening there. 
The end-result may be coherent and plausible, but the way it was arrived at invites scepticism, particularly during election season. For, arguably, the biggest surprise in the data were the revisions to August’s gloomy numbers: remember, August’s original data showed non-farm payrolls up only 96k – the revision pushed that up to 142k, that’s a revision of just under 50%. August’s original 8.1% unemployment rate hinged on a jump of 581k mom of those ‘not in employment’: September apparently saw a decline of 211k mom in that same count. These are surprisingly volatile monthly swings in data which tracks those falling into and out of the labour market.

 2. UK Auto Sales September is the second-most important month of the year for UK car sales, typically accounting for 17% of the year’s total, so the strength recorded this week – sales up 8.2% yoy on a monthly rise 2.5 SDs above seasonalised trends – really matters. This was the strongest monthly performance since Sept 2010, and adds to the mounting evidence that British domestic demand is stronger than popularly  held (for further evidence, see last week’s comment on Britain’s record current account deficit in 2Q, and the erosion of the private sector’s savings surplus.) The sort of cars being bought also suggests that the upper/luxury end of the market is booming whilst the cheaper marques are not. Hence, September’s winners 
were Land Rover, BMW, Audi and Toyota: the market losers were Vauxhall, Nissan and Mini.

3. China Leading Indicators Two sets of leading and coincident indicators for China’s economy are published: the Conference Board publishes one, and the comes from the China Economic Monitoring & Analysis Centre (CEMAC),in association with Goldman. One of the most useful ways to use these is to check how the leading indicator is moving relative to the coincident indicator, since regardless of the overall direction, knowing how economic conditions are looking likely to develop relative to how they are currently actually experienced often gives an early warning of inflection points in the cycle. 

In the last two weeks, both the Conference Board and CEMAC reported gains in August’s leading indicators sharply higher than expected, based primarily on the rebound in real estate prices and (slightly) easing credit conditions, whilst industrial conditions remain depressed. But note that both sets of data are showing a major and sustained inflection in the leading/coincident indicator ratio. In other countries, where we have a longer track record to demonstrate the effect, this would be taken as a significant early indicator of a cyclical upturn.