Two weeks ago we were musing here about the whole “good news is bad news” meme, reflecting on the market crash that week as a “bad news is bad news” event. The Fed was set on tightening mode and the market thought it wouldn’t even be deflected by bad news, so bad news was really bad.
On Friday the job creation was a bit worse than expected, the unemployment rate was a tick lower than expected and wage growth was stronger than expected, especially given the previous month’s growth rate had been revised up. The result was that the good news on the unemployment rate and especially on wage growth meant we had a “good news is bad news” as it would most likely incite the Fed to raise rates. So US equities fell heavily.
Equities had already fallen on the very important Non-Manufacturing ISM data where the composite index had come in much lower than expected and the employment intentions were also bad, clouding the future outlook, but not a data set that much influences the Fed.
Normally when central banks are thought to be tightening policy and the economy is thought to be reasonably robust the currency will strengthen, further tightening policy in a passive way. Indeed, this is what happened to the US Dollar through most of 2015.
Last week on the Non-Manufacturing ISM news the US Dollar fell heavily too, which would normally have been quite good for equities. Bad news being taken as bad news, and therefore bad for both the currency and equities. New York Fed President Dudley made a comment that settled equities a little but did help to weaken the Dollar as he hinted the Fed would look at current markets for guidance [see Blog post of February 3:Dudley is the markets man on the FOMC, a small mercy]
On the jobs news the US Dollar reversed a modest part of those losses in a normal response to expectations of tightening.
The rates markets seemed to move yields up a touch on the wage growth scare but then yields fell back to the start of the day’s levels, and thus back to the lows of recent months.
The 12-month rate had spiked up five basis points earlier in the week in what looked like a piece of discipline by the Fed to ensure rates out to one year were at least in line with the Fed Funds Target Rate. The NY Fed must have not liked the 12-month rate dipping below 0.5% as it was beginning to show a lack of control of “market” rates by the Fed.
Topic of the week: Manufacturing ISM is pointless, as manufacturing is only 10-15% of GDP. But it is much followed because it covers tangible output, like agriculture and mining. Stuff! But services are 80% of GDP. However, they are extremely hard to monitor as they are so diverse and there is little tangible output. Little stuff, mostly services. Hence the Non-Manufacturing (i.e. services) ISM is a very useful survey, unique almost.
We aren’t too interested in the split between services and the rest as we focus on Aggregate Demand or NGDP. That said there remains no direct way to measure NGDP growth expectations, so the growth expectations of the service sector are a guide to the growth expectations of the whole economy.
Dissection of GDP can be useful for trying to assess productivity trends. Again, however, these trends are relatively easy to see in industries that grow, pump or knock together real stuff, but very, very tricky in services.
Almost all productivity discussions I read fail miserably to even address the measurement issue let alone discuss it, preferring windy, macro-level, Total Factor Productivity (TFP) aggregations that reveal little.
This quote from a speaker at a very high-powered recent conference pretty much sums up the utterly hopeless state of the productivity discussion game (p. 27):
So, some tentative conclusions. And first, I think we really do need to emphasize and recognize that productivity is difficult to analyze. I think of it as an emergent property of a complex adaptive system which everything that goes with that implies. So, real uncertainties remain around our analysis and we should always bear that in mind. I do think that micro and rim-level analysis has huge potential to enlighten our understanding. And that’s been a common theme of this morning. I think measurement is a genuine issue. I think policy responses need to be multifaceted and adaptive.
For once it’s been a data-driven week rather than markets being moved around by notions of Chinese growth (or lack of it) or utterances from the monetary policy gods.
The Non-Manufacturing ISM and the jobs numbers have already been discussed. The Atlanta Fed GDPNow model was clearly impressed by the wage growth number and over the week the 1.2% QoQ annualised RGDP growth became 2.2%. The less well known “Final Sales” GDP estimate, excluding changes to volatile inventories, started higher at 1.7% and rose to 2.4%, due mostly to higher Personal Consumption Expenditures (PCE) growth than expected, thanks to the wage growth number.
It is interesting that the disappointing ISM Non-Manufacturing number appears to have had zero impact on the GDPNow estimate, emphasising that it is capturing current expenditure rather than future confidence in that expenditure
As we also said above, the market should have welcomed the wage growth number but took fright at what the Fed might think. In historical context, 2.5% YoY growth in average hourly earnings is still pitifully low, if not quite as pitifully low as the last seven years. “Normal” should be well above 3%, well above 5% might even justify monetary tightening.