Recent Updates

  • US: New Residential Sales (Mar), FHFA Monthly HPI (Feb)
  • Euro area: Flash Consumer Confidence (Apr)
  • Canada: Wholesale Trade (Feb)
  • Canada Regional: Wholesale Trade by Province (Feb)
  • Luxembourg: Employment (Mar)
  • Mexico: Employment (Mar)
  • Brazil: IPC-S (Apr)
  • Greece: Non-financial Accounts (2018)
  • Hong Kong: CPI (Mar)
  • more updates...

Economy in Brief

German IP Follows German Orders Down the Rabbit Hole; IP Drops 1.9% in November
by Robert Brusca  January 8, 2019

German IP moves lower and faster than its diffusion reading
German output is falling over 12 months; however, on a diffusion basis, the German reading in November is 51.8, pointing to ongoing expansion. This creates some dissonance among German economic data.

However, the message here is clear if we look at the readings in the table sequentially from 12-months to six-months to three-months. We find a steady deceleration for:
* Total orders
* Consumer goods orders
* Intermediate goods orders
* Total manufacturing orders
* The average level of the manufacturing PMI index
* Real manufacturing orders
* Real sector sales in manufacturing

Do you see a pattern there?

Capital goods are an exception because they do not decline faster over three months than they do over six months, but the output of capital goods is still declining over three months. Construction is an exception with a gain over three months after a drop over six months. But even construction logs two monthly declines in a row for October and November.

In the quarter-to-date, all these same items show that deterioration is ongoing except for a gain at a 2.2% annual rate in real manufacturing orders.

The fact of a less weak manufacturing PMI than an IP index is interesting- but is it telling? The IP index is about the output of real stuff and is output-weighted. The diffusion index is about the breadth of the trend and its direction. Having diffusion above 50 tells us that the decline in output is not occurring across most industries just yet. But in comparison to IP, industrial production is telling us that there are more declines in large industries or more severe declines in larger industries and that is still a matter of concern. There is no solace in a PMI reading above 50.

In this cycle, output has been mostly rising (everywhere) in slow motion. I have argued to be wary of the message in the diffusion indexes in this environment because diffusion is about breadth, not strength. If output is advancing rapidly (I argue) diffusion is likely to be not as high valued as it would be if output got to the same level by rising more slowly over time. Why? The reason is that with slow output gains, the trend has a greater chance to be recognized and for other industries to join in. Thus, slower IP gains may lead to higher diffusion values being hit earlier. The caution here would be that diffusion may look strong relative to output in such circumstances. And when output turns lower, diffusion may take a while to unwind. The leading sectors will lead and by definition the lagging sectors will follow and do so more slowly. So we should be wary of how diffusion readings react especially when IP pulls back sharply as it is in November with a drop of 1.9%

Severe decline in progress
Output is falling at a 10.5% annual rate over three months, led by a 26.7% annual rate collapse in consumer goods output. This pullback is quite dramatic. Capital goods, while not part of the 'all but all-in deceleration parade,' are nonetheless weak at a 4.5% annual rate decline over three months and an 8.5% pace of decline over six months. We do have the German manufacturing PMI for December (at 51.5); it is only marginally lower than November (51.8) and still above 50. But evaluating the level of the PMI from January 2012 to date, the German manufacturing PMI has been this low or lower only 15.8% of the time. So if we step back from glorifying the signal from the absolute level of the PMI and evaluate it compared to recent the history of data, there is no disconnect; the PMI reading is exceptionally weak. The EMU manufacturing PMI in December has a 12.3 percentile queue standing, in line with the German weakness.

Germany and Europe and beyond...
Europe is showing a great deal of manufacturing weakness. China is weakening. There is weakness in other-Asia as well as globally. Even the U.S. that has bucked the trend is giving off uneven signals with the December job report having been exceptionally strong against a backdrop of weakening manufacturing and nonmanufacturing PMI readings. The U.S. stock market shows the signs of having been beaten up too and the bond market's yield curve is singing 'Danger Will Robinson' like the robot in 'Lost in Space.' But clearly, few are listening because it does not suit them. Central banks are still braking or preparing to brake.

The Fed apparently plans to continue to hike rates-although it will do it while running in Jello- at a reduced speed. Its gazillion equation model of the U.S. economy, perhaps supported by demand in the Lego™ sector, shows strength in the economy will continue. Yes, job growth is 'strong.' Although it produces few really good jobs as the U.S. economy is specializing in mediocrity and the Fed is dead set to see that that does not get out of hand. There are some wage gains. But there is very, very, little inflation.

The Fed argues that its shrinking balance sheet is not an issue for the economy as the Street argues that the balance sheet is an issue because it still distorts yields and is a factor making the signal from the yield curve robot (Danger Will Robinson, Danger!!) a moot point. Ignoring the yield curve is the one thing that the Fed and many market participants can agree on...why?

In the past, policy has had a very hard time twisting the yield curve. Bond market assets simply have too high a degree of substitutability for yield curve twisting or distortion to take hold and stay. Yet, we have disbelievers who want to throw out the yield curve signal, perhaps the most consistently accurate recession-predictor in the last half century.

I understand the problem… what good is a track record like that when pitted against the dogma of those in markets and at the Fed and a truck load of economists with PhDs that want to believe their complex impenetrable matrix of equations (that have never been right!) above the signal of the difference of the yields on two securities whose prices are set in the marketplace and anyone can read off the screen or in the financial papers? Clearly, the yield curve is outnumbered. And no one shaves with Occam's razor anymore (here).

However... if Board of Governor economists adopted the hockey playoff practice of growing a beard and not shaving until they were done with the Stanley Cup playoffs, every economist at the Fed would have a beard down to his toes (well, the male economists anyway). The Fed has yet to successfully forecast a recession. So rejecting the one thing that does successfully forecast recessions is a very strange practice. I think we call it 'denial.' Yet, here we are. And since the U.S. economy is globally important, that signal also matters to Europe and to Asia. To be fair, the signal is not yet switched 'on.' But it is very close to being 'on' and the sharp move to flatten by the yield curve is an odd thing to relegate to scrap heap.

large image