There was a lot of up and down action this week in most world markets, but by the end, no net change in prices. U.S., European, and emerging markets were all virtually unchanged, a highly unusual state of affairs.
Most world markets had one of their best days of the year Friday and one of their worst on Thursday. The very widely anticipated European Central Bank monetary announcement on Thursday underwhelmed traders, setting off a chain of trades to unwind big ECB-related bets. In reality, the ECB monetary policy changes weren't all that far from expectations. The ECB actually raised its economic growth forecasts, and the U.S. jobs report was at least a little better than expected, causing everyone to dive to the other side of the boat on Friday. The S&P 500 fell 1.4% on Thursday and then surged over 2% on Friday. For the whole week, the S&P was up just 0.1%. A lot of heartburn for not much gain. Even commodities and interest rates showed little change for the week.
The U.S. economic data for the week was mixed. Manufacturing looked terrible, according to the Institute for Supply Management's report on Monday, slipping below the closely watched 50 level. However, there were enough inconsistencies in the report and a belief that the manufacturing sector is too small to affect the broader economy that the market had little reaction to the news. We also think the ISM news was irrelevant, but for much different reasons. We explain why later in this report.
The auto sector, which has been a key driver of the current recovery, remained on fire in November, with the third month in a row of annualized, seasonally adjusted sales of over 18 million units. The auto industry is now within spitting distance of the previous full-year auto sales record of 17.4 million units. At the beginning of the year, most motor vehicle forecasts were for just 16.5 million to 17.0 million units.
The U.S. jobs report was also modestly better than expected, and two previous months of data were revised higher. Combined with hourly wage growth, the employment data suggest that consumers have plenty of money in the tank for the holiday season. Still, consumers continue to hunt aggressively for bargains, much to the chagrin of retailers. Pending-home-sales data showed that the existing-home sales data will stabilize soon, but growth rates in 2016 are likely to be lower. Hopefully new-home sales and home remodeling will fill in some of the void from slowing existing-home sales growth.
Employment growth still relatively stable
The employment report for November was very close to expectations and to the trends we have seen going back to 2011. For November, 211,000 jobs were added, compared with the 12-month average of 220,000 jobs and market expectations of 200,000 jobs. That was down from 271,000 jobs for October. However, we believe that most of the November drop-off was the result of some unusually good behavior in the business-services category in October, when it added a "too good to be true" 90,000 jobs, way above its 50,000-per-month trend. The business-services category added only 28,000 jobs in November.
In any case, the year-over-year nonfarm percentage payroll growth rate remains relatively stable at 2.0%. The small sojourn below the 2% trend line in 2013 and the mirror image jump in early 2014 is a statistical artifact caused by negative weather effects in 2013 and a return to normal in 2014. Today's 2% or so employment growth rate remains consistent with GDP growth of 2.25% and 2.5%. Over an extended period, employment growth should be slower than GDP growth because of improved worker productivity.
Outside of business services and construction, the sector data was more of the status quo
The broad patterns of past employment reports continue: no net additions in manufacturing, declines in mining, strong growth in many service categories, and limited growth in government. However, the construction industry did have a great month, adding 46,000 jobs. The category has gone from basically no jobs added in June all the way to 46,000 jobs in November, with every month adding more jobs than the previous month. Those are good-paying jobs with long work hours, so we are very glad to see that category do well.
Some the housing data and recent construction reports suggest there might be more good news ahead. That said, unusually warm and dry weather this fall could be helping the data just a bit, too.
Also worth noting is that, despite some generally grim news out of conventional retailers, retail employment growth continued in line with recent trends, just under 2%. That is, nevertheless, smaller than the overall private payroll sector growth rate of 2.3%, perhaps reflecting the industry's struggles.
Hourly wage growth rates continue to pick up
After an unusually large jump (0.4%) in the hourly wage rate for October, November still managed to post a 0.2% growth rate. The year-over-year private sector, averaged nominal wage data has continued to demonstrate relatively steady improvement in 2015 and is now at a recovery high of 2.4%. With headwinds from mix (lower-paying services-related jobs), a younger workforce (younger workers tend to earn less), and the move to more bonus-based compensation (think autoworker bonuses and limited increases in base wage rates), recent hourly wage results are even more powerful than expected. Indeed, evidence of our labour shortage thesis may be finally beginning to play out.
Adjusted for inflation, the wage data looks downright stunning, but the low inflation rate that we are seeing now is probably not sustainable over the longer term. Still, the wide gap between wage growth and inflation should be very beneficial to consumer spending in the short run. It will be very interesting to see if increasing wages because of labour shortages can keep real wages growing as inflation returns to a more normal 2% rate in the years ahead.
Auto sales show the consumer is alive and well
We have been keeping a close eye on auto sales as a key indicator of consumer confidence, along with restaurant sales and new housing permits. We were very pleased to see auto sales in November hold steady above 18 million units on a seasonally adjusted annualized rate basis for the third month in a row. The great number came despite the relatively small number of selling days in the month. Sales almost always drop after a month that includes five Sundays (which tend to pull more sales back into the earlier month) and that did not happen this time around. However, warm weather, cheap interest rates, and an extensive ad campaign in November may have provided at least a small boost to the results. Although sales growth has slowed from the early years of the recovery, the further slowing that we had been expecting has not materialized.
Even stepping back to full-year data, results still look good but not quite great.
However, the National Automobile Dealers Association is expecting the 5% rate of the past two years to fall to a crummy 2.5% in 2016 as interest rates increase, and cars-to-population rates are approaching normal levels. It's hard to imagine that the average term of an auto loan can get much longer, either (they are over 65 months currently). Lengthening loan terms have been a key to keeping monthly auto payments in check despite rising transaction prices. In fairness, NADA (and us, too) also expected growth to slow in 2015, so maybe they (and we) will prove to be too pessimistic again for 2016.
One other note: For some quirky reason, October was a great month for fleet sales and just so-so for consumer sales, suppressing the October consumption report. Those consumer sales came back in November, which will help push up the consumption numbers.
ISM purchasing manager data puzzlingly bad
We will always have a very fond place in our hearts for the ISM report because it was just about the only metric that helped spot the end of the recession in early 2009. We also always look at the report because the methodology allows for a very early look at the data for the most recent month, as much as three weeks earlier than some of the more conventional manufacturing reports. However, the report has done a much better job at the bottoms of recessions than at tops, where the index has been a bit trigger happy.
Note that between 1991 and 2001, the PMI reading dipped below 50 four times, with only one of those, the last one, resulting in a recession. Although the market thinks that the 50 level (which means that more businesses are reporting decreases than increases) is magic, the ISM is always careful to assert that it is a reading of 43 or below that is typically associated with recessions.
With that warning, the PMI for November did indeed drop below 50 as shown on the graph above. And both new orders, the most forward-looking of the subcomponents, and production also dropped below 50.
However, the puzzle is that the employment index had a nice jump, and low inventories readings seemed to suggest that more product is called for even with an increase in end demand. The export reading was unchanged at 47. None of this seems consistent with any kind of manufacturing disaster. Durable goods and industrial production (manufacturing) suggest stabilization in the manufacturing sector. We wonder if it is low prices that are raising havoc with this statistic. The price reading dipped yet again, to 35, one of the lowest readings we have seen, suggesting that perhaps unit orders might be OK while the dollar volumes of order and shipment are eroding because of declining prices.
In any case, absent some truly awful data, we don't believe that manufacturing is large enough to move the economic needle. After a series of years when we were constantly wishing that manufacturing was a bigger part of the economy, we are relieved that it represents less than 10% of all employment in the U.S. In addition, with autos still doing OK, 
Boeing (BA) in ramp-up mode yet again, and new-home construction set to improve, it's hard to imagine a free-fall in manufacturing.
China PMI data looks better again
For a number of reasons, the China PMI data didn't seem to get as much attention as usual. The Caixin data that we use showed a modest improvement in November and its best reading since June. U.S. and Europe's Markit PMI readings were little changed since last week's flash reading.
Pending home index stabilizes, but growth rates still slowing
A lot of commenters were very excited about the fact the pending-home-sales index managed a small increase, moving from 107.5 in September to 107.7 in October after four straight months of decline beginning in June. Furthermore, the same folks were excited that the year-over-year growth rate was 3.4% for a single month. The bulls went on to note the year-over-year growth rate in pending sales now stretched for 14 months. We are a little surprised at how many news outlets spun the pendings report just that way. While we, too, were pleased that the index didn't continue down, there wasn't a lot to be excited about. The year-over-year growth rate in pending sales has been slowing for some time, and the October report did not provide any relief on that front, as shown in the middle column below.
Also, the poor pendings report is slowly filtering into the more important existing-home sales data. For the full year 2015--that is, all 12 months of 2015 compared with the 12 months of 2014--we suspect existing-home sales will increase from 4.91 million units to 5.26 million units, a 7% increase. Unfortunately, the National Association of Realtors is expecting slower growth in existing-home sales of just 3.5% in 2016, to just about 5.45 million units. That is very close to what the current pendings report would suggest. Qualitatively, the Realtors presentation cites low inventories, high prices, and potentially higher mortgage rates as the reasons for their relative pessimism. If they turn out to be right, which seems to be a decent bet, existing homes will be a much smaller contributor to GDP growth than they were in 2014.
We don' want to be all doom and gloom here. Certainly, existing-home sales shrank in 2014, and the economy still managed to do all right. Maybe smaller-than-expected interest-rate increases and better-than-expected wage growth may make the Realtors forecasts too pessimistic. Also, the realtors are much more optimistic about new-home sales and housing starts than existing-home sales, though that is getting away from their core expertise.
Import/export data not as positive as last week's advance report suggested
Just when we thought the import/export statisticians had gotten their act together, the U.S. Census Bureau threw analysts another curveball on Friday. The preliminary October data, released last week, suggested the trade deficit in goods showed a narrowing in October compared with September, indicating that the trade deficit would not be a big factor in the fourth-quarter GDP report. However, the final version of the October report showed the goods deficit widened a bit in October and did not shrink after all. The report was certainly no disaster, but it probably means that instead of trade being neutral in the fourth quarter, it could end up taking off a couple of tenths. However, we are not changing our fourth-quarter GDP estimate of 2.9% just yet, as there are two months of data and many revisions yet to come.
The overall trade data paints a picture of small export declines and still positive import growth. While many lay the blame on the strong dollar, we surmise that falling commodity prices, excess capacity in commodity processing, and slow world growth are much bigger parts of the trade issue. We hope to have more time to explore trade issues in future columns.
Retail sales top next week's data
Retail sales are a key element in total consumption report, the major driver of GDP growth, so it is a very closely watched report. Unfortunately, widely divergent inflation rates for various categories have diminished the usefulness of this report, at least until consumer price data is released the following week. Big declines in some retail sales report categories often turn into gains in the consumption report, which, unlike the retail sales report, is adjusted for inflation/deflation. Lately, falling prices of electronics, gasoline, and apparel have made the retail sales report look depressing. However, once adjusted for falling prices, things don't look quite as bad. Gasoline prices, which were down almost 6% in November, are likely to depress the headline figures for retail sales growth.
Auto sales are always tricky in this report, too, because the Bureau of Economic Analysis uses a different set of auto numbers to calculate GDP. In addition, the headline auto sales numbers the automakers provided this week include fleet sales and government sales. So while total auto sales were flat sequentially, the retail auto numbers might look a little better.
In total, most analysts are looking for sequential growth of retail sales of about 0.3%, which would represent a nice acceleration from the 0.1% rate of the previous month. On a more useful basis that excludes autos and gasoline, we still expect year-over-year growth of 3.5% to 4.0%, which is still in line with recent results and relatively strong consumer spending.
We suspect the job-openings report will continue to show openings outpacing hiring; the only question is by how much. The NFIB Small Business Report should give us some sense of how small-business owners are feeling and more importantly, what they think about the labour market. Data on the federal budget for the first two months of the fiscal year should also be available next week. We will be particularly interested in trends in tax collections. In the just-completed 2015 fiscal year, collection growth far exceeded spending growth, and we are not expecting the news to be quite as good in 2016. The years of sharply improving budget deficits is probably behind us, but we suspect the level of the deficit will remain relatively constant even as the economy continues to grow.
