It was a bad week for markets across the board, with world stock and bond markets down, as well as commodities. U.S. and European equity markets were both down over 2.5% for the week, with emerging markets doing somewhat "better" with declines of just over 1.5%. Better-than-expected GDP growth also pushed up bond rates slightly with the 10-year U.S. Treasury rate moving up from 2.46% to 2.5%.
While commentators are falling over each other trying to explain why the market suddenly fell off a cliff this week, it seems to me that the market was tired after accumulating some very healthy gains over the past 18 months. There was a lot of news on many fronts, but there wasn't anything truly earth-shattering, new, or revolutionary. A lot of earnings reports from around the world, but especially from Europe, were soft this week, which didn't help matters in an already-skittish market. Throw in the Argentinean bond default, new Russian sanctions, and an escalating situation in the Middle East, and there was plenty for the worrywarts to fuss about.
In economic news, the GDP report for the second quarter was better than expected. Even more important, the first-quarter contraction was revised to a considerably smaller contraction. I fully expected to be reducing my full-year forecast when the week began. (I had the new models all set to go with a sub-2% growth rate.) However, this week's GDP report now makes my original full-year 2.0%-2.5% GDP forecast a real possibility. This growth rate has remained virtually unchanged for almost three years.
However, to get to the middle of my forecast range, the economy needs to grow faster than 3% in the back half of the year. A large increase in consumer incomes over the first half remained largely unspent, providing the potential rocket fuel for second-half gains. Consumer spending in June already showed some signs of that additional accelerant.
Countering some of that good news was a jobs report that came in slightly below consensus, but the private-sector growth rate of the past three years remains intact at 2.1% on an averaged, year-over-year basis. Auto sales also looked great at 16.5 million units, but they, too, were off the recovery high of 17.0 million units sold in June.
Second-quarter GDP growth accelerates to 4% on better weather, inventory restocking
Headline GDP grew 4% in the second quarter when compared with the first quarter and annualized, in contrast to a decline of 2.1% in the first quarter. I would argue that to understand the true state of the economy, you would probably have to look at those two numbers in aggregate, with the first quarter probably not as bad as the numbers suggest, and the second quarter not as good. Still, the annualized first-half results show growth of just under 1%, slightly less than the 2% long-term trend. However, using the more accurate year-over-year methodology, growth in the first half compared with a year ago remained on its long-term trend of 2.2% growth.
I am growing increasingly wary of looking at sequential quarterly data and then annualizing it. It puts way too much weight on short-term swings in inventories and net exports, which are volatile, hard to measure, and quickly reverse themselves in subsequent quarters. And those wild swings are multiplied by four in the sequential growth calculations. The sequential data seem to make the economy much more volatile than it really is.
Looking at the past data in the table, you can see that the sequential data has experienced massive swings from a 4.5% increase to a 2.1% contraction in just six quarters. Meanwhile, comparing the current quarter with the same quarter a year ago produces a much narrower range of just 1.3%-2.6%, which is in line with consumption and employment data that shows similar rates of growth and even greater stability.
The reasons for the wild swings in sequential quarter-to-quarter GDP are on clear display in the contribution table shown below.
It's glaringly evident that inventories and imports/exports are the root cause of all this volatility. Excluding those two categories, the second quarter looked good with all the other major sectors making a contribution to GDP growth, including housing and government, which previously had been detractors. It's a more balanced performance than we have seen in some time, which is probably more important than the magnitude of the GDP headline growth figure.
The only troubling portion of the report was growth in services continued to fall. Again, softer utility figures and surprisingly soft health-care numbers are holding back this sector. Unfortunately the poor health-care spending numbers have been validated by lackluster employment growth for the health-care sector.
Sticking with my 2.0%-2.5% growth forecast yet again
I was all prepared to trim a little bit from my full-year GDP forecast of 2.0%-2.5%, based on the originally estimated 2.9% contraction in the previous revision to the first-quarter GDP estimate. I was particularly scared when the IMF took its U.S. forecast down to 1.7%, which now looks overly dour. However, I wanted to hold off on my adjustments to see if some type of miracle might happen in the second quarter.
We didn't quite get our miracle, but at 4%, the second-quarter GDP growth rate was better than the 3.2% consensus, primarily because of a complete reversal (and then some) of the decline recorded in the first quarter's inventory contraction. However, what really saved the day was a revision in the first-quarter GDP estimate that now showed the economy slowing only 2.1% instead of the nasty 2.9% decline that the government estimated just a month ago.
I do note that to get to my 2.2% full-year GDP forecast, growth in the back half of the year needs to be over 3%. That's very possible if we get an import/export swing. Though I've modeled 3.5% growth for each of the past two quarters, it's more likely that one of the two quarters is substantially better than that and one is worse. It's also possible that the major shrinkage in health-care expenditures is eventually revised out of the first half of 2014's GDP estimates.
Employment growth reverts to mean
Headline employment grew by 209,000 jobs during the month of July, which is exactly equal to the average of the past 12 months. Some analysts found that number disappointing because it was a lot lower than the 298,000 jobs that were added in June. However, I always viewed the June number as a fluke, perhaps a delayed catchup from the severe winter weather when job growth hit a measly 84,000 for the month of December.
The year-over-year percentage change, averaged over three months for both private sector and nonfarm payrolls, remain stuck very near their 12-month averages of 2.1% and 1.7%, as shown below. This stability is in direct contrast to GDP growth, which has been all over the map and is quickly losing economic relevance, at least in the short run.
July's employment report had a certain internal consistency that we haven't seen in months. The month-to-month data is consistent with the year-over-year data, the revisions to past months were tiny, and the seasonal adjustment factor for July was next to nonexistent.
Sector data seems to indicate improving job quality
This month the construction sector and durable goods manufacturing, as well as mining and extraction, all produced job growth substantially above their 12-month averages. These are all generally jobs that offer higher wages and longer work hours. Government jobs, generally decent payers, also showed some growth after years of losses. Meanwhile, retail, education, health, and leisure all produced job growth well below the 12-month averages. Leisure and retail are two sectors offering lower wages and hours.
The slowdown in hiring in health care is both disappointing and surprising. It's disappointing because these are decent-paying jobs, especially by the standards of the service sector. I had hoped that health-care hiring would have a great year because of the Affordable Care Act, which should have theoretically boosted health-care usage for those who were previously uninsured.
Fortunately, one of the major potential negative effects of the Affordable Care Act has not happened, either--at least on a major scale. Namely, the widespread fear that one full-time or high-hour job would be split into two or more part-time jobs. How do we know this? The Labor Department tracks hours worked per week. If this job-splitting had occurred on any truly systemic basis, we should have seen at least some decline in the average workweek. This just hasn't happened. The average workweek has been stuck between 34.4 hours and 34.5 hours over the past year. Even in the retail sector, where the job-splitting phenomenon was expected to be widespread, hours worked slipped only modestly from 31.5 hours last July to 31.3 hours in July 2014. Hardly catastrophic. The leisure sector, which restaurants dominate, saw hours worked increase from 25.9 hours to 26.2 hours over the same period. I am not denying that some employers have split jobs, and we have certainly heard a number of these stories ourselves, but it doesn't seem that the practice is widespread. And while very volatile month to month, part-time jobs fell by 100,000 from July to July while all jobs grew by over 2 million for the same period. In other words, the percentage of workers employed part time has fallen (from 18.9% of the work force to 18.6%).
The raw jobs number was in line with trend, the hours worked flat and acceptable, but the news on hourly wages continued to be depressing, at least on a monthly basis. My associate on the markets team, Roland Czerniawski, summed up some of the hourly wage data released in Friday's wage report. His punch line is that while hourly wage rates are up nicely, they are just about flat when adjusted for inflation:
"Despite clear signals that the labour market is tightening, as more job openings and steady employment growth seem to indicate, a meaningful pickup in wage growth hasn't materialized just yet. The Bureau of Labor Statistics reported on Friday that monthly wages increased by 0.04%, which makes the headline number rounded to one decimal point (0.0%) somewhat misleading. Nonetheless, this equals to a monthly increase in hourly wages by just a penny to $24.45. Before diving into the actual interpretation of the wage report, it is important to notice that the headline monthly wage data is extremely volatile, and due to the aforementioned rounding technique, it can often lead to slightly misguided conclusions. The chart below shows monthly hourly wage growth since 2006."
"In order to gain more insight into what is really happening with wages, it is necessary to smooth out the data by converting it into a year-over-year, three-month average format. On that basis, the trend is crystal clear. Wage growth has been disappointingly slow, but fairly steady since 2010. The chart below illustrates this. The July report shows that despite the immaterial monthly increase, the year-over-year numbers are still growing at about 2%, which equates to about half a dollar more an hour compared with a year ago. While this paints a slightly better picture than the monthly 0.04% growth, it is still way below the pace that we were experiencing prior to the recession. What is even more alarming is that on an inflation-adjusted basis, workers are earning basically the same wages as they were a year ago.
"Considering the various signals of tightening in the labour market that we've been reporting over the past weeks and Thursday's encouragingly strong Employment Cost Index, the continued slack in wage growth remains puzzling. One potential explanation might be that the shift in working-age demographics might be keeping the hourly earnings growth at bay. As baby boomers (who tend to be a more-experienced and higher-paid part of the labour force) retire, the less-experienced replacements are simply not receiving the same amount of compensation compared with their retiring predecessors. Regardless of the reasons, wage growth should finally pick up some steam in the months ahead, given the current direction of the labour market. Considering that all eyes are now focused on inflation and the timing of the Fed's withdrawal from its accommodative monetary policy, the wage data might easily become the focal point of future monthly jobs reports, and we will continue to monitor it very closely."
Year to date, income growth has far exceeded consumption growth
Year to date, consumption is up its typical 0.94% (1.9% annualized), while income (as measured by inflation adjusted total income, less taxes) is up a stunning 2.1% (4.2% annualized). This means that consumers, for one reason or another, have spent less than half of the income that they have gained so far in 2014. That may explain why consumer confidence is at a recovery high. That gives consumers a lot of firepower to spend more in the second half to make up for the lackluster first half.
In the most recent month, income growth was finally more in line with income. Year-to-date income growth has exceeded spending growth in five of the past six months, a rather remarkable string of consumer restraint. It's a tough call to determine whether that holding back was intentional, or bad weather creating a literal inability to get to the store and the lack of need for certain products.
With any of these explanations, though, the outlook for more spending increases is better than it has been for some time. Recent consumption growth in May and June suggests that some of the extra cash is finally being spent. Keep in mind that income growth is more a result of additional people working, while the hourly wage rate, adjusted for inflation is little changed, as Roland notes above.
All of this is very important because consumption represents almost 70% of the GDP calculation. Improving consumption generally leads to more employment, which leads to more income, which leads to more spending, and that virtuous cycle reinforces itself. Given real problems with the GDP report, the consumption and income report is probably the most important economic statistic we get each month. Unfortunately, it is just about the last of the monthly indicators we get, and it is subject to a lot of revisions because it is dependent on the volatile and oft-revised monthly payroll report.
Another good month for autos
U.S. automakers reported another great month, with 16.5 million units sold in July (seasonally adjusted annual rate), following June's 17.0 million units. David Whiston, who is a senior equity analyst at Morningstar, offers the following commentary:
"U.S. automakers reported another strong month for July new light-vehicle sales on Friday. Total sales rose 9.2% from July 2013 to 1.44 million units while Automotive News put the seasonally adjusted annualized selling rate, or SAAR, at 16.5 million. July 2013's SAAR was 15.80 million, and this July's SAAR was the best July since 17.16 million annualized units were sold in 2006. We continue to see full-year sales coming in close to or exceeding the high end of our estimate of 15.9 million-16.2 million. On Friday's sales call, 
Ford F economist shared an interesting statistic on the age of the fleet. In the mid-2000s, about 40% of the fleet was at least 10 years old, but that ratio grew to an estimated 50% in 2013. We continue to expect continued growth in U.S. auto sales until the next recession, however, we expect a slower growth rate of about 2%-3% per annum compared with the double-digit and high-single-digit rates seen following the U.S. bottoming out at 10.4 million units in 2009. With great product available, easy credit, and an old fleet, we are not concerned about a bubble forming in auto sales at this time."
Manufacturing extends its hot streak (written by Roland Czerniawski of the Morningstar markets team)
The manufacturing sector continues to perform exceptionally well as the ISM Manufacturing Purchasing Managers Survey indicator increased to 57.1 in July (from 55.3 in June), which is the highest reading since April 2011 and indicates an expansion in manufacturing for the 14th consecutive month.
New orders remain one of the strongest parts of the report, increasing sharply from 58.9 in June to 63.4 in July, which is a good indication that the outlook for production, shipments, and employment is positive. In line with Friday's jobs report, ISM also confirmed that manufacturing employment is on the rise, with an increase of 5.4 percentage points in July. Production also ticked up, accelerating from an already-high level, suggesting that July's industrial production number should also be very good.
On the downside, however, it seems that some of the geopolitical risks are beginning to weigh on the U.S. manufacturing sector as Russian demand for medical devices has plummeted, and purchasing managers in some industries are worried about the impact of geopolitics on their businesses. Despite the great continuous progress manufacturing has made, it is important to note that it is now only a relatively small part of the U.S. economy, and the positive impact on GDP and overall employment will, unfortunately, be limited.
Pending home sales slip just modestly
The last pending home sales report (June) wasn't terribly helpful, as the index barely budged in June, decreasing 1.1% between May and June after four consecutive months of improvement. Two of the four regions showed improvement, the West and the Midwest, while two saw declines, the Northeast and the South.
The little-changed pending home sales index most likely means that existing-home sales for July will be little changed from June levels. For the record, existing-home sales in June were 5.04 million units. That existing-home sales reading for June was on par with existing homes in April 2013, which is just before existing-home sales data soared as buyers rushed to close mortgages before rates raced even higher. Something right around 5 million existing-home units appears to be the central tendency of this key metric. The mortgage-related spike in July 2013 saw a high of 5.38 million units (and a pending home sales index of 110). But it looks as if that merely stole sales from later periods, as existing-home sales dropped to a low of 4.59 million in March of this year (and a pending home index of 94.2 in February).
The current existing-home and pending homes readings are both exactly in the middle of those highs and lows. For the first six months, existing-home sales averaged 4.7 million units. The National Association of Realtors is expecting full-year sales of 4.95 million units, which implies a pickup to 5.2 million units in the back half of the year.
CoreLogic and trade report on tap next week
After the flood of data this week, the next week will be pretty tame, with only two important indicators being reported. The first one is CoreLogic, which is expected to show continued moderation of home prices across the nation.
The second one is the international trade report, which remains an exceptionally tricky data point to forecast. I am expecting some improvement from last month's 44.4 million deficit as energy exports grow and energy imports slow as the effects of a cold winter recede (cold U.S. weather meant there was less product available to ship and more imports were needed to meet peak demands). On the other hand, a potential West Coast dock strike is causing some importers to rush holiday shipments into the United States that would normally be made in the months ahead. Net exports have hurt the GDP report two quarters in a row, and it seems like some relief is overdue, if the potential strike doesn't mess us up. A rush of new Apple AAPL products due next month won't be particularly helpful to upcoming trade reports, but the timing of those new releases remains unknown.
