It was yet another week of high volatility, where markets didn't do quite as badly as some of the midweek headlines would have you believe. Even emerging markets managed a small rebound for the week, down "just" 2% after having been down as much as 3% at midweek. Driving emerging-market results were poor economic data out of China as well as China's moves to devalue its currency. Given the data, one might have expected worse, but a least some of the results were already anticipated, with major emerging-market indexes down almost 20% since April, and the Shanghai down even more. Worries about a slower China also hurt European markets, which were down about 1% even as some of the cloud over Greece has lifted.
The U.S. managed to do the best of the bunch, with a gain of about 0.7%. Most of the economic data this week strongly suggested that turmoil in Greece and China had little impact on the more insular U.S. economy in general and consumers in particular. Given the China news, commodities remained soft, down another 0.5%, led by yet another decline in oil prices. The stronger relative U.S. economic performance this week probably explains a small increase in the 10-Year U.S. Treasury rate to 2.2%.
Given the slowing worldwide economy and extremely cautious businesses, all eyes were on the U.S. retail sales report, and the news here was excellent. The report was incredibly solid, if not spectacular, with a broad base of categories doing well, and past months being revised up.
In fact, a lot of economic data has been quietly revised up over the past week or two, suggesting that the next revision to the GDP report may prove to be unusually large. Second-quarter growth was originally reported as 2.3% at the end of July, and now it appears that may need to be revised up to a range of 3% to 3.5%. Retail sales, inventories and construction data have all been revised sharply upward.
In other economic news this week, industrial production also proved unusually strong on a month-to-month basis, perhaps inflated a bit by strong auto production data aided by seasonal factors. The Congressional Budget Office also reduced its budget deficit estimate from US$486 billion to just US$425 billion on the back of stronger tax collections and spectrum auction fees. Alas, there is no budget yet for fiscal 2016, which starts in October.
Both the Job Openings and Small Business sentiment reports suggest that businesses are still having a hard time filling openings, but have yet to pull hard on the salary lever to fill those positions. After years of cheap and abundant labour, businesses need to come to grips with the fact that times have changed. They are going to need to pay up if they want to get some of their positions filled, or at least do more than post a job online and wait for the résumés to flow in.
After moving sharply higher over the past several years, the Chinese government devalued its currency by 1.8%, which is just a drop in the proverbial bucket. If it did one of these each week for the next five, then we might be more concerned. Nevertheless, it is recognition that Chinese growth is not nearly as robust as the government had hoped. And that weakness is not great news for other emerging-market trading partners or commodity producers. Meanwhile, the U.S., which receives just 1% or so of its GDP from sales to China, stands to benefit from cheap commodities more than it will be hurt by weak sales to China. Also, because China has kept such tight control over foreign investment, financial dislocation in China will not reverberate around the world financial markets in the way that U.S. subprime mortgages did.
Retail sales report shows the U.S. consumer is alive and well
As we noted above, the performance of the U.S. economy really comes down to the U.S. consumer. Therefore, this week's retail sales report, which comprises about a third of total consumption, was extremely important. The report did not disappoint on any level. The headline growth figure of 0.6% month-to-month growth annualizes to over 7% and met consensus expectations. Even taking out the strong auto numbers and volatile gasoline sales dollars, month-to-month growth annualized was still an impressive 4.2%.
The year-over-year averaged data both on a nominal and inflation-adjusted basis were healthy and stabilizing at a very acceptable rate. Shifting weather patterns and a temporary, unsustainable jump in many economic metrics spiked results last fall and have brought the average year-over-year growth rates down in spring 2015. Current levels feel steady and sustainable.
The strong report will be good news for the July overall consumption report and gets the third quarter report off to a very good start. The way the numbers are falling currently, consumption growth in the third quarter could easily exceed the 2.9% consumption growth rate of the second quarter. So far, all the Greek and Chinese headlines and volatile U.S. stock markets have yet to scare off the U.S. shopper.
Retail sales report stronger than just the growth figures
We always caution not to read too much into any single retail sales report. The reports are volatile, subject to revision and not adjusted for inflation. That said, there was a lot to like about the July report. Data for May and June was revised upward, probably at least enough to push second-quarter GDP up from 2.3% to at least 2.5% all by themselves. (Including inventory and other adjustments, other recently revised economic reports suggest that second-quarter growth may turn out to be over 3%.)
In addition, the month-to-month improvements in retail sales were unusually broad-based. Of the 19 categories in the report, which includes some overlaps, only three were down: electronics (mainly because of falling prices and a strong performance the previous month), department stores (the 
Amazon AMZN effect; Wal-Mart WMT and Target TGT slow results) and general merchandise (driven by those department store sales that overlap with general merchandise).
So many categories did so well that it's hard to single out one particular theme. Autos, online retailers, sporting goods and furniture all did well, as did building materials. I suppose stronger furniture and building materials may indicate that the stronger housing market is finally having knock-on effects on retail sales.
We were also exceptionally pleased to see restaurant sales move up a strong 0.7%. Restaurant sales are often a strong litmus test of short-term consumer sentiment. Plus, it's unusual to see restaurant sales do so well in a month auto sales are soaring. Time spent shopping for cars and the need for down payment money often spells bad restaurant sales. That was not the case in July.
However, all those restaurant meals did limit grocery store sales, which were one of the few weak categories in July. The online sales number was one that might not have been quite as good as it looks. That number was likely to have been inflated by Amazon Prime Day, which promised Black Friday prices in July.
Whenever we have an unusually good month of retail sales, we like to take a good look at the weekly shopping centre data to make sure things didn't fall off a cliff the following month. So far, the data looks relatively benign for early August, suggesting that July sales didn't take a plunge in August. Some of the nasty headlines from overseas don't seem to be scaring the U.S. consumer just yet.
Industrial production good, but not quite as good as it looks
Headline industrial production surged 0.6% in July, handily beating forecasts of just 0.3% growth. Looking at manufacturing alone, excluding utilities and mining and oil extraction, growth was even more impressive at 0.8% month to month.
A big piece of that surprisingly large surge was a 10% month-to-month "seasonally adjusted" increase in auto production. The auto numbers always get really goofy this time of year as shifting summer auto production line shutdowns (for vacations and retooling) totally confound statisticians. Without the auto production surge, month-to-month manufacturing growth would have still been a quite acceptable 0.3%. There is no simple way to adjust the auto industry summer shifts out of the statistics, although our year-over-year, averaged data softens that statistical blow a bit.
Still, the year-over-year manufacturing data is slowing and looks likely to shrink to a growth rate of 2% or less over the next two or three months, until some of the really strong months of the second half of 2014 begin to drop out of the averages. This should really come as no surprise to our readers as we have been warning about a slow manufacturing sector for some time. However, the data is certainly not in a freefall--things are just growing more slowly. Very cold, terrible weather in 2013 artificially boosted 2014 growth rates while shifting auto plant shutdowns affected the summer months, and oil-related issues crept into the data sets very late 2014 and early 2015, making analyzing manufacturing data a real challenge.
Right now it does appear that things are neither surging nor collapsing, and growth has slipped below the 2.5% trend as some oil-related activity moved from being a big boost to the numbers to a real detriment. We take some solace in the fact that manufacturing industrial production has been up on a month-to-month basis for four of the past five months, despite the oil situation and the strong dollar that has limited exports growth.
Plus, just four of 20 categories were down month to month in July: machinery, aerospace, miscellaneous and textile. A bunch of categories beyond the auto sector grew faster than 1% in July, including wood products, apparel, paper and plastics. Certainly not a world that is falling apart. Manufacturing employment data and purchasing manager data suggest that things are still growing, but not as fast as the heady days when a lot of oil-related activities (that don't show up directly in the mining or energy results) were boosting the data.
We suspect that a return to 4% or 5% growth in manufacturing industrial production is unlikely. Getting back to the trend line growth of 2.5% or more over the next year is a real possibility as growth shifts from energy- and commodity-based activities to housing, plastics, chemicals (the latter two being helped along by lower energy prices), and even some consumer goods.
Job openings and labour turnover report affirms an improving labour market
Job openings have surged over the past year but many openings remain unfilled. The surge is unprecedented in the metric's brief 15-year history. While we have trumpeted the rise in past reports, sooner or later employers will have to pay more to entice workers, employees will need to acquire more skills, or employers will have to lower their standards and offer more training. So far it seems to be a bit of a stalemate as hourly wages have increased ever so slowly.
The report for June, released this week, suggests that slightly stronger hiring trends at least brought the new openings level down modestly. The number of openings dropped from the recovery high 5.35 million in May to 5.25 million for June, which is still miles above the recovery low. A further drop, if not too extreme, could be a good thing as it would indicate a breaking of the logjam that has kept both employees and employers a little too cautious.
One other theory floating around relative to job openings is that employers now require new employees to jump through a lot more hoops before they are hired. These include more background checks, testing, and even special pre-hire projects. This means that every new opening stays open for longer, potentially inflating the statistic. Though there may be some truth to this, the increase in opening longevity suggests that the longer hiring process is certainly not the whole story. At some point, when all these new standards were adopted, openings should have begun to level out. So far, they have not.
The small-business report, discussed below, suggests that indeed, businesses have been a little stingy, contributing to their inability to find new employees.
Small-business compensation plans prove to be a growing dilemma (by Roland Czerniawski)
The NFIB Small Business sentiment report pointed to overall improvements in small- business conditions in July. The overall index rose from 94.1 to 95.4. This number isn't indicative of a boom, but demonstrates that small businesses are doing fine, and definitely much better when compared with years prior to 2014. What we particularly like about this report is its employment portion that tracks a number of important labour market metrics (small businesses employ almost half of all private employees).
Specifically, we like to analyze the proportion of job openings with no qualified applicants, which is indicative of the level of labour market tightness. Additionally, we pay attention to the compensation plans portion of the report. The two measures combined can paint a picture of how difficult it is for small businesses to fill new openings, and how willing they are to raise wages in order to attract more talent.
From the post-recession portion of the data, we see that small-business employers have been slow to announce compensation increases despite the increasing difficulties in attracting qualified candidates, especially as the recovery ages and labour markets tighten. This is portrayed in the chart below by the growing gaps between the two lines, during the 2012–13 period, for example. Since the beginning of 2015, small businesses found themselves in a similar situation where the difficulty of hiring qualified people is now higher than ever, yet compensation plans have actually gone down.
This is somewhat puzzling, but it might explain why job openings in general have risen to historic highs. Higher compensation is the inevitable solution to tighter labour-market conditions, but relatively few businesses are actually acting to remedy this problem by raising their competitiveness and improving the compensation packages. The gap between the difficulty to hire and the compensation plans is now the highest it's been in a few years, but the discrepancy between the two metrics can grow only so wide before converging back again. Given the current high demand for skilled and semi-skilled labour, it will be just a matter of time before workers realize that their leverage to demand higher pay is now the highest it's been in many years.
The U.S. federal government acknowledges its budget deficit estimates were too high
Last month we wrote that many estimates for the U.S. federal budget were too high and that a range of US$385 billion-US$435 billion for fiscal year 2015 versus US$483 billion in 2014 made the most sense. With one more month of official data in hand, the Congressional Budget Office finally threw in the towel and reduced its deficit forecast from US$486 billion to US$425 billion, now in range of our most recent thinking. The data so far in 2015 show spending is up just 5% while revenue is up 8% when adjusting for August payments made in July because of a normal payment date falling on a weekend. Better revenue is a result of higher tax collections due to higher corporate incomes and more bonus and stock market income for individuals.
Notice the other category, which is almost everything except social programs and transfer payments, is actually down so far this year (with the help of revenue from spectrum sales, which shows up as a reduction of costs, and not revenue).
CBO plans to release a new set of projections later this month, which should also show new expectations for 2016 that are likely to reflect the increased revenue pattern, lower-than-expected interest rates, and potentially a very small, if any, increase in Social Security payments in 2016. We hope to report on those new projections when they become available.
Paradoxically, the good short-term news on the deficit could cause both Republicans (who want more defense spending) and Democrats (who want more social spending) to dig in their heels when negotiating the 2016 budget that takes effect in a little over a month. There are some fears that this could cause another government shutdown, or alternatively, both sides getting their increases because the budget deficit has improved so much. We doubt that a shutdown will materialize this time around, and both parties have pledged to avoid this outcome. It's a little harder to rule out more spending increases.
Next week is all about housing, the other key variable for second-half growth
With net trade likely under pressure from a combination of a strong dollar, terrible commodity markets, business investment spending remaining under pressure from slow growth, and government spending still under the gun, economic growth comes down to just the consumer (discussed above), and housing.
Next week should shed a little light on the housing market with builder sentiment, housing starts and permits, and existing-home sales all due. Housing data has been trending better lately, but affordability and lending standards are still not exactly where we need them for a robust recovery. Next week should reveal whether higher prices and increased interest rates have had much of an impact on the market. The builder sentiment report has generally been creeping up and has generally been a little too bullish. At 60 in July the index was at a recovery high and I suspect it may stay there for the August report.
Starts and permits both had an incredible June, generally driven by some expiring tax credits in New York City. Strong permits data reported for June is likely to mean a decent month for starts in July. With the New York City tax credits now extended, it will be interesting to see if the permits data continues to be strong. Permits grew to the 1.34 million level in June, and we would expect to see at least a small decline here.
Starts, on the other hand, should accelerate further as those previously permitted homes are started. Starts are projected to grow from 1.17 million units to 1.24 million, which would be the best level of the recovery. Now, the starts number is still likely to be less than permits, providing further upside potential in the months ahead. We would check out both starts and permits data, excluding the Northeast, to get a clearer gauge of overall strength of the housing market. However, even unbalanced growth will boost the GDP data and other metrics of economic activity.
Existing-home sales, which are also important in driving economic activity, have also been strong recently. The 5.49 million units sold in June were a recovery high. A slight decline in the pending home sales data for June suggests that home sales in July are likely to slip a bit. The consensus is for a drop to 5.4 million units, which strikes me as just a little high. At 5.4 million, the result would still be the second-best level of the recovery.
CPI data is also due next week. The consensus forecast for overall CPI growth from June to July is 0.1%. Reported gasoline prices are likely to show an increase of 1% in July after increasing 3.5% in June, helping to lower the expected inflation rate from 0.2% in June to 0.1% in July. If this consensus forecast is correct, year-over-year inflation could increase to as much as 0.4% as the U.S. begins to lap the start of gasoline price declines. Year-over-year inflation has been close to or below zero for six consecutive months. Housing has kept core inflation quite high and we aren't expecting housing prices and apartment rents to moderate just yet. We will also be watching health-care prices, which have been volatile lately.
