Despite a happy ending on Friday, markets had another tough week, with stocks down around the world. Emerging markets were again hit hard, down over 4% as a poor purchasing managers' report from China panicked already frightened investors. Europe did just a little better, down over 3%, and the U.S. did best of all, down just under 2%. Bond yields on the 10-year U.S. Treasury were up ever so slightly from 2.13% to 2.17% as the bond market digested stronger U.S. economic data. Commodities also moved modestly higher.
In addition, Fed speakers, including the Fed chairman, strongly suggested that interest rates need to be higher by year-end. An upwardly revised second-quarter GDP report along with the Fed comments suggested the U.S. economy wasn't on the edge of an economic precipice, about to be pushed off by a falling Chinese economy. We believe there was some fear last week that the Fed might have had access to undisclosed data of some type that presaged weakness in the U.S. economy.
As we mentioned, the world PMI data was the economic highlight of the week, with continued disappointing news out of China that showed continued contraction at accelerating rates. Europe remained solidly in the growth camp, although results weren't as strong as the previous month, while the U.S. remained the strongest of the bunch with no degradation in the September flash data from Markit.
In other U.S. economic news, the second-quarter GDP report was revised up for the second time despite a downgrade in the contribution from inventory building. As usual, the consumer saved the day. In other news, housing remained healthy, although the new home sales data was more impressive than existing-home sales, which are feeling at least a little heat from higher prices and lower inventories. Meanwhile, durable goods orders weren't down as much as expected, and excluding transportation, managed their third straight improvement.
GDP gets a high-quality revision, officially boosting our full-year GDP estimate
Growth in the second quarter was revised up from 3.7% to 3.9%, beating the consensus estimate of 3.7% growth. It was what economists like to call a high-quality revision as the contribution from inventory expansion was now just 0.2% instead of 0.4%, meaning fewer goods were piling up on shelves.
More than offsetting the downward inventory calculation was a large revision to consumer spending, primarily the services-related category. Consumer spending grew at a 3.6% rate in the second quarter, one of the best performances of this recovery and constituting almost two thirds of total GDP growth. Most of the construction-related categories were also revised higher.
The new, earlier flash trade report seems to be helping as there were almost no revisions in net exports, versus the massive revisions under the old regime. Interestingly, despite all the haranguing on the strong dollar and weak world economies, net exports overall were surprisingly a net contributor to GDP. Also of note is that government spending has at least temporarily pulled out of its slump, making its first meaningful quarterly contribution of the entire recovery.
Overall, it was one of the most balanced GDP reports we have seen in a long time, with every category with the possible exception of business equipment doing well in the quarter.
We do need to caution that just like in 2014, part of the reason for the very high GDP growth rate in the second quarter was a weather-related slump in the first and the subsequent good weather bounce. Recall that the first quarter grew by only 0.6%. Combining the two quarters, growth was about in the middle of our 2.0%-2.5% long-term forecast.
We are raising our 2015 GDP forecast from a range of 2.0%-2.5% to 2.2%-2.6%
It's not much of an increase, but we would put most of our emphasis on the high end of the new range, with 2.6% growth now the most likely case. That would require growth of 2.9% over the next two quarters. Given what we have seen of the third quarter so far, the third-quarter portion looks likely to beat that expectation, excluding whatever happens to inventories. Retail sales and auto sales growth look to be better than they were in the second quarter. Autos in particular are on a roll. The September auto number will be out next week, which should help firm up estimates even further. Unfortunately, our longer-term, five-year estimate of GDP growth remains 2.0%-2.5%.
U.S. durable goods orders not as bad as headlines suggest
Manufacturing has been the poor stepchild of 2015 after a fantastic 2014. At less than 10% of U.S. employment, we spent most of 2014 wishing it were a bigger part of the economy. Now in 2015 we are glad it isn't huge.
In 2014, everything that could go right did, and just the reverse is true of 2015. In 2014 the energy industry was still booming, using a ton of metals and pipes and even computer equipment. The auto industry did great, too, helping further drive manufacturing numbers. Industrial production in total was up 3.9% in 2014, well above the 2.5% long-term average. This year, industrial production isn't likely to grow at all, although despite all the turmoil, an outright decline in production appears unlikely.
The durable goods order report is one early gauge of the manufacturing sector. Durable goods, because of long production cycles, need to be ordered well before they are manufactured and shipped. Therefore, orders can provide a window into the future of the manufacturing sector.
At first blush, the news on durable goods was not good. The headline number showed total durable goods orders down 2.2% in August after two relatively healthy months of increases. However, these include aircraft orders, which are highly volatile. In addition, aircraft manufacturers are working off decade-long backlogs, so aircraft production is more tied to 
Boeing's (BA) manufacturing schedule than to incoming orders. And it is production that drives the health of the manufacturing sector and the GDP calculation. This week's report makes it crystal clear that including aircraft orders in the new orders is silly.
Year to date, year over year (that is, the first eight months of 2015 over the same eight months of 2014), aircraft orders are down a stunning 44%, which drives the headline order growth numbers through the floor. Meanwhile, in the real world, aircraft shipments are up a stunning 15%. For slightly different reasons, we toss out the rest of the transportation sector, primarily autos, when discussing durable goods orders, too. Here in the year-over-year data, few new orders are negative and continue to deteriorate.
However, the ray of hope is that, excluding transportation, sequential month-to-month orders have improved in each of the past three months. In addition, the year-over-year comparisons will get easier very soon as the oil-related slump began in the last quarter of 2014. Between the sequential improvement and easier year-over-year comparisons, we suspect that year-over-year order growth could be positive by as early as December. The only potential fly in the ointment is that commodities seem to slipping again, which will potentially cause some slippage in some order categories. However, I suspect some of the recent strength in the housing sector may mitigate some of the commodity-related weakness in the fourth quarter.
Business spending on equipment not likely to be a black hole in the third quarter
This subsection of the durable goods orders report provides an excellent proxy for business spending on equipment as reported in the quarterly GDP report. We have been complaining about poor business investment spending for months, noting that a lot of corporate cash flows have been going to acquisitions and stock buybacks and not for expansion. Now with more fears of worldwide growth slowing, many commenters seem to be expecting a complete collapse in spending. On that count, this week’s report was reassuring. If the July and August shipping data is not revised, and things just hold flat in September, it looks as though capital goods shipments will have a net neutral effect on the GDP, hardly the disaster that some envision.
World PMI data OK in developed world; China still a problem
The flash world PMI has become the second-most watched economic indicator around the world, perhaps behind only the Federal Reserve Open Market Reports. And it is almost always market-moving.
It is watched closely because it is one of the first available reads on the current month, and it puts a lot of worldwide manufacturing data on a common base. This week’s flash data showed China's manufacturing still shrinking and continuing to get worse. Europe is still growing, and the decline in the September manufacturing reading was small. The U.S. data continued to lead the pack with the highest percentage of companies reporting growth, and it was the only major region reporting no deterioration between August and September. Readings over 50 mean more companies are reporting growth than shrinkage. Although the absolute number is probably the most important number, the trends in the month-to-month change are important, too.
The China numbers did have stock market effects across all of Asia on Wednesday, as the broadest Asian stock market indexes were down over 2%. The weak China numbers also helped push oil down close to 3% on Wednesday. The raw Chinese overall reading was now at its lowest level in six and a half years. Most of the individual subcomponents looked equally bad, including current production, new orders, and employment. The press release made the points that the benefits of increased fiscal stimulus had not really kicked in just yet and that at least some of the decline in manufacturing was due to some needed structural changes. While this all may put a little too much positive spin on the numbers, the manufacturing numbers don't seem catastrophic just yet.
On the other hand, European markets had a good day when the manufacturing data was released, with everyone pleased that the index was still near its 2015 high as looser monetary policies and cheap oil continue to work their magic. Investors were particularly pleased that new orders, the most forward component, remained close to their 16-month high. Backlogs also soared, suggesting more room to raise production in the months ahead. The service sector did even better, with new orders at their highest level since April. Overall, the Markit analysis suggested that the September data supported another quarter of 0.4% GDP growth, in line with the second quarter and putting the eurozone on pace to grow at 1.6% for the full year.
France continued to do poorly and Germany didn't do as well as it has been. The rest of Europe is growing somewhat faster, however, the improvements also slowed in September. We are modestly fearful about what the Volkswagen (VLKAY) scandal will do to Europe's results as it is one of Europe's largest employers.
The U.S. data continued to look good, though we always prefer to use the ISM data, available next week. That series has a much longer track record.
Housing data continues mostly positive
Residential housing activity has been an important contributor to GDP growth in 2015, kicking in 0.3% to GDP growth in each of the first two quarters of 2015. We think the full-year contribution could be just a touch better; however, more second-half growth is likely to come from new construction versus real estate commissions on existing homes, which were particularly strong in the first half.
This week's data for August bear that out with decent levels for both new and existing-home sales, though the data clearly showed that new homes data is holding up better than existing-home sales as higher prices and lower inventories begin to slow the excellent growth seen previously in existing-home sales. From an economic viewpoint we are pleased to see excellent growth in single-family construction as it is potentially a bigger driver of economic growth than the sale of an existing home.
New home sales hit a recovery high
At 552,000 annualized units, new home sales set a new recovery high and were up 5.7% sequentially and single month, year over year by 21.6%. Growth in transaction value was slightly less, as the average price of new homes is being pulled down by more interest and availability of smaller, more affordable homes. The year-over-year regional data showed low-double-digit growth rates in all regions, although the very large Southern region and the tiny Northeast region showed the best growth rates.
We were also pleased to see that almost half the month and year-over-year gain in new home sales came from the sale of homes still on the drawing board and not even started (versus partially and fully completed homes). Not-started homes now account for 38% of all home sales versus just 33% a year ago. Not-started homes are now the largest slice of the new home market. We love to see that because it will lead to more housing starts and economic activity later in 2015. It also indicates a combination of tight markets and enthusiastic homebuyers when buyers are willing to purchase a home without even seeing the actual building. The lower selling price points combined with higher inventories (though still not high enough) also bode well for sales for the rest of the year.
Existing-home sales take a break
As signaled by pending home sales data, existing homes were off their previous highs. Sales dropped from 5.6 million units in July to 5.3 million units in August, which was below the consensus of 5.5 million units. A combination of low inventories and higher prices may be beginning to take its toll. Unlike new home sales, inventories are down from a year ago, and prices are up for the average existing-home sale versus down new home sales. The table shows that on a moving-average basis prices are up 4% while inventories are down 2.4%, not a particularly good omen for accelerating growth rates in the back half of 2015.
Again, 5.3 million units sold is nothing to cry about. Though it is down month to month, the August sales data is still the fifth-highest month of the recovery. It also pulls up the year-to-date average sales level to 5.22 million units, which is within spitting distance of our 5.3 million unit full-year forecast.
FHFA points to price pressures brewing in the existing homes market (by Roland Czerniawski)
The FHFA home price index increased 0.6% in July, marking a 20th consecutive monthly increase. Year over year, price growth stood at 5.8%, up from 5.5% recorded in June. When averaged for three months, year-over-year growth was at 5.7% in July, pointing to a continued trend of rising home price growth.
One of the reasons behind the increasing pace of this price growth is stronger demand while existing-home inventories remain below the 2014 levels for a number of months now. We forecast FHFA price growth to reach 5.0%-6.0% at year-end, and the breach of the upper end of this range now appears possible.
Regionally, the Pacific, Mountain, and West South Central census areas drove the faster price growth. In addition, Colorado and Nevada are two states that continue to grow double-digit year-over-year growth, while overall U.S. composite growth remains shy of 6%.
The FHFA Home Price Index, which tracks prices on homes sold to or guaranteed by Fannie Mae and Freddie Mac, has been consistently increasing for more than a year now, and as a result, the index is just now 1.1% below the March 2007 peak. It's worth noting that the index tracks higher-quality mortgages, which were subject to less severe price drops during the collapse of the housing market. Nonetheless, it appears that in absence of a major cataclysm between now and the end of the year, national home prices might actually mark a full recovery--at least according to the FHFA metric.
The week ahead
There is a lot of data next week, and it is hard to pick which piece is the most important. The least important is probably the ISM purchasing manager reports. Manufacturing is having issues again, and the problems are well-known. The consensus is for the report, which is been eroding for almost a year, is for a little stabilization, as housing does better and some non-energy sectors show signs of life. However, the regional purchasing manager reports so far in September have been a little soft. The one reason that this report could be important is if there is a reading below 50. A lot of reporters read a lot of magic into crossing that demarcation. Historical studies generally indicate that somewhere in the mid-40s might be the time to panic, but not when a reading cracks below 50. My best guess is we won't fall below 50, but if we do, I am not going to bail on my current forecasts.
Personal income and consumption data will prove to be quite important only because the data will give us two months of the consumption data that drives the consumption portion of the GDP report, which affects two thirds of the entire GDP calculation. The retail sales part of the equation, so far, seems to suggest that goods consumption could accelerate in the third quarter compared with the second. Still, the consumption report is inflation-adjusted and the retail sales report is not, creating the potential for surprises in either direction. The consensus is for 0.4% personal income growth and a slightly slower consumption growth rate of 0.3%, both healthy numbers and consistent with quarterly consumption growth of 3% or more. And with the income numbers in excess of consumption, there is potential for still more growth in consumption in September--that is, if the consensus is correct. If anything, the consensus looks low to us.
Autos have also been a key driver of recent U.S. economic strength. Autos have had some help from the calendar and new products, and limited competition from used cars. It will be interesting to see if autos can extend their string of successes into September. The official consensus is for sales of 17.4 million units in September, off modestly from August's record pace of 17.7 million annualized, seasonally adjusted units. However, the trade rags are pointing to a late Labor Day and continued drops in oil prices with most forecasts centering on another month of 17.7 million units. We will see who gets it right: the industry experts or the economists. Whoever is right, those are impressive numbers and suggest autos will continue to help out the GDP calculation.
Employment data will probably get the most press next week but will shed little light on the economy. Employment is at best a concurrent to slightly lagging economic indicator, not the driver. For some yet-to-be-determined reason, employment surged in the fourth quarter of 2014 and the first quarter of 2015, while the economy slowed. That set us up for relatively disappointing employment reports in the second half of 2015. There is no need to panic, but the consensus is for job growth on a month-to-month basis to remain below 200,000 (190,000 to be exact) following just 173,000 jobs added in August, which was viewed as disappointing. That would pull the average job growth over the past 12 months down to 243,000 jobs, which is still above the 220,000 level that we believe represents sustainable employment growth given our GDP forecast and some productivity improvements.
Pending home sales and construction data will also provide some interesting reads next week and give us a little better handle on residential investment spending for the third quarter.
