It wasn't a great week for world markets, with most equity markets down 0.2% to 2.0%, and bond yields down also, with the 10-year U.S. Treasury yield sliding to 1.82% from 1.89%. These moves seemed to be a result of an uninspiring corporate earnings season and soft economic news out of the U.S. Poor AppleAAPL results and down iPhone sales certainly didn't help technology stocks. Oddly, commodities, which usually move markets in the same direction, had a great week, with commodities, gold and oil all up over 3%. We didn't see any real news on that front. Nevertheless, some decent weekly oil price gains added up to a full 20% jump in oil prices in April. Good news for oil companies and equipment producers--not such great news for consumers.
Central banks were in the spotlight again this week, too. The Bank of Japan refused to provide more monetary easing at its most recent meeting, and the U.S. Federal Reserve managed to not say much at all at this week's meeting. The Japanese announcement had meaningful market impact and caused the U.S. dollar to weaken further. Unfortunately, central bankers are having increasing difficulty fighting slowing demographic growth pressures with easy money policies. Also, with so many countries adopting similar policies, it hasn't provided some of the currency depreciation benefits that helped earlier adopters of easy money policies.
The GDP report was about on target with recent whispers, with sequential, annualized growth of just 0.5% in the March quarter. That extended the run of poorly performing first quarters to potentially four of the last five years. Some of the forecasts from the Atlanta Fed helped cushion expectations. Its exceptionally detailed projections, more of a compilation of already-released data than a from-scratch forecast, were awfully close to the mark, with an overall forecast of 0.6% growth versus the actual 0.5% result. The category data was almost as equally correct. Year-over-year GDP growth of 1.9% was much more benign. Still, as some of this recovery's rocket engines fade (shale oil, autos, smartphones), even year-over-year growth has softened some. We are holding fast to our full-year forecast of 2.0% to 2.5% growth, with the high end of that range looking to be at risk.
Housing, which looked good in this week's reports, should continue to be helpful. It was one of the few real bright spots in the first-quarter GDP report. New durable goods orders continued their overall malaise this week, indicating that the manufacturing sector is basically treading water, not really getting worse or better from a relatively depressed level.
The U.S. economy: A one-trick pony named the consumer
U.S. GDP growth between the fourth quarter and the first quarter came in at 0.5% annualized, largely as expected and in line with we have been preaching about for weeks.
For some perspective, economic growth has averaged 2.0%-2.5% during this recovery and just over 3% for the last 50 years. Sounds scary, right? Because of sources like Atlanta GDPNow forecasts and a few vocal economists, the weak report came as absolutely no surprise. U.S. equity markets were up on GDP day after some modest weakness early. That early weakness was due more to a lack of new monetary action from the Bank of Japan and disappointing Apple earnings than the GDP report.
The fine work of the Atlanta Fed combined with a hard-learned lesson that the first quarter is almost always weak and statistically flawed may have kept markets unusually calm. Also, the first quarter still showed relatively reasonable growth of 1.9% compared with last year's first quarter, though we admit that glacial slowing in this metric is a reason for at least some modest caution. And with slower (or similar) economic activity, employment growth, the markets current solace, looks ahead of itself.
Sequential first-quarter data terrible, but year-over-year GDP data not bad
The worthlessness of the quarterly sequential data becomes immediately clear from the table above. Over the nine quarters of sequential GDP growth shown, the range has been an astounding negative 0.9% to as high as 4.6%. However, the more reliable year-over-year data shows a more logical range of 1.7% to 2.9%. The year-over-year growth was even more stable at 2.4% in both years. Year-over-year employment growth has also been amazingly stable at around 2%, suggesting that the full-year numbers or the quarter-over-quarter data is the right way to analyze this data set.
The slowing year-over-year trend bears watching
The slowing year-over-year pattern has been visible since the first quarter of 2015, as a number of readers have pointed out in the past. We had not been terribly alarmed because first-quarter 2015 growth (2.9%) looked unusually and unsustainably high because of a more normal winter after a snow- and cold-addled 2014. Second-quarter 2015 (2.7% growth) benefited from an end to port-related shipping issues. Now, though, we have had three quarters centered around 2% and showing the glacial slowing that we mentioned earlier.
So while most economists are pooh-poohing the soft first quarter, we are at least modestly concerned about the year-over-year trends. This recovery will be seven years old in June. While not a barn-burner, it's been a relatively consistent grower. Instead of the more normal huge bounce in interest-rate-related sectors, housing and autos, those gains have been spread over many more years. Autos just got back to prerecession levels in 2015, and housing still isn't even close. That has made for a more sustained recovery and the potential for even more gains.
This recovery has been driven by several smaller micro trends and not one theme
If one thinks of recent recoveries, many had one key driver that was quite visible. In the 1990s, it was the Internet boom, and in the 2000s, it was real estate. This recovery has instead had a collection of smaller drivers, the gas and oil shale boom, Boeing's BA big 787 ramp-up, and the auto recovery with its very favourable mix shift to large motor vehicles and maybe Apple smartphones (new slowing there was evident in Apple's earnings release this week). These have been slowly brewing improvements that each added a bit more to GDP growth. Now the trends in at least some of those categories are going the wrong way. What will replace these drivers remains more murky than we would like.
We have been writing about the potential harm of a slowing auto industry that will increasingly depend on population growth now that we have recovered to previous highs of about 17 million units. The end of the heady days of growth got a bit of a reprieve in 2015, as lower gas prices drove demand for larger, more profitable vehicles. Motor vehicles drove a massive hole in both consumption and business investment data in the first-quarter report, taking as much as 0.8% off of sequential growth rates compared with trend (though production rates and inventory building mitigated some of the sell-through issues in the GDP report). Although faulty seasonal factors and antiquated business-day calculations drove a bunch of that decline, the days of heady growth from this sector are behind us.
The gas/oil shale boom, especially in its early days, went almost unnoticed by a lot of macroeconomists. While it didn't necessarily drive a needle-moving direct contribution in one year, it did touch a lot of industries--computers for pinpointing drilling, railroads for shipping, quarries for sand and metals for drilling to name just a few. Many of these beneficiaries have now seen long-simmering growth go dramatically in reverse. The oil-well portion of business structures collapsed in the first quarter, taking business investment spending with it. Mining structure spending, on an annualized basis, fell 86% in the first quarter. Without this dramatic decline, spending on structures such as office buildings and factories would have been up, not down as reported in the GDP report. Mining equipment has also been on a steep descent.
With some of the key growth sectors of this recovery in decline, it might be more realistic to think in terms of the very low end of our 2.0%-2.5% GDP forecast. Even that might be slightly optimistic. The good news is that continued spending on housing, which is benefiting from demographics and a natural ramp-up in government spending, may offset some of that weakness. With consumer incomes continuing to outrace spending, we are hopeful that this sector will continue to hold its own. That could hold true even with a potential for slowing employment growth.
First quarter driven by consumer, not out of line with history
We have been struggling for some time with how to present some of the sequential GDP data. We have decided to look at the data using the average contribution by sector (the contribution is the sector growth rate times its relative size in the GDP calculation; sectors sum to total GDP). That way, it is easier to spot outliers and change. We have done this for the last 16 and the last eight quarters to help demonstrate the dramatic shift to services. We have also averaged the last five first quarters, which shows that despite seasonal adjustments, the first quarter is far weaker than any other quarter of the year. Some early low numbers for the government do skew this calculation to the downside, but even without the government, GDP growth has been regularly lower in the first quarter. The average of all the quarters is close to 2%, while the first quarter barely manages to get to 0.6%. That hard-learned fact is why the markets are yawning about the soft first quarter.
First, note that adding the two consumer categories produces 1.2% GDP growth, well above the total GDP growth rate of 0.5% because of three notable negative categories. That's why we are calling this a one-trick-pony economy. In fairness, housing was and continues to be a meaningful contributor to growth, though that is really about the consumer, too. Government is finally back in the contribution category, but it is still not a huge number.
Two categories, investment spending and consumer goods, disappoint
Next , the two key problem categories are consumer goods and non-residential investment. The consumer goods short is half-related to autos. Consumer motor vehicles took 0.4% from normal GDP growth trends. While the group is indeed slowing, faulty seasonals are the key contributor to the poor quarterly performance. Even in a slow auto world, autos should not detract from GDP. Warm weather and poor apparel sales and winter sporting goods also temporarily pressured the goods category.
Interestingly, motor vehicles were also a key part of the business investment spending problem. Of the 1.4% swing in investment spending from trend, almost a third was due to transportation equipment. Oilfield equipment and structure spending also weighed on the business investment spending. Without these two energy-related issues, business investment spending might have been close to flat. Add back the artificial auto decline, and business investment spending would have been nicely in the black.
Interestingly, the market was warmed up for some very nasty export and inventory data. Those did indeed hurt, but a last-minute soft import report and large build in auto inventories limited the damage to relatively normal first-quarter levels. Inventories and net exports subtracted about 0.3% each from the GDP calculation. Unfortunately, the GDP problems were more related to fundamentals, including motor vehicles and oil issues.
Both new and pending homes suggest continued shift to new home market
New-home sales continue at a brisk if not necessarily accelerating growth rate. New-home sales, which include only single-family homes, are the bedrocks of residential investment. Recall that was one of the real bright spots in the most recent GDP report. Apartments rack up unit growth numbers but contribute less per unit. More favourable demographics have added some new life to new-home sales. The trend isn't perfect here, with growth appearing to slow, but some of the 2015 growth numbers were inflated a bit because of cold-weather effects on building in early 2014.
With builder sentiment higher than a year ago and demographics (more 31- to 33-year-olds), we think full-year growth will be around 10%, a forecast we have carried for some time. While the data for March was below February's report, we note that a pattern of large upward adjustments continues.
Pending home sales, which drive existing-home sales, performed admirably in March, too.
With a reading of 110.5 pending home sales were up 1.5% from both a month ago and a year ago, suggesting that existing-home sales should continue to moderate as we have expected.
Based on the moderating pending home sales, the existing-home sales-growth rate should continue to drop back to our forecast 3% to 5% rate, slower than last year's 6% rate and lower than the surprisingly strong first three months of 2016.
Our broken record on manufacturing continues: The bottom still appears at hand
For probably the last six months, we have indicated that the year-over-decline in new orders and industrial productions was moderating and that we would see modest growth in the near future. Unfortunately, each month, the prior month's data gets revised lower, but the moderating slowdown seems tantalizingly close. (It's the reverse of what we have been seeing in the new-home market, where the data is revised higher every month.)
The year-over-year moderating trend is continuing, as declines have slowed from a high of 4% to just 1%. Also, we have been faked out on occasional huge pops in the monthly sequential data, but in retrospect, those look to be statistical mirages. There is almost no growth in the monthly data when three months are averaged together. We have to go all the way back to August to find a three-month average of monthly gains that is greater than zero. If that doesn't turn around soon, the improvement in manufacturing growth is likely to stall out slightly below zero. That will put a little more emphasis on the ISM purchasing managers' data due next week for some additional clues.
We are a little worried that auto inventories are too high and suspect that as those work their way down, industrial production and durable goods orders could have another round of disappointments. We know Chrysler has shuttered a few plants and suspect others may need to take action. On the other hand, inventories of crossovers and SUVs are probably too low. And supply chain issues from the Japanese earthquakes will continue to bounce through the data, further confusing matters. For now, the manufacturing sector continues to weigh on broader economic data sets, though not disastrously so (with the exception of oilfield equipment).
On a brighter note, four of the seven categories of durable goods orders showed growth in March: primary metals, transportation, machinery and other. And it's worth mentioning the headline number for all durable goods, which included large Boeing orders, was up 0.8% for the month. Excluding transportation, sequential durable goods orders were down 0.2%.
Favourable trade deficit reductions contain some very dark clouds
The new advance report on trade for March showed a shocking drop in the trade deficit in goods from US$63.4 billion in February to US$56.9 billion in March. That is one of the larger drops we have ever seen in a single month. We suspect that eventually some the huge drop will get revised away or show up as an unusually wide number in a future month.
How we got to such a large drop is also a bit suspect. Headline goods exports were down 1.7% sequentially, which was near expectations and recent trends. Imports, instead of growing, also fell 4.3% across a very broad range of categories. Consumer imports fared worst of all, dropping close to 10% month to month. We will be racing to check cellphone imports when the full report is issued next week. Whether the sharp drop in imports was part of a short-term inventory correction in a world of too much inventory or the start of something more ominous is yet to be determined. A broader report on trade due next week, this one adjusted for inflation and including services and product details, will provide some answers, but we suspect that economists will have to wait a few more months before they can draw any firm conclusions on trade. A weaker dollar should have been some help on the trade front, but not as much as we got in this oddball report.
Payroll report and motor vehicle sales key next week; manufacturing and trade data will need a look
With GDP growth in a funk the last two quarters and year-over-year growth settling at 2%, the year-over-year trend employment growth rate of 2% looks suspect and vulnerable. Because of productivity growth, employment growth should run at least a quarter to half a percentage point below GDP growth, or 1.5% to 1.75%, which equates to about 175,000 to 205,000 jobs per month. That's not far off the 200,000 job increases forecast for April. However, it is still below the 12-month average of 234,000.
Given the low unemployment claims rate, it doesn't seem like April should be the month for anything particularly shocking on the job front. However, we will need some months of growth below 200,000 jobs to pull the average, not just a one-month report, below the 200,000 level. Other sloppy economic data has increased the focus on the often volatile employment report. A poor report would have a disproportionate effect on markets. As we approach normally strong hiring months, we remain concerned. Recent winter months are known more for layoffs and less for hiring. We know that employers have been loath to fire anyone, which has helped the employment reports. Now, as we approach big hiring months, we'll see if there is renewed interest in new hires. If not, we are likely to see a poor employment report somewhere in the near future.
Falling new-vehicle sales to individuals and corporations have driven giant holes in consumption and business spending reports. Complicated seasonal and selling-day calculations have hurt auto sales over the last six months after providing an artificial boost the prior six months. The April sales report will contain another one of those extra selling days that has a tendency to depress the monthly sales counts. If total sales were identical in both April of 2015 and 2016, government statisticians would turn that into a 4% decline (one less day divided by 25 selling days). That said, hopes are still high that auto sales will break out of their bad mood in April. The economist consensus is that sales will rebound sharply from their poor 16.5 million unit annualized selling rate in March to a whopping 17.5 million units in April. The industry trade rags are slightly more pessimistic at 17.4 million units. Even that number would be higher than the strongest number of the first quarter, setting us up for a rebound in quarterly sales and GDP growth.
We have already mentioned that the full trade report should provide some clue as to why imports fell off a cliff in March, and which countries may have been the victims. The combined services and trade deficit is expected to fall sharply from US$47 billion to a mere US$40 billion based on already-released goods data.
The ISM manufacturing report, which has shown some nice improvement over the last few months, will provide a few clues on U.S. manufacturing. The consensus is for a small drift down to 51.4 for March from 51.8 in April. Some auto-related issues could affect the report (Chrysler shutdowns and earthquake effects). Auto issues always seem to have a disproportionate effect on ISM reports. Of course all eyes will be on the Chinese PMI data, with markets sensing and expecting some improvement in the latest data. Poor U.S. import data does raise some questions about Chinese export potential.