Manufacturing Picking Up Steam; Housing Isn't
Another rise in new-orders data suggests the industrial economy is on solid ground, but demographics are holding housing down.
Overall, it wasn't a great week for U.S. equity markets, with the S&P 500 down 1.75% and U.S. small-cap losses approaching 3% as doubts about the amount and timing of potential federal government policies rose.
Confusion as House Republican leaders pulled their healthcare bill certainly didn't help matters. Given the market's post-election boom had been premised on the quick adoption of several Trump initiatives, the market's reaction, especially on Friday, is not surprising.
European markets were flat on the week, and emerging markets were up almost 1% as these markets could be beneficiaries of delays in those same policies. Not surprisingly, the bond market was up as the feared stimulus measures and the growth and inflation that they might engender moved further into the future. Commodities, another beneficiary of changed government policies, also had a small down week.
After last week's flood of data, there wasn't much news this week. The durable goods report released on Friday suggested the manufacturing sector was picking up steam, with a few caveats regarding backlogs and results from a couple of key sectors: computers and electronics, as well as machinery.
Taking a longer year-over-year view, not much has changed with existing-home sales growth faltering under low inventories and higher prices. New home sales looked great on a month-to-month basis, but the year-over-year data suggested the sector is stuck in a rut of 8%-10% unit sales growth. However, falling growth rates in inflation-adjusted housing dollars (as the units being produced today are smaller and at lower price points) will likely limit housing's contribution to GDP growth in 2017.
This week's data didn't do much to change the Atlanta Fed's dour assessment of first-quarter growth prospects with first-quarter growth now projected at 1.0% versus 0.9% a week ago. Consensus estimates are down to 1.7% after starting the year at well over 2%. For the full year, we are sticking with our GDP forecast of 1.75%-2.0%.
New-Orders Data Suggest the Manufacturing Economy Continues to Pick Up Steam
Sometimes the manufacturing sector gets swept under the rug here in the U.S. because it is viewed as a small and shrinking sector of the economy. That is perhaps an oversimplification.
True, manufacturing only accounts for 8.5% of all nonfarm U.S. payrolls. However, excluding the government sector, it makes up just over 10% of private sector employment. Furthermore, given the slightly higher hourly wage for this sector ($26.35 per hour versus $26.09) and more important, higher average hours worked (40.8 versus 34.4), manufacturing accounts for 12% of weekly private sector wages. In addition, recent quarterly BEA data suggests that manufacturing contributed over 20% of corporate profits in the third quarter of 2016.
With a lot more sentiment markers for manufacturing diverging, we are now focusing on new orders for durable goods orders, excluding transportation, as our best early warning metric for the manufacturing sector. As the graph below shows, new orders have tended to lead industrial production by three to four months. The blue new-orders line almost always makes its move before the red industrial production figure.
The new-orders data for February showed yet another improvement in year-over-year growth rates with even the averaged data set now approach 4% after a bottom near negative 4% in fall 2015, when oil prices began to tumble. New-order growth of 4% implies that year-over-year industrial production growth should exceed 2%, on a three-month averaged basis, by midsummer. The unaveraged new-order growth for February approached 5%, implying more potential improvement in the averaged data sets.
Though we aren't fans of month-to-month numbers because of poorly understood seasonal factors and weather, the month-to-month data looks quite good, too.
The new-orders data has now strung together eight consecutive months of gain since August, which is a very impressive run for a volatile indicator that routinely goes negative, even in a strong economy. However, the category factors were sending mixed messages. Only two of seven sectors were down sequentially in February, after being stuck at three for several months. That's the good news. The bad news: Only two of seven factors showed improvement in the growth rate from month to month. The swings in the categories, primary metals and electrical equipment, were so large as to move up the month-to-month growth rate.
Unfortunately, slowing in two categories, machinery and computers and electronic gear, which have been key drivers of industrial production growth, have seen substantial slowing in recent months. So the manufacturing sector isn't completely rosy, either. Plus, backlogs have continued to erode, extending their losing streak to eight months. Inventories were up a bit, too, on a month-to-month basis. And with prices stabilizing or even up a bit, the order dollar numbers could be a bit inflated. None of these metrics is all that worrisome, but some of the recent improvements in unit-based output may begin to back off from recent highs.
Markit Flash Purchasing Manager Data Continues to Slow a Bit in March Report
The Markit PMI flash data released this week was a bit discouraging, showing its second monthly decline in a row. Given that the ISM data increased in February and Markit data declined, we were hopeful that the Markit Flash data for March would converge on the higher ISM readings.
The new ISM data for March won't be available for another 10 days or so. Today's Markit data suggests that the unusually strong prior ISM number will likely fall back to earth for the March reading. The new-order data discussed above, which is generally a more reliable metric for forecasting, suggests that neither the jubilant ISM data nor the dour declines of the Markit report are a good representation of reality. The new-orders report data suggests steady improvements for aggregate numbers over the months ahead, with some limited worries in a couple of sectors.
World Data Has Generally Done Better, Potentially Pulling Up the U.S.
Perhaps paradoxically, the U.S. manufacturing economy is generally helped by a stronger manufacturing economy around the world. One couldn't be blamed for thinking that if other countries were doing well that it might be at the expense of U.S. manufacturers. However, the U.S. exports a lot of capital goods, used by other manufacturers. Indeed, with some delay, a stronger world manufacturing sector has recently led moves in the U.S. In the most recent manufacturing upturn, both China and Europe preceded the U.S. In the latest flash data, the eurozone continued to gain substantial strength. Over the past several months, European manufacturing and service industry data have performed better than most analysts had expected.
The 'Hotel California' Syndrome Hits the Housing Sector
In one of the most famous lines of the Eagles hit "Hotel California," Don Henley belts out, "you can check out any time you like, but you can never leave." We have used this line before to describe labor market participation rates that count even 90-year-olds as potential workers, substantially understating the real participation rate.
In the case of houses, most retirees are choosing to retire in place. They have checked out of work life but just aren't moving in mass quantities to retirement playgrounds. Doctors, friends, and finances have tended to keep people in their homes until the very end of life.
What's happening now is that the age cohort of a typical first-time homebuyer (31 or so) is about 3.8 million people; meanwhile, the number of people who die each year is considerably lower at about 2.6 million. (While this is a gross simplification because of relative household size and the impact of trailing spouses, it hints at the problem).
Existing-Home Sales Suffer as New Home Sales Outperform
The gap between new-home sales growth and existing-home sales growth has continued to widen. Because of different seasonal factors for new and existing homes, we are forced to use less sensitive rolling 12-month data in order to draw strong conclusions. At the moment, single-family new home sales are growing at about 12% while existing-home sales are only growing at about 4%.
Without Some Help From Remodeling Spending, Residential Won't Boost GDP Much in 2017
There are two key drivers of this divergence: inventory and price. The inventory picture is quite telling, with new home sales inventory growing in the double digits as builders bring on more supply. However, the Hotel California effect, fewer moves as homeowners age (the planned tenure for a retiree is over 20 years and Gen-Y buyer under 10 years) along with an overall trend toward fewer employee transfers have kept a tight lid on supply.
Existing-home inventory growth has been in a long-term pattern of decline since 2014 and inventories have been outright declining since early 2016. Realtors can't sell what they don't have. Certainly, quicker sales and less time on the market have eased some of the pain, put at 3.6 months of supply, it is hard to imagine that sales outperforming inventories can continue indefinitely. This lack of supply has certainly contributed to higher prices, which are also a problem, creating affordability issues that keep young buyers in apartments longer.
New home prices are growing faster, but these tend to be purchased by move-up buyers who have participated in the recent price rally. However, the wide gap between new and used home prices, one of the widest of this recovery, may be forcing some move-up buyers into the existing-home market.
Wrapping this all together, existing-home sales, with low inventories and slow unit sales will be hard-pressed to add much to GDP growth in 2017. In fact existing-home sales brokerage commissions may be down in the first quarter compared with fourth-quarter 2016 and flat compared with first-quarter 2016.
While the unit sales data for single family have remained rock-solid at around 10%, the new home data used in the GDP report have been in a real funk. Despite our glowing summary of the new home market above, single-family home construction was down 6% or so from a year ago in the fourth-quarter GDP release. A combination of increased sales of lower-priced units as a percentage of the total and skyrocketing construction costs are killing the inflation-adjusted new home sales dollars used in the GDP calculation. Year-over-year construction cost indexes are approaching 7% inflation, which was a real surprise to us. On a year-over-year basis, we have another quarter or two of pain, even if prices are unchanged from here on out. Without a surge in home improvements, we suspect that a combination of slow existing-home sales (which is in broker commissions and counts as a housing investment), a smaller contribution from apartment construction, and flaky inflation adjustments mean that housing will likely be a relatively smaller contributor to GDP growth in 2017 compared with 2016. In 2016 housing investment in total contributed about 0.2% to total GDP growth of 1.6%.