Is it up, up and away for the U.S. housing industry?

Other big news this week was that U.S. housing data, though not fantastic, appears to indicate that the housing industry has established a bottom. Early spring shoots are clearly beginning to turn into new blossoms, especially in the new home sector. Don’t be fooled by some flaky month-to-month data points; the pattern in the yearly statistics appears to be moving up, up and away for now.

Manufacturing going from bad to worse

We have been among the more bearish on the manufacturing sector and even we were disappointed by the durable-goods orders report. The headline number actually seemed quite bullish, with new orders increasing a surprisingly strong 4%. However, that good news was almost entirely from the transportation sector, including autos as well as civilian and defense aircraft, with orders for these goods up a stunning 13%. Excluding the transportation sector, durable-goods orders were down 0.2% between February and March, marking the sixth month-to-month decline in a row. Even the single-month year-over-year data is down 0.2%, and the three-month average is down almost to zero. Five of the seven individual categories were down, with only the aforementioned transportation sector and computers showing gains in March. Even our year-over-year, three-month average isn't painting a pretty picture.

News for the longer-lasting capital goods sector didn't offer any more hope than the entire nontransportation book of business did. It's not just small business or some trimming of short-term business supplies that is hurting the numbers. Big businesses, too, appear to be afraid of the future.

Nondefense capital goods orders have been going down for a month longer than regular orders and are even closer to falling below zero on a year-over-year average basis.

Recall that durable-goods orders of all kinds are important because those orders eventually turn into real-world manufacturing on a production line with workers and robots and eventually are shipped, perhaps months later. Currently shipments of nontransportation are running 5% higher than the order rate, suggesting more bad news for industrial production and manufacturers in the months ahead. Obviously with the West Coast port issues being settled in February and weather improving sharply in March, this week's poor numbers are not a one-off. Weather and special factors were not the driving cause of the manufacturing weakness in January and February.

Not in any particular order, a poor commodities market, a weak housing construction market (at least in the prior six months), less need for oil- and gas-related equipment and supplies and a strong U.S. dollar have taken the wind out of manufacturing's sails. Unfortunately, energy and the dollar are not issues that are likely to be fixed soon. The good news is that the decline in the commodity sector may be slowing, and housing may take the baton from the energy sector as the engine of manufacturing. And civilian aircraft manufacturers have so many years of backlog in front of them that it will be hard for this sector not to grow 5%-10% every year for the next five or 10 years, offsetting some of the weakness in nontransportation goods.

World manufacturing looks even worse than the U.S. data

Markit Flash Purchasing Data was disappointing for not only the U.S., but the rest of the world as well. The data wasn’t awful, but all three major manufacturing regions listed in the following table, plus Japan, registered down readings between March and April despite all the supposed tailwinds including better U.S. weather, the port strike settlement and weaker currencies for many regions (though not the U.S.).

Though the readings did decline sequentially in all markets, the U.S. and Europe both stayed well above the 50 reading that generally separates growth from declines. However, China has been at or below that 50 level for five of the past six months. A strong currency has not helped matters for China, as is the case with a softening real estate market. Prices and orders in the Chinese housing market have been slumping for over a year, and it appears that weakness is finally affecting the manufactured goods that go into building and furnishing a house. The overall China reading was the worst in a year, and most individual categories looked as bad as the headline number. Japan dipped back below 50 for the first time in nine months, capping a string of nice improvements. Europe held up the best and declined the least in April, as quantitative easing probably helped that economy improve. If Europe can hold the April month PMI levels for two more months, an increase in the second-quarter GDP growth rate for the eurozone should be stronger than the 1.2% annualized rate of the first quarter. Some concerns about Europe are that the data out of France still looks terrible and that while Germany remains strong, it saw a decline in the April Flash reading. Outside of France and Germany, admittedly the largest countries, 16 of the 18 other eurozone countries saw improvement in their PMI readings, a pretty amazing feat. The PMI readings outside of France and Germany are at a recovery high and are now at higher levels than at any time since 2008. Our guess is that Germany, the U.S., Japan and China are still hurting from the commodity bust that is now well into its fourth year. The relative lethargy of the manufacturing sector around the world clearly shows that there are problems beyond currencies and the U.S. energy sector.

Housing to the rescue of the manufacturing sector?

Residential housing still accounts for an anemic 3.1% of the U.S. economy, well below its long-term average of just under 5% as shown in the following chart. This includes remodels, commissions on existing homes and both single-family and apartment buildings. While the mix of these various categories is likely to be significantly different in this recovery, we still believe that housing-related investment will approach 5% of GDP over the next five to 10 years, providing a long and sustainable tailwind to the otherwise poor, demographically challenged GDP growth rates. That's why housing is so important to this recovery.

Thus far the housing recovery has been slow, and 2014 was particularly slow and very disappointing. At the moment everybody seems to have kind of forgotten about or given up on housing as an engine of growth. While there are a number of tactical short-term reasons that housing has been weak (affordability, student loans, tough loan standards), these aren't the whole story. The entire housing industry was totally trashed. People who built the homes, bankers that lent to the industry, readily available tracts of desirable land to build on--all are gone or diminished. This supply side of the issue has kept on the conservative but still-too-optimistic side of the housing industry. Most economists have believed that it was just a matter of time until we got back to needing 1.5 million units of housing per year (versus disappointing housing starts of just over 1 million recently). The demographics seemed irrefutable based on births, deaths, normal destruction trends and household formation rates. In the long run, I still believe that they are. But getting back to those 1.5 million units (which is just two thirds of the previous high of 2.2 million units) is not fast or simple, nor is it likely to be. More likely is a slow improvement with occasional speed bumps like in 2014. And hopefully we will get a potential bump in activity when rates begin moving up and the panic factor sets in.

Existing home sales bounce back

So far, there is no panic-buying in the existing home market, just slow improvement. In March, existing home sales hit 5.19 million, which represents the third-best level of the recovery. Sales were terrible in February at 4.88 million, and the small increase represented in the consensus estimate of just 5.03 million units proved to be way too low. Rolling three months together, the sequential pattern doesn't look all that great, but at least this metric has stopped going and should almost automatically bounce back as the poor winter months drop out of the averages. Always remember that this is not the best way to look at the data. But it is among the more popular ways of looking at housing data in the financial press.

One very bad consequence of the sequential numbers shown in the previous table is that if one looks at total house value sold, the three months ending in March will be down from the three months in December. That means the housing component of GDP is likely to be down between the fourth quarter and the first quarter. This will put even more pressure on the already worrisome potential for a sub 1% GDP growth rate for the first quarter when it is released next Wednesday.

Year-over-year existing home sales data shows a steady run of improvements

As I have said many times, year-over-year data is my preferred way of looking at housing data. Year-over-year data has improved for an impressive six months in a row and is now in what we believe to be a long-term sustainable growth rate of 5% to 10% in units and perhaps slightly higher when including price increases.

New home sales surprisingly strong, up 20% year over year

New home sales growth has strongly outpaced existing home sales recently, after a period of looking similar to or even worse than existing home sales. We like to see new home sales coming in better than existing homes because they make a much bigger contribution to employment and GDP growth rates. In fact, there is even a small chance that new home sales, which have improved sequentially, might be able to bring the residential fixed investment category out of the red in the first quarter despite the poor showing for existing homes noted above.

We couldn't help but peek at the numbers not seasonally adjusted, and to our surprise both February and March showed one-month sales of 45,000 units, both records for the recovery. And while we have seen an occasionally good month, it has been highly unusual to see two of these in a row, which is why we believe the recent housing market increases are the real deal.

Demographics working toward first-time buyer

One little-recognized factor in the housing market collapse beginning around 2006 was the demographics. Most first-time home purchases have been made at age 31, a number that remained incredibly stable over time and even economic conditions. The number of people turning 31 reached a peak in 1992 at 4.2 million people and began a slow slide to 3.1 million people by 2006. While we improved back to over 3.6 million people in 2015, we are only halfway back to peak levels (3.67 million in a range of 3.1 million to 4.2 million). This should mean better news for those serving the first-time buyer in the years ahead.

Home price growth begins to heat up

According to the Federal Housing Finance Agency, U.S. home prices increased 0.7% in February, and the year-over-year growth stood at 5.1%. That’s a sizable increase from the much lower 4.5% pace recorded in September and October 2014. After a brief period of stabilization, the pace of home price increases has begun accelerating again. The year-over-year three-month average, which is typically a powerful trend-telling tool, stood at an elevated 5.3% level for the second consecutive month. We estimated home price growth in 2015 at 4%-6%, and given the relatively high demand coupled with unusually low inventory levels and lethargic housing starts, the high end of this range is now likely. The price growth dynamic is largely dependent on the magnitude of the typical inventory rebound that comes with the spring season. If builders get to work more aggressively than usual, it could put a lid on future price increases. On the other hand, low land availability in regions such as New England and Middle Atlantic will be difficult to overcome. States in those regions might experience more persistent price increases going forward.

Next week: GDP, autos, purchasing managers data and pending home sales

Psychologically, next week's GDP report will be the most important but the most unpredictable. Sequentially GDP was up a stunning 5% in September, which drove a lot of activity, particularly hiring in the fourth quarter. Then growth shrunk back to a more normal and sustainable 2.2% GDP growth rate in the fourth quarter as exports and government became big detractors to the GDP calculation. Hopes were high at the beginning of the year for a quick rebound to 3% growth in the first quarter led by the consumer. However, the weather, West Coast port issues, export issues due to the strong dollar and a stubborn consumer are widely expected to limit growth to 1.5% or less on a sequential, annualized basis. The consensus is lower at 1.2% growth, and we think GDP growth will be lucky to breach the 1% mark. The consumer is likely to grow less than 2% in the first quarter after growing over 4% in the fourth quarter, which is the primary reason for our short-term pessimism. And unless the government makes a heroic assumption about March net exports, they are likely to be a meaningful detractor from GDP in the first quarter (though slightly smaller than in the fourth quarter). Government and business investment aren't likely to be too influential in either direction. Inventory and exactly how inflation flows through the calculations are potentially large and important numbers within the GDP forecasts. These factors are really impossible to project and include government estimates instead of hard data. So we could theoretically have no growth or 2% growth, but check the inventory and export data before panicking or breaking out the champagne. I am sticking with my forecast of 0.5% growth for the first quarter.

Autos are always important to the economy, and with much of the manufacturing sector beginning to fall apart, we could really use some good news out of the auto sector. March was exceptionally good, with 17.2 million annualized, seasonally adjusted units (though aided by a day count issue). Consensus is for minor slippage to 17.1 million units, which would put the second quarter off to a good start. Because the day count factor won't be present this month, I am thinking auto sales could drop back to 16.8 million. As long as we stay above 16.4 million units, I would not be too worried.

Pending home sales will also be important, as we have laid out the significance of a good housing number earlier in this piece. Housing could act as an offset to energy-related issues for the rest of the year. Though new home sales are important, the pending home sales index should at least give us some hints as to the customer's state of mind. Given a large bounce in pending home sales for February (up over 3% sequentially), economists are very wary that the March number could be horrible, with expectations for a 1% decline in pending home sales. I think these concerns are overblown, and with all the activity in new homes, I suspect that the original homesteads (which are almost always sold near the time a new house is purchased) will go on the market soon, creating more supply. Lack of good-quality homes for sale has been one of the factors holding back existing home data. If that is true, I would suspect that we should see some improvement even as early as the March pending home sales data, not the month-to-month decline pending home sales that the market is looking for.

In terms of the ISM manufacturing purchasing manager data, the March report was down after an already lengthy list of monthly declines. The current reading of 51.4 is still in growth territory, but that growth rate has been declining since October. With the consumer still in the early stages of coming out of winter hibernation, the terrible durable-goods orders report and oil and export issues, I am wary of forecasting any improvement for April. I am still hoping for a reading above 50, but if it doesn't happen, I won't be surprised at all. And I won't panic until this metric drops to 45 or below.