Weekly Market Update - April 23, 2015

Global GDP Growth and Forecast through 2020: Figure 1 shows the growth of global GDP since 1970 with this month’s IMF forecast through 2020. We have highlighted the periods of the early seventies and the mid 2000’s when the commodity bubbles prevalent at those times were driven by unusually high and sustained global growth. Global growth has been flat in the years 2012 through 2014 and projected for 2015 followed by a forecasted slow growth through 2020. Prior to the 2001 recession, the US held its own in the economic growth stakes but between 2001 and 2008, growth in the US averaged 1.84% less per year than the growth of the global economy as a whole. In the years 2011 through 2013 that gap narrowed to 1.36%. Global growth is now forecast to gradually climb back to 4.0% in 2020, the discrepancy with US widening again. The US has outperformed the Eurozone in every year since 2008 and is forecast to continue to do so through 2020. The gap will narrow as we approach the end of this decade. Within NAFTA, Mexico had the lowest rate of growth in 2014 but is forecast to re-take the lead next year and to widen the gap with the US through the end of the decade, (Fig. 2). Canada is forecast to be relatively flat in that time period.

There is a relationship between the growth of GDP and steel demand at both the international and national level which also extends to other commodities such as cement. One percent growth in GDP does not result in 1% growth in steel demand. At the global level it takes a 2.8% increase in GDP to get any increase in steel demand, (Fig. 3). This is a long term average over a period of 65 years. If the IMF forecast proves to be correct then the growth in steel demand through 2020 will be around 3%. This is considerably better than current World Steel Association data through February 2015 and will help to mop up some of the global steel overcapacity. In the US steel market it requires about 2% growth in GDP to generate growth in steel demand therefore again if current forecasts prove to be correct regarding the US economy we can expect steel demand to expand in 2015 and 2016 then tail off as we approach 2020.

Fig 1

Fig 2

Housing Permits and Starts through March 2015: Starts in March rose by 14,000 to 926,000 following the decline of 164,000 in February. This data is seasonally adjusted by the Census Bureau and the monthly numbers are annualized. The three month moving average, (3MMA) declined from 1,020,000 in February to 969,000 in March, (Fig. 4). The projection to 1.6 million in 2018 is not looking so good at present. The 3MMA of March starts fell below a million after six consecutive months above that level. The year on year growth rate of the 3MMA of total starts in March was 4.7%. The growth rate by this measure is down from 16.7% in September. Note this is not seasonal because the data is seasonally adjusted and, in addition, we are considering year over year relationships.

Multifamily starts are now beyond the pre-recession high of February 2006 but the growth of starts in this sector has slowed dramatically from a 3MMA year over year of 35.5% in August to 3.1% in March. If we ignore the growth spurt of mid-2014, this sector has been slowing since early 2012 and may be approaching saturation. Single family is still 66.1% below the pre-recession level but growth in this sector has been higher than multi-family for the last five months. Figure 5 shows the trajectory of single family and apartments.

Permit data is useful as a forward look at starts. If permits exceed starts then we anticipate an acceleration in construction starts and vice versa. Table 1 shows total permits and starts nationally and regionally. At the national level the differential between permits and starts for single and multi-family units is suggesting that the shift in consumer’s preference towards apartments is far from over. In March on a 3MMA basis permits of multi-family exceeded starts by 94,000 in contrast to the 4,000 for single family. In total permits were 98,000 more than starts. The differential between permits and starts for multi and single family units was the same across all regions. The permit data suggests that the recent decline in multifamily starts could be short lived and that the improvement in single family doesn’t have legs. Figure 6. shows the regional situation for total residential starts since February 2000. So far in 2015 only the South has held its own.

Fig 4

Fig 5
Table 1

Canadian Metals Service Center Report, (MSCI): Daily 3 month moving average, (3MMA) intake for all carbon long products fell 2.7% y/y, while the 3MMA of daily shipments fell 6.9% y/y. This pushed inventory levels higher, to 3.88 months on hand (MOH), an increase of 24.4% y/y. Table 2 breaks down the Canadian service centers activity by steel category. Shipments of structural steel were off by 10.2% on a 3MMA basis as inventories rose to 3.91 MOH, (3MMA), the highest level since February 2009. Shipments fell sharply at the end of 2014 and have yet to rebound, (Fig. 7). Bar intake fell 5.5% y/y, as shipments fell 2% y/y, pushing inventory to 3.11 MOH, 3MMA, the highest level since February 2009. Shipments have remained largely flat, while inventories have risen, (Fig. 8). Sheet inventories mushroomed by 44% to 4.14 MOH, as pipe & tubing inventories also rose y/y. Plate intake slowed dramatically in March, down 21.3% while inventories fell by 2.6%.

Table 2

US Metals Service Center Report, (MSCI): March data reported the daily 3MMA of all carbon steel shipments fell 1.7% y/y. Intake on a 3MMA fell 3.9% y/y, driving steel inventories to 2.77 months on hand (MOH), a y/y increase of 15.8%, (Table 3). Shipments of structural steel were down 1% on a 3MMA basis as inventories rose to 2.75 MOH, (3MMA), a 15.6%, 3MMA y/y increase. Shipments fell sharply at the end of 2014 and have yet to rebound. (Fig. 9). Bar intake was flat, as shipments decreased by 0.8% y/y. Inventory levels rose 5.5% to 2.70 MOH. Shipments have traded in a narrow range between for all of 2013 and 2014 and are down from the 2011 level, (Fig. 10). Sheet shipments fell by 2.3%, 3MMA y/y as inventories increased by 16.9% to 2.67 MOH. Plate inventories surged by 25.4%, 3MMA y/y to 3.43 MOH, while shipments increased by a much smaller 2.9%.

Table 3

Producer Price Index: March PPI data from the Bureau of Labor Statistics reports that the overall cost of Materials and Components of Construction increased by 1.1% y/y and by 0.4% over each of the last three and six month intervals compared to the same period a year ago. Studying Table 4, we see that the PPI of structural steel shapes fell by 8.5% y/y and by 8.6%, 3 months y/y. Fabricated structural steel on the other hand rose by 1.4% y/y and was off just 0.3% on a 3 month y/y measure. Ready-mixed concrete PPI was up 4.2% y/y, and up 1.5%, 3 months y/y, while fabricated rebar, and bar joists rose 1.7% y/y and fell by just 0.4%, 3 months y/y. The has been a much larger change in the raw material prices of structural shapes vs. ready-mix concrete in recent times which may cause a shift in the selection of framing material for non-residential buildings going forward.

Periodically we at Gerdau like to do a reality check on the accuracy of the Bureau of Labor Statistics technique in generating its PPI value. Figure 11 compares the PPI of structural shapes with the CRU Midwest spot price for beams. The correlation coefficient is a rather remarkable 0.978 indicating that the BLS method for developing PPI is excellent.

The PPI for hot rolled carbon bars was down 7.8% y/y and off 6.2%, 3 months y/y. Carbon steel forgings are also down but not nearly as much, (-1.7% y/y and -2.5%, 3 months y/y). The competing materials of powder metal and aluminum forgings and aluminum die castings have had little change in price over the past year, Table 5 and (Fig. 12).

Table 4

Fig 11

Chicago Fed National Activity Index: The CFNAI is a weighted measure of 85 economic indicators that gauge economic activity and inflationary pressures (Federal Reserve Bank of Chicago). The indicators are derived from production and income, employment, unemployment and hours, personal consumption and housing, and sales, orders, and inventories. The Index has a value of 0 and standard deviation of 1, with a positive reading signaling above average growth and negative reading signaling below average growth.

For the month of March the CFNAI fell 0.42 from a negative 0.18 reading in February. The 3MMA reading also fell, off 0.27 in March on the heels of a negative 0.12 posting in February. The index is still trending in a positive direction when looking at a 6MMA, up 0.05 in March and up 0.17 in February. Production related indicators lead the fall in the top-line sub-index, sales indicators were flat. The employment sub-index contributed -0.03 to the CFNAI down from 0.11 in February as non-farm payrolls recorded just 126,000 new jobs in March which was disappointing compared to the 264,000 positions created in the prior month. The consumption sub-index improved somewhat to -0.13 in March up from -0.13 in February, (Fig. 13).

The West Coast port backlog and harsh winter weather contributed to the decline in production indicators as did the strengthening US dollar which hurt manufacturing export volumes. The consumer is not helping increase retail sales despite having lower gasoline bills. Automotive dealerships and building material outlets saw increased sales in March but not enough to offset the poor weather in February. Inflationary pressures remain tepid as the economy has stumbled in the first quarter with the result that the Federal Reserve will likely put off raising interest rates until September or later to ensure that the economic growth stays on-track.

In Figure 14, we present Long Products steel shipments with the CFNAI. In the period between 2003 and 2009, there was a fairly good relationship between the two data series with a correlation coefficient of 0.615. This relationship fell apart from 2009 to present, with a correlation coefficient of -0.291. The reason being that the non-residential construction market did not return post-recession.

Fig 13

Fig 14

Non-Residential Construction Starts: March data from McGraw-Hill Dodge on construction starts presented in Table 6 shows that non-residential construction (NRC) starts (a start is defined as all the square footage that will be built over the entire project is counted at the ground breaking. This gives us an 8 to12 month forward look at the outlook for NRC). Construction starts totaled 1,229 million square-feet, up 11.3% y/y, but only up 3.8%, 3 months y/y indicating that momentum has slowed substantially.

Manufacturing plants recorded the biggest percentage increase y/y, up 55.5%, but slowed to +9.1%, 3 month y/y. Hotels and motels posted 33.7% y/y growth and continue strong at +25.6%, 3 months y/y. Civic, medical, parking structures and religious buildings are four project categories that are showing positive momentum while the other 10 categories have recording declines in momentum. Among the decliners are apartments (>4 stories) which has been and continues to be the largest aggregate square feet category, representing 21.0% of the total square footage built over the past 12 months. Despite the decline from 10.4% y/y growth to 0.7%, 3 month y/y growth, apartments account for an even larger 23.7% of total NCR. If multi-family construction becomes saturated, this would have a significant impact on the total NCR market unless one or more other project categories “step-up”. Figure 15 compares the recovery trajectory of Dodge starts both with and without apartments. Assuming no recession and no slope change in these trajectories it will take until 2017 to reach 1,600 million square feet of NRC with apartments and to 2020 to reach the same threshold without apartments.

Table 6

Fig 15

Contributors this week include; Laura Remington, Peter Wright and Steve Murphy