Weekly Market Update - November 17, 2016

Growth of US GDP in the 1st Estimate for Q3 2016: The annualized GDP growth rate in the 1st estimate of the 3rd quarter of 2016 was 2.91% which was the best result since the 3rd Q of 2014. This report was not as good as it first appears. It is disturbing that the contribution of personal consumption was down in Q3 and that the construction components were lack-luster. The change in inventories was what made this report appear good on first glance but as we report below this is simply moving growth from one period to another. Historically it has been necessary to have about a 2.5% growth in GDP to get any growth in steel demand so this latest estimate of GDP suggest that at least steel demand won’t contract in the immediate future. This relationship is a long term average and in reality steel is much more volatile than GDP. If GDP takes a dive then steel demand craters and if GDP takes a sudden upturn steel soars. Neither of these extremes is evident at present.

GDP is measured and reported in chained 2009 dollars and in the 1st estimate of Q3 was $16.702 trillion. The growth calculation is misleading because it takes the Q over Q change and multiplies by 4 to get an annualized rate. This makes the high quarters higher and the low quarters lower. Figure 1 clearly shows this effect. The blue line is the trailing 12 months growth and the black line is the headline quarterly result. On a y/y basis GDP was up by 1.5%, an improvement from 1.28% in the 2nd quarter. In the last six years, since Q1 2011 the trailing 12 month growth of GDP has tracked between a high of 2.88% to a low of 1.07% therefore the latest result remains at the low side of this range.

Figure 2 shows the headline quarterly results since 1990 and the October IMF forecast through 2021. In their October revision, the IMF downgraded their forecast of US growth in 2016 from their April 2016 estimate of 2.4% to 1.58% and downgraded 2017 from 2.50% to 2.20%.

The biggest difference between Q2 and Q3 was the contribution of inventories which changed from negative 1.16% in Q2 to positive 0.61 in Q3 a swing of 1.79% and accounted for the whole of the change in Q3 and then some. Declining inventories have a negative effect on the overall GDP calculation and this was the case for the previous five quarters. Over the long run inventory changes are a wash and simply move growth from one period to another. When a decline in one period is subtracted from growth, it can often be a healthy indicator for future periods. Such a decline suggests that there are lower levels of overall inventories which will set the stage for inventory increases - which will then add to GDP growth in the future. This seems to be what happened in the 3rd quarter and can be expected to continue into Q4. Figure 3 shows the contributors to GDP extended back through Q1 2007 and describes the quarterly change in the six major subcomponents. In the 2nd Q the negative contribution of inventories was the greatest since Q1 2014 and in the 3rd Q was positive contribution was the greatest since Q1 2015. The contribution of personal consumption was 1.47% down from 2.88% in Q2 but up from 1.11% in the 1st Q. Personal consumption includes goods and services, the goods portion of which includes both durable and non-durables. Government expenditures contributed negative 0.3% in Q2 and positive .09% in Q3. The contribution of fixed residential investment at negative 0.24% in Q3 and negative 0.31% in Q2 were the first quarters that this component detracted from GDP since Q1 2014. Non-residential which includes infrastructure made a small positive contribution in both quarters. Net exports made a positive contribution in both Q2 and Q3 adding 0.18% and 0.83% respectively.

Figure 1

Figure 2

Figure 3

Architectural Billings Index: The national ABI increased 2.4 points in October to 50.8 as demand for architectural design services rose. The ABI typically leads construction put-in-place (CPPI) numbers by 9 to 12 months. Figure 4 presents a map of the US with the ABI shown as three month moving averages as well as the y/y change for the four regions that the American Institute of Architects (AIA) collects data. Regionally the Southern US recorded the strongest score at 53.7 3MMA, while demand the other three regions declined.

Figure 5 charts the ABI sub-index for commercial / industrial and institutional projects from 2011 to present. Both recorded weaker indexes, with institutional posting a 49.1, its second month in a row below 50, followed by commercial and institutional down to 49.8. Figure 6 presents awards for design contracts and measures the trends in new design contracts at architectural firms which provides a strong signal of future billings. The design contract index as at 48.7 for October, falling below 50 for a second time this year.

Figure 4

U.S. and Canadian Service Center Shipments: October U.S. service center total shipments continued to decline falling 6.2% on a three month moving average basis, (3MMA) y/y and down 15.8%, 3MMA compared to two years ago. Plate shipments fell the most, down 19.0%, while sheet products performed comparatively better, down 2.2% y/y, (Table 1). Overall inventories fell across the board (-19.4% for all products), except for structurals which posted a 1.9% increase 3MMA y/y. Intake was down in every product group category. Intake for all products was down 10.5% 3MMA y/y pulled down by plate off 22.4%, followed by pipe & tube, off 16.0%. Bar intake fell 13.8% and sheet declined 8.5%. Structural intake was off the least, down 5.0%, 3MMA y/y.

A slowing manufacturing base, trade actions, weak energy prices and a pause in the non-residential construction sector coupled with short mill lead times have combined to reduce service center shipments, inventory and intake volumes.

Canadian September service centers total shipments declined 8.8%, 3MMA y/y and were down 19.1% from two years ago. Bar & shapes shipments increased 15.6%, all other product groups recorded declines ranging from -12.9% for plate to -5.6% for pipe and tube, (Table 2).

Overall inventories levels fell 12.4%, 3MMA y/y, pulled down by sheet off 18.1%, 3MMA y/y. The exception was bar & shapes which rose 4.8%. Intake on all carbon products increased 15.0%, 3MMA led by a 32.0% surge from bar & shapes and a 22.4% jump from sheet products. Pipe & tube recorded a decline in intake, off 3.8%, 3MMA y/y, while plate intake was flat at -0.2%.

Table 1

U.S. Manufacturing Capacity Utilization: The 3MMA of capacity utilization fell to 74.80%, right on the threshold of the recession line by historic standards. The index has drifted lower since reaching its short term peak of 76.26% on November 2014, (Figure 7). Manufacturing other than automotive continues to struggle against the headwinds of; low energy prices, a strong dollar which hurts exports and high import levels of semi and finished manufactured goods. An additional hardship for US manufacturers is slowing productivity. Data from the Bureau of labor Statistics shows that unit labor costs are rising at a faster rate than output per hour, Figure 8. From Q4 2011, unit labor costs have increased by 9.7%, while over the same timeframe output per hour has increased just 2.3%.

In Figure 9, we overlay MSCI carbon steel shipments onto manufacturing capacity utilization. The correlation coefficient of 0.714 illustrates that there is a fairly strong relationship between the two data sets. The relationship began to diverge after the recession ended and has widened further in 2015 and 2016 (correlation coefficient measures -0.10 from 2013 to present). There are many factors that contribute to this change. We think the main influences are direct imports to end-user and fast availability of product from steel mills as a result of low mill capacity utilization which equates to short mill leads times.

Looking forward to 2017, Evan Koenig, Senior VP of the Dallas Federal Reserve anticipates stronger GDP growth of 2.7% and the U3 unemployment rate to fall to 4.5%. This is much stronger than the IMF forecast of 2.2% (discussed in "GDP" above). Hurdles for future growth include the low labor participation rate currently at 62.8% and stubbornly low inflation.

Figure 7

Producer Price Indexes of Steel and Competing Materials through October 2016: Figure 10 presents a historic review of all commodities from 2000 to present. Prices have been rising on a 3MMA basis since the recent bottom in March 2016. When examined on a y/y basis in percentage terms, overall commodity prices continue to decline and have now declined for 24 consecutive months. Breaking out the materials used in non-residential construction shows a similar picture to “all” commodities except that the percentage y/y drop was much larger throughout all of 2015. The y/y decline thus far in 2016 has lessened with each passing month reaching -0.2% in October. Prices have stabilized but remain range bound in the 100 range which is the 2005 reference criteria for this PPI chart, (Figure 11).

Table 3 presents construction steels and competing materials referencing 3, 6, 12 and 24 months ago. The structural steel PPI was down by 9.7% y/y ending October, but has risen to +4.8% vs. six months ago. Hollow structural shapes have moved up in price on both 12 and 6 month y/y comparisons by 6.7% and 8.8% respectively. In the most recent 3 months structural steel price was flat (+0.1%) while HSS prices fell 5.1%.

Fabricators making structural steel parts for non-residential construction projects have been able to increase prices in each of the last four time references. This has not been the case for fabricators that produce steel bridge components. These fabricators have seen prices decline for each of the four comparisons. Meanwhile pre-stressed concrete bridge beams have seen strong price increases indicating a strong demand for this product. The resultant differential with bridge structural steel should translate into a cost advantage for structural steel. Wood and ready-mix concrete have enjoyed steady price increases as demand for housing continues to recover. However, the most recent 3 month y/y time comparison are much smaller; flat (0.0%) for ready-mix and just 0.5% for softwood lumber.

Up until the most recent time comparison asphalt paving mixtures has seen price declines for no-doubt a function of the large drop in petroleum pricing. For the 3 month y/y comparison asphalt and tar paving mixture reversed the slide rising 0.9%.

Figure 12 shows the PPI for each of ready-mixed concrete, hot rolled structural steel and softwood lumber from 2010 to present. Concrete has seen slow but steady price improvement since 2011. Softwood lumber moved higher from 2011 through 2014 albeit with a lot of volatility and has been range-bound since then. Structural steel PPI declined from 2011 to 2012 and then was relatively flat in 2013 and 2014 before falling throughout 2015. Prices have rebounded somewhat in 2016. Note that the PPI for each of these three materials in January 2005 were: Concrete = 105, wood, 113 and structural steel 82. As of October 2016 ready-mix concrete = 153, softwood lumber = 108 and structural steel = 82. Steel and wood PPI today are very close to where they were in 2005, while ready mix concrete has increased by 47%. Given this relative valuation, structural steel would seem to have a considerable competitive advantage on a raw material basis where either material can be employed.

Table 4 examines the PPI for hot rolled steel bars vs. its competition. Hot rolled bars for engineering applications have suffered a much greater price erosion than downstream forgings, castings or powder metal parts. Carbon steel forgings have gained a competitive advantage over aluminum forgings over the last two years and to a lesser degree over powder metal parts. Forgings have gained a slight competitive advantage over castings in the last 12 months. We have monitored the Bureau of Labor Statistics PPI reports for many years and as far as we can tell from comparison with known transaction prices, these PPI are a good representation of the real world. However our observation is that the actual values of the PPIs of different products cannot be compared with one another because they are developed by different committees within the BLS. We believe that this data is useful in comparing the direction of prices in the short and medium term but not the absolute value.

Figure 10

Figure 11

Table 3

Steel Demand Indicators: Table 5 is a snapshot of the market situation on 11/17/2016. Indicators updated since we last published four weeks ago are shaded beige. In most cases this is not November data but data that was released in this month for previous months, the actual month to which the data relates is shown in the second column. Of the 27 indicators under consideration, the present situation of 6 are positive by historical standards, 11 are negative and 10 are neutral. There was an increase of one negative and a decrease of one neutral since our last update on October 20th. Our intent in using the word neutral is to say that this indicator is considered to be in the mid-range of historical data. The changes that occurred in the last month were as follows. We re-classified our perception of GDP growth from negative to neutral based on the first estimate of Q3 as reported in this issue of our WMU. Based on the latest Federal Reserve senior loan officer survey for Q4 2016 we reclassified loan demand for commercial and industrial loans from neutral to negative. We also reclassified manufacturing capacity utilization from neutral to negative when it fell below 75% in October.

In our trends analysis, most of the values reported are three month moving averages to avoid the knee jerk reactions that are characteristic of most economic reports in the press. Please note that there is nothing subjective about this trends analysis. The numbers presented here are the facts available as of November 17th 2016. The number of indicators trending positive in this latest analysis was 10 with 17 trending negative. This was no net change since October 20th however there were several changes in the detail.

In the general economy both the Chicago Fed National Activity index and loan demand for Commercial and Industrial loans reversed direction and trended negative. In the long products section both the Chicago shredded scrap price and the rebar import price reversed direction and trended positive. In construction, nonresidential starts deteriorated and housing starts improved. There were no changes in trends for the manufacturing section.

We have separated the leading indicators from the main table for a closer look, these are shown in Table 6. Of the twelve leading indicators four are trending positive, and eight negative. This was a move of two from positive to negative since our last update. These were the Chicago Fed data and loan demand as described above.

In summary the present situation is historically weak and deteriorated slightly in this update. Trends have weakened in the last eight months and this tendency has accelerated in the most recent two months of data. At the end of September 51.9% of indicators were trending positive and 48.1% negative. Currently 63% of indicators are trending negative. This is the worst trend result since January 2010.
(Explanation of Indicators).

Table 5

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