Weekly Market Update - January 7, 2016

Construction Put-in-Place: The U.S. Census Bureau revised construction data from January 2005 through October 2015 because of a "processing error in the tabulation of data." Fig. 1 illustrates the magnitude of change in percentage terms over this timeframe (Chart provided by the US Census Bureau). The revisions, which show that total construction spending was not as strong as previously reported for much of 2015, could prompt economists to lower their fourth-quarter gross domestic product estimates. Digging deeper we find that much of the revisions had to do with residential construction. Fig. 2 examines total building expenditure not including single family residential. The orange line is last October data and the blue line is November data which includes the revisions. From this we see that the revisions were almost entirely isolated to 2015 and that the revised spending level was higher than previously thought. Note that the total construction spending chart (Fig. 1), shows overall construction spending to be revised downward in 2015.

Table 1 (after revisions), shows that total CPIP has performed admirably over the past 12 months with total spending up 8.6% y/y. The picture is even more rosy when measured on a 3 month y/y basis, at +32.9%. Private spending grew 3.9% on a 12 month comparison and accelerated to a 35.4%, 3 month y/y metric. State & Local total CPIP spending posted sharply higher numbers for both 12 and 3 month y/y evaluations, up 22.3% and 27.7% respectively. Momentum (3 month minus 12 month) was solid overall at +24.3%. Of concern is that non-residential buildings recorded negative 2.3% momentum, more on this concern will be discussed in the section on Dodge non-residential starts.

Table 2 breaks out non-residential (NRC) CPIP and expands it into 11 project categories as well as the broad sectors of Private and State & Local construction. Private NRC spending rose 36.8%, 12 months y/y and 32.0%, 3 months y/y resulting in negative 4.8% momentum. Conversely State & Local spending saw its momentum rise by 3.4% as the three month percentage increase of 24.6% was higher than the twelve month value of 21.2%. Seven of the eleven project categories recorded negative momentum with the most notable being a -14.7% for manufacturing which despite the negative momentum continues to post impressive growth values, up 62.3% 12 months y/y and up 47.6%, 3 months y/y. The US manufacturing sector appears to be slowing based on the ISM manufacturing index which slipped again in December, now down six months in a row.

Fig 1

Table 1

Dodge Non-residential Square-Foot Starts: While the numbers for CPIP are terrific, the numbers for Dodge starts are not good. Recall that a “start” includes all of the square feet to be constructed on the day the project begins. In this way we get a window into the future of about one year. Table 3 presents the Dodge starts data for 3 and 12 months y/y in fourteen NRC project categories. For the month of November, starts declined 2.1%, 12 months y/y and were off 16.0%, 3 months y/y. On a 3 month y/y basis starts have declined for four months in a row and two consecutive months on a 12 month y/y comparison. Fig. 3, illustrates this decline graphically. Ten of 14 project categories recorded 3 month y/y declines in November, one more negative than for 12 months y/y. The project category with the largest decline was manufacturing, off 23.5%, 12 month y/y and down 54.5%, 2 months y/y. Manufacturing starts have now declined for eight months in a row on a 3 months y/y basis after expanding for twelve months consecutively. Apartments (> 4 stories) recorded 16 consecutive monthly increases before declining for two months on a 3 month y/y measure. November’s decline was -11.3% resulting in -26.6% momentum. On a positive note, transportation buildings posted growth of 4.4%, 12 months y/y and accelerated to +20.3%, 3 months y/y.

Fig. 4 presents non-residential construction starts in dollar terms for both 3 and 12 month y/y change employing percentage change on the Y axis. Non-residential starts were growing steadily through 2014 and through the early months of 2015 before reversing course abruptly and starting to decline on a 3 month y/y basis. The “blue” line has now been negative for four consecutive months and will continue to pull the “red” line (12 months y/y) down going forward unless y/y growth increases. The fact that NCR CPIP growth and momentum are strong indicates that there is work in the pipeline that will continue for a period of time. The weakening starts data suggests that future NCR CPIP will wane. Giving that starts have been negative for four months and that starts tend to look-out approximately 12 months into the future, we should begin to see the impact on NRC CPIP in about eight months’ time.

Table 3

Fig 3

U.S. Industrial Construction Starts: Data released by Industrial Resources Inc. indicated industrial construction starts slowed through December 2015. Total dollars spent, are counted in their entirety, during the month that a project begins. Using a longer rolling mean, we view this data at 3 month & 12 month moving totals to smooth out short term fluctuations & highlight any long term trends or cycles.

Fig. 5 illustrates how volatile spending can be. Total US construction starts were down 39.3%, 3 months y/y and off 10.5% y/y. Comparing 3m y/y change to the 12m y/y, reflects a steep slowing of industrial construction starts. Transmission oil & gas and alternative fuel were the only project categories to post positive numbers for both 3 and 12 month y/y comparisons. Production oil and gas fell 7.2%, 12 months y/y, plummeting to -87.2%, 3 months y/y. Chemical processing and industrial manufacturing also recorded weak growth numbers: Chemical; -46.7%, 12 months y/y and -73.3%, 3 months y/y. Industrial; -29.9%, 12 months y/y and -54.1%, 3 months y/y, Table 4.

Fig. 6 maps the US census regions, cataloging 3 month total expenditures and 12 month y/y percentage change through December 2015. On this basis only the West North Central and Southeast are seeing growth. Table 5 presents 3 and 12 month y/y detail by census region. Momentum was positive in four regions and negative in five. The Midwest showed the strongest growth for both 3 and 12 months metrics of +28.8% and +95.5% respectively. The Rocky Mountains also recorded solid growth on a 3 month y/y basis, up 19.4% despite a -32.9%, 12 month y/y performance. The Southwest, which has been the perennial growth leader saw growth plummet to -62.6% on a 3 month y/y basis, no doubt a function of the sharp decline in oil pricing. The Mid-Atlantic, New England and West Coast also posted strong negative growth numbers.

Fig 5

Table 4

Imports Long and Flat Products through December: Total rolled product licensed imports in the single month of December were 1,955,625 short tons up by 6.1% from the preliminary November volume. The 3MMA through December was 1,996,248 tons, down by 3.1% from the 3MMA through November. On the same basis long products were up by 4.5%, flat rolled was down by 2.8%, and pipe and tube down by 11.3%.

Fig. 7 shows the 3MMA through December licenses for semi-finished, long and flat products. Long includes structurals, bar size shapes, hot rolled bars, cold finished bars, rebar and wire rod. Flat includes all hot and cold rolled sheet and strip plus all coated sheet products including tin-plate plus both discrete and coiled plate. The 3MMA of total long product imports has been stuck in the range 519,000 tons to 772,000 tons since March 2014 with no particular trend evident. In December the 3MMA of long product imports was 582,654 tons. The 3MMA of flat rolled imports peaked at 1,634,000 tons in November 2014 and declined to 1,066,419 tons in December 2015. In November 2014 the 3MMA of semi-finished imports was 884,641 tons, this volume declined to 544,106 tons in December 2015. It is therefore evident that long product imports are not following the rest of the steel market.

Fig. 8 shows the trend of long products since January 2011. Imports of rebar in three months through December were higher than at any time since the recession. Imports of wire rod have been steadily increasing since Q3 2014.

Table 6 provides an analysis of major product groups and of long products in detail. It compares the average monthly tonnage in the latest three months through December with both three months through September, (3m/3m) and three months through December 2014, (y/y). On a y/y basis all rolled products were down by 33.2% but longs were down by only 0.7%. The color codes in Table 6 for the three month and year over year change show which products are improving and which have still experienced import volume increases in these two time frames. It is immediately obvious that the situation of longs is far different than for flat rolled. For total long products the tonnage was down by only 3,873 tons y/y which was 0.7%. In 3m/3m the volume actually increased by 1.0%. Rebar and wire rod both experienced a more than 20,000 ton increase year over year.

Fig. 9 shows the import market share of sheet and long products through November which is the latest data available for total steel supply. The import market share of sheet products peaked at 24.3% in March, fell to 21.3% in October then rose again to 22.5% in November. For long products the import share peaked at 30% in May, declined to 24.7% in October then in this latest data bounced back to 27.3%.

Fig 7

Fig 8

Table 6

Unemployment Claims: New unemployment claims fell to 277,000 for the week ending January 2nd, down 10,000 from the previous week, Fig. 10. This figure has been below the 300k benchmark since February 2015, a signal of a further strengthening labor market. The seasonally adjusted four week moving average is 275,750, 6% lower on a y/y basis. The 4WMA figure has been among the lowest on record and has also remained below the benchmark since March of this year. The non-seasonally adjusted figure rose above the 400k mark, however, this is typical as part time employment for the holiday season goes away. A stronger labor market was conducive to the Fed raising interest rates in December.

Seasonally adjusted continuing claims rose to 2.23 million, up 25,000 from the previous week. The figure for December 26th is 10% lower than the same period last year. The four week moving average reported 2,218,000; under the 3 million mark since the middle of 2013, yet another signal of a tightening labor market, Fig. 11. The number of persons claiming any type of unemployment benefit fell this week to 2.2 million. This figure has fallen more than 100,000 persons on a y/y basis.

Fig 10

Institute for Supply Management - Manufacturing Index: The ISM manufacturing Index slipped again for the sixth consecutive month and is currently at the lowest level since the end of the Great Recession in June 2009, Fig. 12. The Index is down 12.5%, nearly 7 percentage points, on a y/y basis. Production, New Orders, and Inventories all rose slightly while Supplier Deliveries and Employment sunk in December. Employment dropped to 48.1 from 51.3 conceivably leading up to a poor jobs report on Friday, January 8th. Manufacturing has lost jobs in three of the last four months from August through November. The Manufacturing sector declined for the second straight month, however the overall economy expanded for the 79th straight month. (Institute of Supply Management).

Fig 12

Institute for Supply Management - Non-Manufacturing Index: The ISM nonmanufacturing Index dipped again in December to 55.3 from 55.9, Fig. 13. This is the lowest reading since April 2014 and has dropped 4 percentage points since July’s ten year high of 60.3.Business Activity, New Orders, and Employment both rose in December, however, the large decline in Supplier Deliveries, 48.5 from 53.0, drove the readings down. Supplier Deliveries hasn’t been below the 50.0 benchmark since October 2014. Prices paid declined below the benchmark threshold for the third time in the last four months. (Atlanta Fed GDP Estimate).

Fig 13

Steel Demand Indicators: Table 7 is a snapshot of the market situation on 12/31/2015. Indicators updated since we last published two weeks ago are shaded beige. In most cases this is not December data but data that was released in the first two weeks of December, the actual month to which the data relates is shown in the second column. Of the 26 indicators under consideration, the present situation of 4 are positive by historical standards, 8 are negative and 14 are neutral. This was a decrease of one negative and an increase of one neutral since we last published on December 17th. We have made a change to our analysis this month by removing the capacity utilization of the long product mills and replacing this with shipments. We have done this because the calculation of capacity utilization is problematic and we find shipment data to be much more factual. This resulted in a decrease of one negative and an increase of one neutral in the latest analysis. There were no other changes to our view of the present situation.

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 December 31st. The number of indicators trending positive in this latest analysis was 12 with 13 trending negative and one unchanged. Overall this was an increase of one negative since December 17th. The Industrial Production index published by the Federal Reserve began to trend negative year over year for the first time since February 2010.

We have separated the leading indicators from the main table for a closer look, these are shown in Table 8. Of the twelve leading indicators three are trending positive and nine negative which was no change since December 17th. (Explanation of Indicators).

Table 7

Table 8

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