Morgan Stanley’s dry van truckload index indicates that van capacity-demand balance shifted at the end of 2016, with capacity becoming tighter than in 2015 and very close to the 11-year trend line for year-end. As we begin 2017, capacity is tighter than in early 2016 and again is very close to the longer-term trend line for January. Capacity started to tighten up a little through July of 2016, unlike 2015 when the index was declining through July.  The index then flattened out in 2016, but in December capacity tightened up and the index hit its highest level of the year.

While the negative impact on capacity of ELDs will not be felt until late 2017, any acceleration in the freight economy could lead to tighter van capacity earlier in 2017. While capacity is not nearly as tight today as was the case in early 2014 or 2015, the index is very close to early 2011 and 2013 and to the 11-year trend line. Any surge in freight volumes as we get closer to Q2 could lead to significantly tighter capacity and upward rate pressure. The index measures incremental demand for dry-van truckload services compared to incremental supply. The higher the index the tighter is capacity relative to demand when compared to a prior period.

image

Graph reproduced with permission from Morgan Stanley. *2006-2014 average trend line excludes financial crisis years of 2008 and 2009. For more information contact: Alex Vecchio at Alexander.Vecchio@morganstanley.com

Morgan Stanley’s refrigerated freight index indicates that refrigerated capacity ended 2016 slightly tighter than in the prior year but more readily available than normal for year-end. For most of 2016, refrigerated capacity was more readily available than in 2015, but in Q4 the lines crossed indicating a tighter capacity environment than in the prior year. As we start 2017, capacity is tighter than was the case in 2016, but very close to the 11-year trend line for January.

Refrigerated capacity began 2015 the same way it ended 2014, significantly tighter than normal. Throughout Q2 of 2015, the market shifted with the result being that capacity was not nearly as constrained as normal. That was even more so the case in Q3 and Q4, as the index dropped to a level lower than in any recent year, including 2009. That trend continued through the first half of 2016, but then we saw a slight shift, moving towards a more balanced market, but not quite getting there. For the last 18 months, the index has been very stable at a level reflecting excess refrigerated capacity.

We do not believe that refrigerated rates will increase much in early 2017, but later in the year we could see some increases as smaller refrigerated carriers adopt ELDs and see equipment utilization drop by 4-5%.  Demand for refrigerated transportation is less correlated to economic fluctuations than dry van or flatbed freight, so the future robustness of GDP growth will not determine demand growth in this market. The index measures incremental demand for refrigerated truckload services compared to incremental supply. The higher the index the tighter is capacity relative to demand when compared to a prior period.

image

Graph reproduced with permission from Morgan Stanley. *2006-2014 average trend line excludes financial crisis years of 2008 and 2009. For more information contact: Alex Vecchio at Alexander.Vecchio@morganstanley.com

Morgan Stanley’s flatbed freight index indicates that flatbed capacity finished 2016 just slightly tighter than at the end of 2015 and just slightly more more readily available that the 11-year average at year-end. As we start the new year, flatbed capacity-demand balance is right on the 11-year trend-line and capacity is a little tighter than was the case in early 2016. Flatbed capacity modestly tightened through June of 2015, but eased substantially after that. In 2016, flatbed capacity was more readily available than was the case in 2015 until the last 6 weeks of the year, when the lines crossed. 2016 was a year of remarkable excess capacity in the flatbed market, quite similar to 2009. The last twelve months have seen unusual stability in the flatbed index considering the historical seasonal volatility of capacity-demand balance. With the ELD mandate not coming until December (well after the peak time of year for flatbed demand), it seems likely that we will not see significant tightening of flatbed capacity in 2017.

Flatbed capacity-demand balance favored the carriers over the shippers more so in 2014 than in the prior two years, but capacity never became as tight in 2014 as in 2010 and 2011. Low oil prices and a falloff in drilling activity meant that capacity never did tighten in 2015. The index measures incremental demand for flatbed truckload services compared to incremental supply. The higher the index the tighter is capacity relative to demand when compared to a prior period.

image

Graph reproduced with permission from Morgan Stanley. *2006-2014 average trend line excludes financial crisis years of 2008 and 2009. For more information contact: Alex Vecchio at Alexander.Vecchio@morganstanley.com

The U.S. Bureau of the Census reported that seasonally adjusted real retail and food service sales decreased 0.1% to $192.1 billion in November. (Note that actual sales are deflated using CPI 1982 – 1984 = 100). November sales were 2.0% higher than the prior-year period.

Nominal (unadjusted for inflation) seasonally-adjusted retail sales totaled $465.5 billion in November (second graph), up 0.1% from prior month, and up 3.8% from prior year results. Total nominal sales are up 3.1% year-to-date.

The results were below consensus expectations and October was revised downward. Nominal retail and food services sales excluding gasoline were up 3.8% year-over-year. Nonstore retailer sales (e-commerce and mail order) were up 11.9% and food services were up 4.9%. Gasoline station sales were up 4.0% year-over-year, a reversal from most of 2016 with YTD sales down 7.3%. Department stores sales fell 6.4% year-over-year, while electronics and appliances store sales were down 3.8%, clearly reflecting e-commerce gains. We focus primarily on real retail sales because they are a better indicator of freight volumes than the inflated figures.

image

image

According to the U.S. Census Bureau and the Department of Housing and Urban Development, single-family new home inventories increased to 250k in November (seasonally adjusted). Inventories had leveled off over the 12 months ending in October (low of 230k and high of 246k), but are now at the highest level since September 2009. November’s new home inventories were 20k (8.7%) above the prior-year level of 230k. New home inventories still remain low by historical standards.

The growth in inventories in the last year (seasonally adjusted) has been driven by homes under construction (+9K), more so than by homes not yet started (+5k) and homes completed  (+6k).

Inventory levels slowly increased throughout 2014, peaking at 212k in December. Inventories rose from 149k at the beginning of 2013 to 187k by December, following a year of remarkable stability in 2012 where the absolute inventory of new homes remained within a consistent range of 142k – 150k.

Seasonally adjusted new home inventories held relatively steady at 5.1 months of supply in November, from 5.4 months a year ago and 5.2 months in October. Sales of single-family houses rose to 592k (seasonally adjusted annual rate), up 5.2% from revised prior month sales and up 16.5% from prior year. Year-to-date absolute sales are up 12.7% at 522k.

Full year 2015 new home sales were up 14.6% to 501k. For the full year 2014, new home sales only grew by 1.9% to 437k units. The months of supply figure remained below 5 months between February 2012 and June 2013, but was 5.0 or more from that point through the end of 2014 with only one exception. In 2015, eight months were at 5.0 or greater months of supply, including each of the last 7 months. The average months of supply over the last 50 years is 6.1, so current new home inventory are well below “normal” levels. For the 9-year period of 1997 through 2005, the inventory level averaged 4.1 months with relatively little volatility, despite the dot-com boom and subsequent recession, and we are above that level today.  The vertical bars in the graphs represent recessions.

image

image

Housing starts totaled 1.090 million in November (seasonally adjusted annual rate – SAAR) down 18.7% from last month’s revised figures, and down 6.9% from November 2015 results, and were well below expected levels. Single family starts totaled 828k (SAAR), down 4.1% from October but up 5.3% year-over-year. Starts of multi-unit structures plummeted in November to 259k (SAAR) from 462k in October. November multi-unit permits were down 15.8% from October so it was a tough month overall for multi-family units. Total starts exceeded a 1.0 million unit annual pace for the 20th straight month, but November had the second fewest number of housing starts in 2016. YTD total starts are up 4.8% over 2015 with single family starts up 9.6% and 5+ unit starts down 4.1%.

In comparison to the last couple of years single-unit starts are growing faster than multi-unit starts. For the full year 2015, total starts were 1.112 million, up 10.8% over 2014. Single unit starts were up 10.3% in 2015, while 5+ unit starts were up 12.9%. For the full year 2014, there were 1.003 million total housing starts, up 8.8% from the 925 thousand starts during 2013. Single family starts were up 4.9% and multifamily starts were up 16.4%. Total 2013 housing starts were up a robust 18.5% from the 781k housing starts recorded in 2012 and in 2012 starts were up 28.2%.

There remains a lot of ground to cover for the housing sector to fully recover from the recession. Housing starts are still far below the average of just over 1.5 million per year over the last 40+ years, and even farther below the 2.2 million peak of the most recent housing boom. Since 1968, the U.S. population has grown from 200 million to over 300 million. Some economists believe that slower population growth and household formation in the U.S. mean that housing starts will not recover to 1.5 million units for a long time.

Total starts reached a low point of 478k (SAAR) in April of 2009, while single unit starts bottomed out at 353k in March of 2009. A low housing starts figure not only impacts transportation demand for building products but also for appliances, furniture, and other related items, so continued improvement in the housing sector should lead to rising freight volumes. The ATA estimates that each housing start generates 8 truckloads of freight.

image

FMCSA received more than 4,500 comments on its joint proposal with NHTSA for mandatory speed limiting devices, and motor carriers and drivers were unified in their opposition. A coalition of five organizations – AEMCA, AHFA, AHAA, TEANA and TLP&SA – filed in opposition to the NPRM for three reasons:

1. NHTSA and FMCSA have failed to make a case for speed limiters producing safer highways;

2. The agencies have failed to adequately consider the effect of speed limiters in permitting

commercial motor vehicles to avoid accidents; and

3. Imposition of speed limiters by the federal government is inconsistent with the preemption of

state law exceptions found in 49 USC 14501.

A copy of the coalition’s comments is available at https://www.regulations.gov/document?D=FMCSA-2014-0083-4494. Other comments and the NPRM are available at https://www.regulations.gov/docket?D=FMCSA-2014-0083

Weekly retail on-highway U.S. diesel prices fell 2.2 cents to $2.421 per gallon on November 21. Diesel has eased in each of the last three weeks but is only down 5.8 cents over that time frame. In February, diesel prices reached their lowest level ($1.980) since January 2005, dropping below the recessionary trough, before rebounding to a year-to-date high of $2.481 on 10/17. The last year that diesel did not hit $2.50 per gallon all year was 2004. Diesel prices are almost identical to their level one year ago; down 2.4 cents or 1.0%. Diesel prices had stabilized between August 24 and November 16 last year, with a high of $2.561 and a low of $2.476 during those 13 weeks, but were in steady decline between then and February 2016 before beginning a steady climb through June of this year. Since then diesel has again fluctuated in a fairly narrow range of 2.310 to 2.481 per gallon. On November 8, the Energy Information Administration (EIA) held its pricing forecast level at 2.69 per gallon for 2017.

A view of weekly prices over the last 6+ years (second chart) indicates fairly stable prices between Q2 2011 and the start of the 2015 slide (min of $3.65 and max of $4.16). We remain well below that range, but have caught back up to late 2015 price levels. Diesel is well below the price level in each of the last five years prior to 2015 for November.

Diesel experienced a high but narrow pricing environment throughout 2013, fluctuating between a low of $3.817 on July 1 and the high of $4.159 on February 25. In 2014, diesel prices remained within the 2013 range until early September, but then began a steep decline.  In 2012, diesel exceeded $4 per gallon for a total of 26 weeks but only reached that level for 8 weeks during 2013, and only 4 weeks in 2014. The recessionary low price point for diesel was $2.023 in March 2009. Diesel prices peaked at $4.771 per gallon in July 2008.

image

image

Stephens Inc. released their Q3-2016 update on publicly traded TL carriers reporting that rates per loaded mile excluding fuel surcharge decreased by 2.1% over the same period last year, and were unchanged from Q2 of 2016. The most recent two quarters have seen year-over-year price drops, following 24 consecutive quarters of year-over-year TL rate increases.  TL rates fell more than normal between Q4 and Q1 and have not rebounded since then. It seems unlikely that Q4 of 2016 will see the same kind of upward TL pricing movement as was seen in each of the last two years.

Stephens expects continued TL rate pressure as we enter bid season, but does think that rates may start to rise in the second half of 2017. Quarterly data shows how weak TL pricing has been for the last 3 quarters (second and third graph). The data does not necessarily represent the entire TL industry as the publicly traded carriers tend to be larger and more successful in general, so are also more likely to successfully raise rates.

Average length of haul fell to 573 miles in Q3 from 609 in Q3 2015. This is an interesting trend given how weak the domestic intermodal market has been in 2016. It is likely indicative of  more fundamental supply chain network changes (more DCs, closer to customers) than modal shifts from TL to intermodal. Revenue per tractor per week was down from Q3 2015 but remains high by historical standards. Miles per tractor per week were up year-over-year, which is surprising given the decline in length of haul. Mileage utilization remains well below historical averages due to shorter lengths of haul and tighter hours of service regulations.

TL Q3 16 rates TL Q3 16 rates qtr TL Q3 16 rates qtr b TL Q3 16 lohTL Q3 16 rev per tractor TL Q3 16 miles per tractor

 

Stephens Inc. reported that LTL yields (revenue per hundredweight) increased by 1.0% from Q3 2015 to Q3 2016, and increased by 2.0% from Q2 2016. The year-over-year percentage increase for Q3 was the lowest since 2010. The LTL rate index hit a new all-time high point (series began in 1996) in Q3, indicative of pricing discipline by the large LTL carriers. For 2017, Stephens is expecting low single-digit LTL rate increases.

For the full year 2015, LTL yields were up 5.9%, but the average quarterly year-over-year increase in 2016 is only 1.4%. LTL yields dropped in Q4 of 2013 and Q1 of 2014, before resuming their post-recession climb, but that climb was interrupted in Q1 of 2016 as rates dropped from Q4 of 2015 (second graph).

Tonnage (graph 3) was up 0.7% in Q3 ‘16 over Q3 ‘15 for the group of carriers reported on by Stephens. For the full year 2015, tonnage was down 0.4%, indicative of the softness in the freight market in 2015. Weight per shipment (graph 4) was down 1.6% year-over-year in Q3, likely due to plentiful TL capacity taking more of the heavier weight LTL shipments. Note that rate per hundredweight is higher at lower shipment weights so the weight drop is responsible for some of the yield increase.

From their previous Q4-2007 peak level, LTL rates dropped 11.2% to their trough in 2010 but have now surged 24.8% from Q2 ‘10 to the current all-time high. Despite the improving trends, the challenges facing LTL carriers remain apparent as the current pricing levels remain only 10.8% over the previous peak in 2007 despite the realization of significant cost increases over that 9-year period. Some of the capacity issues that impact the TL segment, like CSA and Hours-of-Service rules, are not as relevant to the LTL segment. Industry concentration and consolidation does provide LTL carriers better pricing power than is the case for TL carriers.

LTL Q3 16 yield

LTL Q3 16 yield b

LTL Q3 16 tonnage

LTL Q3 16 wgt per ship

Graphs reproduced with permission from Stephens Inc. For more information contact: Jack Waldo at jwaldo@stephens.com or Brad Delco at brad.delco@stephens.com.