Home » 2015 » October

Monthly Archives: October 2015

Heard at Dodge Data Outlook 2016, Oct. 30, 2015

Dodge Data & Analytics Outlook 2016 event held in Washington DC, October 30, 2015.

A brief summary of comments heard and information from my notes.

Art Gensler – Founder Gensler

How do you control 5000 people?  Hire good people and get out of their way.

People value what they pay for and ignore what they get for free.

Beth Ann Bovino – U.S.Chief Economist, Global Economics & Research, Standard & Poor’s

Domestic economy is strong and strengthening.

Jobs are stronger – Quits rate is at a 7 year high.

Housing starts are up – Home prices are up.

Wages are struggling and we have a historical 38 year low labor participation rate.

Ted Hathaway – CEO Oldcastle BuildingEnvelope

We increased wages significantly to keep people from leaving.

The cost and disruption is huge if you lose a valuable member of a team.

Dan McQuade – President, Construction Services, AECOM

Three emerging trends

Global collaboration

Investing capital with clients and partners

Better collaboration with vendors & suppliers. Treat subs and vendors as partners.

Larry Kudlow – Economist and Senior Contributor CNBC

Our biggest problem – We do not have strong steady economic growth.

Corporate profits were high after recession but have declined last three quarters. Profits were likely responsible for the stock market rise.

Bob Murray – Vice President, Economic Affairs, Dodge Data & Analytics

The DMI is reflecting the institutional dip has ended and now beginning to grow, although slowly.

New construction starts 2013 = 11%, 2014 = 9%, 2015 = 13%p

Actual $ put-in-place 2013 = 7%, 2014 = 5%, 2015 = 10%

New starts that declined in 2015 Warehouses, Stores, Public Bldgs, Manufacturing

New Starts that increased in 2015 Residential, Hotels, Highway, Electric-Gas-Power

Expectations for 2016

Total new construction starts up 6%.

Residential up 16%, single family will grow faster than multifamily.

Commercial up 11%, led by warehouses and stores

Institutional up 9%, led by educational

Manufacturing down 1%, but from very high 2014 and 2015

Power down 43% from extreme high starts in 2015

Construction cycles may be indicating we have years of growth left in the current cycle.

Nonresidential Buildings Construction Spending 2015 – How Do Industry Predictions Compare?

Throughout the year a number of firms provide predictions of various construction data.  Some firms provide estimates for all segments of construction.  More firms provide estimates only for spending on nonresidential buildings.  This is a summary of various firms estimates published in the 2nd quarter and also for those who’ve updated their estimate recently.

The current available spending data through August allows an analysis of a select data set that gives a prediction of the year end result within +/- 1.5%. My current data predicts 2015 will finish with nonresidential spending at $393 billion, with a potential range between $387 billion and $400 billion. We will have even better data on November 2nd when the US Census publishes construction spending for the month of September. Once the September data is incorporated into the monthly totals, an analysis of a select data set provides a prediction of the year-end totals that has not varied more than +/- 1% from the end-of-year actual since 2002, as far back as the market data is available.

Comparison Nonres Bldgs

Construction Spending or Construction Starts – Media Headlines That Get it Confused

Here’s a headline published recently.

“Construction spending rose year-over-year during first 9 months of 2015”

What they really meant to say

“Construction STARTS rose year-over-year during first 9 months of 2015”

Why make a big deal out of this.  Well, here’s an example.  The brief states “nonbuilding work climbed 35%” during the first nine months of 2015 compared to same months 2014.  TRUE, if we look at new construction STARTS, but construction spending for non building work is down 2.5% during the first 8 months of 2015 vs 2014 (Sept spending won’t be available until November 2nd).

Construction starts help us understand where future spending is headed.  Construction spending is based on how much is occurring this month from all the projects that started in the previous 12 to 30 months.

Let’s use an example to understand new construction starts vs construction spending:

Assume:

All the nonresidential building projects that start in a given month have a schedule duration of 20 months.

The total starts in this month is $60 billion.

Over a duration of 20 months they will average a put-in-place construction spending of 60/20 = $3 billion a month. Spending follows a typical bell curve, starts out slow, gets real strong in the middle and finishes slow. The spending cash flow from these new starts will begin slowly at a rate of perhaps $1 to $2 billion/month, it will peak in about 10 to 12 months at perhaps $4-5 bil/mo, and then will taper to the finish again at about $1-2bil/mo. It will take 20 months to spend this $60 billion in new starts from this single month.

The point here is this:  Current spending is based on the last 2 to 3 years of starts.  New starts will generate the next 2 to 3 years of spending.  Don’t get the two confused.  Big increases in new starts is not telling us how much spending is going on right now, it’s telling us how much spending to expect in the future.  And to get the real picture of what to expect in the future we need to look at the cumulative cash flow that occurs in any given future month from all the previous starts that are still ongoing. Furthermore, the construction starts values that are published every month represent a sampling of new construction activity, approximately 50% to 60% of total construction activity.  While the total reported construction starts for a year might be $500 to $600 billion, the total spending in that year will be more like $1.0 to $1.1 trillion.

Residential Construction – Not All Data Tells The Same Story

The latest New Housing Starts numbers were released today.  Residential growth is looking good and based on several inputs, I’m predicting an increase in residential construction spending next year.  But let’s take a look at the variance you might get when looking at different data sets.

All the data below represents residential construction growth for the period from January 2011 until current, the last 4 years 8 months

New Construction Starts in $ (by Dodge Data Analytics) +19%/yr

Snip Constr Starts DDA RES Jan11 Aug15

New Housing Starts (number of new housing units) +20%/yr

Snip Housing Starts Sept 2015

Total Construction Spending +12.5%/yr

Snip Constr Spend RES Jan11 Aug15

Volume (construction spending minus construction inflation) +7%/yr

Snip Constr Spend minus inflation RES Jan11 Aug15

The obvious first question is why don’t all the data agree?  Without a lot more information on housing that cannot be answered here, but there are a few reasons that can be considered as cause for variation;

  • the average size of housing units being built
  • the quality of the components built into the housing units
  • the cost to the contractor for the materials used
  • the cost of labor wages to build the housing unit

I’m sure there are other reasons to consider as this is not intended to be a complete list of what might cause variances between starts and spending, but it does highlight that starts does not give an exact indication of the growth in spending.  There is a fairly consistent growth rate in starts of 20%/year and yet construction spending in current dollars has been growing at only 12.5%/year.  Furthermore, a sizable portion of that spending growth is just for inflation.  After inflation is taken out we see real construction volume in constant 2015$ has been growing at only 7%/year.

I don’t have an answer to explain these variances.  I’m highlighting the data to show these variances exist and we can’t always rely on one data set exclusively.  Perhaps this will initiate a discussion as to why these data vary by so much.

Construction Spending Market Performance of Major Nonresidential Buildings 2015-2016

The Construction Spending BOOM in 2015 is being led by spending on nonresidential buildings.  Spending on nonresidential buildings year-to-date (YTD) is +20%, +$41 billion. For housing the YTD is +11%, +$24 billion and for nonbuilding infrastructure projects YTD is -2.5%, -$5 billion.

Let’s take a look at the current growth trends to find out where they are headed.

In 2004-2006, residential spending was 55% of all construction spending. The annual growth in 2004 was 19% and in 2005 it was 15%. For the last 5 years residential spending has been only 32%-37% of total spending.  In 2012 & 2013, residential led with annual spending gains of 13% and 19%. In 2014 & 2015, nonresidential buildings, also at 37% of total spending, led the gains at 9% and 19% growth. In 2016 the lead shifts back to residential with a projected growth of 14%. Infrastructure has not led growth since 2007 and 2008 when that sector had growth of 19% and 10%, at a time when residential spending was declining by 19% and 28%.

We can get a very good idea of nonresidential buildings spending and growth by looking at the five major markets. These five markets make up 85% of all nonresidential buildings construction spending and half of total 2015 construction spending growth.

Snip PCT of Spending 5 markets Oct 2015

See my blog post on October 11, 2015. I wrote:

“New nonresidential buildings construction starts cash flows indicate spending will continue to grow until Feb-Mar 2016, then drop consistently each month until Q3 2016.  The decline is almost entirely due to big starts from Q3-Q4 2014 finishing and dropping out of the monthly spending numbers.”

Snip STARTS YTD Aug2015Snip Spending Growth 5 markets 2015 2016 Oct2015 

More detail of how each market will perform, and why, follows.

Educational Construction Spending 2016

Spending in 2016 is projected to grow +5% over 2015. Other industry projections for educational spending in 2016 range from 1.5% to 12% growth over 2015, with the average of those seven estimates at 6%. As of August 2015, project starts that will generate 60% of all spending in 2016 are already booked.

Starts for the first 8 months of 2015 were up 12% from the same 8 months of 2014.  Educational spending increased only 4% year-to-date 2015 from the same period 2014, but the current annual rate of growth is 11%. Monthly spending is increasing and should continue to do so at least until mid-2016 before dropping off slightly into year end.

Healthcare Construction Spending 2016

Spending for healthcare is expected to remain flat with no growth in spending in 2016.  Other industry projections for healthcare spending in 2016 range from 3% to 12% averaging 6% growth. As of August 2015, project starts already booked will generate 60% of all spending in 2016. New starts in 2016 generate about 25% of the total spending in 2016. If we get some very large new starts in the next few months, that could change total spending projections in 2016. Starts would need to increase 20% ( every month) over my projections for the next 16 months to reach 6% growth in spending next year.

Starts for the first 8 months of 2015 were down 4% from the same 8 months of 2014 and most recently have been declining. 2014 starts grew only 2% over 2013. Healthcare spending had an annual growth rate of 5% in the first eight months of 2015. The decline in new starts signals a projected decline in spending for the next 8 months. Spending growth resumes in mid-2016 but at a very low 3% annual rate and that from an already low rate of spending at the start of the year.

Snip Constr Spending Plot Educ Hlthcr Oct15 2015

Commercial/Retail Construction Spending 2016

Spending in 2016 is projected to grow +7% over 2015. Other industry projections for office spending in 2016 range from 5.5% to 15% growth over 2015, with the average of those estimates at 10%. As of August 2015, project starts that will generate 55% of all spending in 2016 are already booked.

Starts for the first 8 months of 2015 were up 17% from the same 8 months of 2014.  Commercial spending increased 15% in the first half 2015 from the first half of 2014, but then spending declined by 8% in the last three months and may continue to decline for the next few months.  Spending will resume a growth rate of 15% annual in the first 8 months of 2016. Commercial spending will peak in the second quarter 2016 before dropping again into year end.

Office Construction Spending 2016

Spending in 2016 is projected to grow +8% over 2015. Seven other industry projections for office spending in 2016 range from 7% to 18% growth over 2015, with the average of those seven estimates at 12%. As of August 2015, project starts that will generate 55% of all spending in 2016 are already booked.

Starts for the first 8 months of 2015 were 23% lower than the first 8 months of 2014  Spending from 2014 starts will start to drop off in late 2015 and early 2016 and based on new starts in 2015, by mid-2016 the monthly rate of spending will start to decline, keeping totals for 2016 to less than 10% growth. Spending on office buildings in 2016 will peak in the 1st half year with the 2nd half coming in 10% lower.

Manufacturing Construction Spending 2016

Spending in 2016 is projected to grow +9% over 2015. Seven other industry projections for manufacturing buildings spending in 2016 range from 5% to 18% growth over 2015, with the average of those seven estimates at 11%. As of August 2015, project starts that will generate 70% of all spending in 2016 are already booked.

Starts for the first 8 months of 2015 were only 6% lower than the first 8 months of 2014. However, even if starts for the next 4 months increase each month by 50% they will still not equal the amount of starts in the last 4 months of 2014.  Total starts for 2015 are projected to finish 20% lower than 2014.  That’s probably a good thing since 2014 starts were up 87% from 2013, the highest annual growth ever recorded for any market sector.

Spending from 2014 starts will start to drop off in late 2015.  Spending reached a peak this year in the 2nd quarter but is expected to drop for the next five to six months. Spending on manufacturing buildings in 2016 will again peak in the 2nd quarter and then drop off into the end of the year.

Snip Constr Spending Plot Mnfg Offc Comm Oct15 2015

Nonresidential Buildings Construction Spending Through 2016

New nonresidential buildings construction starts cash flows indicate spending will continue to grow until Feb-Mar 2016, then drop consistently each month until Q3 2016.  The decline is almost entirely due to big starts from Q3-Q4 2014 finishing and dropping out of the monthly spending numbers. New starts in 2015 did not grow as much as in the previous two years. Although the predicted decline in monthly spending over 6 months is 8%, 2016 may finish with a rate of monthly spending higher than when it started.

The drop and recovery can vary from the predicted shown here and it’s not likely to be so smooth, but new starts from here on forward would really have to skew from a normal growth pattern by a lot to change this pattern by a little.  Nonresidential buildings on average take about 20 to 24 months to complete, so every month we move out adds about 4% to 5% uncertainty to future spending.

This prolonged period of spending declines is sure to cause alarm in the headlines in mid-2016, but the decline and the reversal are supported in large part by starts already booked.  Unless something dramatic and unexpected comes along to throw a wrench in the works, I’m expecting a pattern like this for 2016.

Total nonresidential buildings spending in 2016 will finish the year about 10% higher than 2015.

Snip Constr Spend Nonres 4yr oct15

Welcome to my Construction Economics blog

Welcome to my new blog.  Here I will expand on current issues of construction economics.  On Twitter @edzarenski, I will tweet updates to my most recent Construction Economic report and out of necessity I will keep it short.  When issues demand further explanation, you will find it here.  Thanks for visiting. edz

%d bloggers like this: