The demise of retail real estate, mostly at the hands of e-commerce, has had one positive side-effect for the industry: it’s been a boon for owners and developers of warehousing space, and industrial real estate more broadly. Numbers speak louder than words: according to CBRE, the average cap rate for acquisitions of industrial space reached a record low of 6.3% in the first half of 2019 (as a comparison: office stood at 6.7%, neighborhood retail at 7.5%, multifamily at 5.2% and hotels at 8.3%), down from 7.5% in 2013.
In some markets, such as LA’s Inland Empire, cap rates are now consistently below 4%. That’s not much higher than what you would get on a virtually risk-free savings account. Public market investors have also taken notice: the stock price of Prologis, the largest industrial REIT in the U.S., is up 27% for the year (compared to 2.6% for the S&P) and up 121% over the past 5 years (compared to 52% for the S&P).
A wall of capital looking for opportunities
The attractiveness of industrial and warehouse assets is, of course, no state secret, and investors of all colors and stripes have piled into the sector. Real estate giant Blackstone has amassed 800 million sqft. of industrial space, as mentioned in this WSJ article. Prologis, another gorilla in the industrial space, just acquired Liberty Property Trust, for USD 10 billion. The “average” private equity investment manager has also significantly increased his or her allocation to real estate. Constituents in the MSCI PREA/ACOE Index (the ACOE Index includes the major commingled private equity real estate funds, such as those managed by JP Morgan and Morgan Stanley) increased their allocation to industrial real estate from a low of 12.5% in 2015 to 17% in Q2 2019.
As capital continues to flow into the real estate space (Preqin estimated the “war chest” or “dry powder” of private equity real estate funds at some USD 333 billion in March 2019), and industrial real estate remains the “flavor of the day” (or perhaps: the decade), many real estate investors and developers are looking for opportunities in the industrial real estate space. But equally, investors that developed or acquired industrial assets over the past years may consider capitalizing on the current craze by offloading some of their assets, or their full portfolio of assets.
That leads to the question: which markets will do well, with significant rental growth ahead, and which markets will see rental growth tapering off, or perhaps even decreasing?
Understanding industrial markets using (big) data
Forecasts of rental growth (or decreases in rental rates) are typically inaccurate (this is an interesting paper that studies the forecasting performance of real estate professionals). Today’s plethora of big data can help investors spot the causes of future rent change. For the forward-looking investor, it’s just a matter of sifting through those mountains of data, but that’s not exactly what an investor is looking to do.
Enter GeoPhy Value Drivers—a set of curated market and location attributes that help to quantify the demographics, freight outlook, and hyperlocal characteristics of the area surrounding a property. ValueDrivers come at four levels:
- Macro or nationwide
- Regional or metro area
- Neighborhood or location
We will look into a selection of these Value Drivers using an overlay of all REIT-owned U.S. assets with an “industrial” focus—assets owned by the likes of ProLogis, Duke Realty, W.P. Carey, First Industrial, etc. Given that all of these assets are in the U.S., macro isn’t necessarily worth discussing, as it refers to macroeconomic conditions, such as core inflation, the FED interest rate, stock market performance, and sentiment indicators such as consumer confidence. These factors mostly affect capitalization rates, for all property types, and across the country.
Comparing industrial properties, from Kansas to Carolina
In contrast to macroeconomic indicators, regional or metropolitan area indicators are highly relevant for industrial assets. Which metro areas will see demographic growth in the next 5-10 years? Which markets provide a deep pool of (affordable) labor—an increasingly important question in the current tight labor market?
One source of data that we uncovered at GeoPhy is the Freight Analysis Framework, developed by the Bureau of Transportation Statistics. The graphs below provide a snapshot of the Freight Analysis Framework data (an online, interactive version can be found here). There are some clear differences between areas—for example, the amount of freight in the Raleigh-Durham metro should decrease (clearly, the region is shifting to higher-value freight), whereas freight volume in the Kansas City metro should increase both in tonnage and in dollars. Although high level, this data provides a good first insight into a ranking of regions by freight flows and of growth therein.
The Freight Analysis Framework graphs above underscore just how diffuse industrial investment opportunities are. Given the wide variety of metros that fit this initial criteria, additional analysis is clearly needed to assess a property in one metro relative to a seemingly similar property in another metro.
This is where an investment mandate or thesis becomes an important consideration. For purposes of illustration, we evaluate two different purposes:
- Industrial properties that will benefit from the continued growth of e-commerce, so those that are in close proximity to last-mile markets, with high density of consumers, especially those with higher income; and
- Industrial properties where light manufacturing or regional distribution takes place, requiring (light) industrial job skills and a deeper pool of affordable labor.
Zooming in: demographics and socio-economic indicators
Data at the metro area provides pretty good insight into demographics—both to understand access to a skilled and willing workforce, as well as access to a large catchment of potential consumers (most relevant for regional distribution centers, and critical for last-mile delivery, as laid out nicely by Prologis in this article). How many people live in an area? What’s the average household income? What’s the cost of housing?
We pick two industrial properties at two totally different places:
- A 340,000 sqft. asset located in Greensboro, North Carolina, owned by Liberty Property Trust
- A 314,000 sqft. asset in Olathe, Kansas, owned by Monmouth Real Estate Investment Corporation.
Both are larger facilities, both “multi-market” properties at key transportation hubs. Using Census ACS and Zillow data, we analyze the nearby area. It’s important to look beyond the immediate catchment or nearby area of a property. We measure every indicator both at the ZIP level, as well as within a region covered in a 30-minute and 60-minute drive, using GeoPhy Reach.
At first glance, Olathe, located just south of Kansas City, seems more populated, with higher incomes. The total population in the ZIP code is 80,000 people, (in 27,385 households), but within a 60-minute Reach, there are more than 2 million people (in 806,000 households). The median household income is USD 92,000 (relatively high compared to the U.S. average)—the aggregate household income within a 60-minute drive is some USD 64 billion. That’s great for assets focused on local distribution (there are many willing buyers for goods stored in distribution centers), but of course, for assets focused on regional and pan-regional distribution, different things matter (those facilities need people to work).
Greensboro seems to be better positioned here. Within 60 minutes, there are 1.6 million people. But, the labor market participation rate is relatively low, house prices are low (which implies higher affordability), and the population is not as highly educated (logistics facilities typically offer few high-skilled jobs). We also have data on hourly wages for different job types. For industrial/logistics, we’re specifically interested in jobs related to Production, Transportation, and Material Moving (PTMM). In Greensboro, hourly wages are at USD 16.65, decreasing slightly to USD 16 at a 30-60 minute driving distance. In Olathe, where the labor market is tighter, wages are about 10% higher.
Other nearby points of interest: Amazon, airports, and ports
Using other sources of data, we evaluate the distance of a property to the following key points of interest that are relevant for tenants in (and thus owners of) industrial space:
• Airports, ports, rail terminals
• Amazon fulfillment centers (presumably Amazon has a superior location selection strategy)
• Nearest highway on-ramp
• Presence of public transit (bus stops matter for industrial facilities, at least for their employees)
• And more
The pictures below (online, interactive versions can be found here) show both Olathe and Greensboro are far inland and don’t have direct access to seaports. But the Greensboro asset is very close to an airport, whereas the Olathe asset is in close proximity to Amazon.
A hyperlocal perspective: from fast food to financial services
A final layer of data that is often overlooked is where (hyperlocal) location data comes in: the micro level. Here, we investigate the presence of amenities for employees within walking and driving distance—a place to eat, to drink, or do some quick shopping after work. Of course, there are many other data layers available at the local level, such as the incidence of crime, but also more sophisticated data such as real-time mobile phone data to better understand local dynamics.
At the hyperlocal level, Kansas and the Carolinas (Olathe and Greensboro) are quite different. For example, there is a relatively high density of amenities in Greensboro, as opposed to the asset in Olathe. Whereas restaurants, shops, and other amenities are not typically associated with industrial real estate (much more so with multifamily and office properties), such amenities are increasingly important for the workforce. As the battle for talent shifts to all jobs, not just white-collar jobs, employers compete for labor, and the presence of amenities close to the job may be a competitive advantage, as access to amenities could increase employee morale and retention.
Closing thoughts: data driving investment decisions
As e-commerce continues to become a more important part of retail sales, competition for industrial space won’t be abating anytime soon. The war for “last mile, last touch” locations is just starting, and it is in these locations that land and space are especially scarce. But whether it is for regional distribution centers or properties used for last-mile delivery, investors will find themselves increasingly in markets that they previously did not operate in. Local knowledge sourced through brokers is useful, but hard to scale and often hard to quantify.
Relevant data can help investors and lenders objectively assess the merit of locations, supporting location selection as well as underwriting decisions. Of course, the use of a property may require being in close proximity to an urban area, such as in the Olathe example above, or it may rather require the presence of a large workforce that may be available to work in the industrial facility. In the era where bricks-and-mortar retail is replaced by data-driven e-commerce platforms, using data for investment, lending, and location decisions for industrial and logistics real estate is the best way forward.