Logistics demand: The trade effect

Introduction

How much does demand for logistics space depend on international trade—and how might tariffs impact logistics markets across the U.S.? In this paper, we examine the connections between changes in freight movement and logistics demand across 69 U.S. markets and develop a framework to predict the marginal change to logistics demand resulting from changes in international trade.

To conduct this analysis, we draw on the Freight Analysis Framework dataset, or FAF5, from the Bureau of Transportation Statistics (BTS) with support from the Federal Highway Administration (FHA). The current version of FAF5 provides estimates for freight tonnage by origin-destination pair of FAF regions, commodity type, and mode from 2017-2023 at a regional level for 69 metropolitan regions. Appendix 1 provides details on this dataset and our methodology.

Foreign trade into and out of the U.S.

Relatively few U.S. metros have direct international trade links. The vast majority of imports enter and exports exit the country through just ten markets (see Figure 1). The twin ports of Los Angeles and Long Beach serve as the single largest trade gateway, through which a quarter of East Asian (which includes China, South Korea, and Japan) trade enters the U.S. Greater New York, which includes the ports of New York and New Jersey, ranks second and clears 17% of European imports. Houston’s production of petroleum-based products for export has facilitated the growth of its port since ships can leave laden with resin. About a third of Mexican imports enter through the Laredo, Texas border crossing, and Detroit serves as the primary entry point for Canadian imports. Other key international trade gateways include Savannah, the Bay Area, Chicago, Dallas-Fort Worth and South Florida.

Figure 1: Foreign trade by 2023 tonnage (millions), 10 largest markets

In the largest gateway markets, international trade represents a relatively small share of overall freight movements—about 31% in Greater Los Angeles and 25% in Greater New York. In Houston, Dallas, Chicago, Atlanta, Philadelphia and Boston, international freight traffic makes up 20% or less of the total. Detroit, the Bay Area, South Florida, Seattle and New Orleans have higher exposure to international trade (Figure 2).

Figure 2: Highest 2023 freight tonnage, top 25 markets (millions of tons)

Foreign trade and logistics demand

To estimate the impact of reduced trade on logistics demand at a market level, CBRE Investment Management’s Americas Research team modeled the annual change in logistics space per capita from 2018 to 2023 across the 69 regional areas included in the FAF5 dataset as a function of:

We shocked the raw FAF5 data based on international region of origin (e.g., trade to and from East Asia falls by 20%) and used the model coefficients to estimate impact on logistics demand at a market-level. We used per capita measures to normalize trade and logistics demand across markets. The results (shown in Figure 3) can broadly be interpreted as changes in the occupancy rate arising from reductions in trade.

• The logistics markets in Laredo, TX and Buffalo, NY have the largest exposure to foreign trade and a 20% decline in Mexican and Canadian trade respectively would increase the Laredo vacancy rate by some 800 bps and the Buffalo vacancy rate by 300 bps. Neither are major logistics markets, however.
• Savannah, GA is home to the U.S.’ third largest port but just 430,000 residents, resulting in one of the highest levels of logistics space per capita in the U.S. Despite its location on the east coast, more than half of Savannah’s port traffic is from Asia, such that a 20% reduction in East Asia trade would likely increase the vacancy rate by 150 bps. A similar reduction in European trade would increase the vacancy rate by another 80 bps.
• Charleston, SC has a similar profile to Savannah as a large port in a relatively small market, but with relatively more dependence on European trade. A 20% decline in trade with Europe would likely increase the vacancy rate by 150 bps. A similar reduction in East Asian trade would increase the vacancy rate by another 80 bps.
• Seattle serves as a secondary west coast gateway for Asian trade, as well as a primary entry point for Canadian goods. The model estimates that a 20% reduction in East Asian trade would result in a 70 bps increase in Seattle’s logistics vacancy rate, and a 20% decline in Canadian trade would increase vacancies by an additional 30 bps.
• Mobile, AL serves as a key rail transshipment hub between east and west coast freight rail lines. Rail traffic requires much less logistics space per ton, but the sheer amount of international freight traffic passing through the market’s relatively small logistics inventory magnifies its exposure to international trade to and from all regions.
• Greater Los Angeles, which includes Los Angeles, Orange County and the Inland Empire, is the single largest logistics market in the U.S. with nearly 1.5 billion SF of inventory serving more than 18 million residents. The twin Ports of Los Angeles and Long Beach handled nearly 13 million loaded TEUs in 2024, more than double the Greater New York ports. More than 80% of port traffic is to and from Asia, and per our model we estimate that a 20% reduction in East Asia trade would increase Greater Los Angeles’ logistics vacancy rate by 30 bps.
• Greater New York, which includes the New York, Northern New Jersey and Allentown (Lehigh Valley) MSA Divisions, is home to the largest population in the U.S., the second largest logistics inventory, and the second busiest port complex. Port traffic is diversified across Europe, Asia and Canada, and a 20% reduction in trade would likely increase logistics vacancies by 40bps, 30bps and 20bps, respectively.

Figure 3: Predicted change in occupied logistics space given a 20% decline in trade, most affected markets

On the other hand, other large logistics markets including Chicago, Dallas-Fort Worth, Phoenix, Houston and Atlanta have lower exposure to international trade, and many smaller Midwest markets are unlikely to be affected by reductions in trade. Figure 4 shows predicted outcomes for logistics demand for the 25 largest logistics markets in the U.S. by inventory.

Figure 4: Predicted change in occupied logistics space given a 20% decline in trade, largest logistics markets.

Summary and conclusions

The Trump administration’s imposition of extraordinary tariffs will likely decrease international trade into and out of the U.S. and as a result also decrease demand for logistics space in markets most dependent on trade—Savannah, GA; Charleston, SC; Laredo, TX; Buffalo, NY and Mobile, AL. None of these metros are major logistics markets and pose relatively little risk to institutional investors.

Reduced trade is likely to further the demand losses already sustained in Southern California and Greater New York, the two largest logistics markets and the two largest ports in the U.S. Other large logistics markets, however, have minimal exposure to international trade. And many large logistics markets feature relatively healthy fundamentals, like Chicago, Kansas City, St. Louis, Memphis or Minneapolis, or resilient demand—albeit amidst heavy supply—as in Dallas, Houston, Phoenix and Nashville. Tariffs add a significant new headwind to logistics demand in major port markets, but could also create investment opportunities:

• Gateway markets: The heavy demand losses in Greater Los Angeles and New York over the past two years and now concerns around reduced trade create opportunities for contrarian investors with long time horizons to acquire irreplaceable assets in these markets at reset bases.
• Midwest markets: Often dismissed for their lack of growth and new inventory, Midwest markets—anchored by Chicago but including Kansas City, Louisville, Memphis, St. Louis, and Cincinnati—feature low vacancy rates relative to the national average and, importantly, limited exposure to international trade and minimal oversupply. These markets also have outsized manufacturing sectors potentially poised to benefit from increased domestic production.
• North American trade partners: Thus far, Trump has levied the highest tariffs on Europe and Asia—and especially China. The 10% tariffs on Canada and Mexico exempt USMCA goods, making imports from North American trade partners relatively cheap and potentially increasing trade. Texas markets are entering a period of much reduced supply and offer growth profiles and exposure to Mexican trade, while Detroit, Minneapolis-St. Paul and Seattle could benefit from increased imports from Canada.

Appendix 1

To conduct this analysis, CBRE Investment Management’s Americas Research team drew on two key data sources—CoStar, for logistics supply and demand by market, and the Freight Analysis Framework (FAF) dataset on freight movement from the Bureau of Transportation Statistics. Per the FAF website, “The Freight Analysis Framework creates a comprehensive picture of freight movement among states and major metropolitan areas by all modes of transportation. The FAF integrates data from a variety of sources. Starting with data from the Commodity Flow Survey (CFS) and international trade data from the Census Bureau, FAF incorporates data from agriculture, extraction, utility, construction, service and other sectors. The FAF is produced by the Bureau of Transportation Statistics (BTS) with support from the Federal Highway Administration (FHWA).”

“The current version of FAF5 (FAF5.6.1) provides estimates for tonnage (unit: thousand tons), value (unit: million dollars) and ton-miles (unit: million ton-miles) by origin-destination pair of FAF regions, commodity type, and mode for the base year (2017), the recent years (2018-2023), the forecast year estimates (2025-2050), and the state level historical trend estimates (1997-2012). The information may be accessed through the Data Tabulation Tool and downloaded as either a complete database or in summary files.”

CBRE IM first removed all “non-warehousable” commodities from the FAF dataset, which includes all pipeline freight traffic, as well as the following FAF commodity categories:

We then matched CoStar geographies to the 69 regional areas used in the FAF dataset. In particular, the FAF helpfully creates market areas that correspond to logistics trade areas in Greater Los Angeles, which includes Orange County and the Inland Empire, as well as Greater New York, which includes the New York, Long Island, Northern New Jersey and the Lehigh Valley MSAs.

We ran models on both the change and level of logistics demand from 2017 to 2023 (the last year available in the FAF) as a function of mode of transport (truck, rail, air, water or multi-mode), provenance (import, export or domestic) and whether the traffic was inbound to the market, outbound from the market or intra-market. The large number of combinations, however, resulted in too many independent variables.

To solve this issue, we ran two separate regression models on the change in logistics demand per capita. The first model considered the change in freight tonnage per capita by mode of transport, as well as control variables including the share of new supply in the market, the change in the total log household income in the market and rent growth in the prior year. The second model considered the provenance of freight traffic—domestic, imports and exports—and its direction (intramarket, outbound or inbound), as well as the same set of control variables.

Both models are statistically meaningful and have r-squared measures above 0.50. To arrive at our estimates used in this paper, we predicted outcomes using both models and then took the average.