Module #5
Module #5 Industrial real estate.pdf
279
The Environment and Performance of Industrial Real Estate John D. Benjamin,* Emily N. Zietz** and G. Stacy Sirmans***
Abstract
This study examined over seventy industrial real estate research papers and categorized them according to: 1) property characteristics, 2) demand determinants, 3) rent and income determinants, 4) return and valuation issues, 5) environmental concerns, 6) international issues, and 7) management and financing concerns. The findings indicate that industrial properties tend to be segmented and clustered by use with greater demand for flexibility and smaller size, yet with reduced demand for warehouse space. Employment, age, and percentage of office space affect industrial rents while industrial land values indicate future rental rates. In addition, local market factors, physical characteristics and location affect industrial property values, but environmental contamination has an adverse effect.
Introduction Interest in industrial real estate is increasing because it is a key sector of the real estate market in the United States. In 2001, for example, industrial properties accounted for almost 20% of the NCREIF Property Index with a market value of almost $21 billion dollars. Over the past twenty years, industrial properties have been a relatively stable investment for many investors, providing solid returns and attracting high demand, but the industrial real estate sector continues to be an evolving and dynamic market.
In this review of the empirical, theoretical and descriptive literature concerning industrial real estate, over sixty recent studies are examined by addressing seven questions: (1) What are the characteristics of the industrial real estate market? (2) What are the determinants of demand for industrial real estate? (3) What are the determinants of rent and income for industrial real estate? (4) What are the return and valuation issues in industrial real estate? (5) What are the environmental concerns in industrial real estate? (6) What is the international perspective of industrial real estate? and (7) What are the management and financing concerns in industrial real estate?
The article is organized as follows. The next section defines industrial real estate and identifies the environment for industrial properties. Succeeding sections survey the literature by addressing the above seven questions. The examination of categorical topics is presented both in a discussion and tabular format. The last section provides a summary and conclusions.
*American University, Washington, DC 20016 or [email protected]. **Middle Tennessee State University, Murfreesboro, TN 37132 or [email protected]. ***Florida State University, Tallahassee, FL 32306-1110 or [email protected].
280 JOURNAL OF REAL ESTATE LITERATURE
VOLUME 11, NUMBER 3, 2003
What is Industrial Property?
Expanding technology use in many industries is causing changes in the makeup of industrial real estate. Many industrial properties such as those used for research and development, for example, have benefited from improved technology. The performance of warehouse properties serving as large-scale distribution centers is similarly attracting attention.
In his seminal hedonic study of industrial real estate pricing, Ambrose (1990) defines industrial property as the land, structural improvements and equipment connected with a particular property that is being used for the conversion of materials into finished manufactured products. In general, warehouse property and minor processing plants of all kinds are included in the category of industrial property.
What specific types of properties are considered industrial real estate? According to the 1998 Real Estate Information Standards disclosed on NCREIF.org, there are five types of industrial properties.
1. Warehouse properties with at least 50,000 square feet with up to 15% office space, and the balance of the structure having 18- to 30-foot ceiling height; all loading must be dock-height.
2. Manufacturing buildings with 10- to 16-foot ceilings or sufficient height to accommodate overhead cranes must provide floor-height and dock- height loading.
3. Office showrooms that are single story (or mezzanine) with 10- to 16- foot ceilings with frontage treatment on one side and dock-height loading or grade level roll-up doors on the other. Less than 15% of these structures may be used as office space.
4. Flex space structures are single-story buildings with 10- to 18-foot ceilings with both floor-height and dock-height loading, including wide variation in office space utilization, ranging from retail and personal service to distribution, light industrial and occasional heavy industrial use.
5. Research and development industrial properties are one- and two-story, 10- to 15-foot ceiling heights with up to 50% office/dry lab space (with the remainder in wet lab, workshop, storage and other support), and with specific dock-height and floor-height loading.
Commercial properties used for other purposes including agricultural, offices (not mentioned in type 3 above), apartments and retail endeavors are not considered industrial properties.
Characteristics of the Industrial Real Estate Market
This section provides an overview of the literature discussing the nature and characteristics of industrial real estate relative to other types of properties while Exhibit 1 presents summaries of studies that focus on the characteristics of the industrial real estate market. Some of the earlier research on industrial properties
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
281
Exhibit 1
Characteristics of the Industrial Real Estate Market
Author(s) Title Source Date, Location, Time Summary of Article
Mitchell and Jucius (1933)
Industrial Districts of the Chicago Region and Their Influence of Plant Location
The Journal of Business of the University of Chicago, 6:2, 139–56
Three manufacturing districts in Chicago, early 1900s
Evaluates the services rendered by three Chicago manufacturing districts: the Central Manufacturing District, Clearing Industrial District and the Kenwood Manufacturing District. Finds that the impact of manufacturing districts includes valuable transportation connections, financial advantages and cosmetic benefits to both neighboring property owners and to businesses and consumers nationwide who gain by having efficient and economical transportation resources.
Mayer (1942) Patterns and Recent Trends of Chicago’s Outlying Business Centers
The Journal of Land and Public Utility Economics, 18, 4–16
Location of business subcenters within the city of Chicago; changes in land values 1931–1941
Highlights the impact of business subcenters on economic activity and land values. Finds that land values may be used as an index of the importance of business subcenters. The median decline of land values of business corners in outlying business intersections (blighted areas) was between 20% and 29.9% during the period examined.
Shenfield and Florence (1944– 1945)
The Economies and Diseconomies of Industrial Concentration: The Wartime Experience of Coventry
The Review of Economic Studies 12:2, 79–99
Industrial growth inside Coventry and Birmingham, early 1900s to 1940
Examines industrial growth inside Coventry, England, specifically the automobile manufacturing, artificial silk yarn and electrical equipment industries. Finds that the effect of the war altered and unbalanced the industrial structure of the city, expanded the motor, and aircraft and associated engineering industries, and lead to difficulties but ultimately additional contracts for textile and service industries.
282 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 1 (continued)
Characteristics of the Industrial Real Estate Market
Author(s) Title Source Date, Location, Time Summary of Article
Grissom, Hartzell and Liu (1987)
An Approach to Industrial Real Estate Market Segmentation and Valuation Using the Arbitrage Pricing Paradigm
Journal of the American Real Estate and Urban Economics Association, 15:3, 199–219
Quarterly return data from a large commingled real estate fund consisting of 382 properties of various types and locations with a market value of $2.6 billion
Evaluates the segmented market within industrial real estate in relation to risk versus return factors, and considers whether a submarket analysis or an integrated market analysis provides more accurate and useful results regarding rate of return and the Arbitrage Pricing Theory. Using the Chen methodology, concludes that the industrial real estate market is indeed segmented between various regions and a submarket analysis is better for predicting returns, but does not establish a link between general economic factors and industrial real estate returns.
Moriarty (1991) Urban Systems, Industrial Restructuring and the Spatial-Temporal Diffusion of Manufacturing Employment
Environment and Planning, 23:11, 1571–88
Manufacturing employment density distribution patterns in the U.S. from 1947–1982
Finds that geographic sorting of manufacturing operations based on marginal returns has led non-production workers to become more concentrated in large urban centers such as in the Northeast and Eastern North–Central U.S. This sorting process has helped influence geographic expansion of development but the influence is beginning to weaken.
Graham and Bible (1992)
Classifications for Commercial Real Estate
The Appraisal Journal, 60:2, 237–46
No empirical data Develops a systematic approach for classifying apartments, office buildings, retail centers and warehouse buildings.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
283
Exhibit 1 (continued)
Characteristics of the Industrial Real Estate Market
Author(s) Title Source Date, Location, Time Summary of Article
Harrison (1992) Industrial Districts: Old Wine in New Bottles?
Regional Studies, 26:5, 469–83
No empirical data Analyzes a recent trend of economic growth where similar types of companies are locating in close proximity to one another and contrasts this pattern with other trends that have led to geographic dispersal of some production facilities. Suggests that governmental incentives may encourage similar businesses to form ‘industrial districts’ by providing infrastructure and other services. Notes that the interaction of independent firms located closer together may be able to lower overall costs and that trust can be built more easily in face-to-face meetings. Concludes industrial district prototypes involve more than simply agglomeration economies.
Bruce (1994) Industrial Goes Upscale Journal of Property Management, 59:2, 14–7
Industrial property near the U.S. –Mexico border, May/June 1994
Evaluates the changing and higher technology standards that industrial property tenants demand as they consolidate operations into a few major locations. Finds that the elimination by NAFTA of duties and tariffs on goods being sent between the U.S., Canada and Mexico has influenced the demand for industrial property. Documents a shift from industrial to office / industrial spaces demand as tenants who previously sought cheaper office buildings move to save money and to reduce rents and commute times for employees.
284 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 1 (continued)
Characteristics of the Industrial Real Estate Market
Author(s) Title Source Date, Location, Time Summary of Article
Franklin (1994) Industrial Inspirations Journal of Property Management, 59:3, 18– 20
No empirical data Asserts that industrial properties are strong performers in the real estate industry and that with innovative management, even older industrial properties may be excellent investments. Provides three case studies on how to make older industrial properties better investments.
Myers (1994) Fundamentals of Production That Influence Facility Designs
The Appraisal Journal, 62:2, 296–302
Utilizes information dated back to 1985
Suggests that valuing a production facility presupposes evaluation of industrial production methods. States that it is essential to understand the productive capacity of a facility and the history of the property before trying to assign a value to a property. Evaluates various advantages and disadvantages in facility design: capacity of a property allows evaluation of sales of similar buildings based on design capacity, potentially the most important unit of industrial engineering measurement. Evaluates the factors the companies should consider when searching for an industrial property including property flexibility and the capability of the site to handle future technology, capacity and plant design.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
285
Exhibit 1 (continued)
Characteristics of the Industrial Real Estate Market
Author(s) Title Source Date, Location, Time Summary of Article
Bruce (1995) Recycling Your Industrial Properties
Journal of Property Management, 60:4, 44–7
No empirical data Discusses obsolescence strategic planning and environmental concerns as the commercial and industrial real estate markets suffer, suggesting that alternative uses must be found or property may be abandoned. Categorizes three types of properties: those affected by central business district (CBD) expansion due to their downtown location, CBD-related or waterfront-related residential projects and properties further from desirable CBD or waterfront. Discusses managing approaches for handling vacant properties. Evaluates the decisions involved in a potential recycling project, such as one that is converted from commercial to residential.
Venables (1996) Localization of Industry and Trade Performance
Oxford Review of Economic Policy, 1996, 12:3, 52–60
No empirical data Examines the determinants of a country’s industrial structure and trade patterns. Finds that ceteris paribus, firms are more profitable and are able to pay higher wages in locations that have the benefit of an agglomeration of complementary activities. Finds that comparative costs of different sectors determine the pattern of trade and that agglomeration forces are only one factor determining trade patterns.
286 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 1 (continued)
Characteristics of the Industrial Real Estate Market
Author(s) Title Source Date, Location, Time Summary of Article
Black, Wolverton, Warden and Pittman (1997)
Manufacturing versus Distribution: Implicit Pricing of Real Property Characteristics by Submarket
Journal of Real Estate Finance and Economics, 15:3, 271–85
331 industrial property sales in southeastern US from 1987–1995
Evaluates differences in distribution activity and manufacturing activity, and proposes that these differences may lead to pricing differences. Provides evidence that there are indeed two submarkets in the industrial real estate environment—manufacturing and distribution. Enumerates significant variables for manufacturing properties (age, year of sale, MSA distance and lifting crane) and significant variables for distribution properties (air-conditioning, rail and excellent condition rating). Concludes that industrial submarkets should be considered when pricing a property and finds that of these fourteen characteristics, four proved to be different between the two submarkets. Provides analysis showing that the existence of these submarkets is a result of production process differences.
Cleminshaw (1997)
Measuring Superadequacy in Older Industrial Plants: A Case Study
Assessment Journal, 4:1, 37–40
No empirical data; Case Study
Defines superadequacy as ‘‘an element within a building’s design which creates an undesirable excess’’ and provides procedures and a case study of computing functional obsolescence due to superadequacy.
Thompson (1997)
Industrial Employment Densities
Journal of Real Estate Research, 14:3, 309–19
510 U.K. industrial properties
Evaluates long-term floor space-to- employment ratios in the U.K. and seeks to identify densities formed by specific industrial building types and to explain these densities and how they change over time. Attempts to create a method for predicting demand for various types of industrial property by linking employment data with property data. Concludes that there is no identifiable relationship between macroeconomic trends and employment densities.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
287
Exhibit 1 (continued)
Characteristics of the Industrial Real Estate Market
Author(s) Title Source Date, Location, Time Summary of Article
Hanson (1998) North American Economic Integration and Industry Location
Oxford Review of Economic Policy, 1998, 14:2, 30–44
Canadian, Mexican and U.S. manufacturing employment and wages, 1850–1997
Finds that in Mexico, closer economic ties with the U.S. appear to have contributed to a contraction of employment in the Mexico City manufacturing belt, an expansion of manufacturing employment elsewhere in the country and an increase in wages of skilled workers. Concludes that with the exception of U.S. cities on the Mexican border (whose employment growth is strongly related to export production in neighboring Mexican cities), the effects of economic integration on industry location in Canada and the U.S. appear to have been much weaker.
Ellison and Glaeser (1999)
The Geographic Concentration of Industry: Does Natural Advantage Explain Agglomeration?
American Economic Review, 89:2, 311–16
Four-digit manufacturing industries from 1987
Notes that agglomeration may be caused by local industry spillovers or as a result of natural cost advantages present in a specific area. Evaluates the effects of natural advantage location, suggesting that industrial location is a product of resource availability and labor-market advantages. Presents a model that explains industry clusters in three ways: observed cost advantages, unobserved costs advantages, geographically localized industry spillovers. Considers sixteen interactions relevant to the advantages provided from natural resources, labor and transportation costs. Concludes that at least 50% of the industry concentration is a result of natural advantages present in specific regions of the country.
288 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 1 (continued)
Characteristics of the Industrial Real Estate Market
Author(s) Title Source Date, Location, Time Summary of Article
Hill and Brennan (2000)
A Methodology for Identifying the Drivers of Industrial Clusters: The Foundation of Regional Competitive Advantage
Economic Development Quarterly, 14:1, 65–96
194 industries in the Cleveland-Akron, Ohio
Presents a methodology for determining clusters of industries for which a region holds a competitive advantage—the ‘‘driver industries—because they have a significant economic effect on the region. Identifies ‘Foundational clusters’ for areas such as metalworking, chemicals and allied products, as having important relationships with other industries in the area. Concludes that there is a strong competency in materials industry especially metals and specialty chemicals, suggesting that the region must be innovative.
Crawford and Slade (2001)
Appraising Industrial Special-Purpose Properties
The Appraisal Journal, 69:2, 161–73
Hypothetical asset value Illustrates a new, convenient method, based on utilization rates, for estimating economic obsolescence in the appraisal of industrial special-purpose properties. Stipulates that this utilization-based method consider the operating leverage of the property, allowing an adjustment for obsolescence. Suggests that in order to accurately estimate economic obsolescence in industrial special purpose properties, an industrial appraiser must have access to information on underutilization, operating leverage, replacement cost new, current age and the total estimated economic life of the property.
THE ENVIRONMENT AND PERFORMANCE OF INDUSTRIAL REAL ESTATE 289
examined the economic and general impact of manufacturing districts and industrial parks. A study by Mitchell and Jucius (1933) evaluates the service benefits of three manufacturing districts in the Chicago Central Manufacturing district. This study highlights the unique services that industrial districts provide, including transportation and financial benefits to many businesses and consumers in these areas, as well as consumers across the country who receive shipments conveniently and cost efficiently. Mayer (1942) also examines major industrial centers in Chicago and finds varying rates of change in land values when specifically examining important outlying business corners. Industrial concentration in England was examined by Shenfield and Florence (1944–1945). This study highlights the effect that the wars had on several key industries, including altering the industrial structure of the city, expanding the motor and aircraft industry and associated engineering industry, and causing difficulties but ultimately facilitating additional contracts for textile and service firms.
A systematic classification approach for defining various commercial properties, provided by Graham and Bible (1992), includes apartments, industrial properties, office buildings, retail centers and warehouses. Consolidation of industrial real estate into a few major locations owing to higher technology standards is found by Bruce (1994). Examining the issue of market segmentation within the industrial real estate market, Grissom, Hartzell and Liu (1987), for example, find empirical evidence to support a segmented market within the industrial real estate environment, and conclude that looking at regional submarkets allow better predictions of returns. Black, Wolverton, Warden and Pittman (1997) show that the industrial real estate market can be divided into at least two subcategories: manufacturing and distribution. Their study suggests that property pricing is still most feasible when measured in aggregate, but that adjustments should be made for variability between subcategories.
Several studies examine the clustering effect of local industrial property markets. Moriarty (1991) attributes the concentration of nonproduction workers in large urban centers to the geographical sorting of manufacturing operations based on marginal returns. The advantages and disadvantages of companies locating within close proximity are discussed by Harrison (1992). He also examines governmental incentives and infrastructure that encourage businesses to form ‘‘industrial districts.’’ Firms are more profitable, according to the findings of Venables (1996), in locations that have an agglomeration of complementary activities.
Natural cost advantages are used as an explanation by Ellison and Glaeser (1999) for the geographic concentration of industrial properties. Their study suggests that firms can save on labor and other resource costs through industry clustering and that transportation and marketing costs can be reduced by locating closer to target markets. This concept is further developed by Hill and Brennan (2000) who examine 194 industries in the Cleveland-Akron, Ohio area in an attempt to determine clusters of industries or ‘‘industry drivers’’ for which a region holds a competitive advantage. Their study concludes that there is a strong tendency in the materials industry, specifically metals and specialty chemicals, for driver industries to add value to a region. In Mexico, Hanson (1998) finds that closer economic ties to the U.S. caused
290 JOURNAL OF REAL ESTATE LITERATURE
VOLUME 11, NUMBER 3, 2003
reduced employment in the Mexico City manufacturing belt, but that there were advantages of economic integration for cities along the U.S.-Mexico border.
Several studies have focused on the characteristics of industrial properties that enhance the utilization of property. One study examines obsolescence and suggests ways to recycle industrial properties to prevent them from becoming abandoned (Bruce, 1995). In examining changes in central business district (CBD) properties, this study elaborates on how to handle vacant industrial properties in ways such as converting them to residential. While providing case scenarios on how to make older properties good investments, Franklin (1994) notes that industrial properties are one of the stronger performers in the real estate industry. His study suggests that innovative management could facilitate older industrial properties becoming outstanding investments. Crawford and Slade (2001) illustrate a new method, based on utilization rates and operating leverage, for estimating economic obsolescence in special-purpose industrial properties.
Some studies have examined physical characteristics such as capacity and design. Myers (1994) discusses the importance of understanding the productive capacity and design of the facility, arguing that design capacity is the most important measurement for an industrial engineer. Other factors important in industrial property design include flexibility in technology and capacity expansion. Cleminshaw (1997) argues that a building’s design can create functional obsolescence by creating undesirable excess.
To evaluate long-term floor space-to-employment ratios in England, Thompson (1997) uses employment densities. His study finds no significant relationship between macroeconomic trends and employment densities.
The major findings from this section are:
� Industrial markets are segmented (e.g., manufacturing and distribution) and investment returns are better predicted in sub-markets. Studies have provided a systematic approach to classify industrial property types.
� Clustering is present in industrial properties due to infrastructure, government incentives and natural cost advantages (e.g., labor, transportation, marketing costs).
� There is some advantage to economic integration.
� Recycling industrial properties to other uses is important because obsolescence can occur due to design and age. Studies have suggested ways to revitalize industrial properties in order to avoid abandonment.
� Capacity and design are important. Properties must be flexible for technology and capacity expansion.
Determinants of Demand for Industrial Real Estate Studies that examine the demand patterns and determinants and the supply of industrial properties are synthesized below and reviewed in Exhibit 2. Using Detroit
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
291
Exhibit 2
Determinants of Demand for Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Wheaton and Torto (1990)
An Investment Model of the Demand and Supply for Industrial Real Estate
American Real Estate and Urban Economics Association Journal, 18: 4, 530–47
Fifty-two major metropolitan areas with 130,000 structures from 1972–1989
Develops a model that is useful in estimating excess plant capacity in particular markets. Suggests that many tenants actually use industrial property for some mix of uses including manufcturing, distribution storage and research. Suggests that construction lag time may cause a temporary inefficiency in the supply and demand for industrial property.
Hartman (1991) Industrial Real Estate: Go Figure
Real Estate Issues, 16:1, 23–7
Detroit Market Data from Michigan Society of Industrial and Office Realtors (SIOR) from 1982–1989
Examines the growth in average sales prices per square foot of industrial buildings and land and compares city values to those of the suburbs. Finds that the relationships between the number of market transactions and the sizes of industrial buildings did not change over the time period examined: sales and leases of buildings of 25,000 square feet or less regardless of market conditions are dominant. Shows that unit prices consistently decrease as building size increase.
Rauch (1993) Does History Matter Only When It Matters Little? The Case of City- Industry Location
The Quarterly Journal of Economics, 108:3, 843– 67
Theoretical model Examines the historical role of city-industry locations and the traditional incentive businesses have to remain subject to agglomeration economies. Finds that since World War II, industries have had a tendency to cluster together and are reluctant to move from an old, high-cost site to a new, low-cost site. Notes that industrial park property developers may alleviate this inertia through discriminatory pricing strategies.
292 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 2 (continued)
Determinants of Demand for Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Hughes (1994) Determinants of Demand of Industrial Property
The Appraisal Journal, 62:2, 303–09
Data for industrial and office properties in Atlanta, Chicago, Boston, New York, and Washington, DC from 1984–1988
Examines the difference in demand determinants for industrial properties versus office properties, noting that simple population changes may not be sufficient to determine current and future demand for industrial property. Finds that general population growth is only one factor that determines demand, while other factors include: technological changes, natural resources, labor, infrastructure and exchange capability.
Wheaton (1996) A Perspective on Telecommunications Technology and Real Estate: Office, Industrial, and Retail Markets
Real Estate Finance, 13: 2, 13–7
Data used are national Predicts enhanced technology will reduce industrial space demand by improving inventory management and thus reducing warehousing needs, but the effects of improved technology will not be too quick or drastic. Evaluates the effects of corporate downsizing, noting many companies use more subcontracted services, thus shifting rather than reducing real estate demand. Suggests that ‘‘Hoteling’’ may not be a suitable strategy to save on real estate costs unless the time spent in the office is random and not determined by cycles or seasons.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
293
Exhibit 2 (continued)
Determinants of Demand for Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Christensen, Wisener and Campos (1997)
Attributes of Tomorrow’s Warehouse Structures
Real Estate Review, 27:3, 51–7
No empirical data Assesses the changing spacing needs of warehouse tenants, but warns that these needs do not validate the views of either modernists or traditionalists. Consistent with modernists, tenants perceive a need for slightly more building square footage. Tenants’ expectations for change in dock doors vary with building size of their warehouse spaces. Users of properties with at least 50,000 square feet expect to add doors more quickly than users of smaller properties. Concludes that landlords’ belief that tenants prefer buildings with higher ceilings may be inaccurate.
Rabianski and Black (1997)
Why Analysts Often Make Wrong Estimates About the Demand for Industrial Space
Real Estate Review, 27:1, 68–72
No empirical data Offers three approaches to estimate industrial space demand: forecast industrial space relative to office space, estimate future manufacturing space demand and forecast the demand for business parks. Factors sometimes overlooked include the performance of the regional and national economy and overall economic growth. Proposes that the single most important problem is that analysts often assume that analysis of the demand for industrial space is synonymous to the analysis of demand for office space.
294 JOURNAL OF REAL ESTATE LITERATURE
VOLUME 11, NUMBER 3, 2003
market data from the state Society of Industrial and Office Realtors (SIOR), Hartman (1991) examines the growth in average sales prices per square foot of industrial buildings and land, comparing city to suburban values. This study finds that prices consistently decline as building size increases. Hartman also finds that demand is greatest for buildings of 25,000 square feet or less regardless of market conditions.
Examining the tendency of various firms within an industry to cluster together and often form the nuclei of a city, Rauch (1993) examines the historical role of city- industry locations. This study notes the historical incentive businesses have to remain subject to agglomeration economies. Rauch finds that firms are reluctant to move from an old, high-cost site to a new, low-cost site, but that property developers may alleviate this inertia through pricing strategies over time.
Evaluating data relevant to plant completions, Wheaton and Torto (1990) provide a model that is useful in estimating excess plant capacity in a particular market. They also stress that many tenants use industrial properties for mixed purposes such as manufacturing, distribution, storage and research.
Two studies discuss the impact that changes in technology are having on the demand for industrial properties (see Hughes, 1994; and Wheaton, 1996). Both studies concur that large companies are looking to consolidate operations into a few key locations. Hughes, in examining the determinates of industrial space demand, reports that firms are shifting from industrial to industrial /office properties, often as a way to meet their more advanced warehousing and distribution needs. These needs include more automated inventory tracking and ordering systems such as JIT inventory management. Wheaton (1996) supports the popular view that enhanced technology will reduce the demand for industrial properties as inventory management is improved, and thus the demand for warehouse property is shrinking. Another study addresses the changing demand for certain physical attributes of industrial space. Christensen, Wisener and Campos (1997) discuss the changing space needs for warehouse tenants such as square footage, dock doors and higher ceilings.
Noting the difficulties in analyzing industrial space that arise from the assumption that demand for industrial space is synonymous to the demand for office space, Rabianski and Black (1997) suggest three approaches. First, they estimate the demand for industrial properties relative to office space; second, they forecast manufacturing space demand; and third, they forecast the demand for business parks. It is important to not overlook the performance of the regional and national economies, overall economic growth and overall population change as demand drivers for industrial space.
The major findings about the demand for industrial real estate are:
� Price per square foot declines as building size increases, and demand is greatest for buildings of 25,000 square feet or less regardless of market conditions.
THE ENVIRONMENT AND PERFORMANCE OF INDUSTRIAL REAL ESTATE 295
� Estimating the excess supply of industrial properties is confounded by their use for mixed purposes such as manufacturing, storage, distribution and research.
� Large companies generally seek to consolidate operations into a few key locations and are shifting from industrial to office/industrial properties.
� The demand for warehouse property is decreasing due to the improved inventory management that results from enhanced technology.
� Demand for industrial space should be estimated relative to demand for office and manufacturing space, regional and nation economies, overall economic growth and population changes.
Determinants of Rent and Income for Industrial Real Estate
The determinants of rent and income for industrial real estate are examined in the articles grouped in Exhibit 3. Topics range from geographic differences in rental rates to the factors that best forecast rents and vacancy rates. Bible and Hsieh (1996) argue that Geographic Information System (GIS) technology may be useful in industrial real estate research. Where property characteristics in general are not important in determining warehouse occupancy rates, GIS technology may be used to examine warehouse occupancy rates by applying distance variables along with conventional property variables. Where location is important, Reenstierna (1997) discusses the advantages and disadvantages of using a ‘‘baseline’’ valuation model in lieu of the traditional sales comparison method. His study also uses a GIS map, indicating the change in value for a ‘‘baseline’’ property in different locations throughout the area examined.
Studies that seek to explain rental rate patterns and the determinant factors of industrial rent variations include those by Sivitanidou and Sivitanides (1995), Buttimer, Rutherford and Witten (1997), Jones and Orr (1999), Thompson and Tsolacos (1999) and Thompson and Tsolacos (2000). Sivitanidou and Sivitanides use alternative empirical models to examine the long-term rents of similar properties within competing industrial areas. Their study finds that several property-specific variables are relevant to the firm and other variables are relevant to workers; moreover, both have significant effects on the rental rates of industrial properties. Firms value access to raw materials and other markets, available services, freeway, intersection and major airport proximity as needed amenities whereas workers value education, low crime rates and proximity to shopping areas.
Presenting an empirical analysis of the determinants of pooled industrial warehouse rents, Buttimer, Rutherford and Witten (1997) note characteristics that positively influence rent per square foot as well as characteristics that are negatively related to rent per square foot. Using Dallas/Fort Worth data for 848 industrial properties from 1989:4 through 1992:4, they find positive and significant variables for the number of ground-level doors (a proxy for accessibility and building size) and the annual change in net employment. Ceiling height, percentage of office space, building age and a
296 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 3
Determinants of Rent and Income for Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Sivitanidou and Sivitanides (1995)
Industrial Rent Differentials: The Case of Greater Los Angeles
Environment and Planning, 1995, 27:7, 1133–46
1990 industrial (production space) rents in Greater Los Angeles
Explores the relationship between location and industrial real estate rental rates within decentralized metropolitan areas. Indicates that amenities for both the firm and for workers can have significant effects on the rental rates achieved by industrial properties. Emphasizes the importance of access to raw materials and other markets on the rents received for a property and notes other variables including: services available, freeway proximity, nearest major intersection, proximity to a major airport, education within the area, crime rates, and proximity to shopping areas. Concludes that all property-specific variables are statistically significant. Explains the pricing pattern as a result of firm and worker preferences for amenities.
Bible and Hsieh (1996)
Warehouse Buildings and Geographic Information Systems
The Appraisal Journal, 1996, 64:4, 416–22
No empirical data Suggests using Geographic Information System (GIS) for estimating occupancy rates, producing maps to depict where buildings are clustered and showing locations of buildings with different occupancy rates. Posits that property characteristics in general are not important in determining warehouse occupancy rates.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
297
Exhibit 3 (continued)
Determinants of Demand for Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Chai (1997) Determinants of NOI for Warehouse Properties
Real Estate Finance, 14: 2, 48–54
Chicago, Dallas, LA, 1979–1996
Suggests that although net operating income (NOI) is often determined by forecasting rents and vacancy rates, this may not be the best way of determining NOI for warehouse properties. Indicates that while industrial property data are available, they may not be segmented by warehouse, manufacturing and R&D facilities. Indicates that NOI is determined by variables that can change from one market to another, thus it would be misleading to use a single model across various markets.
Buttimer, Rutherford and Witten (1997)
Industrial Warehouse Rent Determinants in the Dallas /Fort Worth Area
Journal of Real Estate Research, 13:1, 47–55
848 buildings (about 60%) in the industrial warehouse market in Dallas /Fort Worth, 1989– 1993
Uses a multivariate random effects model to examine the determinants of market rents for industrial warehouse space. Finds that characteristics relating positively to rental price per square foot include number of grade high doors and the annual change in net employment, and that characteristics relating negatively to the rental PSF include ceiling height, percentage of office space, building age and existence of a sprinkler system. Reports also that rent rates are impacted by location variables and general market conditions. Suggests that owning smaller, standardized structures would best suit landlords.
298 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 3 (continued)
Determinants of Rent and Income for Industrial Real Estat
Author(s) Title Source Data, Location, Time Period Summary of Article
Reenstierna (1997)
Industrials: The Baseline Method
The Appraisal Journal, 65:3, 291–300
More than 500 one-story one-tenant industrial buildings in the Boston area from 1992–1995
Employs the baseline methodology to create a valuation model for single tenant industrial buildings. Method can be used in markets where location has a significant impact on the selling price of a property. Discusses using the baseline valuation model versus the traditional sales comparison method. Creates a GIS map that shows the change for a typical baseline building in different locations throughout the region studied.
Jones and Orr (1999)
Local Commercial and Industrial Renal Trends and Property Market Constraints
Urban Studies, 36:2, 223–37
Data are from local rental information since 1977 for a large number of urban centers in the U.K.
Evaluates the effect of market constraints on the long-term rental rate trends in the various sectors of the commercial and industrial real estate markets. Finds that the inelastic supply of retail units in town centers leads to the likelihood of long-term local retail rental premiums in growing centers; the large increase in demand for offices could create long-term local rental premiums; many of the influences on rents are the localized implications of long-term macroeconomic trends, especially the growth of services and the decline of manufacturing industries. Notes short-term or cyclical macroeconomic influences, which will influence output and sales via incomes. Concludes that office rents reflect local fixed effects, and that the effect of national economic variables is strong.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
299
Exhibit 3 (continued)
Determinants of Rent and Income for Industrial Real Estat
Author(s) Title Source Data, Location, Time Period Summary of Article
Mueller (1999) Real Estate Rental Growth Rates at Different Points in the Physical Market Cycle
Journal of Real Estate Research, 18:1, 131–50
Occupancy and rental growth rates, 1967–1997
Finds that national average rental growth rates vary significantly at different points in the physical market cycle (occupancy rates). Shows that local demand and supply are major determinants of rental growth rates.
Thompson and Tsolacos (1999)
Rent Adjustments and Forecasts in the Industrial Market
Journal of Real Estate Research, 17:1 /2, 151–67
Quarterly national industrial market of Great Britain taken quarterly from the of 1977 of 1997
Seeks to forecast industrial rent variations by utilizing the double-smoothing approach so that older data is given less weight in the model. Concludes that changes in GDP have an influence on industrial rental rates by virtue of increasing demand.
Thompson and Tsolacos (2000)
Projections in the Industrial Property Market Using a Simultaneous Equation System
Journal of Real Estate Research, 19:1 /2, 165–88
Data from the national industrial market in the U.K. from 1984–1998
Using a three-equation simultaneous system to model the industrial property market in U.K., suggests that real rents and construction costs in the industrial building supply are correlated. Suggests that simultaneous equations can be effective as an alternative method in analyzing industrial property markets, both at the local and aggregate market levels. Forecasts market growth and new industrial building supply from 1999-2003 providing a general scenario as well as a ‘Golden Scenario’ and a ‘Recession Scenario.’
300 JOURNAL OF REAL ESTATE LITERATURE
VOLUME 11, NUMBER 3, 2003
sprinkler system are inversely related to rent (these results seem counter intuitive, but may be data specific). Location variables and general market conditions also influence industrial property rent. Buttimer, et al., summarize their findings by supporting Ambrose’s (1990) results that property owners would be best served by owning smaller, standardized facilities given that the market appears to reward these types of properties with higher rent.
Chai (1997) confirms the finding that location and general market conditions influence income for industrial properties. This study cautions that since variations in rents and vacancy rates affect net operating income, a single model across various markets may not be the best way to estimate warehouse NOI. Variables that change from one market to another should be incorporated into forecasting. Local demand and supply, argues Mueller (1999), are major determinants of rental growth rates.
Thompson and Tsolacos (1999, 2000) suggest models for estimating industrial rents. Their earlier study uses rent indices to estimate changes in real rents, seeking to provide empirical evidence of determinant factors of industrial rent variations over time and to provide forecasts of future rents. Results indicate that changes in GDP have an influence on industrial rents due to increasing demand. Thompson and Tsolacos conclude in their 2000 study, also using data from the United Kingdom, that there are relationships for real rents and construction costs with industrial building supply. The belief that supply variables are more elastic for industrial properties than for retail spaces is addressed by Jones and Orr (1999) who confirm this hypothesis with empirical results from U.K. data.
The major findings from this section are:
� GIS technology may be useful along with conventional property variables in estimating warehouse occupancy.
� Variables relative to the firm that affect industrial rents are access to raw materials and other markets, available services, freeway, intersection and major airport proximity.
� Variables relevant to workers that affect industrial rents are education, crime rates and proximity to shopping.
� Variables that have a positive effect on industrial rents are the number of grade high doors and the change in net employment.
� Variables that have a negative effect on industrial rents are ceiling height, percentage of office space, building age and sprinkler systems.
� Supply variables are more elastic for industrial properties than for retail space.
� Changes in GDP have an effect on industrial rents due to increasing demand.
� Forecasting industrial property NOI needs to consider location and general market conditions that affect income.
� Local demand and supply are major determinants of rental growth rates.
THE ENVIRONMENT AND PERFORMANCE OF INDUSTRIAL REAL ESTATE 301
Return and Valuation Issues in Industrial Real Estate
Exhibit 4 provides a survey of the literature relating to return and valuation issues. Several of these studies look at the pros and cons of using traditional appraisal methods in valuing industrial properties.
In one of the first multivariate hedonic pricing studies for industrial real estate, Ambrose (1990) uses fifty-seven industrial property listings in the Atlanta, GA from 1986 to 1987. He finds that the market for industrial property is priced rationally based on property physical and locational characteristics. Between the market for industrial property listed for sale and property listed for lease, a clientele effect exists: industrial properties listed for sale are more than twice as large and have less finished office space than comparable properties listed for lease. Thus, participants in the industrial real estate market value property characteristics differently when they are purchasing from when they are leasing.
In an important series of papers using Dallas/Ft. Worth area industrial real estate data, Rutherford examines a number of issues associated with the pricing of industrial properties. Fehribach, Rutherford and Eakin (1993) estimate multivariate hedonic pricing models for industrial real estate sales price using physical, financial, locational and economic variables for each property. They show that a number of variables are significant in determining industrial real estate selling price. Ceiling height, office space, size of building, number of dock-high loading doors and single tenant occupant are positively related to sales price, while distance to airport and age are negatively related to sales price. These results differ slightly from Buttimer, Rutherford and Witten’s (1997) later hedonic rent per square foot study of the same market where they find negative relationships between rent and ceiling height and between rent and percentage of office space.
To examine the efficiency of the industrial real estate market, Atteberry and Rutherford (1993) develop a multivariate hedonic pricing model using 764 industrial property sales in the Dallas/Ft. Worth area from 1983 to 1991. They find a significant lag relationship between past and current industrial real estate prices and conclude that industrial real estate markets may be inefficient because current prices do not fully reflect past available information.
Local market factors, physical characteristics and location, according to Lockwood and Rutherford (1996), affect value. Using a factor-analytic linear structural relations model to test for the effects on industrial property value of various factors, they find that the log of sales price or value is related most strongly to local market factors, physical characteristics and location. National market interest rates, however, have little influence on the sales price of industrial properties.
After evaluating issues relevant to using the cost approach in appraising industrial properties, Treadwell (1988) suggests considering the actual costs. Such cost analysis would require distinguishing the cause and effect of such items as depreciation and functional obsolescence. While the cost approach may sometimes be the only tool
302 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 4
Return/Valuation Issues in Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Treadwell (1988)
Intricacies of the Cost Approach in the Appraisal of Major Industrial Properties
The Appraisal Journal, 56:1, 70–9
No empirical data Suggests an approach to accurately apply the cost approach in the appraisal of conventional industrial properties and limited-purpose properties. Considers questions: ‘‘What are we finding the cost of?’’ and ‘‘At what point in time are we calculating the cost for?’’ and suggests distinguishing between various causes and effects of depreciation and functional obsolescence. Concludes that the cost approach may be the only tool available to appraisers for limited-purpose industrial properties, and warns that amounts for depreciation may be far greater than those normally used.
Ambrose (1990)
An Analysis of the Factors Affecting Light Industrial Property Valuation
Journal of Real Estate Research, 5:3, 355–70
Fifty-seven industrial property listings in the Atlanta, 1986–1987
Provides a seminal study of industrial property pricing using hedonic methodology and asking prices. Asking prices are highly correlated with selling prices (R2 � .99). Finds the market for industrial property is priced rationally and property physical and locational characteristics determine most of its value. Further finds that there is a clientele effect between the market for industrial property listed for sale and property listed for lease: industrial properties listed for sale are more than twice as large and have less finished office area than comparable properties listed for lease. Concludes that participants in the industrial real estate market value property characteristics differently when they are purchasing from when they are leasing.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
303
Exhibit 4 (continued)
Return/Valuation Issues in Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Zimmer (1991) Avoiding Traps When Using Sales Comparison to Value Storage and Distribution Facilities
The Appraisal Journal, 59:3, 390–94
Six comparable sales in 1990
Evaluates the appropriateness of using price per square foot when also using the sales comparison approach. Suggests this can lead to the over- or under-valuing of property especially when there are varied land-to- building ratios. Presents an alternative method for valuation to the sales comparison approach. Concludes that components of sales value should be isolated because properties have varying characteristics and land values, and thus their overall prices vary.
Atteberry and Rutherford (1993)
Industrial Real Estate Prices and Market Efficiency
Journal of Real Estate Research, 8:3, 377–85
764 industrial property sales in the Dallas /Ft. Worth area, 1983–1991
Focuses on the efficiency of the industrial real estate market. Estimates multivariate hedonic pricing models for industrial real estate sales price per square foot. Finds a significant lag relationship between past and current industrial real estate prices. Concludes that the industrial real estate market may be inefficient because current prices do not fully reflect past available information.
Fehribach, Rutherford, and Eakin (1993)
An Analysis of the Determinants of Industrial Property Valuation
Journal of Real Estate Research, 8:3, 365–76
170 industrial property sales in the Dallas /Ft. Worth area, 1987–1991
Estimates multivariate hedonic pricing models for industrial real estate sales price using physical, financial, locational and economic variables for each property. Finds a number of variables are significant in determining industrial real estate selling price: ceiling height (�), office space (�), size of building (�), number of dock-high doors (�), distance to airport (�), age (�) and type of tenant (single (�) vs. multiple (�)).
304 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 4 (continued)
Return/Valuation Issues in Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Alvayay, Rutherford and Smith (1995)
Tax Rules and the Sale and Leaseback of Corporate Real Estate
Real Estate Economics, 23:2, 207–38
No empirical data, but examining the effect of sale and leaseback of corporate real estate on stock price, 1982–1989
Develops a theoretical model to examine the effects of tax law changes on sale and leaseback transactions. Explains tax effects and potential savings of sale and leaseback agreements. Examines the impact of the Tax Reform Act of 1986 on sale and leaseback transactions. Argues that sale and leaseback transactions positively impact the equity value of the firm.
Lockwood and Rutherford (1996)
Determinants of Industrial Property Value
Real Estate Economics, 24:2, 257–72
Sales of 308 industrial properties in the Dallas / Fort Worth area from 1987–1991
Employs a factor-analytic linear structural relations (LISREL) model to test for the effects on industrial property value of various factors including physical property, national market, local market, interest rate and location. LISREL lessens problems with multicollinearity. Finds that the log of sales price or value is related most strongly to physical characteristics and location factors. Concludes that national market factors and interest rates have little influence on the industrial sale prices.
Myer, Chaudhry and Webb (1997)
Stationarity and Co- Integration in Systems with Three National Real Estate Indices
Journal of Real Estate Research, 13:3, 369–81
Industrial, office, retail and aggregate properties for the U.S., Canada and the U.K, 1987–1992
Reports that commercial real estate returns show evidence of variables with drift and time trend (nonstationarity) components. Notes that apparently the long-run forecasts of U.S., Canadian and U.K. real estate indices (with the exception of industrial properties in nominal terms) do not diverge significantly. Concludes that the exchange rate-adjusted and real series are co-integrated, thus a common link may be explained by inflationary expectations.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
305
Exhibit 4 (continued)
Return/Valuation Issues in Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Ling (1998) The Fundamental Determinants of Commercial Real Estate Returns
Real Estate Finance, 14: 4, 13–24
Quarterly data from 1978–1994
Shows how to quantify the risk premium that investors receive in commercial real estate returns. Concludes that interest rate related variables have the most significant amount of risk. Finds that the risk associated with consumption is significant. Finds that unobservable variables and investor preferences are important determinants in real estate returns.
McGrath (2000) Urban Industrial Land Redevelopment and Contamination Risk
Journal of Urban Economics, 47:3, 414–42.
Chicago industrial land development and contamination risk and demolition costs
Concludes that contamination liability may be considered a land demolition cost and capitalized into the bid price of industrial properties, thus providing a way to incorporate contamination risk into land values.
Young (2000) REIT Property-Type Sector Integration
Journal of Real Estate Research, 19:1 /2, 3–21
Monthly total returns and capitalization rates of apartment, industrial, office and retail REITs 1989–1998
Reports that various property types in equity REITs are becoming more integrated. Examines pairs of equity REITs containing mostly apartment, industrial, office and retail properties for correlations over rolling sixty- month returns. Concludes that property- specific REITs show a similar trend over rolling sixty-month return correlations.
Thompson and Tsolacos (2001)
Industrial Land Values-A Guide to Future Markets?
The Journal of Real Estate Research, 21:1 /2, 55–76
Quarterly data from six markets in the U.K., 1987–1999
Provides empirical evidence from the U.K. that industrial property prices are linked to other economic occurrences. Presents several methods that can be used to measure land value. Concludes that industrial land values have an effect on industrial property prices, but the effect is not always constant.
306 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 4 (continued)
Return/Valuation Issues in Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Derbes (2002) Non-Comparable Industrial Sales
The Appraisal Journal, 70:1, 39–45
No empirical data Posits that the lack of bona fide comparable market transactions prevents the sales comparison approach from being a viable approach to appraising special- and single- purpose properties. Argues that, to accurately use the sales comparison approach in industrial properties, the sale properties should have the same highest and best use.
Ellsworth (2002)
Industrial Facility Valuation: Electric Generating Projects
The Appraisal Journal, 70:1, 34–8
No empirical data Suggests that valuations of industrial facilities such as electric generating facilities should compare more than price per kilowatt of generating capacity. Competitive and economic environment should be examined.
He (2002) Excess Returns of Industrial Stocks and the Real Estate Factor
Southern Economic Journal, 68:3, 632–45
Stock Market returns, U.S. and corporate bonds, treasury bills, new house price index and S&P Industrial Stock Index, 1963–1997
Examines major risk factors in pricing industrial stocks, and estimates various explanatory models. Shows that the real estate factor plays a significant role in explaining excess returns on industrial stocks. Concludes that when examining subperiods, results indicate that the real estate effects are stable and second only to the overall stock market effects.
THE ENVIRONMENT AND PERFORMANCE OF INDUSTRIAL REAL ESTATE 307
available for appraisers, its calculations may often use amounts for depreciation that are in excess of the amounts normally included.
Additionally, Zimmer (1991) warns of traps when using the sales comparison approach to value certain industrial properties. This study suggests that components of sales value should be isolated because properties have varying characteristics and land values, and that their respective prices vary accordingly. The lack of true comparable market transactions, Derbes (2002) notes, prevents the sales comparison approach from being a viable technique to valuing special and single-purpose industrial properties.
Alvayay, Rutherford and Smith (1995) focus on the impact of the Tax Reform Act of 1986 on sale and leaseback transactions. They argue that sale and leaseback transactions positively impact the equity value of the firm.
Thompson and Tsolacos (2001) later show that industrial land prices have an effect on local rental rates and that appraisers may use land values as an indicator of future rental rates. Valuing specific industrial properties containing electric generating facilities is addressed by Ellsworth (2002) who argues that valuations of these special properties should address factors related to the competitive and economic environment within which the property operates rather than just addressing price per kilowatt of generating capacity.
There is a significant co-variation between the overall stock market and the real estate market. Young (2000) finds significant integration of returns of various types of equity REITs. In examining excess returns of industrial stocks, He (2002) finds that the real estate factor plays an important role in explaining excess returns of industrial stocks.
Ling (1998) examines non-diversifiable or systematic risk through a multifactor asset pricing model and seeks to identify the most important macroeconomic drivers that affect real estate returns. Results suggest that the largest risk premiums in commercial real estate returns are associated with interest rate-related variables and with consumption. Myer, Chaudhry and Webb (1997) also use macroeconomic factors to examine the long-run forecasts of U.S., U.K. and Canadian industrial real estate data. They find that industrial property variables have drift and time trend components. McGrath (2000) focuses on contamination risk and finds that contamination liability may be considered a land demolition cost and factored into the bid price of industrial properties, thus directly affecting land values.
The major findings concerning the return and valuation of industrial real estate are:
� There is a clientele effect in light industrial properties so that properties for sale tend to be larger and to have less office space than properties for lease.
� A number of variables are significant in determining industrial real estate selling price including ceiling height, office space, size of building,
308 JOURNAL OF REAL ESTATE LITERATURE
VOLUME 11, NUMBER 3, 2003
loading door height, distance to airport and type of tenant (single vs. multiple).
� Local market factors, physical characteristics and location affect industrial property value.
� In using the cost approach in appraisal, accurate cost analysis requires distinguishing cause and effect of items such as depreciation and functional obsolescence.
� When using the sales comparison approach, components of sales value should be isolated because properties have varying characteristics and different land values.
� Valuation of special purpose industrial properties should address factors related to the competitive and economic environment.
� The largest risk premiums in commercial real estate returns are associated with interest rate-related variables and with consumption.
� Industrial property returns appear to have nonstationarity.
� The industrial real estate market may be inefficient with regards to incorporating past information.
Environmental Concerns in Industrial Real Estate?
Articles that have looked at the impact of environmental issues on industrial real estate (reviewed in Exhibit 5) include studies of landfills, past contamination, clean-up risk and regulation. In an examination of 153 industrial land transactions in Phoenix, Arizona, for the years 1984 to 1994, Guntermann (1995) focuses on nineteen properties located near fourteen landfills. He finds that industrial land prices around open, solid waste landfills are reduced by the presence of the landfill and its inherent stigma, but for a closed solid waste landfill or a refuse landfill, there is no reduction in the sales price. Conklin (1996) provides a checklist to help property owners identify and eliminate illegal dumping. Piatkowski (1997) suggests methods of evaluating, negotiating and alleviating environmental problems through environmental Risk Assessment Surveys (ERAS).
In several studies, Jackson (2001a, 2001b, 2001c, and 2002) presents empirical and survey evidence on how prices of industrial properties are affected by the presence of past contamination. Using paired sales and hedonic regression analyses, Jackson (2001a) makes a comparison of improved industrial sites that were once contaminated to similar noncontaminated sites to determine the effect of contamination on price. He concludes that the sales prices of previously contaminated properties are not affected by the prior presence of environmental problems.
In his literature review, Jackson (2001b) centers on the effects of environmental contamination on real estate prices and values, but concludes that most of the studies show an adverse impact on the sales price and on the value of non-remediated industrial and commercial real estate. The effects include reduced sale prices, reduced transaction rates and increased incidence of seller financing. No-further-action letters
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
309 Exhibit 5
Environmental Concerns in Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Guntermann (1995)
Sanitary Landfills, Stigma, and Industrial Land Values
Journal of Real Estate Research, 10:5, 531–42
Data on 19 sales around 12 landfills and 153 total industrial property transactions in Phoenix, Arizona, 1984–1994
Focuses on open and closed solid waste landfills that contain various non-hazardous materials and examines the effect on the surrounding land prices. Shows that industrial land values around open, solid waste landfills are reduced by the presence of the landfill, but for a closed solid waste landfill or a refuse landfill, there is no reduction in sales price.
Boyd, Harrington and Macauley (1996)
The Effects of Environmental Liability on Industrial Real Estate Development
Journal of Real Estate Finance and Economics, 12:1, 37–58
No empirical data Evaluates the effects of current liability regulations with respect to property transactions involving environmental contamination concerns. Suggests that market inefficiencies are more related to asymmetrical information and insufficient detection methods than to legal uncertainty. Proposes that pollution in itself does not make property transactions inefficient because the price asked for polluted, or formerly polluted property, should be discounted to account for potential cleanup costs.
Conklin (1996) Operation Dump-And- Run
Journal of Property Management, 61:1, 30–3
No empirical data Provides a checklist for better identifying and eliminating illegal dumping which is becoming a great problem for industrial property owners.
Piatkowski (1997)
Beyond Compliance: Environmental Risk Assessments at Industrial Properties
Journal of Property Management, 62:2, 72–5
No empirical data Highlights common environmental exposures of industrial properties, discusses various responsible parties including tenants and neighboring property owners, and suggests methods of negotiating and alleviating environmental problems through Environmental Risk Assessment Surveys (ERAS).
310 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 5 (continued)
Environmental Concerns in Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Powell (1997) Preemption of Local Zoning Ordinances Under Federal hazardous Waste Laws
Real Estate Law Journal, 26:1, 101–05
No empirical data Examines the conflict between state and federal environmental law and outlines the circumstances where federal law will preempt state action.
Jackson (2001a)
The Effect of Previous Environmental Contamination of Industrial Real Estate Prices
The Appraisal Journal, 69:2, 200–10
Data for the paired sales analysis include 8 properties from southern California area, while data for the hedonic analysis contains 13 contaminated properties and 122 total sales from Los Angeles, San Diego, and Orange Counties during 1998 and 1999
Examines by paired sales and hedonic pricing analyses how the prices of industrial properties are affected by the presence of past contamination problems. Concludes that properties and prices are not affected by the prior presence of the environmental conditions.
Jackson (2001b)
The Effects of Environmental Contamination on Real Estate: A Literature Review
Journal of Real Estate Literature, 9:2, 91–116
Provides a literature review on the effects of environmental contamination on real estate including industrial properties
Shows that all the studies on the effects of contamination of commercial and industrial real estate sale prices and values find significant adverse impact. Includes reduced sale prices, reduced transaction rates and increased dependence on seller financing as effects of contamination. Finds that no-further- action letters from regulators do not increase property marketability. Finds that prices of properties near closed landfills are not affected while properties near active landfills are adversely affected. Finds that groundwater contamination adversely affects industrial properties.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
311
Exhibit 5 (continued)
Environmental Concerns in Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Jackson (2001c)
Environmental Risk Perceptions of Commercial and Industrial Real Estate Lenders
Journal of Real Estate Research, 22:3, 271–88
Survey responses from 238 out of 453 (52.5%) national industrial and commercial mortgage lenders during late 1999
Examines the risk perceptions of mortgage lenders concerning environmentally contaminated industrial and other commercial properties. Finds that perceived lending risk changes before, during and after cleanup. Reports that over 90% of mortgage lenders will not make a loan before cleanup and without an approved remediation plan, but over 60% would lend at normal market rates after cleanup.
Jackson (2002) Environmental Contamination and Industrial Real Estate Prices
Journal of Real Estate, 23:1 /2, 180–99
Data on 140 industrial property sales from Southern California from 1995 to 1999
Estimates using hedonic pricing analysis the impact of environmental contamination on industrial real estate sales prices. Finds statistically significant impacts on industrial real estate sales prices before and during remediation, but not afterward.
312 JOURNAL OF REAL ESTATE LITERATURE
VOLUME 11, NUMBER 3, 2003
from regulatory authorities do not increase property marketability. Groundwater contamination and location near active landfills adversely affect industrial property values, but properties located near closed landfills are not affected.
Both loan risk and real estate prices according to the time schedule for contamination cleanup. Using survey responses from 238 out of 453 (52.5%) national industrial and commercial mortgage lenders during late 1999, Jackson (2001c) finds perceived lending risk changes before, during and after cleanup. He reports that over 90% of mortgage lenders will not make a loan before cleanup and without an approved remediation plan, but over 60% would lend at normal market rates after cleanup. In recently published study, Jackson (2002) uses 140 Southern Californian industrial property sales from 1995 to 1999 in a hedonic pricing model to estimate the impact of environmental contamination on industrial real estate sale prices. The author finds statistically significant negative impacts on industrial real estate sales prices before and during remediation, but not afterwards.
Some of research on environmental concerns addresses regulation and compliance. Boyd, Harrington and Macauley (1996) examine the effects and implications of the 1980 Comprehensive Environmental Response, Compensation and Liability Act (CERCLA) and discuss the effects of asymmetrical information on property transactions. The authors suggest that market inefficiencies are more related to asymmetrical information and insufficient detection methods than to legal uncertainty. They suggest potential legal reforms that may make these property transactions more efficient. Powell (1997) discusses how state and federal environmental laws often conflict and outlines scenarios in which federal law preempts state regulation.
The major findings about environmental concerns are:
� Landfills sell for much less than other industrial sites.
� Property prices are not affected by past contamination for closed landfills, but unremediated contamination has an adverse impact on industrial property values by reducing sale prices and transaction rates, and by increasing the incidence of seller financing.
� Property prices near closed landfills are not adversely affected while prices of properties near active landfills are negatively affected.
� Groundwater contamination adversely affects industrial property values.
� Market inefficiencies in industrial property valuation are more related to asymmetrical information rather than to legal uncertainty.
� State and federal environmental laws often conflict, and federal law often preempts state regulation.
The International Perspective of Industrial Real Estate
Much of the literature in international industrial real estate research relates to issues involving foreign investor choice, industrial property rights and the influence of a
THE ENVIRONMENT AND PERFORMANCE OF INDUSTRIAL REAL ESTATE 313
country’s laws regarding industrial development or clustering. Exhibit 6 provides an overview of the studies addressing international industrial real estate issues.
International legislative and regulatory issues are addressed in several studies. Bumbak (1984) argues that property rights, such as patents and trademarks, need to be more clearly defined and regulated by the European Court. Larger issues such as Non-tariff Barriers (NTBs) are examined in some U.S. manufacturing industries in a study by Busch and Reinhardt (1999). They suggest that geographically concentrated but politically dispersed industries are most likely to receive relief from imports, but that large industries benefit from being politically concentrated. Further, Ietto-Gillies (2000) proposes that countries be defined by the regulatory regimes to which they adhere. Also, this study explores the role of multinational firms in the location of economic activity and in the geographical concentration of productivity. Comparing the location of companies under several political regimes, Martin and Rogers (1995) determine that tenants will likely seek a location with the best possible infrastructure, highest per capital income and the largest market. They conclude that governments can use infrastructure to attract industries.
Many other divergent topics in international industrial real estate exist. Examining the wealth effect of international versus domestic real estate joint ventures, He, Myer and Webb (1997) find that international real estate joint ventures have smaller wealth effects compared to domestic ventures. Ford, Fung and Gerlowski (1998) look at factors that influence foreign investor choice in types of U.S. real estate and find that foreigners investing in industrial real estate are most influenced by changes in capitalization rates, market activity and current rent levels. In an earlier study of foreign investments in the U.S. (mostly Japanese), Porter (1986) predicted that federal aid and the economic base of some industrial cities would decline with increasing foreign competition, specifically in the steel and automobile manufacturing sectors.
Several studies deal with city size and industrial diversification. Begovic (1992), using data from Yugoslavia, examines the relationship between industrial diversification and city size. He concludes that there is a significant relationship between these two factors in Yugoslavia but there is no significant relationship between industrial share and city size. He argues that this is partially due to transportation and tourism development being based on spatial distribution rather than city size. Porter (2000) notes the role of clustering, the geographic concentration of related firms in a country’s economy, as an influence on an industry’s productivity, innovation, new business formation, competition, government support and overall economic policy. Waits (2000) reviews the influence of clusters and finds that clusters of global firms in similar industries are vital to the economic development of a global economy. He further notes that clusters of industries rather than individual firms or industries are a significant source of jobs, income and export growth. In examining a specific industry, Imrie (1989) provides a case scenario of industrial restructuring in the British pottery industry, focusing on the employment and geographic impact of economic change in this sector.
314 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 6
International Perspective of Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Bumbak (1984) Industrial Property Rights and the Free Movement of Goods in the European Communities
Case Western Reserve Journal of International Law, 16:3, 381– 436
No Empirical Data Examines the European Court’s interpretation of article 36, which relates to most all forms of industrial property rights, including patents, trademarks, copyrights, performance rights and industrial designs. Argues that given the uneven nature of decisional law and the pace of technological change, there is a need for European Community wide legislation regulating intellectual property.
Porter (1986) America’s Industrial Cities: Jolted but Resilient
Annals, AAPSS, 488, 77–84
No Empirical Data Predicts that federal aid and the economic base of many industrial cities, especially the steel and automobile manufacturing, will decline with the increasing foreign competition. Suggests that because their human and physical resources are greater than their capacity, industrial cities have an opportunity to become stronger than they were prior to increased foreign competition.
Imrie (1989) Industrial Restructuring, Labour, and Locality: The Case of the British Pottery Industry
Environment and Planning, 21, 3–26
British Pottery Industry, 1975–1986
Provides a case study of industrial restructuring in the pottery industry. Examines the background to crisis and restructuring in this industry as well as the employment and geographical effects of economic change in this industry.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
315 Exhibit 6 (continued)
International Perspective of Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Begovic (1992) Industrial Diversification and City Size: The Case of Yugoslavia
Urban Studies, 29:1, 77– 88
Census and employment data for sixty-eight Yugoslavia cities, 1981
Examines the industrial diversification-city size relationship. Shows results supporting the hypothesis (agglomeration economics theory) that a statistically significant relationship exists between city size and industrial diversification or industrial share. Finds a significant relationship between industrial diversification and size of cities in Yugoslavia. Finds no statistically significant relationship between industrial share and city size, possibly due to the fact that the development of the transportation and tourism and leisure industries is based on factors whose spatial distribution is unrelated to city size. Concludes that city size variation may explain only a small part of industrial diversification /share.
Martin and Rogers (1995)
Industrial Location and Public Infrastructure
Journal of International Economics, 39:3–4, 335– 51
No empirical data Compares the locations of companies under several political regimes and notes that some countries impose rules that restrict or forbid plant closures. Determines that the incentives to inhibit firms’ relocation will be more prevalent in areas with strong economies of scale and where the infrastructures of the current and target countries are advanced. Suggests that tenants will likely seek a location that is in a market with the best possible infrastructure, highest per capital income and largest market. Provides a model to predict the effect an infrastructure change will have on the location tendencies of industrial tenants. Concludes that with the model presented, governments can evaluate policies and use public infrastructure to attract industries from other areas.
316 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 6 (continued)
International Perspective of Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
He, Myer and Webb (1997)
The Wealth Effects of Domestic vs. International Joint Ventures: The Case of Real Estate
Journal of Real Estate Research, 13:3, 349–58
Real estate joint ventures, with hotels 1975–1990
Examines the wealth effect of international versus domestic real estate joint ventures. Finds domestic real estate joint ventures incur a significant increase in value but international real estate joint ventures have a much smaller wealth effect. Finds hotel joint ventures to have a weaker wealth effect than non-hotel real estate joint ventures.
Ford, Fung and Gerlowski (1998)
Factors Affecting Foreign Investor Choice in Types of U.S. Real Estate
Journal of Real Estate Research, 1998, 16:1, 99– 111
Apartment, office, retail and industrial investment transactions from 1980–1991
Reports that foreign investor choice of industrial real estate (as well as apartment, office and retail properties) is most sensitive to changes in capitalization rates, market activity and current rent levels.
Busch and Reinhardt (1999)
Industrial Location and Protection: The Political and Economic Geography of U.S. Non- tariff Barriers
American Journal of Political Science, 43:4, 1028–50
Non-tariff Barriers (NTBs) in 363 U.S. Manufacturing Industries, 1990
Considers correction for inherent problems with endogenous protection research, such as erroneously using geographic concentration as a proxy for political concentration, and finds that geographically concentrated but politically dispersed industries are most likely to receive relief from imports, but very large industries benefit from being politically concentrated. Reports results that show a way to reconcile differing hypotheses via endogenous protection theory.
T H
E E
N V
IR O
N M
E N
T A
N D
PE R
FO R
M A
N C
E O
F IN
D U
ST R
IA L
R E
A L
E STA
T E
317
Exhibit 6 (continued)
International Perspective of Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Ietto-Gillies (2000)
What Role for Multinational in the New Theories of International Trade and Location?
International Review of Applied Economics, 14:4, 413–26
Various data sources from the 1990s
Explores multinational corporations’ role in the location of economic activity and in the geographical concentration of production. Proposes that nations can be defined by the regulatory regimes they encompass and that more qualitative features should be considered in theories of multinational production and location. States that the implications of operating between various regulatory regimes will affect the pattern of location as well as concentration of activity.
Porter (2000) Location, Competition, and Economic Development: Local Clusters in a Global Economy
Economic Development Quarterly, 14:1, 15–34
No empirical data Discusses the role of clustering (geographic concentrations of interconnected companies) in metropolitan, state, regional and national economies. Notes how clustering influences an industry’s productivity, innovation, new business formation, competition, government support and overall economic policy.
Waits (2000) The Added Value of the Industry Cluster Approach to Economic Analysis, Strategy Development, and Service Delivery
Economic Development Quarterly, 14:1, 35–50
No empirical data Case study
Reviews the influence of cluster-focused economic analysis and strategy development. Finds the clusters of world-class firms in related industries are the most important economic development customers in the global economy. Suggests that clusters of industries rather than individual firms or industries, are the source of jobs, income and export growth.
318 JO
U R
N A
L O
F R
E A
L E
STA T
E L
IT E
R A
T U
R E
V O
L U
M E
11, N
U M
B E
R 3,
2003
Exhibit 7
Management and Financing Concerns in Industrial Real Estate
Author(s) Title Source Data, Location, Time Period Summary of Article
Simons (1994) Industrial Real Estate Mortgage Default Experience of the New York State Job Development Authority Second Loan Program: A Preliminary Investigation
Journal of the American Real Estate and Urban Economics Association, 22:4, 631–46
New York State Job Development Authority data 1962–1988
Explores the loan default history experienced by an industrial lending company. Concludes that firm bankruptcy and negative net equity have a strong correlation to loan defaults among the properties studied. Concludes that restructuring within companies is typically unsuccessful as the median survival time after such activity is about one year.
Klein (1999) Monitoring Industrial Tenants
Journal of Property Management, 64:4, 76– 80
No empirical data Addresses issues relating to the ongoing monitoring and management of industrial properties. Argues that the importance of this research arises from the financial liability that management can face with regard to code and ordinance violations by their tenants. Suggests that more time is needed to evaluate the safety of hazardous waste management with increased environmental concerns from the government and community. Gives various measures management can take to become a more effective and safe property manager.
THE ENVIRONMENT AND PERFORMANCE OF INDUSTRIAL REAL ESTATE 319
The major findings for industrial real estate from an international perspective are:
� Property rights such as industrial designs need to be more clearly defined and regulated in Europe.
� Geographically concentrated but politically dispersed industries are most likely to receive relief from imports, but large industries benefit from being politically concentrated.
� Governments can use infrastructure to attract industries and tenants will likely seek a location with the best possible infrastructure, highest per capita income and the largest market.
� International real estate joint ventures have a smaller wealth effect than domestic real estate joint ventures.
� Foreigners investing in industrial real estate are most influenced by changes in capitalization rates, market activity and current rent levels.
� There is a significant relationship between diversification and city size in Yugoslavia but there is not significant relationship between industrial share and city size.
� Geographical clustering influences an industry’s productivity, innovation, new business formation, competition, government support and overall economic policy.
Management and Financing Concerns in Industrial Real Estate A final category of the literature examines the management and financing concerns in industrial real estate. This research is presented in Exhibit 7. Examining borrower net equity in industrial real estate mortgage default in New York, Simons (1994) finds a strong negative relationship between net equity and default, with the greatest probability of default being when net equity is negative. Examining general industrial property management issues and the financial liability that management can face if code and ordinance rules are violated, Klein (1999) notes several measures that management can take to become more effective and safe property managers.
The major findings from this section are:
� In industrial real estate mortgage default, there is a negative relationship between net equity and default with default probability being greatest when net equity is negative.
� Management should be careful to minimize the financial liability that it can face with regard to code and ordinance violations by tenants.
Summary and Conclusions Industrial real estate research is receiving increased academic attention. The goal of this study has been to survey the empirical, theoretical and descriptive literatures concerning industrial real estate by examining over seventy recent studies and
320 JOURNAL OF REAL ESTATE LITERATURE
VOLUME 11, NUMBER 3, 2003
addressing seven questions: (1) What are the characteristics of the industrial real estate market (i.e., property characteristics)? (2) What are the determinants of demand for industrial real estate (i.e., demand determinants)? (3) What are the determinants of rent and income for industrial real estate (i.e., rent and income determinants)? (4) What are the return and valuation issues in industrial real estate (i.e., return and valuation issues)? (5) What are the environmental concerns in industrial real estate (i.e., environmental issues)? (6) What is the international perspective of industrial real estate (i.e., international issues)? and (7) What are the management and financing concerns in industrial real estate (i.e., management and financing issues)?
Some major findings are: (1) industrial properties tend to be segmented (e.g., manufacturing and distribution) and clustered; (2) design and capacity are important and properties must be flexible; (3) demand tends to be greatest for smaller properties (25,000 square feet or less); (4) enhanced technology has reduced the demand for warehouse space; (5) industrial firms are concerned with access to raw materials, freeway access and airport access; (6) some factors that affect industrial rents include changes in employment, age and percentage of office space; (7) industrial land values may serve as a good indicator of future rental rates; (8) local market factors, physical characteristics and location affect industrial property values; (9) contamination has an adverse effect on property values and lenders’ willingness to finance; (10) foreigners investing in industrial real estate are concerned with capitalization rates, current rents and market activity; and (11) there is a negative relationship between net equity and default in industrial mortgage loan default.
References
Alvayay, J. R., R. C. Rutherford and W. S. Smith, Tax Rules and the Sale and Leaseback of Corporate Real Estate, Real Estate Economics, 1995, 23:2, 207–38. Ambrose, B. W., An Analysis of the Factors Affecting Light Industrial Property Valuation, Journal of Real Estate Research, 1990, 5:3, 355–70. Atteberry, W. L. and R. C. Rutherford, Industrial Real Estate Prices and Market Efficiency, Journal of Real Estate Research, 1993, 8:3, 377–85. Begovic, B., Industrial Diversification and City Size: The Case of Yugoslavia, Urban Studies, 1992, 29:1, 77–88. Bible, D. S. and C-H. Hsieh, Warehouse Buildings and Geographic Information Systems, The Appraisal Journal, 1996, 64:4, 416–22. Black, R. T., M. L. Wolverton, J. T. Warden and R. H. Pittman, Manufacturing versus Distribution: Implicit Pricing of Real Property Characteristics by Submarket, Journal of Real Estate Finance and Economics, 1997, 15:3, 271–85. Boyd, J., W. Harrington and M. K. Macauley, The Effects of Environmental Liability on Industrial Real Estate Development, Journal of Real Estate Finance and Economics, 1996, 12: 1, 37–58. Bruce, R., Industrial Goes Upscale (Industrial Property), Journal of Property Management, 1994, 59:2, 14–7. ——., Recycling Your Industrial Properties, Journal of Property Management, 1995, 60:4, 44– 7. Bumbak, D. R., Industrial Property Rights and the Free Movement of Goods in the European Communities, Case Western Reserve Journal of International Law, 1984, 16:3, 381–436.
THE ENVIRONMENT AND PERFORMANCE OF INDUSTRIAL REAL ESTATE 321
Busch, M. L. and E. Reinhardt, Industrial Location and Protection: The Political and Economic Geography of U.S. Non-tariff Barriers, American Journal of Political Science, 1999, 43:4, 1028– 50. Buttimer, R. J., Jr., R. C. Rutherford and R. Witten, Industrial Warehouse Rent Determinants in the Dallas /Fort Worth Area, Journal of Real Estate Research, 1997, 13:1, 47–55. Chai, Y. W., Determinants of NOI for Warehouse Properties, Real Estate Finance, 1997, 14:2, 48–54. Christensen, M. F., B. Wisener and D. J. Campos, Attributes of Tomorrow’s Warehouse Structures, Real Estate Review, 1997, 27:3, 51–7. Cleminshaw, J. G., Measuring Superadequacy in Older Industrial Plants: A Case Study, Assessment Journal, 1997, 4:1, 37–40. Conklin, K., Operation: Dump-and-Run, Journal of Property Management, 1996, 61:1, 30–3. Crawford, R. G. and B. A. Slade, Appraising Industrial Special-Purpose Properties, The Appraisal Journal, 2001, 69:2, 161–73. Derbes, M. J., Jr., Non-Comparable Industrial Sales, The Appraisal Journal, 2002, 70:1, 39– 45. Ellison, G. and E. L. Glaeser, The Geographic Concentration of Industry: Does Natural Advantage Explain Agglomeration?, American Economic Review, 1999, 89:2, 311–16. Ellsworth, R. K., Industrial Facility Valuation: Electric Generating Projects, The Appraisal Journal, 2002, 70:1, 34–8. Fehribach, F. A., R. A. Rutherford and M. A. Eakin, An Analysis of the Determinants of Industrial Property Valuation, Journal of Real Estate Research, 1993, 8:3, 365–76. Ford, D. A., H-G. Fung and D. A. Gerlowski, Factors Affecting Foreign Investor Choice in Types of U.S. Real Estate, Journal of Real Estate Research, 1998, 16:1, 99–111. Franklin, H., Industrial Inspirations, (Industrial Properties), Journal of Property Management, 1994, 59:3, 18–20. Graham, M. F. and D. S. Bible, Classifications for Commercial Real Estate, The Appraisal Journal, 1992, 60:2, 237–46. Grissom, T. V., D. J. Hartzell and C. H. Liu, An Approach to Industrial Real Estate Market Segmentation and Valuation Using the Arbitrage Pricing Paradigm, Journal of the American Real Estate and Urban Economics Association, 1987, 15:3, 199–219. Guntermann, K. L., Sanitary Landfills, Stigma and Industrial Land Values, Journal of Real Estate Research, 1995, 10:5, 531–42. Hanson, G. H., North American Economic Integration and Industrial Location, (Impact of North American Free Trade Agreement), Oxford Review of Economic Policy, 1998, 14:2, 30–44. Harrison, B., Industrial Districts: Old Wine in New Bottles?, Regional Studies, 1992, 26:5, 469– 83. Hartman, D. J., Industrial Real Estate: Go Figure!, Real Estate Issues, 1991, 16:1, 23–7. He, L. T., Excess Returns of Industrial Stocks and the Real Estate Factor, Southern Economic Journal, 2002, 68:3, 632–45. He, L. T., F. C. N. Myer and J. R. Webb, The Wealth Effects of Domestic vs. International Joint Ventures: The Case of Real Estate, Journal of Real Estate Research, 1997, 13:3, 349–58. Hill, E. W. and J. F. Brennan, A Methodology for Identifying the Drivers of Industrial Clusters: The Foundation of Regional Competitive Advantage, Economic Development Quarterly, 2000, 14:1, 65–96. Hughes, W. T., Jr., Determinants of Demand of Industrial Property, The Appraisal Journal, 1994, 62:2, 303–09. Ietto-Gillies, G., What Role for Multinational in the New Theories of International Trade and Location?, International Review of Applied Economics, 2000, 14:4, 413–26.
322 JOURNAL OF REAL ESTATE LITERATURE
VOLUME 11, NUMBER 3, 2003
Imrie, R. F., Industrial Restructuring, Labour, and Locality: The Case of the British Pottery Industry, Environment and Planning, 1989, 21:1, 3–26. Jackson, T. O., The Effect of Previous Environmental Contamination on Industrial Real Estate Prices, The Appraisal Journal, 2001a, 69:2, 200–10. ——., The Effects of Environmental Contamination on Real Estate: A Literature Review, Journal of Real Estate Literature, 2001b, 9:2, 91–116. ——., Environmental Risk Perceptions of Commercial and Industrial Real Estate Lenders, Journal of Real Estate Research, 2001c, 22:3, 271–88. ——., Environmental Contamination and Industrial Real Estate Prices, Journal of Real Estate, 2002, 23:1/2, 180–99. Jones, C. and A. Orr, Local Commercial and Industrial Rental Trends and Property Market Constraints, Urban Studies, 1999, 36:2, 223–37. Klein, J., Monitoring Industrial Tenants, Journal of Property Management, 1999, 64:4, 76–80. Ling, D. D., The Fundamental Determinants of Commercial Real Estate Returns, Real Estate Finance, 1998, 14:4, 13-24. Lockwood, L. J. and R. C. Rutherford, Determinants of Industrial Property Value, Real Estate Economics, 1996, 24:2, 257–72. Martin, P. and C. A. R., Industrial Location and Public Infrastructure, Journal of International Economics, 1995, 39:3–4, 335–51. Mayer, H. M., Patterns and Recent Trends of Chicago’s Outlying Business Centers, The Journal of Land and Public Utility Economics, 1942, 18, 4–16. McGrath, D. T., Urban Industrial Land Redevelopment and Contamination Risk, Journal of Urban Economics, 2000, 47:3, 414–42. Mitchell W. N. and M. J. Jucius, Industrial Districts of the Chicago Region and Their Influence on Plant Location, The Journal of Business of the University of Chicago, 1933, 6:2, 139–56. Moriarty, B. M., Urban Systems, Industrial Restructuring, and the Spatial-Temporal Diffusion of Manufacturing Employment, Environment and Planning, 1991, 23:11, 1571–88. Mueller, G. R., Real Estate Rental Growth Rates at Different Points in the Physical Market Cycle, Journal of Real Estate Research, 1999, 18:1, 131–50. Myer, F. C. N., M. K. Chaudhry and J. R. Webb, Stationarity and Co-integration in Systems with Three National Real Estate Indices, Journal of Real Estate Research, 1997, 13:3, 369–81. Myers, J., Fundamentals of Production That Influence Industrial Facility Designs, The Appraisal Journal, 1994, 62:2, 296–302. Piatkowski, S. M., Beyond Compliance: Environmental Risk Assessments at Industrial Properties, Journal of Property Management, 1997, 62:2, 72–5. Porter, M. E., Location, Competition, and Economic Development: Local Clusters in a Global Economy, Economic Development Quarterly, 2000, 14:1, 15–34. Porter, P. R., America’s Industrial Cities: Jolted but Resilient, The Annals of the American Academy of Political and Social Science, 1986, 488, 77–84. Powell, F. M., Preemption of Local Zoning Ordinances Under Federal Hazardous Waste Laws, Real Estate Law Journal, 1997, 26:1, 101–05. Rabianski, J. S. and R. T. Black, Why Analysts Often Make Wrong Estimates About the Demand for Industrial Space, Real Estate Review, 1997, 27:1, 68–72. Rauch, J. E., Does History Matter Only When It Matters Little? The Case of City-Industry Location, The Quarterly Journal of Economics, 1993, 108:3, 843–67. Reenstierna, E. T., Industrials: The Baseline Method, The Appraisal Journal, 1997, 65:3, 291– 300. Shenfield, A. and P. S. Florence, The Economies and Diseconomies of Industrial Concentration: The Wartime Experience of Coventry, The Review of Economic Studies, 1944–1945, 12:2: 79– 99.
THE ENVIRONMENT AND PERFORMANCE OF INDUSTRIAL REAL ESTATE 323
Simons, R. A, Industrial Real Estate Mortgage Default Experience of the New York State Job Development Authority Second Loan Program: A Preliminary Investigation, Journal of the American Real Estate and Urban Economics Association, 1994, 22:4, 631–46. Sivitanidou, R. and P. Sivitanides, Industrial Rent Differentials: The Case of Greater Los Angeles, Environment and Planning, 1995, 27:7, 1133–46. Thompson, R., Industrial Employment Densities, Journal of Real Estate Research, 1997, 14:3, 309–19. Thompson, B. and S. Tsolacos, Rent Adjustments and Forecasts in the Industrial Market, Journal of Real Estate Research, 1999, 17:1/2, 151–67. ——., Projections in the Industrial Property Market Using a Simultaneous Equation System, Journal of Real Estate Research, 2000, 19:1/2, 165–88. ——., Industrial Land Values: A Guide to Future Markets?, Journal of Real Estate Research, 2001, 21:1/2, 55–76. Treadwell, D. H., Intricacies of the Cost Approach in the Appraisal of Major Industrial Properties, The Appraisal Journal, 1988, 56:1, 70–79. Venables, A. J., Localization of Industry and Trade Performance, Oxford Review of Economic Policy, 1996, 12:3, 52–60. Waits, M. J., The Added Value of the Industry Cluster Approach to Economic Analysis, Strategy Development, and Service Delivery, Economic Development Quarterly, 2000, 14:1, 35–50. Wheaton, W. C. and R. G. Torto, 1990, An Investment Model of the Demand and Supply for Industrial Real Estate, American Real Estate and Urban Economics Association Journal, 1990, 18:4, 530–47. Wheaton, W. C., A Perspective on Telecommunications Technology and Real Estate: Office, Industrial, and Retail Markets, Real Estate Finance, 1996, 13:2, 13–17. Young, M. S., REIT Property - type Sector Integration, Journal of Real Estate Research, 2000, 19:1/2, 3–21. Zimmer, D. W., Avoiding Traps When Using Sales Comparison to Value Storage and Distribution Facilities, The Appraisal Journal, 1991, 59:3, 390–94.
The authors thank Brent Ambrose and John McDonald for their helpful comments and suggestions.
Module #5 Warehouse demand.pdf
J R E R � V o l . 2 1 � N o s . 1 / 2 – 2 0 0 1
A n A l t e r n a t i v e D e t e r m i n a n t o f W a r e h o u s e
S p a c e D e m a n d : A C a s e S t u d y
A u t h o r s Asieh Mansour and Marvin C. Chris tensen
A b s t r a c t This article is the winner of the Best Research Paper Presented by a Practicing Real Estate Professional manuscript prize (sponsored by the American Real Estate Society Foundation) presented at the 2000 American Real Estate Society Annual Meeting.
A unique approach is used to assess the demand for warehouse space. Typically, demand for warehouse space has been modeled using population or employment measures. Unlike previous work, warehouse inventory, rather than employment, is used to model space demand. Warehouse inventory is proxied by data on freight shipments. Detailed information on the location and freight activity of manufacturing plants and distribution centers across the United States, Dallas, Los Angeles, and Seattle is used. Warehouse employment is then compared to freight shipments in determining demand for warehouse space. Preliminary results are mixed and the sample size of the metro areas examined should be increased in future work.
I n t r o d u c t i o n
In this article, we develop an alternative determinant of warehouse space demand. We argue that demand for warehouse space originates more from the volume of inventories stored, rather than from the workers used to move this material around. Mueller and Laposa (1994) developed a conceptual framework for understanding the determinants of warehouse space based on the path of goods movement. They argue that industrial employment bears little correlation with absorption of warehouse space, rather that the demand for warehouse space is predominantly influenced by the movement in goods from the place of production to the place of consumption.
We provide empirical evidence to support the methodology developed on a conceptual basis by Mueller and Laposa (1994), using a recently discovered data source that provides detailed information on the location and freight activity of manufacturing plants and warehouse/distribution facilities across regions.
7 8 � M a n s o u r a n d C h r i s t e n s e n
� P r i o r R e s e a r c h
Past empirical research that focuses exclusively on the demand for warehouse space has been limited. Most of the attention has been centered on modeling the demand for office, retail and hotel real estate. As such, traditional approaches to modeling demand for warehouse space have adopted similar approaches developed for office space. Demand for warehouse space has been modeled either using population measures or using relevant employment measures in broad sectors of the economy, including manufacturing, distribution and wholesale trade (see Wheaton and Torto, 1990).
More specifically, Rabianski and Black (1997) argue that industrial real estate demand is usually performed by either emulating office demand analysis, estimating future demand for manufacturing space or forecasting total demand for business park land. They argue that the first approach is misleading because there are major differences between demand for office and demand for industrial space. They further argue that the second and third approaches produce estimates of the demand for industrial land as opposed to industrial space. In this case, the problem of type of land use must be addressed because different categories of industrial uses require different locations.
Wheaton and Torto (1990) also concede that employment may not be the best gauge for warehouse space demand as it is for office demand. In the case of the office sector, the use of an office-intensive employment measure has been quite appropriate and has made significant contributions in assessing future office space needs. In the case of the warehouse market, however, space demand has grown noticeably faster than would be expected based on employment measures. This brings into question the use of employment measures as a proxy for warehouse space demand. As a practical matter, on a metropolitan level, employment data are more readily available than series on output or inventories. Actual data about inventories and warehouse production are available only nationally and costs of obtaining inventories/output data on a more disaggregated sub-national basis have been prohibitive. Employment simply has been the only readily available and low- cost series that could be used to estimate demand for warehouse space.
Population has also been an important variable used to determine regional warehouse space needs. Hughes (1994) went beyond looking at population, however, and focused on the economic activity of warehouse tenants, which he concluded is almost always involved in export activities. He noted that since population-serving activities are housed in office buildings, population changes might be considered the primary factor of demand for office property. By contrast, industrial property demand cannot be determined by changes in population alone. This demand, Hughes concluded, is derived from the export activities within the regional economy.
Chai (1997) argues that the main data problem in many warehouse studies is that we typically have real estate information on the total industrial market, which
W a r e h o u s e S p a c e D e m a n d : A C a s e S t u d y � 7 9
J R E R � V o l . 2 1 � N o s . 1 / 2 – 2 0 0 1
Exhibi t 1 � Basic Transearch Framework
Time Period: Annualized Freight Traffic
Unit of Freight: Short Tons (2,000 lbs)
Geography: 172 Business Economic Areas 50 States
Commodity: 4-Digit STCC Detail Standard 5-Digit STCC Available for Rail and Water
Modes: For Hire Motor Carrier: Truckload For Hire Motor Carrier: Less-Than-Truckload Private Motor Carrier Rail Carload Rail TOFC/COFC International Waterborne Air
includes warehouse, manufacturing and R&D properties. He argues that data on the total industrial property sector does not properly reflect the behavior and performance of warehouse properties.
Research by Kling and McCue (1991) focuses more specifically on the factors that determine the supply, in contrast to demand, of industrial space. They focus on the relationship between industrial property construction and the macroeconomy in order to determine economic variables that explain the variations in industrial construction and how the supply of industrial property reacts to economic shocks. They find that employment accounts for the majority of the variation in industrial property construction and that shocks to output and interest rates affect the supply of industrial space only through its impact on employment.
T h e R e e b i e D a t a b a s e 1
The Reebie TRANSEARCH database of freight movements provides detailed descriptions of freight traffic shipments for different geographic markets, commodities, units of measure and seven modes of transport, including rail intermodal, rail carload, truck, waterborne and air (see Exhibit 1). The database captures domestic freight activity in the United States and has become a standard source of freight flow information used primarily for transportation planning. The TRANSEARCH data has been gathered since 1980 but due to significant data re- definitions, the time series we use begins in 1985.
TRANSEARCH is constructed from a large number of separate, partially overlapping sources. A major component in the development of TRANSEARCH
8 0 � M a n s o u r a n d C h r i s t e n s e n
Exhibi t 2 � Principal Transearch Sources
Primary Domestic Traffic Flow Sources Annual Motor Carrier Data Exchange (Proprietary Shipment Data) Annual Surface Transportation Board Railroad Waybill Sample Annual Corps of Engineers Waterborne Commerce Statistics Annual Federal Aviation Administration Airport Activity and Domestic Market Statistics Import/Export Trade Statistics Annual Department of Energy Coal Movement Statistics Annual Department of Agriculture Produce Movement Data Bureau of Transportation Statistics-Commodity Flow Survey
Primary Production and Shipment Sources Census/Annual Survey of Manufacturers Annual Mineral Industry Surveys from U.S. Geological Survey Annual Motor Carrier Industry Financial and Operating Statistics Annual Railroad Freight Commodity Statistics Federal Reserve Board Industrial Production Indices Trade Association Production and Shipment Reports Annual County Employment and Population Data State Economic Output by Industry Inter-Industry Trade patterns (Input/Output Table)
is the conversion of many different information sources into a single, common framework. Not all sources are equal; economic modeling is used to aid in the design where data are lacking or confidential, to check such elements as spatial patterns and logic, and to construct forecasts. The domestic database is built from approximately 100 proprietary, commercial and public sources of data, representing domestic and NAFTA trade flows. The data covers the universe of total commodity flows in the U.S., as verified by developers of TRANSEARCH (see Exhibit 2). The data is disaggregated down to the county, Business Economic Area (BEA), five-digit zip code, metropolitan area, state and province (in the case of Canada) level. Goods are defined by commodity (SIC) across all regions, with volumes in terms of loads, tonnage or value.
The data that are used in our analysis is presented in Exhibit 3. Total shipment volume in millions of tons for the period 1985 to 1997 for the U.S., Dallas, Los Angeles/Orange County and Seattle is used as a proxy for warehouse inventory. The data can also be disaggregated across more detailed industrial classifications as presented in Exhibit 4. The sectoral compositions of freight shipments may aid researchers and investors in determining the optimal type of warehouse required across metropolitan areas.
W a r e h o u s e S p a c e D e m a n d : A C a s e S t u d y � 8 1
J R E R � V o l . 2 1 � N o s . 1 / 2 – 2 0 0 1
Exhibi t 3 � Total Shipment
(millions of tons)
Year U.S. Dallas Los Angeles/ Orange County Seattle
1985 4902.1 91.6 260.9 110.1
1986 5099.1 97.3 267.6 118.0
1987 5244.6 102.9 274.3 125.9
1988 5532.3 116.8 277.4 139.5
1989 5459.2 110.4 274.3 152.8
1990 5436.8 110.9 247.3 145.6
1991 5123.0 104.4 240.3 138.0
1992 5153.6 105.6 235.8 136.6
1993 5341.1 111.4 239.1 139.2
1994 5641.1 119.7 261.0 145.3
1995 7546.1 170.7 465.3 181.4
1996 8014.0 292.6 528.8 204.0
1997 8511.6 333.4 523.4 227.3
Source: Reebie Associates.
Exhibi t 4 � Sectoral Composition of Freight Shipments
Sector U.S. Dallas Los Angeles Seattle
Farm products 3.4 2.5 3.1 7.5
Forest products 0.1 0.0 0.0 0.0
Fresh fish or marine products 0.0 0.0 0.1 0.0
Metallic ores 1.5 0.0 0.0 0.6
Coal 12.2 7.3 1.7 0.3
Crude petroleum or natural gas 1.2 0.0 6.9 17.5
Nonmetallic minerals 3.5 3.4 0.3 0.5
Food or kindred products 9.7 18.2 15.0 9.9
Tobacco products 0.0 0.0 0.0 0.0
Textile mill products 0.3 0.1 0.2 0.1
Apparel or related products 0.1 0.2 0.4 0.2
8 2 � M a n s o u r a n d C h r i s t e n s e n
Exhibi t 4 � (continued)
Sectoral Composition of Freight Shipments
Sector U.S. Dallas Los Angeles Seattle
Lumber or wood products 7.6 9.1 7.7 15.7
Furniture or fixtures 0.2 0.2 0.4 0.2
Pulp, paper or allied products 2.4 3.3 2.7 3.7
Printed matter 0.4 0.3 0.7 0.3
Chemicals or allied products 6.9 9.3 5.5 3.6
Petroleum or coal products 9.5 11.5 6.3 5.5
Rubber or misc. plastics 0.7 0.7 1.5 0.4
Leather or leather products 0.0 0.0 0.1 0.0
Clay, concrete, glass or stone 12.5 13.3 16.9 5.4
Primary metal products 3.2 3.6 3.6 1.4
Fabricated metal products 1.2 1.1 1.6 1.6
Machinery 0.5 0.6 0.9 0.9
Electrical equipment 0.4 0.5 1.1 0.5
Transportation equipment 1.4 1.9 1.4 1.2
Instrum., photo, optical equipment 0.1 0.1 0.2 0.1
Mis. manufacturing products 0.1 0.1 0.6 0.5
Waste or scrap materials 1.9 0.3 0.5 2.0
Misc. freight shipments 0.1 0.1 0.1 0.1
Shipping containers 0.1 0.1 0.1 0.5
Mail or contract traffic 0.1 0.1 0.1 0.1
Freight forwarder traffic 0.1 0.2 0.5 0.3
Misc. mixed shipments 1.1 2.1 7.6 5.6
Small packaged freight shipments 0.0 0.1 0.2 0.0
Secondary traffic 17.7 9.8 11.9 13.8
� E m p l o y m e n t D e t e r m i n a n t s o f Wa r e h o u s e D e m a n d
Exhibit 5 provides the historical trends of the warehouse market including percentage change in occupied warehouse stock, warehouse employment and freight shipment volume for the U.S., Dallas, Los Angeles/Orange County and Seattle. Demand for warehouse space is measured by occupied warehouse stock. Change in demand over time is captured by the percentage change in warehouse occupied stock as gathered by F. W. Dodge.2 We define warehouse employment as the sum of employment in the transportation services, wholesale trade and
W a r e h o u s e S p a c e D e m a n d : A C a s e S t u d y � 8 3
J R E R � V o l . 2 1 � N o s . 1 / 2 – 2 0 0 1
Exhibi t 5 � Growth in Occupied Warehouse Stock, Warehouse Employment and Freight Shipment Volume
U.S. Dallas Los Angelesa Seattle
Occupied warehouse stock 3.74 2.40 3.10 2.43
Warehouse employment 0.15 0.92 �0.62 2.78
Total shipment volume 4.71 11.36 5.97 6.23
a Includes Orange County.
manufacturing industries. The data are gathered from the Bureau of Labor statistics based on a two-digit SIC classification.
The occupied warehouse stock measure is based on a composite of the largest fifty-eight metropolitan areas. The Dodge definition includes all warehouse space for which hard costs of construction exceed $50,000.3 The square footage totals for occupied stock include: transportation and distribution centers, self storage warehouses, some R&D warehouse space, maintenance and service buildings, parts depots and public facilities such as roadside equipment storage, salt storage, and maintenance and service buildings for equipment and vehicles.
The annual percentage change in occupied warehouse stock has been significantly greater than the annual percentage change in warehouse employment (see Exhibit 5). Instead, it is more in line with the increases in total shipment volume, which we use as a proxy warehouse inventory.4
� S t a t i s t i c a l R e s u l t s
Simple statistical tests were performed to compare the explanatory power of warehouse employment compared with freight shipments to occupied warehouse space for the U.S. and the three regional economies. Exhibits 6, 7 and 8 present both correlation coefficients and regression results. For the U.S., it does not appear that total shipment volume is better correlated with warehouse demand than is employment by looking at the simple correlation between the percentage change in total shipment volume and the percentage change in occupied stock. Regression results presented in Exhibits 7 and 8, however, show that total shipment volume has a greater explanatory power in determining national occupied warehouse stock. The results from the three regional economies differ somewhat.
D a l l a s
Supported by strong locational advantages and buoyant national and regional economies, Dallas/Ft. Worth is one of the leading warehouse/industrial markets
8 4 � M a n s o u r a n d C h r i s t e n s e n
Exhibi t 6 � Correlation Coefficients
(1985–1997)
U.S. Occupied Stock
Dallas Occupied Stock
Los Angeles Occupied Stocka
Seattle Occupied Stock
U.S. Warehouse employment .323 Total shipment volume �.087
Dallas Warehouse employment .055 Total shipment volume �.095
Los Angelesa
Warehouse employment .707 Total shipment volume .258
Seattle Warehouse employment .397 Total shipment volume �.198
Note: All variables are measured in percentage change. a Includes Orange County.
in the nation. It has evolved as a major regional distribution hub serving the West South Central region. The metro area is centrally located with close proximity to growing consumer markets fueled by above-average population, employment and income growth. Employment is largely concentrated in wholesale trade and distribution/transportation services. The metro area also boasts an extensive transportation infrastructure to support regional distribution activities. Warehouses comprise over 62% of the Dallas inventory of industrial space.
Between 1985 and 1997, occupied stock of warehouse space in Dallas grew 2.4% per year (see Exhibit 5). By contrast, warehouse employment grew 0.9% per year and total shipment volume increased by 11.4%. In this market, the growth in warehouse employment has not matched the change in occupied warehouse stock. The annual percentage change in shipment volume has been very strong over this period since it includes a disproportionate share of crude petroleum and petroleum products. Exhibit 4 presents the sectoral shares of freight shipments disaggregated by industrial groupings for the U.S. and the selected regions.
Data presented in Exhibit 6 show that warehouse employment and occupied warehouse stock in Dallas are not correlated closely. The correlation with the percentage change in total shipment volume actually has the wrong sign. By contrast, the regression results presented in Exhibits 7 and 8 show that shipment volume appears to have a greater explanatory power in estimating occupied warehouse space in Dallas.
W a r e h o u s e S p a c e D e m a n d : A C a s e S t u d y � 8 5
J R E R � V o l . 2 1 � N o s . 1 / 2 – 2 0 0 1
Exhibi t 7 � Regression Results—Part 1
(1985–1997)
U.S. Ostockt � 26.3 � 1.173 EWAREt
(1.9) (�0.72) Adj. R 2 � �.042 Ostockt � 3.68 � 0.502 Shipmentst
(1.04) (3.20) Adj. R 2 � .43
Dallas Ostockt � 4.25 � 1.27 EWAREt
(1.61) (2.89) Adj. R 2 � .38 Ostockt � 8.86 � 0.16 Shipmentst
(12.93) (4.37) Adj. R 2 � .6
Los Angeles/Orange County Ostockt � 19.71 � 0.94 EWAREt
(7.7) (�2.76) Adj. R 2 � .36 Ostockt � 8.74 � 0.2 Shipmentst
(4.67) (2.1) Adj. R 2 � .22
Seattle Ostockt � 5.95 � 0.90 EWAREt
(7.44) (6.5) Adj. R 2 � .77 Ostockt � 4.45 � 0.36 Shipmentst
(2.47) (3.72) Adj. R 2 � .52
Notes: Both exogenous and endogenous variables are in logarithmic form. t-Statistics appear in parentheses. Ostock � Occupied warehouse stock EWARE � Warehouse employment Shipments � Total freight shipments
L o s A n g e l e s / O r a n g e C o u n t y
Los Angeles is the largest warehouse market in the U.S., catering to local, regional and national warehouse users. During most of the 1980s, the Los Angeles warehouse market was influenced largely by the manufacturing sector. In 1980, the share of manufacturing employment in the metro area was 3% higher than the U.S. average. Recently, growing international trade has created significant demand for warehouse space. Metro employment is highly concentrated in manufacturing
8 6 � M a n s o u r a n d C h r i s t e n s e n
Exhibi t 8 � Regression Results—Part 2
(1985–1997)
U.S. Ostockt � 23.5 � 2.79
(2.2) (�2.45) EWAREt � 0.87 Shipmentt
(5.48) Adj. R 2 � .71
Dallas Ostockt � 13.88 � 1.20
(3.5) (�1.29) EWAREt � 0.28 Shipmentt
(2.85) Adj. R 2 � .62
Los Angeles/Orange County Ostockt � 15.84 � 0.75
(4.26) (�2.12) EWAREt � 0.13 Shipmentt
(1.39) Adj. R 2 � .41
Seattle Ostockt � 4.86 � 0.74
(4.0) (3.8) EWAREt � 0.11 Shipmentt
(1.18) Adj. R 2 � .78
Notes: Both exogenous and endogenous variables are in logarithmic form. t-Statistics appear in parentheses. Ostock � Occupied warehouse stock EWARE � Warehouse employment Shipments � Total freight shipments
and wholesale trade. In contrast to Dallas, the Los Angeles industrial base is somewhat equally divided between pure warehouse distribution space and manufacturing space.
Between 1985 and 1997, occupied warehouse space in the Los Angeles/Orange County metro area grew 3.1% per year. This is in line with the 6.0% growth in shipment volume and far different from the actual decline in warehouse employment. Correlation measures and the simple statistical results also show that demand for warehouse space in Los Angeles/Orange County appears to be predominantly inventory driven. In the case of Los Angeles/Orange County, the regression coefficients on warehouse employment even have the wrong sign as reported in Exhibits 7 and 8.
S e a t t l e
The Seattle-Bellevue-Everett metropolis is the largest urban area in the Pacific Northwest, serving as the financial, commercial and transportation center for the state of Washington and the entire region. The metro area employment base is
W a r e h o u s e S p a c e D e m a n d : A C a s e S t u d y � 8 7
J R E R � V o l . 2 1 � N o s . 1 / 2 – 2 0 0 1
dominated by the aircraft/aircraft parts industry and computer and data processing. Warehouses comprise the largest part (71%) of the industrial base, followed by manufacturing space.
In contrast to both Dallas and Los Angeles/Orange County, warehouse employment correlates more significantly with occupied warehouse space, based on preliminary results presented in Exhibits 6–8. Seattle, in contrast to the other metro areas cited in this case study, is the only region to exhibit continued growth in its manufacturing employment base. Between 1980 and 1999, manufacturing employment averaged 0.7% per year in Dallas, declined 1.7% per year in Los Angeles and averaged 0.4% in Orange County. By contrast, manufacturing employment grew 1.2% a year in Seattle.
� C o n c l u s i o n
This is an exploratory case study trying to address the empirical problem of adequately measuring demand for warehouse space. Employment-based determinants of demand for warehouse space have been found to be lacking on both a national and regional basis. Thus, we use the volume of freight shipments as a proxy for warehouse inventory to provide an alternative measure of warehouse space demand. We focus on three regional distribution centers: Dallas, Los Angeles/Orange County and Seattle. Preliminary results suggest that in metro areas where manufacturing employment has been declining, warehouse employment has been a poor measure of space demand as measured by occupied warehouse space. In those cases, Los Angeles and Dallas, freight shipments seem to be a better measure of warehouse demand over time. By contrast, in the case of Seattle, where manufacturing employment has actually been contributing positively to the employment base, warehouse employment measures do adequately track warehouse demand.
In order for commercial real estate investors to assess the strength of a warehouse/ distribution market, the use of simple employment measures may be misleading. Further research is required. For example, subsequent versions of this case study should broaden the array of statistical diagnostics to test for the presence of other important time series relationships such as co-integration. Furthermore, out-of- sample forecasting tests should be performed in order to determine whether the model provides useful predictions for practitioners. There is also a need to broaden the number of markets in the sample in order to expand the comparative analysis and to enhance the robustness of the results.
� E n d n o t e s 1 More information on the Reebie TRANSEARCH database can be found by contacting
the company headquarters at 203-705-0455 or www.reebie.com. 2 We use a stock variable, i.e., occupied warehouse stock, rather than net absorption, the
first difference in occupied stock as a proxy for warehouse demand. This is consistent with total shipment volume, which is also a stock variable.
8 8 � M a n s o u r a n d C h r i s t e n s e n
3 The Dodge data used are based on square footage of stock for the top fifty-eight metropolitan areas. It is not based on the government’s construction put-in-place data, which is in value terms.
4 Using total shipment volume as a proxy for the inventory of warehouse goods across the three markets and the U.S. assumes that velocity, or the speed, at which goods move in and out of the space, is equal. With improving distribution technology and just-in-time inventory management pervasive across the U.S. and specifically these three regional distribution hubs, this assumption is not altogether unrealistic.
� R e f e r e n c e s
Chai, Y. W., Determinants of NOI for Warehouse Properties, Real Estate Finance, 1997, 14:2, 48–54.
Hughes, W. T., Determinants of Demand for Industrial Property, The Appraisal Journal, 1994, 62:2, 303–09.
Kling, J. L. and T. E. McCue, Stylized Facts about Industrial Property Construction, Journal of Real Estate Research, 1991, Fall, 293–304.
Mueller, G. R. and S. P. Laposa, The Path of Goods Movement, Real Estate Finance, 1994, 6:3, 42–50.
Rabianski, J. S. and R. T. Black, Why Analysts Often Make Wrong Estimates About the Demand For Industrial Space, Real Estate Review, 1997, 27:1, 68–72.
Wheaton, W. C. and R. G. Torto, An Investment Model of the Demand and Supply for Industrial Real Estate, Journal of the American Real Estate and Urban Economics Association, 1990, 18:4, 529–47.
Asieh Mansour, RREEF, San Francisco, CA 94111 or [email protected].
Marvin C. Christensen, RREEF, San Francisco, CA 94111 or [email protected].