"GIS: Meeting the Information Demand"

The following text is Gil Castle's final draft of the real estate column appearing in Valuation Insights and Perspectives, Fourth Quarter 1998

A sea change in the real estate industry is undermining traditional appraiser revenue sources. This article discusses several salient characteristics of that sea change, the ramifications for appraisers and other real estate professionals, and the increasing importance of Geographic Information Systems (GIS). The article concludes with a brief review of a new Appraisal Institute book entitled "GIS in Real Estate."

For those not familiar with the technology, GIS is essentially a marriage between computerized mapping and data base management systems. Anything that can be recorded on a map can be replicated in a computer—using longitude-latitude or other coordinate system—and then interrelated to everything else that can be mapped. Since approximately 85% of all computerized data bases have some type of geographic reference (e.g., a street address, zip code zone number), GIS technology can integrate more information than relational, hierarchical, network, or other data base managers. Moreover, people like pictures—"worth a thousand words"—and a GIS definitely draws compelling pictures/maps. A GIS thus uses "location, location, location" to compile and display all the information needed for sound real estate decision making.

What Is Happening?

Particularly noteworthy are these four trends in the real estate industry:

(1) Internet Broker Listings. A large and ever growing percentage of both commercial and residential properties are being listed on literally thousands of brokerage Web sites. Of even greater significance are the efforts of several major players to become the predominate, national Multiple Listing Service (MLS):

On the commercial side, the leading contenders include LoopNet (www.loopnet.com) and CoStar (www.rig.com).
On the residential side, the National Association of Realtors® has more than one million listings at its site (www.realtor.com), but faces formidable, non-traditional competition from Microsoft (www.homeadvisor.com), YAHOO! Real Estate (http://realestate.yahoo.com), Cyberhomes® (www.cyberhomes.com), etc.

Several of these organizations are already trying to gain a substantial competitive edge by using GIS to display the locations of properties satisfying their customers' selection criteria, as well as other important types of data. On the residential side, for example, potential home buyers want to see not only the locations of available houses but also the locations of schools, parks, shopping centers, etc. "Exhibits 1 and 2" show the results of a typical Internet home search in the Bay Area on www.cyberhomes.com.

(2) REITs et al. Real Estate Investment Trusts (REITs), Commercial Mortgage Backed Securities (CMBSs), and other publicly traded real estate securities are becoming major forces within commercial real estate. For example:

The market capitalization of REITs has been increasing exponentially—from $13 Billion in 1991 to $141 Billion in 1997;
REITs accounted for half of the institutional commercial real estate investments in 1997;
REITs control nearly 9% of the institutionally owned U.S. commercial real estate.

Publicly traded real estate securities should continue to gain in importance as:

Investors come to realize that such securities are a credible asset class alternative to stocks and bonds;
The choices among, liquidity of, and information availability on these securities continue to increase;
The age demographics of the U.S. population continue to cause a ballooning of monies going into pension funds, which favor public real estate investment vehicles.

The number of properties encompassed by the largest 135 REITs is approximately 14,000. Sooner or later REIT investors, Wall Street analysts, and other industry professionals will be demanding current GIS displays of the locations of properties owned by any given REIT, not only to just keep track of REIT holdings but also to better understand and predict the financial performance of the REIT. Additionally, as with home purchasers, REIT industry players will want to see GIS displays of other variables, such as submarket-level variables affecting the demand-side for a REIT's properties. "Figure 22" is an MAI appraiser's analysis of the changes in employment opportunities in the vicinity of an apartment property.

3) Computerized Analyses Generally. Real estate is essentially a game of information arbitrage. The likely winner of the game is whoever has the most comprehensive, timely and accurate information. Serious competitive maneuvering is underway by various members of the Real Estate Information Providers Association (REIPA, at www.reipa.com) and other data providers to meet the ever more pressing informational demands of the real estate industry. Since data is to software as fuel is to an engine, the number and sophistication of computerized analysis tools is keeping pace.

Information arbitrage directly affects individual appraisers, brokers, and other industry professionals. When authorizing a five-figure commercial appraisal, for example, a bank officer is likely to be most impressed by the MAI with demonstrated access to the greatest amount of pertinent information—and GIS is an attention-getting way of showing what you know. "Figure 24" is a further analysis of "Figure 22's" apartment property; this time the appraiser is showing the precise locations of the current and potential employers of the property's tenants. Why should the bank officer hire the MAI who lacks access to this level of information?

Given the ever greater size, complexity, and especially cost of data bases, the environment of choice is shifting from stand-alone desktop computers to local area networks—and to the Internet. Given the choice of:

Buying and maintaining a huge data base (much of which might never be needed) on a desktop computer, versus
Accessing that data base on the Internet and acquiring under a pay-per-view arrangement only the data needed,

who among us would continually choose the former option over the latter?

For example, a retail site selection analyst can purchase 200+ demographic variables for every census block group in the nation and load that information into a desktop GIS for approximately $10,000 to $20,000 per year; alternatively, the analyst can have unlimited access to those variables in map form for $595 per year from Market Statistics (www.sbponline.com). Similarly, real estate professionals in Dallas-Fort Worth can access numerous demand-side, supply-side, and environmental data bases for the entire Metroplex in GIS form on a subscription basis from Bruton Information, Inc. (www.brutoninfo.com).

(4) Automated Property Appraisal Specifically. Innumerable individuals and institutions have experimented with automated appraisal systems using statistical formulas, artificial intelligence methods (e.g., neural nets), and other pattern recognition techniques. The research has focused primarily but not exclusively on residential properties. Some banks, such as Seafirst and Norwest, have largely supplanted appraisers with automated appraisal systems. The industry is still waiting for a standard, most likely to come from the research that both Fannie Mae and Freddie Mac have been focusing on such systems. Initial estimates are that approximately 85% of residential appraisals will ultimately be conducted by automated appraisal systems.

GIS is especially relevant here, because GIS:

Allows a much greater number of independent variables to be included in the statistical or other analyses than is available from assessor files, MLS records, and other widely used data sources;
"Knows" exactly how far a subject property is from various candidate comps, measured as-the-crow-flys or as-the-car-creeps;
Produces accurate, full-color maps showing the subject, selected comps, and any other relevant information a reviewer would like to see for audit trail purposes.

In short, GIS-based automated appraisal systems—on desktop computers and over the Internet—will inevitably and profoundly change the appraisal industry. Stated differently, if roughly 85% of residential appraisals are conducted by GIS-based and other automated appraisal systems, then (as discussed more fully below) roughly 85% of appraisers will have to modify their careers.

What Does It Mean?

What, then, are the implications of the above observations vis-a-vis the real estate industry, real estate professionals, and GIS?

Real Estate Industry Overall. Commercial real estate will increasingly be bought and sold in a manner similar to stocks and bonds. Real estate assets will be controlled by ever fewer and ever larger players, operating from the nation's (if not the world's) capital centers. In parallel, the percentage of securitized commercial properties will continue to grow beyond the current 10% level. For both commercial and residential real estate, computerized data bases and analysis tools will increase in number, size, and sophistication—emulating information systems utilized by investors for evaluating stocks and bonds. The Internet will be the predominate information distribution method. As discussed below, GIS is the logical technology to facilitate the real estate industry's information revolution.

Real Estate Professionals. For both commercial and residential real estate, fewer traditional real estate professionals will be needed, especially brokers and appraisers. The survivors of the shake out—especially the 85% of current appraisers who will likely have to change their business practices—will gravitate toward two principal types:

Knowledge workers—with a high degree of computer literacy in both software and data bases (including, of course, GIS), who handle complicated transactions, appraisals, highest and best use analyses, etc;
Data collectors—with a marginal degree of computer literacy, who continually feed updates on sales prices, property listings, asking rents, vacancy rates, and so on to the extensive real estate data bases resident on the Internet, as well as conducting drive-bys and other data verification procedures.

A useful model of real estate brokerage and appraisal firms of the future may well be the Wall Street brokerage firms of today (Morgan Stanley Dean Witter, Merrill Lynch, and so on). Like stock brokers, real estate brokers/appraisers would provide both research and transactional services, helping clients to assemble and revamp portfolios of real estate equity and debt investments.

Role of GIS. No other data base management system uses "location, location, location" to organize and portray information. Accordingly, in the ever more data-intensive and software-intensive world of real estate, GIS is the logical technology to become the industry's informational May Pole. This will be true both on personal computers and the Internet.

Even today, for example, one can:

Log onto one of the above mentioned GIS-driven Web sites;
Display a map of a city and "zoom down" to a specific area or site;
Click on various map "layers" of interest, such as current and projected population, median household income, traffic counts on arterials, vacancy rates, asking rents, flood zone boundaries, etc.;
Prepare a compelling presentation map with the selected map layers, title block, legend, scale bar, and so on, to print immediately in small or large size, cut-and-paste into a report or brochure, or save in an electronic file for future use.

To date the most frequent use and cost justification of GIS has been in data presentation—primarily because relative few professionals have invested the time and effort needed to completely understand and benefit from the full spectrum of GIS capabilities; see "Side Bar One." History will ultimately show, however, that GIS' greatest value-adding capabilities are in data integration and analysis.

GIS will not supplant all other data base managers (e.g., Access, Oracle, RBase), much less spreadsheets, wordprocessing, drawing programs, etc. The reasons lie in the innumerable differences in what these various types of software are designed to do, and in the preferences of individuals and organizations to stick with what they've already installed as long as possible. However, GIS is and will remain a fundamentally complementary technology for assisting real estate professionals in performing their jobs cheap-faster-better.

"GIS in Real Estate" (The Book)

"GIS in Real Estate: Integrating, Analyzing and Presenting Locational Data" is the seminal book on the application of GIS technology to both commercial and residential real estate. The 200+ page book is being published by the Appraisal Institute in November 1998.

After a Forward and a Preface, "GIS in Real Estate" devotes two chapters to an overview of GIS, the technology's relevance to real estate, and the benefits enjoyed by practitioners. The next two chapters explore a broad spectrum of GIS case studies in commercial and residential real estate—including brokerage, mortgage underwriting, institutional investment, corporate real estate, and so on. Chapters 5 through 7 focus exclusively on GIS in appraisal, both in traditional appraisal assignments and in new areas such as automated appraisal systems. The next chapter covers the basics of implementing and maintaining a GIS, including software, hardware, data bases, personnel, etc. The last chapter predicts the future of GIS in the real estate industry. Finally, for further information, the Appendix provides the Web sites of leading GIS software vendors, data sources, service providers, glossaries, periodicals, conferences, etc. "Side Bar Two" provides some of those sources. Finally, note that this article's GIS graphics all come from the book "GIS in Real Estate."

In addition to serving as overall editor, I wrote the Preface and the chapter on GIS applications in commercial real estate. The other authors of the book are:

Anthony H. Hutchinson—Research Analyst with Fannie Mae, Washington, D.C.
John D. Landis, Ph.D—Associate Professor in the Department of City and Regional Planning, University of California at Berkeley, Berkeley, CA
Bill Mundy, Ph.D, MAI, CRE—President of Mundy & Associates, Seattle, WA
Michael L. Robbins, Ph.D, CRE—President of GRAASroots Real Estate Consulting, Denver, CO
Stephen E. Roulac, Ph.D, CPA—President of the Roulac Group, San Rafael, CA
L. Ried Schott, MAI, SRA—President of Schott Consulting Services, Los Angeles, CA
Grant Ian Thrall, Ph.D—Professor in the Geography Department of the University of Florida in Gainesville, Gainesville, FL
Marcie B. Wallis—Managing Director of Marketing Research at the National Association of Realtors®, Washington, D.C.
Bruce R. Weber, MAI—President of Market Foresight, Newport Beach, CA

* * * * *

The Bottom Line issue is not whether you should re-invent yourself, but how—especially with the benefit of GIS technology.

"SIDE BAR ONE"
ILLUSTRATIVE GIS FUNCTIONS, FROM SIMPLE TO COMPLEX

Retrieving and displaying one or more maps (essentially just opening a file, looking at the map, perhaps zooming in-out and scrolling-panning, perhaps laying one map on top of another).
Finding an address, the shortest route from A to B, adding simple annotation, and similar options available at frequently visited mapping sites on the Internet.
Retrieving certain types of information from the map by:
- Clicking on a map location to display attribute data about that site, including text labels, numbers, photographs, other images;
- Selecting a pre-stored query from a menu to see a map of all sites fulfilling the query's criteria, such as "all census tracts with a median household income greater than $20,000 and an annual population growth rate greater than one percent";
- Entering an ad hoc query, probably in an SQL-like format, again to display all sites on the map fulfilling the user's specific criteria.
Producing various types of thematic maps including:
- Shading polygons by different colors and patterns (e.g., census tracts colored in five shades of green based on median household income);
- Using dot density functions (e.g., the number of dots placed in each census tract is proportional to the population);
- Changing the width, color, and pattern of lines when thematically mapping, say, traffic counts on a given city's arterials;
- Modifying the size, symbol, and color of individual points, such a map of homes available for sale.
Defining one or more regularly or irregularly shaped polygons (e.g., the primary, secondary, and tertiary trade areas of a retail outlet) and calculating statistics about those polygons.
Generating buffer zones and proximity calculations, such as all sales "comps" within three miles of a subject property or within one mile of all the freeway interchanges.
Creating accurate, meaningful business graphics (bar charts, pie charts, etc.) from mapped information.
Importing and exporting tabular data from/to such widely used non-GIS software packages as DBase, Excel, Lotus 1-2-3, Oracle, FoxPro, Paradox, etc.
Developing real estate-specific models with outputs that can be mapped, e.g., combining census tract-level data on the number of households and household incomes into a retail buying power index.
Employing network analysis functions, including:
- Finding the fastest and/or cheapest path from one one site to many others;
- Identifying the most efficient route connecting several destinations (the "traveling salesman" problem).
Accurately, completely, and efficiently geocoding (address matching) properties on a map, such as all the locations of all the mortgages a bank has underwritten.
Producing professional, compelling presentation graphics of cartographic and related data, including the use of different colors, patterns, symbols, fonts, raster images, layouts, etc.
Storing and accurately registering to maps multiple types of raster images, including simple photographs, high resolution aerial photography, satellite data and videos.
Digitizing (with a digitizing table or tablet) polygonal or linear features not available from data vendors, such as proprietary submarket boundaries.
Establishing and using real-time "hot links" (using DDE, OLE and other protocols) between GIS and other software packages such as spreadsheets and DBMSs.
Customizing a GIS with scripting languages provided by GIS vendors, such as Avenue from ESRI or MapBasic from MapInfo.
Integrating third-party packages with the GIS, such as linking a GIS to a Computer Aided Mass Appraisal (CAMA) package.
Setting up pattern recognition modules (based on statistical and/or artificial intelligence software) and "red flag agents" that search data bases periodically for predefined circumstances, e.g., any new Class A office lease comps.