Industrials: The GIS Map

For a long time, most of what it took to make a good appraisal were a skilled, experienced appraiser and well researched market data. Then came computers, enabling analysts to handle large volumes of complex cash flows. Then came G.I.S.

A G.I.S., or Geographic Information System, is a representation of information in the form of a map. We have all had some exposure to the color coded maps that show us, for instance, per capita income town by town or the distribution of two tone Buick Rivieras by census tract. G.I.S. brings into focus what once was hidden. Early applications have been to assist in matching product to consumer for instance, Starbuck's outlets to the high income, young adult neighborhoods in which they are popular. G.I.S. can be paired with multiple regression analysis (number crunching of large volumes of data) for appraisals: on the West Coast, systems are already in place to provide valuations of single family dwellings with accuracy, according to G.I.S. guru Gilbert Castle, in minutes. Given the slower turnaround time and largely unknown accuracy of standard appraisals, the product has proven attractive to clients.

Reenstierna Associates has been at work to develop a G.I.S. based industrial valuation model. Regression analysis is applied to a large data base. Regression alone can produce price estimates. But our experience is that location (more than age, construction quality, or land ratio) is the single variable with greatest influence on value, and its influence can most easily be shown by map.

Our map and model produce a series of statistically supported observations:


- the peak of the local market is at Needham on Route 128; an industrial building here achieves a price twice that of an identical building on the Industrial Parkway in Woburn and three times that of one on Turnpike Street in Stoughton
- buildings in well planned industrial parks can achieve prices as much as double those of similar buildings in nearby inferior neighborhoods
- appreciation in industrial property in Greater Boston overall has been a negligible 1% per year since the market bottom of 1992
- central urban locations (Chelsea, South Boston) exhibit high variation over short distances, versus greater uniformity over broad stretches of Route 495.

The advantages of G.I.S. based analysis are numerous:

- increased accuracy from a large body of data, versus standard analysis through a smaller number of sales
- shortened turnaround time
- measurement of accuracy by supportable statistical means (currently, a standard deviation of 14.77%)
- greater objectivity, with virtual immunity of the system to manipulation on the part of the operator to produce a "desired" conclusion.

As with any new technology, G.I.S. based valuation is likely to result in both unforeseen hazards and unforeseen benefits. Standard analysis that makes use of a limited number of sales researched in depth will always have an important place. But where computer and G.I.S. based analysis prove more efficient, they too will find a place. Tools like G.I.S. are likely to expand our vision of how markets function in ways that could not have been guessed.

Eric T. Reenstierna, MAI