The Sales Comparison Approach is a method of estimating the market value of a subject property through an analysis of sales of similar properties. Sometimes referred to as the Market Data Approach, this method assumes that sales of comparable properties, having similar physical and area data attributes, provide a range of value for the subject. The credibility of this approach depends upon the accuracy of the data collected, as well as the degree of comparability between each sale property and the subject property. Properties, which are acceptable for comparison with the subject, meet the following criteria:
Only "arm's length" transactions between two unrelated parties are considered, with neither the buyer nor seller being under compulsion to act.
Details of the transactions are reliable and have been verified to the greatest extent possible with the parties involved.
The comparable properties exist in the same competitive market as the subject, and represent competitive investment alternatives. Physical or area data characteristics, income generating capability, functional utility or a combination of these attributes may define the boundaries of the market.
The selection of a unit of comparison provides a common denominator by which market sales can be related to each other and to the subject property. The commonly accepted unit of comparison in the valuation of industrial real estate is the selling price per square foot of building area. While a diverse array of transactions was initially considered, the sales selected for direct comparison to the subject are those transactions, which are most similar to the subject. For features, which are dissimilar between the sales and the subject, adjustments, have been made leading to an indication of the price at which the property being appraised could be expected to sell.
Five comparable sales from a sample of fifty-seven (57) sales of industrial buildings are identified on the following pages and, based on my investigation, these sales are considered to be the most significant transactions for direct comparison with the subject. Detailed information on the comparable sales, regression graphs and photographs of comparable sales is found in the addenda of this report.
Analysis of Comparable Sales
With the foregoing in mind I have analyzed industrial building sales which are considered to represent the best available data. These sales are considered a representative sample, which permitted the use of a statistical technique known as regression analysis. This is defined "a method that examines the relationship between one or more independent variables and a single dependent variable by plotting points on a graph; used to identify and weight analytical factors and to make forecasts." (Appraisal Institute. The Dictionary of Real Estate Appraisal, 3rd Edition. Chicago: Appraisal Institute, 1993, Page 299).
It is important to note that Simple Regression involves the analysis of one independent variable and one dependent variable. Multiple Regression Analysis involves the analysis of two or more independent variables and one dependent variable. Displayed on a graph the independent variable is plotted along the "X" axis (East/west) and the dependent variable is plotted along the "Y") axis (north/south.
Among the improved sales I identified seven (7) independent variables that are considered important in the valuation of industrial buildings. These are the sale price per square foot of the building area in relation to building age (years), ceiling height (feet), building area (square feet), distance to nearest expressway interchange (miles), lot size (acres), exterior wall (metal/masonry), and market conditions (date of sale). The Sale Price Per Square foot of building area is the dependent variable.
I applied multiple linear regression analysis to my sample of industrial properties. This analysis resulted in the following summary statistics:
Multiple R 84.9%
R Square 72.1%
T Statistic 1.61 to 6.30
P Values 0.000008% to 39.4%
F Statistic 18.1103
The "F" statistic is an indicator of overall "good fit" for the entire sample. At the 95% confidence level should be more than 2.2032 for seven factors with 49 degrees of freedom (57 sales less 7, minus 1). The sample has a very high 18.1103 F statistic indicating that the overall sample is reliable. This indicates a confidence level of 99.99999999%.
The T-distribution statistic at the 95% confidence level for 49 degrees of freedom should be more than 1.6766 (one sided) or 2.0096 (two sided). The sample has a "T" statistic range of 1.61 to 6.30 indicating that most of independent factors are significant. The only factor with a low "T" is lot size.
The "P" value indicates the probability that null hypothesis is zero. The "P" values are low ranging from .000008% to 11.3% (excluding lot size).
Hence, the sample of sales is considered to be excellent indicators of value and can be relied upon for estimating the value of the subject property. The multiple linear regression model calculated the "Y" axis intercept and the sum of least squares of the independent variable coefficients from a statistical sample of building sales. The dependent variable is the sale price per square foot of the building area. These factors were applied to the subject resulting in an indicated value via the multiple regression model.
The same mathematical process can be applied to all of the comparable sales. This produces a predicted value for each of the comparable sales based on the sample of market data. These predicted values are listed in the addenda. The difference between the actual sale price per square foot and the predicted sale price per square foot is the "residual" which the multiple regression model does not directly explain. The multiple linear regression sample produced an 84.9% correlation factor and a coefficient of determination of 72.1%. Real estate is an imperfect marketplace within which there is some degree of random deviation from the market mean price. This is normal. The only comparable that would have no residuals is a property that is identical in every way to the subject.
The net difference between the subject's estimated value per square foot and the predicted value per square foot of the comparable sale price per square foot are the market-derived adjustment from the comparable sale to the subject. The addenda of this report contain the linear regression graphs for each of the seven- (7) dependent variables. The subject is identified at $15.00 per square foot
Building Size: The graph is downward sloping trend line resulting in a lower price per square foot as the building size increases.
Miles to Expressway: The graph is downward sloping trend line resulting in a lower price per square foot as the distance to an expressway interchange increases.
Building Age: The graph is downward sloping trend line resulting in a lower price per square foot as the building age increases.
Exterior Wall: Metal exterior walls are rated a 1.00 and all masonry 2.00. The subject is 50% metal and 50% masonry and is rated a 1.50. The graph is upward sloping trend line resulting in a higher price per square foot for masonry exterior walls.
Ceiling Height: The graph is upward sloping trend line resulting in a higher price per square foot as the ceiling height increases.
Lot Size: The graph is a slightly downward sloping trend line resulting in a lower price per square foot as the lot size increases.
Market Conditions: The graph is downward sloping indicating sale prices have decreased from 1987 to 1998.
Final Conclusion - Sales Comparison Approach
Based on these criteria I selected five comparable commercial sales, which I consider most similar to the subject. The indicated price per square foot of the subject building was $15.00/SF which when multiplied by the subject building area of 200,000 square feet results in an estimated value of $3,000,000.