- 10 - a sewer and the other does not. This difference is important: in reconciling the sales comparables to the subject property, Cobb used an adjustment factor of plus 60 percent or minus 30 percent to correct for inferior or superior utility service relative to the subject property. Yet, if tract B were correctly valued at $2,200 per acre, application of Cobb’s adjustment factor would yield a value for tract A of $3,520 per acre. On the other hand, if tract A were correctly valued at $4,800 per acre, application of Cobb’s adjustment factor would yield a value for tract B of $3,360 per acre. Either way there is a sizable residual discrepancy that Cobb has not explained. That only 70 percent of tract A is currently tillable makes the discrepancy all the more puzzling. Respondent reasons that Cobb has grossly overstated the value of tract A. An alternative interpretation, however, is that Cobb understated the value of tract B. The inconsistency calls into question the reliability of Cobb’s appraisal. The next step in Cobb’s analysis was to calculate the amount of rental income that a hypothetical investor in the subject property would require. This was accomplished by multiplying the estimated fair market value of each tract by a “built-up” rate. This rate is the sum of (1) the risk-free rate of return for a term of years equal to that of the lease, represented by the yield of 8.77 percent on 5-year U.S. Treasury bonds as of April 1990; (2) a minimum risk premium of 2 percent, reflecting thePage: Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Next
Last modified: May 25, 2011