Comparison of the Accuracy of the Minimum-Variance Grid Method and the Least Squares Method – A Non-Linear Extension
Real Estate Valuation Theory, (Wang, K and Wolverton, M. I. ed.) American Real Estate Society Monograph Series: Resear
11 Pages Posted: 14 May 2012 Last revised: 15 Nov 2012
Date Written: August 16, 1999
Lai and Wang's (1996) work suggests that under most circumstances, property valuation based on Vandell's (1991) minimum-variance grid method is more accurate than the valuation based on a linear hedonic pricing model with its coefficients estimated using ordinary least squares technique. This study extends the comparison to the non-linear case. We have shown that Lai and Wang's results also hold even when the hedonic price model is non-linear in its coefficients under the same conditions specified in Lai and Wang. Furthermore, we have also derived the necessary and sufficient conditions under which minimum-variance grid method is more accurate than the least squares method.
Keywords: Valuation, hedonic price model, minimum-variance grid method
JEL Classification: C13, L85, M29, M19
Suggested Citation: Suggested Citation