Tuesday, April 30, 2019

A Good Comparables system must not allow Automated Valuation of Waterfront Homes

(Click on the image to enlarge)

Waterfront is often the most significant variable in comparable sales analysis. Generally, a waterfront home costs 10's of thousands to 10's of millions more than a non-waterfront home. In other words, just move a home one block from the non-waterfront block to the waterfront block and its value could jump many, many fold. 

Given the above axiom, one has to be very careful in relying on the automated comps to value a true waterfront subject, with unobstructed view and access. Unless the selected comps are very similar, the valuation would be off, perhaps way off. That is why, an all-algorithm approach (without the spatial verification) could be a hit-or-miss. Once the algorithm scores and returns the optimal pool, one needs to spatially verify the pool and select the final five, contributing to the subject valuation.

Since Homequant is a self-directed home valuation solution, we urge our users to "experience" the environment (by turning on the Street View and strolling around the neighborhood) before selecting the final five. If the scored pool comprises very similar and clustered comps, we suggest the use of the least adjustment method to pick the final ones. In the above graphic (from the Homequant site), once our algorithm scored and returned the pool of ten best comps, we manually selected the final five after the aforesaid data research and spatial experience. 

Again, we believe that the valuation of waterfront subjects must not be an all-algo hit-or-miss event; it must require proper manual and professional intervention. 

Thank you,

Sid Som, MBA, MIM
President, Homequant, Inc.

No comments:

Post a Comment