GLA adjustments

[EVAUSA]

Hiccup,

this question comes up often with non appraisers, you are not alone. Cost does not = value. It may cost $125/squ. ft to build, but contributory value may indeed only be $50/squ. ft. In my market area (CA), I adjust anywhere from $30/squ. ft. (fair to average quality) to $80/squ. ft. (highly custom/imported/etc).

Keeping it simple, you have three houses that are identical in all respects except #1 is 1500 squ ft ($100,000 SP), #2 is 1800 squ. ft ($115,000 SP) and #3 is 1300 squ ft. ($90,000 SP).

If the subject is 1500 squ. ft, you can see that $50/square foot can be extracted, and all three sales will adjust to $100,000 when adjusted to match the subject.

This same concept applies to adjusting lot size. I typically make $2 - $6/squ. ft. site adjustments for their contributory value as compared to the subject, not to be confused with the $/squ ft. of buying raw land.

 

[Hiccup12]

Assuming your method is attributing value and not cost, your method, simplified, is not that bad. :new_smile-l: And, there may be other influences which are not accounted for, but you seem to have been around the block so I will assume there isn't anything unusual that cannot be accounted for outside of the lot, the garage and the improvement (that is a big if, but I'll make it!).

However, where it runs into a problem is here:
You've compared two sales that are between 1,700sf and 2,500sf and then have established a relationship between the two in terms of living area (which we call GLA-Gross Living Area). Two points of data creates a straight-line between the two. And the best that can be said is that based on interpolation, we can estimate where a value should fall between the two points, if we are assuming that size is the determining variable.

Your house is 3,200sf. To get from the 1,700 to 2,500sf range to your house, one must extrapolate. That's risky business because there is no evidence that the line (which was created only between two points) continues straight to 3,200sf. As a rule (and I think you'd know this), the value of an incremental SF over the mean decreases as the overall size increases. There's exceptions. Move to a different neighborhood and you have a different set of value assumptions. But in the same neighborhood, this generally holds true.
It holds true because there is something called the Principle of Regression at work in most real estate markets.

To answer your question that I bolded above, one method would be to do what you did (with the smaller homes) but get some larger homes to "bracket" your house. Then calculate the slope of the line for GLA and value. Even that isn't perfect, but it is better than only relying on extrapolation.

(And, by the way. I'm not saying your conclusion about the GLA adjustment in the original report isn't correct- i.e. the report has your GLA adjusted lower than what the market indicates. What I am saying is that the analysis between the 1,700sf and the 2,500sf homes doesn't prove it.)

Good luck! :new_smile-l:

All of you, thanks for your replies. From most of what I've seen, I don't think the appraiser did the work that he should have. Bracketing the house with comps is the preferred way to go, but if that is not possible, then interpolation probably is a better way to go than to just say "that's what I think it's worth." I agree that 1700-2500 probably isn't a good match, but it does supply SOME information, whereas the appraiser supplied NO information to back up his statement. I think that interpolating that information is as viable as him using a 2400 SF comp and then saying what he "feels" is the right GLA adjustment. Historically around here, the price per square foot makes a rebound around the 3,000 SF level and usually you see a higher price per SF even with comparable lots and garages. Bottom line, I think that researching and coming up with a legitimate appraisal suddenly became way more time consuming than he thought is was going to be and he just threw down some numbers hoping that I wouldn't know any better.