The Bob
Senior Member
- Joined
- Jan 20, 2002
- Professional Status
- Certified Residential Appraiser
- State
- Virginia
I think I get what you're saying. Whenever I have run iterations of the difference of say, a 3br vs a 4br (all other things more or less equal), I will invariably get a range of "adjustments" from "0" to $zzz.I think it is fine to use 'narrowing the range' as one way to test the reasonableness of an adjustment. But if that's your goal or the only method used for modeling adjustments, then you risk overfitting the sales in your grid using adjustments that, when applied to a larger set of comparable sales, would not fit nearly as well into your model. On the other I agree that if your range widens upon adjustment, the appraisal is not likely reflecting the behavior market participants. And it is always possible that even sales with differences require $0 adjustments.
When I say I "tighten up" the adjustment, what I am doing is using an adjustment within that range that makes sense considering the Comps I used. If I have a difference of $5,000 between a 3br and a 4br, does it make more sense to use $5k, which is within the range of adjustments, or $10k which is also within the range? The other comment I would make is that not every buyer reacts the the same to every feature, so to me it's logical that the premium for a feature will vary within the market and your a adjustments should reflect that.