"Quantifiable Market-Derived Methods" for adjustments required by FNMA/USPAP

[Terrel L. Shields]

Do the people in DC actually believe appraisers can, and will, include quantifiable market-derived support for their adjustments?

Been doing it for years. I include vacant land sales to support my land values. I use sensitivity to support adjustments for effective age and SF adjustments. But I don't work secondary market, and I don't make excessive adjustments.

Yes, to the PTB, "experience" is meaningless. Never mind it is probably the most accurate method of valuation. The same problem of math proving that technically bumblebees cannot fly as their wind resistance and weight ratio to wing size is too high.

 

[Mike Ault]

you side stepped the question, my friend...

I continued on with that post… Feel free to cast stones, interesting discussion!

 

[alebrewer]

I continued on with that post… Feel free to cast stones, interesting discussion!

The question was: If regression is a recognized methodology (along with sensitivity), and the goal of both is to reduce the range, then why would reducing the range not be considered the goal of the exercise? And - while generally not possible - if the goal is reducing the range, then by definition a $0 range is the end goal. I'll say again, though, that it's typically not possible due to the subjective nature of RE transactions. Don't throw the baby out with the bathwater, though...

 

[Terrel L. Shields]

What if there are 20 other signficant differences

Don't overthink it. If you are making 20 adjustments you need better comps. Search more.

 

[Mike Ault]

The question was: If regression is a recognized methodology (and sensitivity), and the goal of both is to reduce the range, then why would reducing the range not be considered the goal of the exercise? And - while generally not possible - if the goal is reducing the range, then by definition a $0 range is the end goal. I'll say again, though, that it's typically not possible due to the subjective nature of RE transactions. Don't throw the baby out with the bathwater, though...

I throw the baby out with the bathwater when I sense somebody is being dishonest. And when I see a "zero dollar value range" in my market, that is prima facie evidence. Whatever they're doing, it doesn't pass muster.

 

[Terrel L. Shields]

t is up to the appraiser to decide the appropriate level of analysis

No, it is the clients who will require you to "show your work."

 

[alebrewer]

Think of it this way: If you have two comps whose only difference is an extra bath, and the comp with the extra bath sold $5k higher than the other, you'd adjust $5k, right? The goal of which is to reduce (or eliminate) the adjusted range. That is nothing more than a very simple single variate regression...

 

[alebrewer]

I throw the baby out with the bathwater when I sense somebody is being dishonest. And when I see a "zero dollar value range" in my market, that is prima facie evidence. Whatever they're doing, it doesn't pass muster.

I don't know that I'd accuse someone of being dishonest when trying to accomplish that, as motivation is typically beyond the scope of any review assignment.

 

[Mike Ault]

Think of it this way: If you have two comps whose only difference is an extra bath, and the comp with the extra bath sold $5k higher than the other, you'd adjust $5k, right? The goal of which is to reduce (or eliminate) the adjusted range. That is nothing more than a very simple single variate regression...

Of course, the unicorn "matched pair". In real life, that one with the extra bath sat on the market for 330 days at $100,000 and sold for $105,000 with a $5000 seller concession. What's the adjustment for the bath then? And to make matters worse, the list price was increased to $105,000 the day the contract was signed.

 

[J Grant]

No doubt mathematical analysis of human behavior is intrinsically incomplete due to exactly what folks are saying - buyers and sellers do not act 100% objectively - there is always subjective motivation in a transaction. Let's say you discount mathematical modeling for said behavior, though. How do you come up with an MV? I mean - as soon as you apply that first adjustment - regardless of the technique, you're using mathematical analysis to model human behavior, right? And once you've crossed that line, folks are going to want you to use modeling techniques that are supportable. And regression is more supportable than experience (IMO).

Experience lets an appraiser analyze the regression results ( and compare the results to other forms of extracting adjustments )

There can be several methods applied, with some more relevant than others depending on the property type, assignment, and available data. Quant and different methods might be more relevant