Appraisal Bias

[Dale Floyd]

Respectfully, some out here would have a coronary if given an appraisal assignment we see commonly in rural Central/Western Tennessee. A point of value, or our professional opinion, is always requested but real appraisers have to dig deeper sometimes in our reporting. Regardless of what the client wants to see on one line, the appraisal is an appraisal and not just a number from the sky.

There are times in our region, that adjusted sale prices aren't as tight as we hope because of multiple factors. We deal with limited sale areas commonly, and many times it takes several sale discussions to discover the market impacts of various components belonging to the subject.

A 75 year old, two-story, log home, situated on 20 acres overlooking the Buffalo River with an easement for access that crosses two other properties.

A manufactured home that has been placed on a basement and modified with a conventional addition, a barn, and horse stables with guest quarters. (Located in a county with only 7,000 in total population.)

The 40 acre cutover property, with a tiny home situated toward a rear boundary with no public electricity available.

These are not the majority of residential properties in our areas, but they exist and we are competent to appraise them.

It's very easy to jump on a forum and declare what everybody should do all the time. But we are all in different areas, and sometimes the norm in our market might seem impossible to other professionals. In many of our cases, regardless of our expertise and experience, we ultimately reveal the arguments that exists within the value range, because they are RELEVANT. That's a word I haven't read much out here. I'm not concerned that my discussions of the range will kick me out of a job, because discussing value and the market IS MY JOB. In the majority of assignments, we ultimately conclude a single value opinion based on comparability factors which might not come down to just the gross adjustment percentages. Sometimes our final opinion deals specifically with the majority of the values shown in that range indicated, as we often use more than just three sales. Skippy does not work for me.

 

[alebrewer]

Respectfully, some out here would have a coronary if given an appraisal assignment we see commonly in rural Central/Western Tennessee. A point of value, or our professional opinion, is always requested but real appraisers have to dig deeper sometimes in our reporting. Regardless of what the client wants to see on one line, the appraisal is an appraisal and not just a number from the sky.

There are times in our region, that adjusted sale prices aren't as tight as we hope because of multiple factors. We deal with limited sale areas commonly, and many times it takes several sale discussions to discover the market impacts of various components belonging to the subject.

A 75 year old, two-story, log home, situated on 20 acres overlooking the Buffalo River with an easement for access that crosses two other properties.

A manufactured home that has been placed on a basement and modified with a conventional addition, a barn, and horse stables with guest quarters. (Located in a county with only 7,000 in total population.)

The 40 acre cutover property, with a tiny home situated toward a rear boundary with no public electricity available.

These are not the majority of residential properties in our areas, but they exist and we are competent to appraise them.

It's very easy to jump on a forum and declare what everybody should do all the time. But we are all in different areas, and sometimes the norm in our market might seem impossible to other professionals. In many of our cases, regardless of our expertise and experience, we ultimately reveal the arguments that exists within the value range, because they are RELEVANT. That's a word I haven't read much out here. I'm not concerned that my discussions of the range will kick me out of a job, because discussing value and the market IS MY JOB. In the majority of assignments, we ultimately conclude a single value opinion based on comparability factors which might not come down to just the gross adjustment percentages. Sometimes our final opinion deals specifically with the majority of the values shown in that range indicated, as we often use more than just three sales. Skippy does not work for me.

Haha! Exactly. As has been mentioned by a few on this thread - the better and more prolific the data set - the tighter the adjusted sales range. - and vice versa.

As a side note - that's typically when I would employ weighting for my reconciliation. If my 'adjusted' data don't show some kind of mathematical central tendency (mode or median), I will weight the individual adjusted values, along with an explanation of why I weighted them as I did.

 

[J Grant]

Ummm - can you provide an example of when this might happen? (assuming you're reporting on the 1004, which requires you to place all your weight on the SCA)....

I have no freaking idea, I was trying to agree with George Hatch's assertion about it. Since he picks on me for everything. I personally do not normally provide a value opinion out of the adjusted range, though on one or two rare occasions I had done it in the past on a very high quality , custom house. and it was accepted. But normally I reconcile at a point within my adjusted value range of the comps

Go ask G Hatch about it, please. He is the one that said it is done, and he called me a form monkey for not doing it , or whatever reason he had for calling me than. Maybe it is because I sign with a banana, idk

 

[J Grant]

Haha! Exactly. As has been mentioned by a few on this thread - the better and more prolific the data set - the tighter the adjusted sales range. - and vice versa.

As a side note - that's typically when I would employ weighting for my reconciliation. If my 'adjusted' data don't show some kind of mathematical central tendency (mode or median), I will weight the individual adjusted values, along with an explanation of why I weighted them as I did.

I do complex work too and the same applies as for more simple work. The question is for the OMV reconciliation what is the most probable price $ in a numerical amount we express as the value,

I personally think it is incorrect to default esp by rote, to a central tendency mathematically to a mean or median as the most probable price. Valuation is not a math problem to be solved, it is a marketability problem to be solved. What it is about the subject, when it is compared to your comps, as well as the rest of it, the quality, of it etc, what makes the subject more probable to bring the higher, lower, or midpoint of price and value? Market conditions and cost things can play a part as well

Buyers do not see any number along a point in the value range, so then neither should an appraiser. If our adjusted value range is 290k to 315k, a buyer very well knows that 310k is not the same as 295k. So we should know it too, and not just say any number along the range is as good or bad as another. It is not just a number it represents the reason the subject should transact for 310k and not 295k( For example, if 310k was our recompilation point.)

 

[alebrewer]

I personally think it is incorrect to default esp by rote, to a central tendency mathematically to a mean or median as the most probable price.

And you've provided a sufficient amount of evidence to cause me to not care what you personally think.

 

[alebrewer]

I personally think it is incorrect to default esp by rote, to a central tendency mathematically to a mean or median as the most probable price.

Quick question: Have you ever run across someone who uses a mean to reconcile? If so, they don't understand outliers and how they work...

 

[Terrel L. Shields]

I did get heated responding to the people who said it should be a value range, not a point value

I don't think anyone has advocated that a range should replace a point value, rather that a range better describes the precision that an appraiser is actually capable of.

I personally think it is incorrect to default esp by rote, to a central tendency mathematically to a mean or median as the most probable price.

When the data are equally good or equally bad why would the average not be equally as likely as any other metric.

And what about those "outlier" properties we all seem to get all too regularly? That log cabin on 10 acres when there is not a single log cabin sale within 20 miles. A geodesic dome. I remember driving 40 miles with my partner to look at a geodesic dome comp after the underwriters demanded one geodesic dome comp from her. The home was twice as large, situated on a lake on a postage stamp lot. The subject had nothing in common with the comp except the style. 100% plus adjustments - adjustments pretty much fairy tales under the best of circumstance. At least the UW didn't demand we weight that comp.

 

[J Grant]

And you've provided a sufficient amount of evidence to cause me to not care what you personally think.

You don't have to care personally, but what I am saying is the reality,of good appraisal practice - we should have a reason to reconcile were we do, and it is inane to say any number along a range is as good or bad as another. clearly, the "numbers" are price $ equivalents and if the market recognizes 290k in a value range is different from 300k, in the range then we should too.

 

[J Grant]

Quick question: Have you ever run across someone who uses a mean to reconcile? If so, they don't understand outliers and how they work...

I don't ask how others reconcile so I would not know teh answer to that. I think in some cases using a mean to reconcile might represent the most probable price per the market and property and in other cases it would not

Reconciling to a mean if rote is averaging and many caution against averaging. using a mean or median to reconcile makes it into a math solution, rather than a market based problem

The appraisal asks for a market value opinion. It does not ask for a mathematical value opinion.

 

[alebrewer]

You don't have to care personally, but what I am saying is the reality,of good appraisal practice - we should have a reason to reconcile were we do, and it is inane to say any number along a range is as good or bad as another. clearly, the "numbers" are price $ equivalents and if the market recognizes 290k in a value range is different from 300k, in the range then we should too.

Wait - are you now saying that the market itself distills property values down to the dollar as well? If so - I simply have no response to that. I cannot make my brain soften that much.