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

[Zoe]

When was "experience and feelz" ever a recognized method/technique for appraisal?

Don't get me going. Do appraisal management companies use experience and feelz?

 

[Zoe]

I feelz like I can make more money on this deal. (appraisal management company).

 

[Terrel L. Shields]

n essence, Realtors use their "experience" and feelz where an appraiser cannot.....anymore.

When was "experience and feelz" ever a recognized method/technique for appraisal?

Many courts have allowed expert testimony that was based upon the appraiser's experience. The same for the old 3 panel system where "three experts in real estate" established the value by agreement among themselves. No grid needed. No "adjustments" established. Were they 'wrong'?

Realtors are acting with a personal interest in the transaction

Even so, what percentage of appraisals are at or close to the contract price? Wouldn't it be more random if the agent's and buyers offers were random unsupported and completely devoid of "adjustments"? I mean what information is the buyer going on except whether they think the property is worth more or less than other similar property they found in the news ads, Zillow or that they have personally inspected during showings?

So, tell me. How much of analysis is the mechanical methods we apply and how much is the appraiser determining in their head what they think the right answer is, then seeking the data that "supports" that judgment. I mean which came first - the chicken or egg argument? I suspect a lot of appraisers have dealt with a complicated problem by determining the adjustment FIRST, then finding that magical "paired sales" or "depreciated cost" value that supports their assumption. And who will be the wiser? After all, no 2 appraisers will have identical adjustments even when the final result is identical or close. Was this an $80k lot, or $90k? Is this a $85/SF adjustment or $87/SF? Does the garage contribute $12,000 or only $11,000. Inquiring minds want to know.

Diamonds also are appraised for their market value

Based upon concrete and identifiable features unlike real estate where "style", "view" and even motivation of the seller is important in seeing a deal consummated.

the determination of whether the adjustments applied draw the range tighter or not.

If done correctly, it should. I mean, just because you are weighing factors you THINK are driving price does not mean the MLR won't suggest that there are other factors more important than what you think.

 

[alebrewer]

Even so, what percentage of appraisals are at or close to the contract price?

Are you arguing here for the elimination of purchase appraisals? Sounds like it...

Wouldn't it be more random if the agent's and buyers offers were random unsupported and completely devoid of "adjustments"?

I've always wondered the same thing - why is the appraiser afforded a PC for a purchase, but not an EMV for a refi? They're either swayed or they're not - and purchase/refi isn't going to change that...

I mean what information is the buyer going on except whether they think the property is worth more or less than other similar property they found in the news ads, Zillow or that they have personally inspected during showings?

You might be surprised at how much homework potential buyers do prior to a purchase, but then again - you might not.

So, tell me. How much of analysis is the mechanical methods we apply and how much is the appraiser determining in their head what they think the right answer is, then seeking the data that "supports" that judgment.

I guess it would depend on the appraiser. For me personally - I try to let the data tell me where to go without forcing my opinion onto the analysis. I'll grant, though, that none of us are able to accomplish that entirely - not even you.

I suspect a lot of appraisers have dealt with a complicated problem by determining the adjustment FIRST, then finding that magical "paired sales" or "depreciated cost" value that supports their assumption. And who will be the wiser?

I mean, any support is better than no support, right? As a reviewer, though, it's pretty easy to see through the forced adjustments.

If done correctly, it should. I mean, just because you are weighing factors you THINK are driving price does not mean the MLR won't suggest that there are other factors more important than what you think.

That's where the experience part can come into play. Of course, you can use math to determine which elements of comparison are meaningful and which aren't, but not many folks on here can perform multi-variate regression, so I'd say p-values and t-tests aren't used too frequently.

 

[Terrel L. Shields]

I have a shop building with monitor style construction. Owner does not know how much it cost because it was there when he bought the place 3 years ago, but it was supposedly only 5 years old then. But it is light tubular steel frame, insulated, with a 4" slab foundation. The sides and roofing appear to be 24 gauge or maybe 26 gauge painted steel. No defects. What was the original cost? My cost books show nothing identical to that style. Do I guess? Or, do I call my guess an "assumption" about the typical per SF cost? Then, I have zero comps with a monitor barn. Do I pretend the barn does not exist - the Sgt. Schultz adjustment. Do I value it like a more typical barn? Do I say it MUST have some sort of functional issue and arbitrarily adjust it down. These are the kinds of problems we often confront, and we have to make a judgment call. What's your call? Is it superior or inferior to another appraiser's call?

 

[alebrewer]

My cost books show nothing identical to that style. Do I guess? Or, do I call my guess an "assumption" about the typical per SF cost?

I generally just pretend it doesn't exist if I don't have what I think is sufficient data to estimate contributory value.

 

[RCA]

Does anyone add "quantifiable market-derived methods" to support their adjustments? I'm in the mortgage business, but also have my appraisal certification, and I don't see the fee appraisers including this. Personally I think paired sales analysis is only possible in theory and all appraisers have to rely on their experience for their adjustments. Do the people in DC actually believe appraisers can, and will, include quantifiable market-derived support for their adjustments?

I am in the process of putting together my final "group" of interlinked websites to support my published documentation, notes, R, Python and C++ programs for MARS regresssion, Neural Networks, Cluster Analysis, Multi-Dimensional Scaling, and other methods for objectively analyzing housing markets.

You can get a peek at https://valuationengineer.com, which is my OJS publication journal. In the menu is "Substack" - which, if you click on it, will take you to my Substack site - which has only one now, but eventually many articles that will then jump back to Quarto code and graph sites for demonstrating various valuation techniques. Right now, there is just one 'Simple R Earth Program". Click on that and then click on "Go To Simple R Earth Program". (A current more direct link would be: https://quarto.valuationengineer.com/analysis.html ). - That takes you to my Quarto website hosted in Lithuania. Then you will see a simple R program that calls the CRAN Earth package to run the Earth version of MARS on some sample data that I had Claude create (almost out of thin air). Now, it isn't the best data, and I had Claude generate (to save time) the sample hyperparameters I would never myself use (e.g. I don't use varmod.method="earth"). So, if you scroll down, you will see some neat-looking diagrams that show/display your "quantifiable market-driven methods". However, Claude seemed to think, for example, it could generate condition data - well, that rarely works. And its models have decent R2 values - but since the data was totally unrealistic, the resulting models are not realistic. That will eventually be modified with more realistic versions of everything. But you can get an idea of what the 'real thing' is.

I keep telling my son-in-law, who has a Ph.D. in GeoStatistics ( and Petro Engineering) from Stanford, that I am the world's best residential appraiser. I am. But he gives me that look of course, - as well he wouldn't know - his "specialty" is becoming a new fangled "Energy Broker" - using AI. And, whereas I use MARS, he uses advanced Kriging. Anyway, that is beside the point. And understand - I don't have any sense of bragging here - it is what it is - I am, as far as I know, the only one in the world with significant experience using MARS for advanced appraisal. I would be happy if there were others, and maybe there are. Eventually, there certainly will be.

But in any case, this is all unfolding on my websites. I have to tie in to Jupyter and some other websites, so subscribers can actually execute code online - so this will eventually become an interactive website with the most advanced valuation methods - for appraisers who can handle non-parametric statistics, programming in R, Python and C/C#/C++, neural networks (AI), matrix algebra and partial differential equations.

 

[Elliott]

I once sat in on a Federal Taking case and the MAI got on the stand and gave his value opinion. There was an attorney question,"In your report you made an adjustment for view, how do you explain that adjustment?"

MAI: "Its based on work our firm had done last year in the The Dalles, OR." No further questions.

 

[J Grant]

I doubt many appraisers literally say, "The adjustments are based on my experience" with no other explanation ( and if they do, inform them they need other means of support).

Experience is invaluable in comp selection, market analysis, making the adjustments , reconciling, etc.

 

[J Grant]

It's interesting that an appraiser has to show proof and illustrate "quantifiable, market derived methods" for adjustments when Realtors, who set the price for properties, use a CMA and their "local market expertise". Realtors provide a snapshot to the sellers of how the market/ buyers perceive the current market.

In essence, Realtors use their "experience" and feelz where an appraiser cannot.....anymore.

I've met Realtors at a lot of purchase assignments and been handed a few MLS listings of how they derived their list price. I've NEVER been handed an Excel spreadsheet of an analysis on how they derived their list price. Just sayin....

REaltors do not need to prove any rational support for their price and they are not under an obligaiotn to price at "market value." They are sales people, first and last.

I understand wanting some qualitative support as ONE indicator of adjustment, but using only that, especially from a computer software click, can produce misleading results. And downplaying appraiser experience is part of the dismantling of the value the appraisers bring to the process.