Adjustment Software

[Mile High Trout]

Check out this data..... big data say increasing or stable.....
I love it when appraisers do not consider active or pending data.....
I want to see how these computer programs derive these adjustments.
Zillow should be sued. It appears Zillow adjusts the value to what the agents list price is and then adjusts it again based on how the agent adjusts the list price. 7230 trifecta was listed at $359.900. Go to zillow to see the z estimate history for this home.
View attachment 93943
View attachment 93946

Great post NC. I'll add to this with the basic iteration of a similar concept that; 'Price is not the same thing as value'.

Not to long ago some of the big name industry publications whom masquarade as independent journalists, but are really just PR mouth pieces for big players pushed some articles describing 'appraier bias' in data sets. They so gloriously proclaimed that AVM tech should substitute manual effort human appraisers, because of the ways human appraisers value opinions 'deviated from price' to such a significant degree. They were saying that if an appraiser deviates more than less than 3%, the appraiser is clearly incompetent. Yet they remained mysteriously ignorant of how price is formed and how price is finalized through the negotiation and listing exposure period process, varied unpredictable motivations from sellers and buyers, all of that.

Check out this MC data chart the RE Colorado MLS system provides with every single data set one may pull from the system. You're observing the 'final listed price' vs the 'final sales price', with an associated percentage ratio indicator. Note the sometimes excessive deviation. And this simplified summary data does not even show the 'initial listed price' and subsequent price additions or subtractions the agents may impose on the listing. As the AVM industries lust for increased participation in the valuation space increased, there was also a concerted effort from associated corporations whom have been actively consolidating MLS platforms under fewer and fewer singular management locations to eliminate and conceal this sort of consideration. The total history of pricing adjustments is now suspiciously absent from many MLS platforms. Wouldn't want the listing agents to get caught red handed with bait and switch approaches to landing the listing by then also showcasing their tremendous, sometimes six digit market pricing reductions over relatively short periods. This type of data is your ironclad defense to 'hit the price number' arguments. The appropriate response being; Which price number specifically are you referring to? Because over the life of most listings, there is quite often one, two, three, or many more 'price numbers' imposed on any given active, under contract, or finalized sale indicator in MLS systems.

 

[OSU Beavers]

Garbage in, garbage out.

 

[Non Sequitur]

Pay attention ladies and gents, Fannie just released an in depth Industry's Guide to the New URAR (link below). So here's the pay attention part, Aloft is one of the adjustment software solutions out there, and you will notice in some of the Fannie slides "Training Materials Provided In Collaboration With Aloft" noted on the bottom right. Like it or not, this stuff is being integrated into the workflow and the days of pounding out your own excel file are all but gone.

The Industry's Guide to the New URAR

For those paying attention, Aloft's Head of Development (and instructor) is now the Lead Associate of SF Collateral Risk at Fannie. Adapt and learn, or risk falling behind.

 

[Sadie]

For those paying attention, Aloft's Head of Development (and instructor) is now the Lead Associate of SF Collateral Risk at Fannie. Adapt and learn, or risk falling behind.

Not sure how I fee about that.

 

[Non Sequitur]

Not sure how I fee about that.

She's sharp, very sharp. Had a very long conversation with her, she's impressive and passionate about the software. Where I disagree, and strongly disagree, is the idea that if it can be measured it should be accounted for with an adjustment or an explanation as to why the adjustment is 0. And what I mean by "if it can be measured", think nearly every single feature UAD 2.0 wants reported; High speed internet, front door elevation which ties into flood zone, etc. If there are differences there will be questions. That's where we are headed with UAD 2.0 and baking photo recognition, scan to sketch software, and adjustment software into the process.

 

[TerryRohrer]

She's sharp, very sharp. Had a very long conversation with her, she's impressive and passionate about the software. Where I disagree, and strongly disagree, is the idea that if it can be measured it should be accounted for with an adjustment or an explanation as to why the adjustment is 0. And what I mean by "if it can be measured", think nearly every single feature UAD 2.0 wants reported; High speed internet, front door elevation which ties into flood zone, etc. If there are differences there will be questions. That's where we are headed with UAD 2.0 and baking photo recognition, scan to sketch software, and adjustment software into the process.

Even more troubling is the notion that it (the data) can be fabricated and rated (as in Condition and Quality) and then adjusted for as if there is no nuance in those aspects. Math works better when the minutia are ignored.

 

[Cob]

She's sharp, very sharp. Had a very long conversation with her, she's impressive and passionate about the software. Where I disagree, and strongly disagree, is the idea that if it can be measured it should be accounted for with an adjustment or an explanation as to why the adjustment is 0. And what I mean by "if it can be measured", think nearly every single feature UAD 2.0 wants reported; High speed internet, front door elevation which ties into flood zone, etc. If there are differences there will be questions. That's where we are headed with UAD 2.0 and baking photo recognition, scan to sketch software, and adjustment software into the process.

I hope all of my fellow appraisers who stick around can get through this, basically residential appraisal reports are going to have to have EVERYTHING explained, even items you DON'T adjust for, I mean an explanation as to why you didn't make an adjustment, previously you would have to keep information in a work file, coming up soon, residential appraisal reports will be 100+ pages, TONS of photos of everything, an explanation as to why there's a pot or pan on the stove top in your kitchen photo.

As I've said before GOOD LUCK to everyone, I already have one foot out the door, I refuse to keep pushing this 900 pound boulder up the side of the hill, as that is what the residential appraising industry has become if you complete GSE assignments. Am I bitter??? Not in the least, to me, it's like I have been in a long term marriage or relationship and have realized that things will NEVER change for the better for me, so time to move on, I simply ignore al the spam emails I get, you know the ones, the ones that tell you to "adapt or die" "the future of appraisaing" "we'll help guide you through the new appraising of the future" yeah, you can have it.

I won't be around for it, after 23+ years, it just amazes me what residential appraising has become, I used make a very decent living at it, now I can barely make enough to pay for everything that is needed to stay in business, time to move on, like a retiring pro athlete that knows it's time to hang up the spikes, it's time for me personally anyway, to hang up the spikes.

 

[alebrewer]

Math works better when the minutia are ignored.

Models certainly do. There are diminishing returns in models just as in elements of comparison. Adding independent variables to a model tightens the model only to a certain point - beyond which the modeler is introducing more variability again.

 

[TerryRohrer]

Models certainly do. There are diminishing returns in models just as in elements of comparison. Adding independent variables to a model tightens the model only to a certain point - beyond which the modeler is introducing more variability again.

Agreed. I was more venting at the notion that a fabricated, bureaucratic, mostly subjective rating is being represented as a factual, discrete variable. I do wonder when they (GSEs) are going to provide appraisers with the chart of weights for all the amenities that make up condition and quality ratings that their software is using to gauge the "accuracy" of appraiser's opinions.

 

[Non Sequitur]

I hope all of my fellow appraisers who stick around can get through this, basically residential appraisal reports are going to have to have EVERYTHING explained, even items you DON'T adjust for, ...snip...

And don't forget, comps considered but NOT USED.

Oh what fun!