GSE Waiver & Data Collection Data

[shrubberyvaluation]

I wonder how many saw this article last week.

https://sf.freddiemac.com/articles/...aiser-the-lowdown-on-property-data-collection

Errors regarding condition ratings are 200% higher in traditional appraisal reports as compared to hybrids.
Errors involving not requiring repairs and/or inspections, in accordance with published policy, are 225% higher in traditional reports as compared to hybrids.

That latter is an especially big deal. There are many reasons why a required repair or inspection might not be called out in an appraisal report, but that type of mistake directly impacts loan risk.

Curious what the parameters were, the sample size, and how they determined there was an error or not.

 

[Joe Flacco]

One thing that amuses me is the number who seem to think that people like me are working with some some pre-ordained outcome in mind. I won't speak for anyone else, but that's just not how I work. As George is so fond of saying, the conclusions are supposed to come as the result of the analysis, not in lieu of the analysis. That is why the process is referred to as "Test and Learn."

Something new is tested and the results are compared to the results from the existing process. Unless the new thing is as good as the old thing, it is not adopted.

I wonder how many saw this article last week.

https://sf.freddiemac.com/articles/...aiser-the-lowdown-on-property-data-collection

Errors regarding condition ratings are 200% higher in traditional appraisal reports as compared to hybrids.
Errors involving not requiring repairs and/or inspections, in accordance with published policy, are 225% higher in traditional reports as compared to hybrids.

That latter is an especially big deal. There are many reasons why a required repair or inspection might not be called out in an appraisal report, but that type of mistake directly impacts loan risk.

For you, your legacy kind of is riding on the "pre-ordained outcome" working out. With the white paper, you are the architect of the whole thing. You really really really want the data to support your vision.

It's more like design the test for the conclusion you need.

As far as condition ratings, they are not perfect definitions and have a wide range. So who was the judge of what is an error and what is not? Different people would have different opinions of what is the wrong rating.

 

[CGinMN]

I wonder how many saw this article last week.

https://sf.freddiemac.com/articles/...aiser-the-lowdown-on-property-data-collection

Errors regarding condition ratings are 200% higher in traditional appraisal reports as compared to hybrids.
Errors involving not requiring repairs and/or inspections, in accordance with published policy, are 225% higher in traditional reports as compared to hybrids.

That latter is an especially big deal. There are many reasons why a required repair or inspection might not be called out in an appraisal report, but that type of mistake directly impacts loan risk.

For my thoughts on that see post 85.

One thing that amuses me is the number who seem to think that people like me are working with some some pre-ordained outcome in mind.

Why did Freddie Mac take down your 2022 research note update? Why did it not account for seller concessions as a confounding variable?

 

[CGinMN]

You want people to believe you're objective, but then you guys frame things in a totally unobjective way. Of course there are concerns and reports of data collectors saying they're appraisers to get in the door.

 

[Joe Flacco]

Even leaning on the appraisal foundation to erase qualitative adjustments three years ago so that today you can be like "adjustments by definition is quanitative" and "there is no such thing as qualitative adjustments" is so methodical and strategic.

I don't like it but I almost admire the strategy and planning of it all.

 

[CGinMN]

Think about this... All the companies hiring PDCs and hybrid appraisers have access to the same computer vision software that the GSEs use to grade quality and condition. The appraisers don't see it! Yeah right. This is how you get to the results you want. If you create the study you get to shape the outcome.

 

[djd09]

that is funny...appraiser miss defects...but become a so called data collector (home inspector) on the cheap

 

[NC Appraising]

One thing that amuses me is the number who seem to think that people like me are working with some some pre-ordained outcome in mind. I won't speak for anyone else, but that's just not how I work. As George is so fond of saying, the conclusions are supposed to come as the result of the analysis, not in lieu of the analysis. That is why the process is referred to as "Test and Learn."

Something new is tested and the results are compared to the results from the existing process. Unless the new thing is as good as the old thing, it is not adopted.

I wonder how many saw this article last week.

https://sf.freddiemac.com/articles/...aiser-the-lowdown-on-property-data-collection

Errors regarding condition ratings are 200% higher in traditional appraisal reports as compared to hybrids.
Errors involving not requiring repairs and/or inspections, in accordance with published policy, are 225% higher in traditional reports as compared to hybrids.

That latter is an especially big deal. There are many reasons why a required repair or inspection might not be called out in an appraisal report, but that type of mistake directly impacts loan risk.

Kinda unfair. Will not argue that appraisers have done this to themselves.

But...let's be honest here.

We have been trained to do it like we are doing it for 30 plus years and the gses and lenders seemed to be OK with it.

I recently appraised a home of a ex-president of a mortgage company. I always knew this but she basically said yeah, if our appraisers didn't make value they wouldn't get paid and we're removed from rotation. Pre 2008. Still happens now. Re agents and lenders...still pick and choose.

Hybrids on the other hand are new, require what 5-15 x more data points And are currently insulated from being blacklusted....?? How long will that last? Apples to oranges almost. AND they have not been brainwashed or held for ransom by the lenders like we have been.

On the Facebook forums lenders are to this day bragging on how they can select their on appraisers.

So let's give it awhile to see if the lenders will get the PCR inspectors by the you know what for killings to many deals.

I really do not have a problem with hybrids...my main issue is how the system is set up. A few will dominate and control the market.

 

[NC Appraising]

How can you do 5 a a day and verify everything. I have trouble sometimes getting information within a couple days from builders and agents.

Most have a team.

I have one right now. Three appraisers on the report. Two are in different states. Funny thing is that they left out the front view of the high voltage pl and one of the comps were located next to the tower with no adjustment.

 

[djd09]

what do you think happens to a data collector (home inspector) who finds defects on the majority of their deals...they will learn no more work