The Race To Automate Appraisal Process

[ucbruin]

The answer in many cases is that the lender is not competent in the appraisal and valuation space and that managing the appraisal process and underwriting appraisals is not easy. Appraisals are by far the most difficult part of the loan file for a lender to evaluate and many of them simply do not do a great job in this area. Another issue, is unlike credit scores, debt to income ratio, serves, etc. (all of which are objectively quantifiable), the market value of a property has a large subjective component.....in most cases, unless the subject property is a complete cookie cutter, it is not at all unusual to get two competent appraisers opining a market value that is more than 10% different from each other.

Is your side of the industry pushing to have appraisers provide proof of their work?

 

[J Grant]

The answer in many cases is that the lender is not competent in the appraisal and valuation space and that managing the appraisal process and underwriting appraisals is not easy. Appraisals are by far the most difficult part of the loan file for a lender to evaluate and many of them simply do not do a great job in this area. Another issue, is unlike credit scores, debt to income ratio, serves, etc. (all of which are objectively quantifiable), the market value of a property has a large subjective component.....in most cases, unless the subject property is a complete cookie cutter, it is not at all unusual to get two competent appraisers opining a market value that is more than 10% different from each other.

Nice try but I am not buying it. Most lenders don't read an appraisal and don't care as long as the value is there...UW are trained to read an appraisal and are pretty good at it, if trained correctly. But UW does not choose the appraiser in most cases. Often large scale selection of appraiser is done by AMC, and the thing the AMC is supposed to be doing is vetting the appraiser and appraisal...but, since AMC are for profit and time is money, the AMC in many cases does not do real reviews, rather a form of pseudo review, a QC checklist or computer scan for errors.

Most lenders who choose their panel direct vet the appraiser more carefully, thus can trust the appraisers adjustments ( and comp selection, which is often the key to the appraisal value being credible...what good are perfectly supported adjustments made on lousy comp choices?) The over emphasis on adjustments shows lack of perception of what really drives the credibility of value which is comp choices, and without a desk or field review by local, competent appraiser the validity of comp choices go undetected.

For example, I signed up for a lender panel, was accepted and started getting work from them. After about a year, I got an email from them I had passed the probationary one year period where every report I submitted was reviewed, and going forward only a portion would be reviewed. I was not even aware of it!....which shows difference between a lender panel and AMC ...if so many poorly supported or lie based appraisals are working their way into the system, that is the price paid by everyone profiting from that system. (and the profit is not to the appraiser )

 

[J Grant]

It's not the low fee or the fastest turn time when it comes to the statements on how adjustments were extracted....
It's common sense....
The industry would grind to a halt (or move at a snails pace) if every appraiser reinvented the adjustment wheel for each assignment....

That is true, and impossible with time constraints. And the over emphasis on "supporting" adjustments, as if showing "more support" is the key to solving issues speaks to lack of perception of what really drives the credibility of value, which is comp choices, and without a desk or field review by local, competent appraiser the validity of comp choices go undetected. What good are well supported adjustments if applied to crappy comp choices?

 

[J Grant]

Is your side of the industry pushing to have appraisers provide proof of their work?

The problem is fast turn time does not allow an appraiser to develop adjustments well...thus the speedsters are using shortcuts of "proof", aka RA and software programs that provide the adjustments....these programs say they provide "support" for an adjustment but they really are doing the adjustment and can auto populate them into a grid or be pasted into a grid- with a nifty chart and graph to show the "support" The fact that the appraiser has not a whit of understanding about the RA is besids the point, now the client request to show "support has been met.

Time pressure, and poor client selection of appraisers due to conflict of interest around profit is the root of the problem.

Competent appraisers choose good comps and make good adjustments that are market derived , no matter what method they use. Typically they use a blend of methods...if they developed the adjustment to reflect the market and qualities of the properties relative it is easy to see on the grid and makes sense reading the comments. The adjustments would hold true among cost approach or income approach as well.

Less competent appraisers , or appraisers with an agenda to make a value can choose bad comps and apply crappy adjustments , but if they use RA / software print outs to "prove" the support then they are good to go . A client is free to choose either appraiser and if bad appraisers are working on assignments then the question has to be directed at the clients.

 

[ucbruin]

The problem is fast turn time does not allow an appraiser to develop adjustments well...thus the speedsters are using shortcuts of "proof", aka RA and software programs that provide the adjustments....these programs say they provide "support" for an adjustment but they really are doing the adjustment and can auto populate them into a grid or be pasted into a grid- with a nifty chart and graph to show the "support" The fact that the appraiser has not a whit of understanding about the RA is besids the point, now the client request to show "support has been met.

Time pressure, and poor client selection of appraisers due to conflict of interest around profit is the root of the problem.

Competent appraisers choose good comps and make good adjustments that are market derived , no matter what method they use. Typically they use a blend of methods...if they developed the adjustment to reflect the market and qualities of the properties relative it is easy to see on the grid and makes sense reading the comments. The adjustments would hold true among cost approach or income approach as well.

Less competent appraisers , or appraisers with an agenda to make a value can choose bad comps and apply crappy adjustments , but if they use RA / software print outs to "prove" the support then they are good to go . A client is free to choose either appraiser and if bad appraisers are working on assignments then the question has to be directed at the clients.

"Competent appraisers choose good comps and make good adjustments that are market derived , no matter what method they use. Typically they use a blend of methods...if they developed the adjustment to reflect the market and qualities of the properties relative it is easy to see on the grid and makes sense reading the comments. The adjustments would hold true among cost approach or income approach as well."

Of course....
And that's why, in my opinion, experience and common sense are key characteristics of the appraisal industry....

And it helps to appraise in areas of high property conformity....

 

[timd354]

Is your side of the industry pushing to have appraisers provide proof of their work?

I don't know what you mean by appraisers having proof of their work...what I do expect is a credible appraisal report with a credible explanation and support for the adjustments, which would include a summary of the technique(s) used by the appraiser to derive the amount of the adjustments, and, which in the case of unusual/atypical adjustments or property attributes and/or very large across the board adjustments, may need to include a summary of the data that supports the adjustments.

 

[RCA]

Singular property appraisal has an effective date. Assessor's property inspections may span years and doesn't always play well with as-of xx/xx/xxxx. Add a deck or inground pool without permit the day after I inspect and that record will lack that information for the next 5 years or so. Different purpose, intended use and scope. A mass appraisal model can account for this issue...a singular property model (for most uses) will be flawed without making sure the data is updated and relevant to the effective date...via MLS, for example. Sale data have a short lag-time...45 days by law in my state. Used as base for most MLS here too.

I have to wonder whether you think pulling 30 comps from a neighborhood or market area and running a regression on them constitutes a mass appraisal falling under Standard 5. That is not the case, of course. A regression can be done for the purpose only of extracting adjustments, rather than valuing the comparables. Otherwise, I am nonplussed as to why you are bringing up mass appraisal with respect to appraising a single subject property.

 

[George Hatch]

"Appraiser judgement" is at least as valid as "buyer judgement", if for no other reason than ours runs so much deeper. Even the Realtors are marketing their judgment with slogans like "Ask a Realtor".

An appraiser's judgement is the net result of all of their exposure to the many iterations of the appraisal process they have previously performed, each involving the research and analysis - to one degree or another - large amounts of raw data that eventually get refined into that set of comps that get presented on the grid. That's not nothing.

I consider the disavowal of the legitimacy and efficacy and the usage of our judgement and the subsequent effect on our work to be a problem.

 

[Artemis Fowl]

I have to wonder whether you think pulling 30 comps from a neighborhood or market area and running a regression on them constitutes a mass appraisal falling under Standard 5. That is not the case, of course. A regression can be done for the purpose only of extracting adjustments, rather than valuing the comparables. Otherwise, I am nonplussed as to why you are bringing up mass appraisal with respect to appraising a single subject property.

Same data...different purpose. What is credible for one use maybe not so much for another. Moreso taking the opportunity to help clarify what limitations can exist with PRD.

 

[Cob]

Here's the question I ALWAYS have when I hear lenders, banks, Etc are going to stop using appraisers and go with AVM's and such, who are they going to go after or blame when the loan goes into default??????