Three days in a row. Different GLA than advertised.

[alebrewer]

Have there not been Forum members who admitted in this forum that they alter measurements so the GLA matches the tax card?

You're way better than that, my friend. Leave the extreme scenarios to the J's of the world.

 

[alebrewer]

It's always struck me as interesting that the use of public records for GLA for the subject is considered insufficient for appraisal purposes, but the use of public records for GLA for the comparables (the ones we're using to establish value) is perfectly fine. Seems inconsistent to me.

 

[DWiley]

I am sure the Freddie Mac waiver system is foolproof.

I am sure that it is not. I am not aware of any system or process that is. And that is more to my point - those who criticize the use of alternatives to traditional appraisals often use an implied assumption that the traditional appraisal is a foolproof system. When alternatives are tested, they are not evaluated to see if they are perfect, or as you said, "foolproof." They are tested and compared to the traditional appraisals process, with no expectation that either is perfect.

Ale suggests that appraiser use of tax info alone for the subject is an "extreme scenario." My years as a state investigator suggests otherwise.

 

[alebrewer]

Ale suggests that appraiser use of tax info alone for the subject is an "extreme scenario." My years as a state investigator suggests otherwise.

Ale did not suggest that. Ale suggested that using an extreme example (e.g. someone who admits to adjusting GLA to match tax records) is a ploy utilized by folks less capable than you at holding their own in a debate.

On the contrary, Ale's next post was this:
"It's always struck me as interesting that the use of public records for GLA for the subject is considered insufficient for appraisal purposes, but the use of public records for GLA for the comparables (the ones we're using to establish value) is perfectly fine. Seems inconsistent to me."

 

[J Grant]

Even USPAP says perfection is not possible.
I have said it before, no valuation method, whether a traditional appraisal, a hybrid or an alternative method such as an AVM is "foolproof." That is because market value is a concept, not a fact. Yet we deal with facts to arrive at market value. However, it is not each fact in isolation which matters, or even a factual error ( unless substantial enough to impact value opinion ) Does a borrower really care if a house is 2100 sf vs 2070 sf? Probably not. Focusing on an error of 30 sf off in measurement or other source is inane then. However, a borrower might care if the house was 1800 sf. And that would be enough to impact value.
The problem is compounded in that no measurable adjustment exists between lot size variances ( 900 sf vs 9800 sf ) or year built when similar ranges ( 1960 or 1967 ) . Yet a software program might be mechanically adjusting for such differences.
Buyers purchase properties typically for the combination of features, it ticks off the boxes for what they want ( a 2000 sf range house on a decent-sized lot in a certain area.) And then, there is often a driving factor about a property as well - such as a lake view. The buyer always wanted a lake view, is willing to pay for it, or has a boat. So the lake view might get a disproportionately larger adjustment than other factors might.

Imo, computer stats are the best at market condition adjustments but might be off in other areas - or they could support adjustments made that the comps itself on the grid show. Which method to rely on the most ( the comps on grid for extraction, statistical models, cost derived, RE agent survey) etc can vary by assignment. And more than one source of support for an adjustment is best.

If quality is the goal, the mad race for turn time needs to slow down. It is antithetical to have the time needed for analysis with the often unnecessary speed race. One day more when the validation is well ahead of what is needed to close won't matter. Breakneck speed all along the line is foolish if the stakeholders are concerned with risk. ( speed in underwriting, income verification, valuation, all of it ) - is being ratcheted up for no particular reason. RE sales are not online sneaker purchases. If a borrower can not understand that, they probably can not function well enough to make payments on a home. Same thing for refinances. If someone is so desperate that they need the money tomorrow, maybe they are not a good credit risk. That said, of course, delivering in a timely manner is the goal.

Lastly, if the GSEs;s and stakeholders are as concerned with risk as they claim, they need to address the AMC fee payment structure. The GSEs;s could simply state that any loan they back if an AMC is used must have the C and R fee from the borrower as the fee paid to the appraiser and the lender pays the AMC fee as a cost. The GSEs;s can do that, they can do whatever they want in setting loan conditions. They had no problem setting a payment condition when they let the lenders know a buyer paid RE broker fee is not considered a concession by them and thus can be placed in the seller column. The lenders can pass on the AMC fee to the borrower or absorb it as a cost. Or they can dump the AMC and form their own panel. I am well aware this action from the GSE's is highly unlikely to happen, but they and others should consider a solution to this problem if they truly believe appraisers have a valuable place in the valuation system - then they need quality people, not just people willing to take the lowest fee and then pump out whatever they can to survive. That is not the way to maintain a viable profession.

In sum, the facts matter, but by themselves or even when organized into a format and adjusted for in an AVM or by an AI system, the facts alone are not enough. An integrated view of the whole of the property, mirrors how buyers consider it, as well as the key driving factors of value ( a view, or close distance to an urban center, etc) need to be reconclied and analyzed and since people are the buyers nd sellers, people are the ones who can decipher their motivaitons and reconlice them with the condions and terms of market value.

 

[alebrewer]

In sum, the facts matter, but by themselves or even when organized into a format and adjusted for in an AVM or by an AI system, the facts alone are not enough. An integrated view of the whole of the property, mirrors how buyers consider it, as well as the key driving factors of value ( a view, or close distance to an urban center, etc) need to be reconclied and analyzed and since people are the buyers nd sellers, people are the ones who can decipher their motivaitons and reconlice them with the condions and terms of market value.

Did you steal this from Kamala, J?

 

[Sandra Koutsopoulos]

IMO, rarely is there only one absolute value for a property using cold data alone and without taking into consideration what the appraiser saw inside the property, and without closely reading the MLS info and reviewing all MLS photos of the comps. Typically, its a value RANGE which then can be analyzed by taking those non-public factors into consideration. Lack of seeing and evaluating those comp conditions is a shortcoming with the AVMs and waiver situations, but as long as the perceived equity position is large enough, they will likely be fine most of the time.

The analysis, the thinking part, takes as long as it takes to come to a supportable value, which is why I'm always particularly peeved when AMC twiddledums want to know what TIME the report will be uploaded. When I'm satisfied with the analysis and conclusion; that's when.

 

[ZZGAMAZZ]

Accuracy and honesty are what is important for a reliable report! Consistency is what is needed from report to report. That is what professional appraisers know.

I often find myself making comments like ".... the GLA of xxxxx sf per the Assessor was confirmed by the appraiser's dimensions that revealed a GLA of APPROXIMATELY xxxx sf...." [Wondering if somewhere in the report I could comment that factors I report based upon my observations are "almost" accurate...

 

[ZZGAMAZZ]

IMO, rarely is there only one absolute value for a property using cold data alone and without taking into consideration what the appraiser saw inside the property, and without closely reading the MLS info and reviewing all MLS photos of the comps. Typically, its a value RANGE which then can be analyzed by taking those non-public factors into consideration. Lack of seeing and evaluating those comp conditions is a shortcoming with the AVMs and waiver situations, but as long as the perceived equity position is large enough, they will likely be fine most of the time.

The analysis, the thinking part, takes as long as it takes to come to a supportable value, which is why I'm always particularly peeved when AMC twiddledums want to know what TIME the report will be uploaded. When I'm satisfied with the analysis and conclusion; that's when.

I wonder whether the proponents of AVM valuation take into consideration the potential error inherent in the absence of condition/location/view/functionality/etc factored against the advantages of dispensing with a real-life appraisal--whatever that is, because it's not like the client/lender saves itself the cost of anything, nor should an AVM save that much time, if appraiser's would commit to realistic, e.g. 48 hour turns...

 

[J Grant]

IMO, rarely is there only one absolute value for a property using cold data alone and without taking into consideration what the appraiser saw inside the property, and without closely reading the MLS info and reviewing all MLS photos of the comps. Typically, its a value RANGE which then can be analyzed by taking those non-public factors into consideration. Lack of seeing and evaluating those comp conditions is a shortcoming with the AVMs and waiver situations, but as long as the perceived equity position is large enough, they will likely be fine most of the time.

The analysis, the thinking part, takes as long as it takes to come to a supportable value, which is why I'm always particularly peeved when AMC twiddledums want to know what TIME the report will be uploaded. When I'm satisfied with the analysis and conclusion; that's when.

Buyers do not buy a collection of "data", they buy an entire property with its features and defects . Usually, one or two property features dominate a choice ( a new home, near a beach, whatever ) and then secondary is X and so on.

Lower price buyers are limited in choice and have a modest wish list compared to a buyer looking at expensive homes. Investor buyers have their own motivations and so on. Then market conditions and financing options affect prices - this is where an appraiser comes in, to analyze the disparate influences on price and develop that into a market value opinion. To do it in a meaningful hour does take some time - hours and not minutes, or a day or a few days vs filling out boxes in a car to rush to the next appointment. Goes without saying a fee needs to compensate adequately for the time.

That is why trying to turn an appraisal into a rote/computer-driven model does not work - or only works with limited success. That is why AVM's and computer models have confidence ranges around their value estimate - there is no universal god like market value out there.

Since the GSE;s can perform their own valuations or reviews using their own data, it acts as a check and balance on the appraisal

Likely in the near future, AI can be developed enough to be integrated into the appraisal process in a meaningful way, or used in the areas where it makes sense to use it.