USPAP 2006 Online

[JSmith43]

"Putting a requirement in USPAP for appraisers to develop "accurate" results can do nothing to obtain more reliable results except to create a moving target that no appraiser could even be confident in hitting -regardless of how their appraisal process is chosen or defined. "Accurate" in terms of an opinion of value is a term that lies within the eye of the beholder and can neither be proven nor disproven on an objective basis. It would not be an enforceable requirement, except to the extent that we would care to encourage the "I know a bad appraisal when I see one" crowd. I can't imagine anyone here would want that. "

Don S. might nod in agreement at this point:icon_smile:

 

[Mr Rex]

On the 7th page of posts on this topic, Roger W. broaches the idea of accuracy. 7 pages of debate, and the real question remains "What's it worth" A well worn tag line around here, says "they don't want an appraisal, they want to know how much its worth "(or similar). While I have great admiration and respect for all the posters, I worry that they live in glass tubes, unaware that the lending world has moved on to the very skippies you rail against. While you debate theory, the real world moves on. I love and respect all of you, but I wish your energy was directed outward instead of inward, these skirmishes here are but wasted time while Skip takes over our business. My 10 cents worth, and I hope I **** enough folks off to make a difference.

 

[Fred]

George,
You know, I still have a copy of “The New USPAP,” by George Hatch, 1999. It concludes,

"For the professional appraiser, this revision [the 1999 USPAP] marks an evolutionary step forward and should be welcomed by all."
​

These changes so mangled the scope rules that the 2002 ASB created a concept paper (published in Jan 2003) illustrating how USPAP connotes “improper meaning,” is “confusing” – in short, the scope rules were wired up wrong.

We already have requirements for using facts that are true and correct, and for using correct methodology

Aw come on. ‘True and correct, given the scope of research’ – is not the same as - ‘true and correct’ without a limiting condition. As to correct methodology, the AF tries to be mum on that.

the SOWR are nearly as "elastic" as Steven portrays

It’s not my portrayal. I am using the AF literature.

How many times, since the instructor’s manual was published, have I had to bring up the illustrations on page 57?
· Appraisal 1: Desktop, because the guy is just curious about his equity and doesn’t want a written report.
· Appraisal 2: Now the bank wants you to driveby to figure out the equity.
· Appraisal 3: Divorce appraisal were one spouse is trying to figure out equity.
All three can be in the same week, reach significantly different conclusion (like because only appraisal 3 finds the place has been trashed for years) – ah, but all three are “credible.” I didn’t make up those case studies and I did not come up with a standards theory, unique in the history of appraisal, that all three of those appraisals would be equally “credible” under the same one rule. It is certainly not my theory that appraisal 1 and 2, in particular, have different “intended uses.” It’s not my silly putty, George. I did everything I could (over 70 pages of responses to the scope project) to get standards that set a higher bar, in which appraisals 1 and 2 above would entail using scope exceptions.

I am not the one claiming public trust is enhanced when you appraise the same house three times in one week for three different amounts. No wonder the ASB is manic about transferring reports. In USPAP, market value vaires depending on who asks to you to estimate it (per page 57 of the USPAP instructors manual).

Take John’s comment about other professions using scope of work the way USPAP does. Do you think it’s acceptable for a doctor to have to diagnose you three times?
- The first time he just pulled the diagnosis by looking up a few symptoms on the internet.
- The second time he drive by your house but didn’t stop and go inside.
- Only the third time, did the doc give you a personall, interior inspection. Your addenda fell off because the medicine prescribed in the first diagnosis, but it was still a “credible” diagnosis, given the scope of work. Right? :icon_smile:

Like I said, if someone else has uniform standards like this, I want to see them. What we have is neither uniform, nor standard, when one silly putty rule can be stretched that far. Even plain-old “do no harm” is a higher standard.

 

[JSmith43]

Mike, I have been limiting my tag line to a SC decision that has 2nd Amendment connotations. That's much less controversial around here than posting about "comp checks", subjective vs objective value, adjusting for sales and/or financing concessions, appraising as an art vs appraising as science (or a blend).

As you can see by the post above, that "nobody wants an appraisal....just want to know what it is worth" is Steve Santora's moniker??

I can remember Steve fighting for USPAP clarity in the NAIFA chat rooms in the '90's. I already was planning my escape and didn't really pay attention to USPAP development. To my amazement, one day I was corrected by Jo Anne regarding use of Extraordinary Assumptions. I respect her posts and so I went and read USPAP...well, that part. Sure enough, I must have been daydreaming in USPAP class.

Since posting here, I have learned quite a bit about USPAP that I wouldn't have learned if I were cranking out appraisals dusk to dawn. These conversations make me want to pay attention, although this whole regulation scheme, on it's own, still almost makes me gag every time I think of the darned thing.

 

[Steve Owen]

Concerning accuracy...

... or if you prefer (and I do) reliability.

We already have requirements for using facts that are true and correct, and for using correct methodology. Whether by Departure Rule or SOWR, we also have requirements to use sufficient diligence in the development of the workproduct to fit the needs dictated by the use/user.

That's a fact!

You must be able to be credible or else you cannot do the assignment. However, within credibility there is a range of reliability. How can we determine that an appariser was reliable (accurate)? The only way to do that is within the framework of the work product and data used.

Some people seem to miss the basic fundamental point... appraisers are hired to give an opinion of value. Even when the end product sought is Anticipated Sales Price (which will be measured for accuracy by the client) it is still stated as an opinion.

Mathematicians use math. Statisticians use statistics. Appraisers should be able to use both, but must do more.

I have rarely appraised a property where it could be proven empirically that the end result, the opinon of value, was accurate or inaccurate. In the real world there is no perfect set of comps (or, at least, only rarely) and that is why AVM's are so often so far off base. Sure, with residential property, there is sometimes a possibility of good enough comparables to prove accuracy (at least from an historical perspective), and on those cookie cutter deals, AVM technology, math, or pure statistics should produce a reasonably good result. In fact, possibly better than an appraiser who is trying to push the deal to meet the requirements of his or her comp check.

But, for all the rest, a certain amount of subjectivity is necessarily going to come into play... that is what separates the appraisers from mathematicians. And, that is what the client/public wants... appraisers are needed to provide an opinion of value (whether market value or some other value definition) when an accountant or statistician cannot accurately determine a theoretical sale price (a fact) from the available data.

 

[George Hatch]

Steven,

Lemme ask you this - when was an exterior-only appraisal ever considered less than credible for certain uses?

 

[Fred]

George.

Steven, Lemme ask you this - when was an exterior-only appraisal ever considered less than credible for certain uses?

Nice riposte?
Considered by whom? Per the instructor’s manual, all three appraisals are “credible.” In the real world - the idea that reported market value varies depending on who asks the appraiser to estimate it and all estimates are in full compliance with the same one rule that governs all practice - is not credible.

Also, you say, “something less,” but the phrase “something less than” is exactly what is coming out of USPAP in July. I am open to suggestion on this. In the 2006 USPAP, is "credible;"
1) binary (0 and 1), either it is credible or not
2) variable one-sided (something ranging from 0 to 1), perfectly not credible ranging up to perfectly credible
3) variable two-sided (something between –1 and +1), perfectly incredible ranging up to perfectly credible
?
You really don’t see what I am saying? The objective or mission of every assignment – credible – is something with no fixed definition and no known scale to measure it. It’s like an archery contest, but without targets. Who wins?

This solves the enforceability problem, because there is virtually nothing left to enforce. Error is now largely comprised of defining something as necessary and not doing it or not doing it “correctly” (if there is such a thing). Borrowing Steve Os word, “opinion,” you now have to review that in your opinion the scope determination was inadequate. There is no more pairing an act or process against a fixed standard of performance to say objectively that, by definition, there is a performance shortfall. You need a supplemental standard to do that.

This is what happens when we lose a fixed or “gold standard.” To paraphrase Nixon, they are all credible now.

---------------------
Steve O
BTW, I mentioned reliability AND accuracy together, which are similar or even overlapping concepts. I notice everyone who attacked the idea of standards recognizing these things parsed off “accuracy” to attack that one by itself.

I have rarely appraised a property where it could be proven empirically that the end result, the opinon of value, was accurate or inaccurate. In the real world there is no perfect set of comps (or, at least, only rarely) and that is why AVM's are so often so far off.

Both of us disagree with the first sentence. You do it in the second sentence. If there is no way to prove “accurate or inaccurate,” then AVM’s are never “off." - and Skippy never inflates value.

Mike

While I have great admiration and respect for all the posters, I worry that they live in glass tubes, unaware that the lending world has moved on to the very skippies you rail against

I don’t miss that. It’s the main thrust of what I am posting about. With “one-rule-fits-all” standards, and a near-zero minimum requirement, what makes Skippy’s appraisal inflation wrong in USPAP? The rule against bias? Given that many would argue there is no “proving” appraisal accuracy and reliability, then there is no proving that Skippy is systematically “high” (biased).

 

[JSmith43]

"The objective or mission of every assignment – credible – is something with no fixed definition and no known scale to measure it. It’s like an archery contest, but without targets. Who wins? "

For a minute I thought I was on a public schools forum (lesson planning?). But then came the weapons analogy and I realized this couldn't be true. It is the appraiser's forum and the new USPAP is being discussed.:icon_smile:

 

[Steve Owen]

Good point

Both of us disagree with the first sentence. You do it in the second sentence. If there is no way to prove “accurate or inaccurate,” then AVM’s are never “off." - and Skippy never inflates value.

And, that is a good point. However, just because a second appraiser has an opinon that the first appraiser is not accurate or that an AVM is off does not mean that such a thing can be accurately measured.

if you go back to what Dennis said about reviewers and how they do reviews you will have the answer. The real question is not whether appraisers (and AVM's) should be accurate/reliable... the question was whether you can measure it. I contend that you can only measure an appraisal's accuracy/reliability through the review standards. That is, if the appraiser uses all the best data, being that most similar, proximate, and recent. And, if s/he uses all the best techniques... that being that approaches or techniques that his or her competent peers would use. Then, the appraisal and appraisal process can be confirmed to be credible. However, I have always felt the reliablility should be determined and opined by the appraiser who did the appraisal. I have often put language in an appraisal similar to "because (reason) the (whichever one) approach is deemed to be much less reliable than if (reason had not existed)." If you have to say similar things about all the approaches, then it is obvious that the end result itself might be not very reliable. But, that does not necessarily mean that it is not credible, or even that it is not the best job anyone could have concievably done.

So... how do you determine that an AVM is off? Basically by looking at the property... unless its a cookie cutter with lots of similar data it's a safe bet that the AVM is off. :icon_smile:

So... how would you measure reliability or accuracy, Steven? Before you answer consider this scenario. An appraiser appraises a house and comes to an opinon of market value of $100k. Six months later, the house sells for $85k. Does that prove that the appraiser was not accurate? Only if you can prove that the house sold at market value and that the market has not changed in six months.

This actual thing happened to me. I appraised a commercial building at about $150k. A few months later, the owner of the building asked me to appraise her house. While I was there I mentioned that I had not seen the building listed in MLS, having known the purpose of the appraisal was to list for sale. She said she had sold it for $100k. I was astounded, why so little. She said "Well, they came along and asked me to buy it, they had the cash, I didn't have to do anything, so it seemed like a good deal to me."

That is the reason you cannot measure accuracy very well... or possibly at all. Because the market is not perfect. People do dumb things all the time... often believing them to be perfectly logical.

You can measure results. The relo people do it all the time. That is similar to measuring accuracy, but even the relo people recognize that just because the results are not good for a single deal does not mean that the appraiser is not accurate overall.

 

[Fred]

Steve O.
I appreciate your thoughtful contribution. Just a couple of things.

So... how would you measure reliability or accuracy, Steven? Before you answer consider this scenario. An appraiser appraises a house and comes to an opinon of market value of $100k. Six months later, the house sells for $85k. Does that prove that the appraiser was not accurate?

I surely didn’t suggest employing that or any other type of fallacious reasoning. And I did answer that same question from John already – concurring with Austin’s mention of residual analysis and adding that the IAAO has produced a lot good work in this area.

Also, don’t lose sight of the main point. Which tools and means are necessary to measure appraisal quality, reliability and accuracy is secondary to the issue that our primary standards don’t explicitly require quality, reliability and accuracy. We don’t have to solve the problem, because it is not a problem that we would have – unless some outside party wants to apply standards that are beyond the SOWR or Std 1.

that the end result itself might be not very reliable. But, that does not necessarily mean that it is not credible,

That is the point. Take that case study from the instructor’s manual I mentioned. Suppose you are hired to review all three. Suppose, you reasonably conclude that appraisal 1 and 2 are inaccurate and unreliable. So what!! That doesn’t make them substandard. They’re “credible.”


So, it follows then that when you say,

its a cookie cutter with lots of similar data it's a safe bet that the AVM is off.

So what!!! It’s still “credible” – given the scope of work. :icon_smile: It is possible to measure appraiser error, but there is just no reason to operating under standards that don't require anyone to get it right in the first place.