TAF and USPAP - great analysis

[TerryRohrer]

I don't have a link, and wouldn't provide it if I did. I will bet I have a copy, and won't provide that either.

I would wager I have put more time and effort toward understanding what Bert is proposing than most, and have a ways to go. But I am sure there are very significant advances in what he has been attempting to share here. That he bothers trying to share it here makes him a better man than I!

 

[Mejappz]

Please point out the post where I made any reference to paired sales (or any other specific technique).

When I look at reports performed near the same time but with a 50% difference in value conclusion for the same property (which I have done numerous times over the past 15 years), the most common finding is that one of the appraisers ignored the most similar comparables sales, choosing instead to use the sales that are very dissimilar and then failing to adjust for differences in location, quality, condition, etc. in an appropriate manner.

If one is appraising an older, non-renovated home, then it is easy to jack the value up by choosing nearby renovated homes and ignoring the fact that they need quality and condition adjustments. That is a problem in executing the methodology, not a fault of the methodology itself.

I didn't say you made a reference to paired sales analysis, but nine out of 10 appraisals I review cite paired sales as their method for extracting adjustments. It's become the de facto choice for most appraisers, even though it doesn't work effectively in the vast majority of cases.

I disagree with the widespread reliance on this method and other methods mentioned in textbooks. To illustrate my point, consider setting up a hypothetical market where you already know the subject's value and adjustments. Program it to run Monte Carlo simulations, adding some market noise (start with a conservative 5% error rate for real estate markets). Run different scenarios using traditional appraisal methods like paired sales, grouped data analysis, and cost approach. Repeat this process thousands of times—it should take about five minutes.

When you aggregate the results, you'll see that the adjustment rates are wrong over 90% of the time. These inaccurate adjustment rates can significantly impact the final value estimate. While the results might be acceptable for some newer homes or cookie-cutter condo projects, such properties are few and far between in my area.

The problem lies in the fact that these methods often fail to account for the complexity and variability inherent in most real estate markets. They assume a level of uniformity and predictability that simply doesn't exist in many cases, leading to potentially significant errors in valuation.

 

[Dublin ohio]

I don't have a link, and wouldn't provide it if I did. I will bet I have a copy, and won't provide that either.

I would wager I have put more time and effort toward understanding what Bert is proposing than most, and have a ways to go. But I am sure there are very significant advances in what he has been attempting to share here. That he bothers trying to share it here makes him a better man than I!

At least you don't belittle everybody that doesn't agree with or possibly understand your point. Bert may have the best method ever for appraising. But are the end users willing to pay for the "new and improved" method. There is significant cost in software and hardware as well as the additional time involved. Is the appraiser supposed to "eat" all of the extra costs. How many clients will be proficient enough to understand the methods. If everyone up the food chain has to be proficient in the use of these new tools to understand there use. Who needs appraisers

 

[Mejappz]

Or the site value is exactly the site value assigned by the tax assessor

Solidifi calls those extraordinary appraisers.Lol

 

[chad hampton]

Solidifi calls those extraordinary appraisers.Lol

probably 5 or so year ago, Solidifi called me up and said if I lowered my fee $50 they'd send me so much work that they would be the only client I'd ever need. They liked the professional service I provided, but they wanted it cheaper. When I declined (not just for the obvious reason, but who only wants 1 client?), they sent me a couple more jobs and then nothing since.

That's a good summary of what the business has become.

 

[TerryRohrer]

At least you don't belittle everybody that doesn't agree with or possibly understand your point. Bert may have the best method ever for appraising. But are the end users willing to pay for the "new and improved" method. There is significant cost in software and hardware as well as the additional time involved. Is the appraiser supposed to "eat" all of the extra costs. How many clients will be proficient enough to understand the methods. If everyone up the food chain has to be proficient in the use of these new tools to understand there use. Who needs appraisers

It seems many these days are asking, "who needs appraisers?" and many are answering, "Not me!"

There is a lot of fertile ground between what PAREA is teaching and what Bert is doing. That there has been no movement off of square 1, as evidenced by discussions surrounding this topic in this forum over the past 25 years, says far more about most appraisers than it does about Bert. He and others had some substantive discussions for and against prior to 15 years ago, but nothing since.

My argument has been that those who understand statistical analysis and real estate have an advantage, but that is disappearing rapidly for most because most sit on their laurels hoping what is happening around them isn't happening around them (if they can even recognize the sledge hammer hints being applied).

But the advancements are taking place, just not among this crowd. Too bad, as there is useful input that could enhance all that, but the value here is in the archives, as no mention of better methods can be broached here, much less lead to a useful discussion.

 

[DWiley]

I didn't say you made a reference to paired sales analysis, but nine out of 10 appraisals I review cite paired sales as their method for extracting adjustments. It's become the de facto choice for most appraisers, even though it doesn't work effectively in the vast majority of cases.

I disagree with the widespread reliance on this method and other methods mentioned in textbooks. To illustrate my point, consider setting up a hypothetical market where you already know the subject's value and adjustments. Program it to run Monte Carlo simulations, adding some market noise (start with a conservative 5% error rate for real estate markets). Run different scenarios using traditional appraisal methods like paired sales, grouped data analysis, and cost approach. Repeat this process thousands of times—it should take about five minutes.

When you aggregate the results, you'll see that the adjustment rates are wrong over 90% of the time. These inaccurate adjustment rates can significantly impact the final value estimate. While the results might be acceptable for some newer homes or cookie-cutter condo projects, such properties are few and far between in my area.

The problem lies in the fact that these methods often fail to account for the complexity and variability inherent in most real estate markets. They assume a level of uniformity and predictability that simply doesn't exist in many cases, leading to potentially significant errors in valuation.

Forgive the fact that I have not memorized the individual positions of all that post here. So, what method do you support?

 

[George Hatch]

An appraiser citing a method of analysis they didn't use is not an example of the methodology not working or otherwise being unusable. It's not even an example of the analysis that actually was performed being insufficient to purpose. No process can work if the appraiser doesn't work it.

I could fill in the blanks in a report and assert it went with some mode of statistical analysis without actually doing any of the work and 99.9% of the users would be unable to discern the difference. Nobody would interpret that to mean that the app doesn't work. Nor is any reviewer going to attempt to replicate that outcome to see if that's the outcome.

What a small percentage of readers might do is compare the range of unadjusted prices to the range of adjusted prices to see if the latter presented a tighter range prior to the appraiser reconciling for a value conclusion. Was that combination of adjustments effective in prompting for convergence or was there an obviously more effective combination?

 

[Terrel L. Shields]

It's become the de facto choice for most appraisers, even though it doesn't work effectively in the vast majority of cases.

I think I decided 'paired sales' weren't all they were cracked up to be about 1993...
That's when I discovered MLR and Sensitivity Analysis. SA was quicker but sometimes complex items could be estimated using MLR.

 

[Joe Flacco]

I think I decided 'paired sales' weren't all they were cracked up to be about 1993...
That's when I discovered MLR and Sensitivity Analysis. SA was quicker but sometimes complex items could be estimated using MLR.

In 1993 you probably only had like 5% of the information we have about properties today and the information was not even easily searchable.