The Issue of Consistency (In work habits)

[RCA]

It doesn't.

You have to remember that this is the person that massages his data to the point where EVERY comp adjusts out to exactly the same $1. Use comps that have a sales price listed generally to the $1,000 level and he likes to claim accuracy to the $1 level based on that data. Sad, clueless, but still funny.

You don't understand the logic. However you want to breakdown the value composition of a comparable that sold for $X, then common sense is that all of your value contributions would add up to exactly $X. That is one constrain 99.9% of traditional appraisers bypass because they don't go through value contributions to calculate adjustments. That value contribution logic implies that when you calculate the feature adjustment as the difference between the feature value contribution for the subject minus the feature value contribution for a given comp, mathematical, all of the comparable adjusted sales values will be the same. That is because they are all adjusted to the same subject. It's just mathematics.

The difficult part to understand beyond that is how we get the subject value, since that is what the whole purpose of the sales grid is.

1. We create a high quality model based on the measured features, e.g. GLA, Lot Size, Room Counts, garage size and so on. If you use MARS, and are experienced in using it, you should be able to get an R2 of around 80% and a CVR2 around 60%+. That leaves you with another 20% to explain with the unmeasured or subjective features like condition, quality of construction, functional vale, aesthetics, style, etc.. Now, using simple math again, the sum of those values is the actual net sale price minus the estimated value from the MARS model. That is your second constraint. You can break that "residual" amount down however you want and it won't impact the value conclusion. That breakdown has no impact on value it is just providing the report user the reasons it is what it is.

2. The one critical point of this kind of valuation is ESTIMATING the residual for the subject, since you don't have a net sale price of that. You do that by ranking the residuals of all the comps that went into the regression, say 150 comps, from largest to smallest. Generally, if you did a good job on creating the MARS model, the appeal of the comps willl correlate with the residual value. The most appealing comps will have the highest residual; that is to say buyers are willing to pay more than what size, age or location would otherwise indicate, because the properties are generally more appealing. If you go through this exercise, and look at the rankings, you will see the correlation. So, then, take your subject and rank it in the 150 comps by find the best comps to put it between. It doesn't have to be exact, but should be within +/- 10% of its ideal location. A 10% error on that remaining 20% is 2%. So, your estimate of the subjects residual is within +/-2%, generally. In fact, in many cases, it can be very clear that it is better than that. So, this method is actually very objective, verifiable and well within +/-2% accuracy.

This is really, from my point of view, rather simple. But it appears far beyond the mental capability of even the smartest MAIs. I guess it requires MENSA level intelligence. Shocking!

Anyway, it appears the WORLD is frigging tired up putting up with asinine appraisers and is going replace them with robots far sooner than expected.

So, why the fuss?

 

[TerryRohrer]

You don't understand the logic. However you want to breakdown the value composition of a comparable that sold for $X, then common sense is that all of your value contributions would add up to exactly $X. That is one constrain 99.9% of traditional appraisers bypass because they don't go through value contributions to calculate adjustments. That value contribution logic implies that when you calculate the feature adjustment as the difference between the feature value contribution for the subject minus the feature value contribution for a given comp, mathematical, all of the comparable adjusted sales values will be the same. That is because they are all adjusted to the same subject. It's just mathematics.

The difficult part to understand beyond that is how we get the subject value, since that is what the whole purpose of the sales grid is.

1. We create a high quality model based on the measured features, e.g. GLA, Lot Size, Room Counts, garage size and so on. If you use MARS, and are experienced in using it, you should be able to get an R2 of around 80% and a CVR2 around 60%+. That leaves you with another 20% to explain with the unmeasured or subjective features like condition, quality of construction, functional vale, aesthetics, style, etc.. Now, using simple math again, the sum of those values is the actual net sale price minus the estimated value from the MARS model. That is your second constraint. You can break that "residual" amount down however you want and it won't impact the value conclusion. That breakdown has no impact on value it is just providing the report user the reasons it is what it is.

2. The one critical point of this kind of valuation is ESTIMATED the residual for the subject, since you don't have a net sale price of that. You do that by ranking the residuals of all the comps that went into the regression, say 150 comps, from largest to smallest. Generally, if you did a good job on creating the MARS model, the appeal of the comps willl correlate with the residual value. The most appealing comps will have the highest residual; that is to say buyers are willing to pay more than what size, age or location would otherwise indicate, because the properties are generally more appealing. If you go through this exercise, and look at the rankings, you will see the correlation. So, then, take your subject and rank it in the 150 comps by find the best comps to put it between. It doesn't have to be exact, but should be within +/- 10% of its ideal location. A 10% error on that remaining 20% is 2%. So, your estimate of the subjects residual is within +/-2%, generally. In fact, in many case, it can be very clear that it is better than that. So, this method is actually very objective, verifiable and well within +/-2% accuracy.

Thi is really, from my point of view, rather simple. But it appears far beyond the mental capability of even the smartest MAIs. I guess it requires MENSA level intelligence. Shocking!

Anyway, it appears the WORLD is frigging tired up putting up with asinine appraisers and is going replace them with robots far sooner than expected.

So, why the fuss?

Some just can't fathom there might be something they haven't yet learned about analyzing real estate data. There is a high probability that they are unable to learn more, for whatever reason(s).

 

[RCA]

Instead of looking at the forest, waste too much time on the trees.
Whether 0.1 or 0.001 foot measurements, the gross area will not make a difference in comparing to comps.
Concentrate on getting the best comps for best consistency.

Another dumb statement from you. Reducing accuracy leaves you with the same number trees.

Of course, some of us do increase the number of measurements as well. So, I measure all room and hallway dimensions - everything needed to make a full floor plan. What does a reviewer think when he sees your floor plan not only with measurements accurate to a hundredth of a food , but all room dimensions as well, with GLA calculated, per ANSI from the CAD - in one report and the other reports, I kid you not. are +/-1 ft accuracy and on top of everything else, they get the room layouts completely wrong. In particular the in-laws.

You know it gets complicated. And I have seen these instances where 6 appraisers were called to appraise a house with an in-law on the top of the garage and a cottage in the back yard. I saw one done by a certified residiential appraiser with a BA I think in Real Estate. Case of the children and their in-laws, with their very old grandfather as the owner wanting the highest possible value for a refinance, apparently that would flow into their pockets. The loans fell through, I don't know the details, so they sued of the appraiser involved with the failed loans. The state investigated and all of the appraisers were investigated as standard practice, including me. My client lender happened to send me one or two of the other appraisals, so I have an idea what happened. The other appraisals were inferior quality - and, well, Carmel is not an easy place to appraise.

...

 

[Fernando]

Another dumb statement from you. Reducing accuracy leaves you with the same number trees.

Of course, some of us do increase the number of measurements as well. So, I measure all room and hallway dimensions - everything needed to make a full floor plan. What does a reviewer think when he sees your floor plan not only with measurements accurate to a hundredth of a food , but all room dimensions as well, with GLA calculated, per ANSI from the CAD - in one report and the other reports, I kid you not. are +/-1 ft accuracy and on top of everything else, they get the room layouts completely wrong. In particular the in-laws.

I never waste my time measuring room and hallway dimensions and can't recall even sketching a layout with interior walls. Waste of time and clients never request it.
Imagine all the time you saved in not dealing with the silly too detailed measurements and drawing interior room layouts. So much more appraisals you could have done.

You know it gets complicated. And I have seen these instances where 6 appraisers were called to appraise a house with an in-law on the top of the garage and a cottage in the back yard. I saw one done by a certified residiential appraiser with a BA I think in Real Estate. Case of the children and their in-laws, with their very old grandfather as the owner wanting the highest possible value for a refinance, apparently that would flow into their pockets. The loans fell through, I don't know the details, so they sued of the appraiser involved with the failed loans. The state investigated and all of the appraisers were investigated as standard practice, including me. My client lender happened to send me one or two of the other appraisals, so I have an idea what happened. The other appraisals were inferior quality - and, well, Carmel is not an easy place to appraise.

...

More complicated layouts will have different gross area from appraisers. As long as the differences not too great, should be no problem.
The problem is with inlaw units which are always difficult to appraise because of lack of good comps especially in the past. Now more properties have inlaws because of tolerance from government pressure to accept them.

 

[djd09]

what is the problem...they get what they pay for...cheap and fast

 

[Fernando]

I'm finding AI is happening in my appraisals and is changing how I can do better appraisals in my investigations.
For example, I can see properties from satellite before going to subject.
I'm seeing a structure in the backyard and is asking AI to tell me what and how big it is. At this point, it's not smart enough but I see the potential. No doubt, Lenders and Insurance companies also see these observations.

 

[Tom D]

Bad appraisers don't become good appraisers with a new form. When i reviewed, the bad ones were always bad.

The fed caused more foreclosures with their inconsistent, at the wrong times, rate changes.

 

[RCA]

I never waste my time measuring room and hallway dimensions and can't recall even sketching a layout with interior walls. Waste of time and clients never request it.
Imagine all the time you saved in not dealing with the silly too detailed measurements and drawing interior room layouts. So much more appraisals you could have done.

More complicated layouts will have different gross area from appraisers. As long as the differences not too great, should be no problem.
The problem is with inlaw units which are always difficult to appraise because of lack of good comps especially in the past. Now more properties have inlaws because of tolerance from government pressure to accept them.

In my opinion you are very bad example of an appraiser. Interesting that you are the most prolific here.

You are a very bad example.

I am pretty sure you will wind up in the burning trenches of hell. Metaphorically speaking of course.

 

[Fernando]

In my opinion you are very bad example of an appraiser. Interesting that you are the most prolific here.

You are a very bad example.

I am pretty sure you will wind up in the burning trenches of hell. Metaphorically speaking of course.

Because I'm a bad example of an appraiser, I should burn in hell. I don't know where that logic came from (a jealous appraiser?).

Anyway, I'm working on this appraisal and I just put in 6 comps including listing.
I haven't see it but by choosing the best comps I can find and doing general adjustment based on absolute number and C condition adjustments even a monkey can do, I come to a narrow range value.
No need for a convoluted RCA system.
Note it's a difficult appraisal since few recent sales but Fernando was lucky having a listing just turned pending.
Tomorrow I'll look at the property and I can fine tune the adjustments to a more accurate point appraisal value.
The point is we don't need to do exact measurements, drawing detailed sketches, or doing the silliness of the UAD 3.6 form.
A good appraiser is able to find the best comps to work with and the numbers fall into place.

 

[Mark K]

You don't understand the logic. However you want to breakdown the value composition of a comparable that sold for $X, then common sense is that all of your value contributions would add up to exactly $X. That is one constrain 99.9% of traditional appraisers bypass because they don't go through value contributions to calculate adjustments. That value contribution logic implies that when you calculate the feature adjustment as the difference between the feature value contribution for the subject minus the feature value contribution for a given comp, mathematical, all of the comparable adjusted sales values will be the same. That is because they are all adjusted to the same subject. It's just mathematics.

The difficult part to understand beyond that is how we get the subject value, since that is what the whole purpose of the sales grid is.

1. We create a high quality model based on the measured features, e.g. GLA, Lot Size, Room Counts, garage size and so on. If you use MARS, and are experienced in using it, you should be able to get an R2 of around 80% and a CVR2 around 60%+. That leaves you with another 20% to explain with the unmeasured or subjective features like condition, quality of construction, functional vale, aesthetics, style, etc.. Now, using simple math again, the sum of those values is the actual net sale price minus the estimated value from the MARS model. That is your second constraint. You can break that "residual" amount down however you want and it won't impact the value conclusion. That breakdown has no impact on value it is just providing the report user the reasons it is what it is.

2. The one critical point of this kind of valuation is ESTIMATING the residual for the subject, since you don't have a net sale price of that. You do that by ranking the residuals of all the comps that went into the regression, say 150 comps, from largest to smallest. Generally, if you did a good job on creating the MARS model, the appeal of the comps willl correlate with the residual value. The most appealing comps will have the highest residual; that is to say buyers are willing to pay more than what size, age or location would otherwise indicate, because the properties are generally more appealing. If you go through this exercise, and look at the rankings, you will see the correlation. So, then, take your subject and rank it in the 150 comps by find the best comps to put it between. It doesn't have to be exact, but should be within +/- 10% of its ideal location. A 10% error on that remaining 20% is 2%. So, your estimate of the subjects residual is within +/-2%, generally. In fact, in many cases, it can be very clear that it is better than that. So, this method is actually very objective, verifiable and well within +/-2% accuracy.

This is really, from my point of view, rather simple. But it appears far beyond the mental capability of even the smartest MAIs. I guess it requires MENSA level intelligence. Shocking!

Anyway, it appears the WORLD is frigging tired up putting up with asinine appraisers and is going replace them with robots far sooner than expected.

So, why the fuss?

The clueless, arrogant condescension is palpable but nonetheless, thanks for making my point, once again, Mr. MENSA candidate. LOL!!!