Square footage adjustments

[Caterina Platt]

I am not going to begin to state I clearly understand the regression speak. However in a rural market when you have to go upwards of 5 to 10 miles away for a sale, I'm going to strongly argue that size of a dwelling on any graph covers all the necessary influences. What you are graphing here, as you stated is a combination of all factors. To label them in one lump sum figure and attribute it to 'size' is incorrect, in my opinion. Just as in any case, we can play with the numbers all we want to justify a particular point. ONLY when the statistics are properly analyzed can we hang our hats on them. How can you defend your 'size only' adjustment?

I can take my own market and show you two relatively equal dwellings in many features. One sits on 2 acres of non-irrigated, desert land, the other on 1.5 acres of green belt property. You will see a vast difference in value and market acceptance due to the location. Size is not the issue here. Should there be a marginal difference in GLA, say 150 SF. (I'm talking 2000 SFish homes) The land value difference is $20,000. Following your example I would have a $133 per sq Ft adjustment. Say my subject is the one on the non-irrigated property. The buyer takes a look at the appraisal and quickly determines he needs to build on a den and give his new home a whopping boost in value. OOOOPS! Didn't happen did it? 'Scuse me Mr. Austin, take a look at your appraisal and the one done 1 year later on Mr. X's property. You made a $20,000 adjustment for GLA. Mr. X has added on a 400 SF den. Shouldn't his home be worth $53000 more than when he purchased last year according to your analysis? What say ye to the Court, Mr. Appraiser???'

Your football analogy is cute, but hardly comparable to real estate. A subdivision of tract homes, one with a fireplace, one without. Same pool of qualified buyers and available mortgage moneys in a given block of time. That's a comparable. That gives you a true substitution and measurable difference. The Broncos in October of '89 with Elway as opposed to the Broncos in October of '02 without Elway. Uh, let's see...completely different team, weather on any given day, new stadium, different coaches, X2 as there is an opposing team that is completely different. Nope. Don't work. Can't measure a thing. Ah, but that gives the announcers mulitudes of nonsense to talk about between plays.....wait what kind of time adjustment should I make here??? :wink: Ah forget it, I'm a real estate appraiser, not an agent for a professional football player.

 

[Fred]

Catherine,
Your earlier post points out the problems. The sample size is (wa-a-a-y) too small. And there is a big assumption that there are no other factors affecting value (so-called explanatory variables).

Using (r-squared) correlation percentages to describe three sales is a form of statistical heresy. With three sales, as long as two are close, the correlation percentage will always be sky high and probably not change regardless of what the third sale is.

Here is what I mean.

Suppose we did not know the GLA of the three sales in the example from the article. Our only size indicator is that sales 1 and 2 have three bedrooms and sale 3 has four bedrooms. Graphing sales price with the number of bedroom, the market trend line has a 96% correlation. As you point out, one should not jump to the conclusion that this one factor accounts for 96% of all variation and cease further analysis – and here is why. I changed sale 3 to five bedrooms and the trend line still has 96% correlation. In fact, I kept changing sale 3. I changed it to six bedrooms, then seven, eight, nine and ten bedrooms; and the correlation never changed from 96%. Then, I changed sale 3 from ten to one hundred to one thousand and one million bedrooms and the correlation still stayed 96%. 8)

 

[Mike Garrett RAA]

Throw this into your analysis....It was proven long long ago that most buyers cannot and do not perceive any square foot differences below 100 square feet in homes of 1500SF or more. 100 square feet could be a 10 x 10 bedroom. On a cost basis could be as much as $7,500 and from an appraiser's point of view be, say, $2500 to $3,000. Go figure?

For years the question has been do we adjust for room count (number of bedrooms), square feet, or both? Is there a functional problem with a two bedroom house of, say, 1,200 square feet or the same house with 3 smaller bedrooms? How does an AVM account for this difference? Aging populations seem to prefer the two bedroom configuration BECAUSE the master bedroom is much larger. Unless the analysis considers the demographics of the buying population...it's value conclusion is usually flawed.

 

[Austin]

You guys are applying a double standard here. Lets stick to the problem at hand, which is the very conventional appraisal theory that you swear by. I have done all of the talking up to now; so let’s hear you guys defend your methods. This appraiser took this data set, which in his rural market was the best available. He made net adjustments for sale 1 of +$14,500; sale 2 of + $6,500; and sale 3 of - $19,000. If the data is limited, then how do you justify using your methods to justify all of these adjustments? I gave you a detailed explanation to justify my determination of the sequence of adjustments and gave you a graphical running account of what I did and why I did it. Now I want to hear Steven, Caterina, or Mike Garrett show me their justification for making adjustments like the following: -$2,500 for a one story vs. an upper story; $2,500 for an age difference with a 7 year age range of 50 to 57 yrs; $4,500 for sale one 1 having no basement; small one car garage $2,000; average size 1 car $3,000; 2 car garage $5,500. This is voodoo.
Explain to me how you justify these numbers in a normal market much less a limited rural market. If your solution is matched pairing, don’t bother. You guys have the same data set I had, so lets see you have at it.

Steven: Your discussion about correlation of factors other than size like bedrooms really raises questions about how you perceive the cost approach. If you build the property in the cost approach starting with a $/sf number loaded with the factors I described above, weight it with number of square feet, then add factors like basements, garages, etc., then land, explain to me why the sales comparison approach should indicate a correlation of factors different than the correlation in the cost approach? If size matters in the cost approach then how can size not matter in the sales comparison approach? If bigger cost more in the cost approach then can’t you see that houses with more bedrooms have to be bigger? How do you expect them not to correlate? Houses with more bedrooms have more bathrooms too. Show me a cost manual that includes room count as an input. This is a common mistake people make; you are entering covariant factors like size, bathrooms, and number of rooms. Any body knows that more rooms must correlate with more size, so when you correlate number of rooms, baths, etc., to price, all you are doing is using a very sloppy vicarious variable to substitute for size, thus you are making a size adjustment multiple times in different forms. Size adjustment, adjustment for number of bedrooms, adjustment for number of bath rooms, adjustment for design and appeal, adjustments for functional utility, all have elements of size. That is why I say they are covariant variables. If what you guys say is correct, then we will have to rewrite the laws of mathematics. Mike Garrett asks how does AVM account for this multiple adjusting for the same factor numerous times, my question is: how does Mike Garrett do it? Please don’t tell me that appraising is an art story Mike. The first time I heard that story I was too young to laugh, so I just rolled over in my crib and filled my diaper up.

 

[Mike Garrett RAA]

I read you post rather quickly so maybe I missed something; however, suggesting that the appraiser compare 2 story to ranch style homes is silly...they are not comparable!

If appraising is simply a mathmatically calculation there is no need for an appraiser. You can poke fun at my "art" comment all you want but it is an art form not a science, thank you God!

You can keep your science...it doesn't fly with the market here.

 

[Caterina Platt]

Austin,

In stating that size alone does not account for all differences, I am not defending the other appraiser. Cooked, 'voodoo', PFA adjustments aren't the answer either. As Mike said, 2 story homes and Ranches aren't comparables as a rule, but when that's all the data you have to work with, that's what you have to do.

From the adjustments you noted, yes, some of those appear 'cooked'. I can't say that I could prove even with reasonable amounts of data in my market that there is a defensible adjustment for a 2 story dwelling being favorable versus a 1 story. In that particular instance, GLA alone should be the determining factor assuming no other functional or demographic issues (aging population as say with a sunbird community may actually prove negative response). Again, credit to Mike G. for the demographics point.

You speak of a correlation to the cost analysis. I would agree that at times this can be a useful tool. Can't extract a basement?? Ok, can you extract a larger storage shed or a workshop??? What is an unfinished basement for all intents and purposes?? A place to put the mechanical, laundry and storage. They are not exactly the same function of course, however you can at least have someplace to start, then tweek with a look at the cost analysis as well. ie, cost of shed vs. market return as a percentage. Can't extract any of those?? Fine, Fannie Mae states (this is hearsay from one of my appraiser/instructor buddies. Disclosure: I gotta look it up for myself but I've not personally had to fall back on it, YET) you can use a figure of 25% of the estimated, depreciated cost to construct in absence of a proven market adjustment. This is where your cost analysis kicks in.

As for the age difference between a 50 year old home and a 57 year old home, voodoo most likely unless one were updated more so than another and in that case, the comments should reflect such. At that age in equal condition, I seriously doubt anyone could prove a difference.

I am fortunate enough to have reasonable market data to work from. I would not begin to state I'm anywhere near an expert on rural situations. You asked for my methods, I give you my best available. The examples I've given are based largely on the training recieved from an AI course in difficult adjustments and extractions. Perhaps Jo Ann Meyer Stratton, Tim (Texas), RStrahn, or others who from what I understand do face many rural situations could chime in with their solutions as well. My main point was and remains, while you state correctly that adjustments shouldn't be dreamed up, you can't determine a lump sum from a small sampling of data and call it all GLA either. Both would be incorrect. Your method of graphing and determining a total amount of adjustments would not be at all lost. It would be an excellent way to help check those items you can determine a market difference for and see if anything else has been overlooked. I would, however, use more than 3 sales.

JMHO

Ok, edited after re-reading Austin's post. The way I understand your challenge Austin, you wanted one of us to try our own hand at adjustments and reconciliation from your small data set. If this were the assignment, I'd have to decline, and so should you. You and I know nothing, zip, nil about this hypothetical market. 3 sales does not give you any where near the amount of data no matter what approach you took. Rural, OK. But at least in a rural market, an appraiser should have a good chunk of history behind him/her. Some data from competing nearby communities, an idea of land values and their influences. Playing with 3 sales and a subject under contract without any background knowledge whatsoever, that is voodoo.

 

[Austin]

Ok, I have read the article that prompted this discourse, and it is absolutely amazing!!! This author used the exact same basic theory as I have outlined above, but he uses intuitive logic whereas I use math, and he goes to hell and back to get to where I started and finished. He even equalizes his equation, which he calls centralizing; he uses iterations to do least sum of the squares, which he calls finding the lowest average deviation. Then he goes through this elaborate scheme, designed solely for the purpose of creating a ruse to justify adjustments to regulators and reviewers and admits it up front, of deriving a cost per square foot number, then removing all influence from all factors except the basic housing value per square foot, then he guesses at the synergy or functional obsolescence by reducing this $/sf number by 50%, then he comes up with a purified size adjustment. Then after he makes a size adjustments, he uses voodoo methods to put back into the mix every thing he just spent three hours taking out with a net result that his final total adjustment number after everything was put back in of $36 per square foot, where as it took me about 30 seconds to find that the mathematically correct adjustment was $33 per square foot to start with. What this means for those off you who live in areas of the country where the laws of science don’t apply and you must survive on artistic analysis, “my method uses math and logic to do exactly what he attempted to do but I presented a graphical picture to described every step of the process.
You people have been arguing that size cannot account for all of these other factors. I explained above that the first thing I normally do is adjust out these factors and bring the sales close to the trend line before I equalize for size. I didn’t do it in this example because the sales were already within 6% of the trend line due to other factors. So he takes all of this other stuff out with some elaborate scheme, makes a size adjustment, then puts all of this stuff back in, and whacko, he is right back where he started. Not to mention the fact that he ignored my perfectly logical, supported, and mathematically correct method of doing all of this.

Here is what bothers me about this: This whole scheme was done to fool regulators, reviewers, but most of all FNMA, HUD, FHA, & VA. The appraisal profession as presently regulated is not concerned or adaptable to discover new methods and ideas. The regulatory and lender review process influence combined preclude that. It all about fooling somebody to keep the thought police off our backs. A sad case in point: I have been fighting this fight for about five years now. I have heard every argument put forth in this thread at least 100 times. “Appraising is an art, science doesn’t apply in my market, etc., until it is coming out of my ears. Out of all the people I have ever discussed this subject with, which included a lot of very smart people, the one person that I felt grasped the subject matter and could understand it, and based on my discussions with him, I feel is one of the most knowledgeable appraisers in the country is Tom Hilderbrandt who just had his license revoked for one year by the North Carolina Appraisal Board after an administrative law judge cleared Tom of all charges brought against him. Why? Because Tom rocked the status quo and in this country, especially in North Carolina, progress is a threat to the powers that be. The worst thing that ever happened to the appraisal profession was when the Federal Government enforced appraisal regulation thereby institutionalizing stupidity and erecting a road block to progress. God help us.

 

[Mike Garrett RAA]

Austin...

Joan of Arch was burned at the stake because she traveled to the beat of a different drummer.

So, let me see if I understand this correctly. The entire industry as it now stands is wrong and Austin and Tom are the only two people right. Of course Tom had his licensed revoked ... wonder what that means? I guess you could say one who is on the cutting edge could slip and get cut.

Until this industry says "change everything to regression analysis" I will continue to do what appraisers have done since the 1930s. Is it voodoo? Well maybe but it is "voodoo that you do so well" to quote an old song.

As a final thought and my last comment in this thread, I am a pilot. I learned to fly in the 1960s At that time the aviation industry had made tremendous strides since Wilber and Orville first pushed their flyer off the hill in Kitty Hawk. My instructor started me out learn the basics and as I progressed from student pilot to private pilot I honed my skills, learned to use things like flight directors, avionics, DME, etc. Each of those things made my job as Pilot in Command abit easier but one thing I learned...never forget the basics because some day you will need them.

That day came one night when at 2AM in the morning I had a complete systems electrical failure. No electronic instruments worked, not even lights. I flew the last hour of the flight with a flash light in my mouth so I could see the instrument panel. I used a compass for direction, something that was invented hundreds of years prior. I used the old needle, ball, and airspeed to get me home safely and when I did get to the airport I manually pumped the gear down, said a prayer, and hped the wheels were under me because I had no panel indication they were down and locked. I survived because I knew how to do it the "old fashioned way".

Good luck in your crusade to change the industry. For me, tilting at windmills is a large waste of time. I am just too busy surviving in a market that has me 15 to 20 assignments behind all the time.

 

[Austin]

Mike: If you ever fly over Virginia, how about giving me advanced warning so I can stay indoors. If the aviation industry had the FAA starting in 1903, you would still be flying the same version of airplane that Orville and Wilbur flew. Besides: “If God wanted us to fly, he would have given us wings.” I am sure the early aviators heard that more than once.
Tom and me make no claim to being the only ones doing it right, but we are reaching for the stars and want it done right. Tom may well turn out to be the Joan of Arch of the appraisal profession. They were both burned at the stake for the same reason; they were a threat to the corrupt powers that be. Unfortunately, the two of us can’t generate the power necessary to overcome the drag imposed by repressive regulations and politically appointed cronies that man the state appraisal boards whose chief aim in life is to maintain the status quo so they won’t have to get off their dead butts and learn something new.
You ask why Tom lost his license: It was because he was a threat to the status quo thought police. Your artistic method of appraising reminds me of a cartoon I saw once. Felix the cat was painting his bedroom floor and painted himself into a corner. When he realized his dilemma, he turned around and painted a picture of a door on the wall, and then he opened the door and stepped into his front yard. Now that is artistic if you ax me, and an excellent example of artistic appraising at its best.