Are You Violating USPAP Every Day?
- Bias. It’s Easier to Blame the Messenger - April 20, 2022
- Hocus Pocus Magic Focus? - March 24, 2022
- Opinion, Estimate, or Prediction? - September 8, 2021
Are you in “violation” of USPAP?
If you pick comps the old way, you may be violating USPAP every day!
The old way of selecting data worked well for the old days. Gather seven or eight that look good, go look at them, make a couple of phone calls, and pick four or five for your report. That was the way I was trained. That was the way the Appraisal Institute courses taught me. When I took the AI demo report course, three words were drilled into me: Explain, Justify, and Support.
Then I took the required USPAP course. (Back then it still belonged to the AI). There was no quasi-governmental organization charged with changing the standards every year or two. Then to now, one USPAP thing never changed—the way I was told to select data pick comps. I was told to pick comps as available. I was ok. No computer to help me. No online data. Just comp books updated every three months, and listing books, and public records, and my trusty landline to call brokers and sellers.
It was Standards Rule 1-4. It said I “must collect, verify, and analyze all information necessary.” Then it went on to say I “must analyze such comparable sales data as are available.”
Again, no problem. Data was hard to get. I was taught it was only necessary to use only three or four comps. And only a few comps were available. I did learn the importance of bracketing from my trainers (it was nowhere in my appraiser education). I was diligent, and of course, I picked my necessary and available comps carefully.
Then things changed. No one noticed. MLS came on line. Income properties came online. Public records came online. All relevant sales became available. Instantly. Without thinking, I ignored the “as available” rule. But stuck to the ‘as necessary’ rule. And heck, everybody used just three comps. In fact, USPAP says I should do what my peers would do. And they all used just three or four.
So, what changed?
Today in most areas, all the sales are available. But are they necessary? Well no. All my peers use just three or four, so it is ok. But what if I want to do more than achieve credible results?
What If I want to provide more reliable results? Not just believable-credible results. What if I want to use all available comparable sales? I know that a comparable sale is one that is competitive to the subject. What if there are 14 sales which competed directly on the date of value? Should I discard all those beyond three or four?
Well, USPAP says “must” use all available competitive sales. Yet USPAP says do what my peers do. Do what your clients expect. And they all use four, not fourteen!
Are you in “violation” of USPAP?
If you use just 4 comps, in instead of 14, you may be in ‘violation’ because STANDARDS RULE 1-4(a) says analyze sales as are available. On the other hand, if you use 14, and your peers do not do that, you do not meet client expectations (as required by the Scope of Work Rule)— you may be in violation.
Is it possible, just possibly possible that I and you are forced to violate USPAP in every appraisal? And evidence can be brought against you either way.
But there is more. In a coming blog and (paid subscription) TAAR* issue, we look at why a credible/believable “standard falls below what is possible, with simple data science tools.
Footnote:
- *TAAR (The Asset Analyst Report) is a paid subscription report, which goes more deeply into topics critical to appraiser survival and prosperity in the coming “data disruption” of the entire industry.
These articles make me wonder how I have managed to stay in this business for 27 years. If you are having a good day, these articles will bring you right back to “eeyore” status-doom n gloom.
Exactly. Just ignore them. Overthinking much?
Stupid Stupid Stupid!!! State search parameters, do the search, save the search results and pick the ones that are applicable. Good God – no brainer!
George…Respectfully, this is what happens when we set aside common sense and look for ‘proof’ rather than credible support. Or, better yet reasonable replication of actual human beings actions and motivations. Respectfully again, science can’t capture all the relative characteristics and motivations. Nor is it the only proper form of analysis.
The problem in the articles analogies is that analyze is being conflated with a report.
I analyze ALL potentially available comparable sales data in my comp selection screening process. My personal bio-based computer applies 68 years of life experience; and more than 40 years of combined RE; finance and appraisal experience and screens out anywhere from dozens to thousands of possible sales in minutes.
It does this first by establishing broad comparability parameters in a public records or mls database. My preliminary analyses is that (normally) I want nothing outside of the same city; where possible I’d prefer to have sales within about 300 sf+- of my subject (more for larger properties) so that I adequately bracket my subject and initially I probably look at everything in the past year.
Once that screening analysis provides a dozen to perhaps a hundred sales, I scan over the data in seconds looking for best comparable or maybe even variable or offsetting features. Something no hardware-based computer can yet do on the fly without additional input.
Depending on my findings, I then select the BEST 3 to 6+- potential sales for a more detailed (gridded analysis). This is where I am looking for truly relevant characteristics of comparability as perceived in the market. There are no adjustments yet.
Once these are inspected from the street; and I drive around THEIR neighborhoods observing and SUBJECTIVELY analyzing (weighing relative NET impact) of the neighborhood and or location features, I review all available written data on them. This may be current mls detailed reports; prior years reports, and in recent years a simple address search online to see what pops up. I look at and consider far more specific variables than any computer can look at.
How good is the site location; site utility, site size, site influences, how fresh is the paint, roof condition, exterior, fenestration or more significant appeal features; yard and landscaping, parking options and bonuses? You get the idea.
UNASKED, my computer is also drawing up monetary comparisons of many of the subject vs comparable sales features. It does it in the background almost automatically. It estimates the cost of that brand new paint job as well as potential impact while it is doing the same for many other features. …JUST LIKE REAL BUYERS DO. It starts with qualitative comparisons which are then analyzed with respect to probable net impact. (That cheap aluminum-roof covered 40-year-old corroded patio may not infact add a penny to value let alone a regression-derived $1,000 or $3,000). My computer will also analyze more abstract influences based on life experiences that few RE analytics programs consider.
How does being close to that ski resort affect millennial motivations? What about that recreational boating lake 10 miles away? Golf course 3 miles away? Perceived great school district (which may or may not be rated as ‘great’in national standings). Driving time to main job centers? Traffic congestion in Los Angeles; San Diego or San Francisco (DC, Pittsburg, NY or Boston?).
In the end I MAY even apply regression to see what the artificial computers and software ‘think’; though it has less relevance to me than what the local agent thinks. I may find paired sales or more likely I’ll consider costs of achieving parity between comps and a subject.
In the end, my 3 (rarely 4) sales comparables will be the best and most relevant comparables available, regardless of FNMAs CU ‘suggestions’ of two-year-old comps; or regression derived adjustment and conclusions that miss 30% of all market characteristics which may not be significantly important to Big Data, but which are critically important to the couples actually buying houses.
There is no chance that I am violating USPAP at either the 4 comp end or 14 comp end of the scale. I have analyzed ALL the available data…to the extent necessary to produce credible results.
In past years it wasn’t necessary to explain the obvious or ‘prove’ that we performed fundamental appraisal steps. As professionals, it was expected. That was before we allowed every non professional parasite, bureaucrat or vested interest hustler to routinely second guess us without pushing back.
I am a highly trained professional artist that applies and uses relevant scientific techniques or derived data in the performance of my art, but in the end I do not pretend subjectivity plays no part. After all, I am dealing with humans as much as real estate rights.
I am also hired for my professional objectively applied OPINIONs derived in part from all these subjective perceptions. (PS – I promise to behave when I finally get to take one of your highly regarded courses in SoCal. I’m really looking forward to it).
Appraising is an art and a science. We don’t have to do statistical analyses on everything because we use math in our work! When I learned the practical things about the procedures we use it was also said that we needed to use common sense, finding our adjustments from interviewing agents to find say, a pool/no pool adjustment or adjustment for a three vs a two car garage. Everything is local, so school district, distance to commute to work, shopping, parks and transportation all contribute to where people buy a home; it is a place to live, not an investment! (even though some people buy homes to rent); it has utility, so there is no REAL way to prove any adjustment unless you are appraising a tract home that is near new with standard improvements and only one difference (say two fireplaces vs one), so there is a lot of hot air going on to force appraisers, intimidate appraisers (and line the pockets of those who play with numbers to impress and profit) and they create and sell us new mathematical programs all the time to “help” us get our work done.
The common sense is that we are supposed to be replicating the process of home shopping that buyers do and we see what sellers agreed to on price on the comps, so it is understanding motivations of the buyers which is something learned with experience, and in keeping with the trends, it is always a judgement call about say, how much to adjust for granite countertops and glass doors on the kitchen cabinets, but those things are all secondary to appeal and location. Besides, it is the wife/woman who is deciding what to buy within the family budget, and it really isn’t brain surgery to appraise, just put on your buyer’s hat and imitate what they do. I guarantee they don’t weigh individual characteristics or features and make adjustments! Requiring us to PROVE our adjustments would be a kin to making it all science and no art.
I agree with just about everything said by Mike above because that’s the way I work also, if I have six comps it is because I want a pending and a listing to supplement four sales; that’s about the minimum, and they MUST be comparable! Artificial Intelligence (like AVMs) and especially MLS which is a sales tool, making the home sound great to get you to look at it and there is wrong information in MLS and public records, so that’s why you need a real, trained, experienced appraiser who is human and you can’t boil it down to mathematics. One exception: statistics are good for forecasting and assessing the market trends and conditions because supply and demand are always working so we can use “bigger” data to help us opine on value because we are making an estimate, or opinion which makes our work only as good as our objectivity which sleeves rolled up, following accepted procedures and using judgement. We don’t have to “hit” the sale price if the comps don’t justify that value, particularly if trends show sales have slowed and prices are dropping…..
I say when robots start buying houses then AVMs will be adequate because then nobody gets hurt or buyer’s remorse. When I appraise homes I walk in and see how I feel, just like the buyer does – who is the typical buyer is a good question, not “How much above the standard deviation is this house from the mean or median price in this neighborhood” which is only a question for the programmer and the appraiser who has to explain the results of these excessively mathematical exercises. Putting these courses into our education are just a way to earn a living rather than appraise. These courses suggest that this is a science when it is only part science, and the most important part is to verify the data with the buyer, seller or agent, and ask them how they decided on price and what was the actual condition. Putting it into digital language for FNMA (UAD, etc) is only digitizing the data, but humans are the ones who plunk their hard earned money down and pay every month for that home, so it needs an experienced, well-educated, objective human to opine on a home’s value.
You’re on fire today! Great commentary.
Just last week some honest to god slimebag realty agent outright lied to me about contractual agreements, replacement of an old roof, and what was really happening. He told me they have an agreement for new roof in addendum, but refused to furnish that. I asked lender for it, they said none existed. List agent claimed it’s the lenders discretion what they share with the appraiser, they must not have wanted to share it. Short cut to the end when I called buyers agent, no such agreement and this guy has been trouble since minute 1 for her as well. He represented a seller whom refused to budge and list agent actually did all my repair items out of his own pocket. Reasons why he could not handle the roof. Actually I’m going to follow up with this roof cert he gave me to make sure it was not a fraudulent document.
Do you think a computer program or non licensed assistant would have been able to catch that and force this guy to be honest and comply with at least minimum standards? I don’t think so. I mention this because I think this is also where a lot of the real value for using full service licensed appraisers really lies. It’s not a perfect world and in real estate, there is another huckster and another fraudster lying in wait like just everywhere and there is really nothing you can do as a consumer or even a licensed or even non licensed worker to avoid that. It’s been 10 years since separation from loan production rules have been in place. Some of these guys have become very adept at using this hindrance to their advantage. Separation from loan production rules have clearly done more harm than good in just every way possible.
Experience matters. Licensing matters. Self proclaimed fancy titles without licensing to back, all hot air. It takes one to know one and that’s why you simply can’t trust the help these days. They’re lucky we choose to be honest and that licensed appraisers are still numerous because it’s extraordinarily easy to scam people in real estate these days.
Wow! Nice. As an aside, in my case against California’s BREA morons Investigator Schmidt actually told me that my contacting a local agent was improper procedure. This was in a highly specialized area for the purpose of obtaining market perceptions information AND a near perfect paired sale (which by time court rolled around two years later – proved to be a near exact reflection of the trend being reported).
ART and science. Not simply science.
I’d liked to hear your answer to Moron Schmidt’s application of your the local agent !
I’m sure it would have been colorful.
Appraisers are educated, experienced, and knowledgeable by LICENSE. TOOO bad many don’t use their balls in standing up to the pseudo authorities
By phone it was. In court it was simply humiliatingly devastating to him specifically and BREA generally.
He was actually PROUD that he has performed “about 500 appraisals” over his 18-year career. (About 27.7 a year for his over $100k salary). Keep in mind NONE were USPAP compliant appraisal reports – all would have been appraisal review assignments with the same quality he demonstrated in ours. Imagine being proud of doing 2.31 non-compliant reports a month!
Keep in mind they are not allowed to do outside work. Probably a good thing too. Imagine pretending any degree of competence when they have NEVER done a UAD appraisal report; and no FNMA work since HVCC or CU. Nor any other recognized, credible USPAP compliant work in nearly two decades.
Peer? My ASS!
This is an interesting article illustrating how the rapid changes in this industry may have unintended consequences. I get around the premise of this article, the need for a certain type of total market analysis, with simple whole market and segmented portion of market analysis. Also utilizing good writing skills which is a cornerstone of logical inference, to be able to prove your position. Also a primary reason why appraisers who hang their entire hat on regression analysis and fill with boilerplate and see addenda may be spinning their wheels.
By shying away from regression analysis, still doing manual research, and then still engaging in logical thoughtful intuitive comparable selection, with more robust and transparent relay of actual raw market data and not just clever interpretive graphs and such, an appraiser gets the best of both worlds. They’re reviewed the whole market and provided written analysis of that. Also reviewing a more pointed segment and detailed comparable analysis of individual data examples, clearly illustrating differences and likely probability of market reaction to the individual differences. Big difference, little difference, matching, paired, unique, it does not matter. The methodology of logical appraisal remains the same. From the shoes of a buyer, what is the fair placement in terms of price and value, the subject property most logically deserves.
It is a good question. When I testify on the witness stand. I never, as in NEVER get asked about the comps I used. I always get asked about the comps I didn’t use. When my reader is going to be for a case, should i use every sale in the neighborhood. No is the obvious answer, but I have to be ready for questioning on them.
Yes Jon. Other appraisers like to use the words cookie cutter to try and explain how suburban appraisers have it easy, but when there’s a $50,000 range in value between 30 like homes in the neighborhood (Say $750,000 to $700,000 / CA PUDs) and let’s say you are in the bottom third in value, there’s nothing like using 3 comps and thus leaving 20 higher priced “sales” to be used against you. Please explain why comps 1 through 20 were not used. Go fly a kite Mary Poppins.
Seek the truth.
True Jon, but when I’m paid $1500 to $10,000+- up front plus a very nice hourly EW rate I don’t mind explaining why every potential sale in a fifteen-mile radius wasn’t used. And, yes, some I did use has been challenged too. It goes hand in hand with “why wasn’t this used”.
In a normal non litigation appraisal (even complex) I’ve either specifying my search criteria (anything outside that is excluded by default) OR discussing a couple “almost made it” to comp” properties…never 20. If I have 20 murky alternatives, then I have not adequately defined my subject or market characteristics. Five? Maybe.
George if we all do what you suggest appraisal fees would be 1-2k+ for all that time. With the industry looking do away with us what you’re suggesting for basic lending work makes no sense. I read these gloom and doom articles all the time take it with a grain of salt, it sounds good in a classroom setting but not based on reality when a borrower is dealing with a Cmt date, rate lock, and is in jeopardy of losing jr home if dates are not adhered to.
That’s the point. We should charge and be paid for the effort, obligation, experience, education, and liability we undertake in each and every appraisal signed. The refusal from an insurance company to carry us, and the blacklisting among regulators would be a greater loose.
The borrower is not our client in a loan appraisal. An attorney-agent may be a legitimate client. Appraisers have the ability to contract with almost any one.
Be careful and identify your client