Is the sale price the best indicator of value for appraisers?
A recent study[1] includes a graph which shows that some 90% of appraisals hit the sale price exactly, or were higher, while only some 10% were below the sale price (when the sale price is known).
Is this a bias on the part of appraisers, or is the bias the cause of the system? What could possibly cause this strong upside skew?
First, ignore the ongoing pressures from the entire ‘loan industry’ to make the loan, make the commission, make the quota, make the bonus, and look successful. Ignore the claimed purpose of the public trust (of our quasi-governmental standards and licensing quagmire).
The goal of protecting the public trust failed, and will fail again— this time with different excuses and blaming— but it will fail again.
Let’s look at some underlying economic truths and social/governmental policy. What economics and public policies come into play here? Three come to mind immediately:
- The fallacy of a supply/demand equilibrium point as the ‘true’ market price;
- The nature of actual sale prices to the “true” market price; and,
- The basic need for housing as a human need.
The market value fallacy
Market value (as a misnamed market price or value-in-exchange) does not exist! When you have 3 or 4 buyers looking at 6 or 7 available unique properties, this is not equilibrium supply/demand economics. This is called game theory. This is a ‘negotiated’ value.
The book, Theory of Games and Economic Behavior was published in 1944 by mathematician John von Neumann and economist Oskar Morgenstern. It was a text for classes I took in the 1960’s, and again in graduate-level classes in the 1980’s and 1990’s. This is core material for econometric analyses— as a fundamental model for nearly all asset assessment and risk management. Yet, game theory is still unheard of in the valuation body of knowledge.
Equilibrium points rarely exist in game theory. Richard Ratcliffe, MAI originally termed the result a “transaction zone” wherein equilibrium economics departs, and game theory is the model to be applied. Equilibrium point values rarely exist in residential/amenity ‘markets’. But equilibriums do exist in some income property market sets (which have large sample sizes).
Sale price reality
The actual contract sale price itself has two considerations, (relative to a ‘true’ market value):
- Is it the ‘cash’ equivalent? (This means equal property rights, adaptation, market level (definition), money measure, financing, motivations, and externals, among other things).
- What is the distribution of likely sale prices, given motives which fall within the range of ‘mutually competitive’?
Within the transaction zone, an equilibrium (most probable price) may or may not exist. Which is what game theory says. For the most part, any sale price within the range is just as valid as any other. The point is that in this context, when an appraiser gives an opinion, there is no real reason why the sale price should not be selected. In fact, it appears to be the most certain of alternative ‘values.’
The need for housing
When appraisers do not have the sale price as ‘loan information,’ the value opinions are quite normally (bell shaped) distributed.
Here’s the issue. If appraisers could somehow be isolated from the sale price information, about half the appraised opinions would be below the sale price. This would imply that lenders, including Fannie Mae and the other GSEs would have to decline to loan on some 50% of all proposed home purchases!
Similarly, the use of AVMs (Automated Valuation Models) will also require the decline of half the possible sales. The difference between AVM results and appraiser opinions becomes moot. Any accommodation of “below price” valuations has to be done in an underwriting or risk algorithm, not by manipulating the type of valuation.
Reality: one optimal conclusion
My underlying motive is to emphasize that if lenders, reinsurers, risk managers, and regulators truly desire to measure risk of all types— the current market price is not best predicted by algorithms alone (such as in an AVM). Also, the market price is not best served by a personal (“professional”) opinion. Neither method alone is worthy of argument. The bottom line: there is only one clear choice for the consumer, the taxpayer, regulators, bank chief appraisers, and risk managers. We have one clear plan for loan industry efficiency and long-term profitability.
The one clear choice is the process integration of market-analytic professionals, armed with appropriate computation tools, and delivery by data-stream visualization. No more and no less.
To serve the public trust. To serve the social requirement of housing for our citizens. We must optimize the blend of competent analysts and simple data science principles. We must.
[1] See Working Paper No. 17-23, Appraising Home Purchase Appraisals, by Paul S. Calem, Laurie Lambie-Hanson, and Leonard Nakamura, July, 2017. [As referenced in commercial product paper Appraisal Accuracy and Confidence Scoring, Dr. Michael Sklarz and Dr. Norman Miller, Oct 30, 2017]
May 29, 2019 @ 7:51 am
You writings are way above my level of comprehension. I am 66 years old and have been in the real estate profession for 44 years both as a PA real estate broker and a PA residential certified appraiser.(Did commercial appraisal and sales as a real estate broker/owner of my own company prior to 1990.) If I want to be frustrated I read one of your articles. Love your titles though. I always have gone from the concept of teaching as if your student (me) is a 4th grade level not an economics professor whom typically do not live in the real word. Someone who is so smart they become stupid, I have taken classes from professors in college like this.My question is, are you trying to educate or impress others with you intelligence. I would truly love to know if others are frustrated by what you are explaining. Teach me just one point but don’t make me feel intimidated. Your welcome.
June 3, 2019 @ 12:16 am
Thank you for your comment. It is always difficult to find a level of writing to fit the intended audience. My topics tend to be around modern technology, critical thinking, and the concepts of data science as they may apply to valuation and asset analytics now and in the future. Colloquial, elementary school words simply do not do. Science requires precise, clear meanings.
I realize that the most “agreeable” words are those we are used to—mirroring the way things have always been. That is easier to ‘comfort level’ as they say. Unfortunately, learning is always a challenge. No pain no gain, as they also say. Yes, 4th grade level words are suitable for elementary school students. And easy for everyone to easily understand. And little learning takes place.
Actually my experience with economics professors has been rather remarkable. They do live in a different environment. Yet I also found an attention to critical thinking, a touch of skepticism, and openness to self-examination and belief-examination too. Quite the opposite is the inertia I find in our established organizations and the resistance to change and growth in individuals.
The answer to your question is that yes, I do want to pass on the things I have learned. To impress you or others is not the goal. I have worked too long and fought enough of that resistance, inertia, and even angry fearful dismissal of new ways of considering things. My IQ is actually not that high. But I have worked hard to understand words, words with precise and useful meanings. And concepts, and even feelings which cannot be expressed properly with 4th grade words.
Feeling intimidated is personal. No one can do it to you. I can empathize. I cannot match every reader’s reading level or education level. I cannot please everyone. Neither can the college professors do it. Their job was to teach and challenge you, not placate or entertain you.
Ironically, one important aspect of my classes, especially Stats, Graphs, and Data Science1 , is to refute much of the difficult, convoluted, and even make-believe misuse of statistics and statistical terminology. The aim is simplicity, understandability, and usefulness.
Much of my motive is driven by my frustration with our professional organizations teaching outmoded, erroneous, and magical “statistics” and “advanced quantitative” methods, while neglecting the beginning and intermediate methods including words, modeling practices, and the simplest ‘statistics.’ All this in a context where “believability” is the overarching standard, rather than reliability or usefulness. All this in a context where users must make numerical risk decisions based on the believability of reviews about the believability of the underlying appraisal.
I give talks regularly all around the country entitled Fallacies, Foibles, Fun, and Future of Appraisal. The greatest fallacy is the use (misuse) of inferential statistical tools in a completely inappropriate setting (the wrong model). If you have any interest, you may want to read the American Statistical Society 2016 paper entitled “Statement on p-Values: Context, Process, and Purpose”. This paper is fairly accessible for readers at the college level. It directly addresses the grossest misuses, and pleads for organizations and individuals to stop the misuse.
I also wrote a journal article you may have missed, Regression, Critical Thinking, and the Valuation Problem Today, published in the Summer 2017 edition of the Appraisal Journal. (However, this is written at the academic journal level, and may not be as much fun . . . )
Thank you for causing me to think about and respond to some of these issues.
George Dell
May 29, 2019 @ 8:17 am
George, Hear! Hear! Thanks!
May 29, 2019 @ 10:17 am
George, those who produce these studies appear to only see the appraisals for the loans that were consummated. I don’t believe they have a source to provide them with the appraisals where the appraiser’s value came in so low the deal died. A final point using humor; a man will pay $50000 more to live across the street from his mother, but will pay $50000 less to live across the street from his mother-in law. Try graphing that.
May 30, 2019 @ 3:38 pm
Hello George,
As always, interesting perspectives … The reality of realty (as you well know) is simple, the conditions and criteria of the definitions (market value, etc.) exist within a “perfect market”, while the real estate market itself is “imperfect”, for many reasons, however mostly due to human nature (the participants, including buyer/seller/broker/lender/etc.).
Always good to read something from you. Hope all is well.
Patrick
May 31, 2019 @ 6:00 am
Over the years, I have had several individuals comment to me, to the effect that the appraiser is “cheating” and will always come in at, or close to, the sales price. I always point out to them the definition of market value.. “Most probable price given a willing & knowledgeable buyer & seller each acting in their own interests, without any undue stimulus…” The point is that if everything is above board, and going as it should, the appraiser’s opinion of value SHOULD be very close to the contract price.
When the appraiser finds their opinion of value differing greatly from the contract price, at that point it is incumbent upon the appraiser to put their detective hat on, to find out why! Is it a distress sale? Non arm’s length transaction? Is one of the parties to the transaction not knowledgeable or acting in their own interest? Or is the market changing so quickly that the historical sales are unable to quantify the change? In this age of AVM’s and Hybrid Appraisal products, we need to understand that any AVM can come up with a number, where we stay relevant is when we do the detective work, reconcile and explain the numbers that otherwise don’t make sense.
May 31, 2019 @ 6:41 am
If there is a 90% chance that an appraisal will be equal to, or exceed, the sale price in a residential sale transaction, what is the value in getting an appraisal? It’s no wonder lenders are going to cheaper bifurcated products where they can get the same results at 1/3 the price.
The fact is that in the residential mortgage world, AMCs, Mortgage Companies, Banks, etc simply do not pay adequate fees, or allow enough time, for an appraiser to do a credible job. The standard has become the cheapest and fastest appraiser gets the assignment. And, if the appraiser “kills a deal” more than a couple of times, that appraiser is gone. I contend that this mortgage world appraisal model is by design. And those who oversee the mortgage industry know exactly how this system works. They are purchasing a piece of paper to stuff in their loan file that rubber stamps their loan and is insured by the appraiser’s errors and omissions insurance policy.
In the end the mortgage industry is a sales based business. Anything or anyone that threatens their bottom line, or commissions, is weeded out. That is why I feel that mortgage appraisals today are not reallly appraisals. They are something else.
May 31, 2019 @ 7:33 am
Who sets the sales price. Is it the seller, buyer, listing agent, selling agent, appraiser or buyers and sellers in the market place? Those that think it should be the appraiser assume that the appraiser is smarter than than everyone else. If the appraisal comes in high it upsets the seller, and listing agent, Low the buyer and sales agent. If the sales price is reasonable who is the appraiser to say it is something else in most cases.
May 31, 2019 @ 2:42 pm
Appraisers do not set the market place price, buyers and sellers do. Many are aided by the real estate agent. If we are doing our job, we are listening to what the market place is telling us. We didn’t develop the marketplace,we are reporting the market as it views the subject’s place within the market. Remember, our responsibility is to the the end investor, those who are at risk, the stock holders, savers, re-insurers and the general overall, public. Licensure says that the public trusts us within certain limitations.
54 years in this business has been an eye-opening experience. The less trust, the more demands. Three similar properties which have sold are now 4 or 5. We are expected to adjust for every nook and cranny. Do you really think that the buyer (in all cases) considers individual component parts of a dwelling in the same manner that we, as appraisers, have been trained to adjust for. Is that extra space in the garage worth $1,000, 2,000 or $5,673 ( a little bit of regression). In most neighborhoods, I will not adjust for the first 200 s/f of GLA since it does not seem to make any difference. AVM’s will have an answer for that because they can, not necessarily because it is required. I don’t believe that most buyers while they consider how much they want to pay, consider all component parts of the dwelling separately. Then total those values to determine their bid.
And your Right, an argument can be made for every point within the range as being the correct value. How many times has the adjusted sales come in at the same value? I know what I want, I’m just not sure what will be acceptable to those who rely on us for honest values.
Oh and another thing Fintech?
June 1, 2019 @ 8:29 am
This is a very important discussion and really needs to drive the appraiser profession. I came from an engineering background so this is somewhat new territory. It’s also the reason many can’t make that transition because you are going from a hard science to the dismal science, ie economics. It is clear to me that the reason the system exists this way is because that is what the banks wanted. They drive the industry because they make the money. I think that the job of appraisers has always been the 10% that needs to be weeded out. We throw up a red flag when it needs to be thrown out. From completely uninformed out of town buyers or just outright cases of potential fraud we are in the best position to know when the transaction is not right.
Appraisers need to start learning how to measure and report errors if we want to be treated as experts. Engineering 101 starts with understanding error and spent tons of time on significant figures because reporting too many is misleading / wrong. For instance, the Value of a house to one significant figure could be $100,000. If we valued a house at $100,001, and all appraisers should inherently see that as wrong, but why? It’s because you’ve now implied an accuracy to SIX significant figures. You’ve implied a range, whether or not you or the readers know, that the value is between $100,002 and $100,000.
This is pretty easy and straight forward in the hard sciences. For instance in measuring the length of something. If I ask someone to measure the length of the nearest pencil on their desk what are some responses? It also depends on the scope of work and the tools / money you have available. Engineering doesn’t require paragraphs for the scope of work, they just need to know the tool you used and the inherent accuracy available using that tool.
By using experience only I could look at the pencil and say that is 5″ +/- 1″.
If I had a ruler nearby and you had the time I could measure the pencil and say it’s 5 1/2″ +/- 1/4″
Now you are not all going to get the same answers but they should almost all be close.
If someone was desperately needing to know the length of the pencil down to the very tip of the carbon you could give a much more precise measurement with perhaps an electron microscope. This is obviously much more time consuming, capital intensive and therefore expensive. You might measure the pencil to the micrometer and give an answer like 5.387″ +/- 0.001″
Then you get into statistics to describe a population of pencils. Economics is much more difficult but at least some of these principles could at least provide a base to move forward.
June 1, 2019 @ 9:55 am
Chad, Since you’re relatively new to the field, you may not have heard the discussion on “appropriate roundedness”. This states that the value(final or adjustment) can and should be rounded to the nearest appropriate number so as to not signify an exactness not normally found in this field. While this is not an avocation for “close enough for government work”, it is more reflective of the human consideration. Therefore, we shouldn’t have that $100,002 final value.
June 1, 2019 @ 2:24 pm
There is one thing that no one has mentioned as a very big factor. The Realtors involved. They overprice and over sell properties. Only interested in the commission they will earn. Buyers agents just don’t assist their clients like they should to make sure they don’t pay to high of a price. It happens all to frequently, the buyers agent sells their client that “if you want this property you had better offer more than list price”. Not doing their homework to know that the list price is way too high to begin with.
All to often, within the immediate subject development, there are more examples of sales proving the sale price is too high.
Remember the definition of market value is NOT the highest price, it is the most probable price.
Appraisers would do well to remember this and appraise properties at the most probable price and not worry about the sale price as I do.
June 3, 2019 @ 12:45 pm
I have been a Realtor for over 30 years, and an appraiser for 26 years. My immediate questions to my peers: why, why, why are we not getting together and fighting back against this mission/goal to get rid of us? Why are we not pushing back? Why are we not getting the message out that to maintain honest and fraud free real estate transactions we are a very needed entity? We need to go back to independence! AMC’s have done NOTHING but cost the consumer money. They should have thrown a few dishonest appraiser’s in jail and made examples out of them instead of painting us all with the wide brush of blame for the crash. The government backed “do-gooders” forcing banks to give loans to people who couldn’t pay the money back caused the crash, not the 99% of hard-working, honest appraisers trying to earn a living! Why are we not fighting back? Let’s start a campaign to take back our profession! Christine Farrar
September 3, 2019 @ 8:08 pm
Christine, you might want to connect with Gynell Vestal’s blog oriented toward consumers.
Also, the Evidence Based Valuation (EBV) (c) focuses on services that can be provided to clients and consumers which go beyond just “a point value opinion two weeks ago.” Our Valuemetrics.info education provides a foundation for a modern profession, designed to provide for modern needs!
March 16, 2020 @ 5:38 am
What’s missing throughout us the subjective factors, the wife wants the large kutchen, the in law suite, the finished basement with the built-in bar, the possible husband’s man cave,etc. All not measurable in the market
November 8, 2020 @ 1:26 am
Is it possible that the scope of work for a purchase appraisal is slightly different than that for a refinance appraisal? In other words, might one causal factor for the peercieved skew be the fact that, in a purchase appraisal, the client is really asking, “Is the purchase price of the subject a reasonable estimate of market value?” And, since there is no such thing as a ‘point’ in value (but rather, a range), that so long as the purchase price falls within that range, it is just as reasonable as any other point within that range?