Wednesday, 27 October 2021 | The Latest Buzz for the Appraisal Industry

An Appraisal is not THE Opinion of Value

Some of you are going to need to hang up your ego hats for this one. You know who you are. I see you hanging out on the online appraisal forums. You are the ones who do not just have an opinion, but THE opinion. Don’t worry, I will let you put it back on before I am done.

As appraisers, we are often asked a question similar to this, “Why can two professionals appraise the same property, yet come up with two different values?” It is an excellent question, and one that has many variables, but I want to focus on just one aspect of the answer today; because appraisers are giving opinions of value not facts of value. There is a big difference (and the latter does not exist).

I am currently working on an assignment in a rural area. When I say ‘rural,’ I mean po-dunk. The subject property has neighbors, but they are miles apart, not feet. There were a total of 22 sales in the entire county last year. That is the kind of rural we are talking about here. Construction costs are high in this area because materials and laborers have to travel from a town 76 miles away.

The appraisal is for a new construction loan on a modular home. The house will be factory-built and trucked in to be set on the foundation which sits on 10 acres. No problem. I have been covering this area for over 15 years. I feel geographically competent. I have taken multiple classes on factory-built housing, have appraised hundreds of modular and manufactured homes over the years, have done thousands of new constructions, and feel competent for the assignment type as well.

Before I even began the assignment, I knew there were no recent, modular home comparables. Even going back 36 months yielded no results. I chose to use manufactured homes as well as stick built homes and make adjustments for each. The appraisal process was no walk in the park. I knew it wouldn’t be and charged my fee accordingly. Once finished, I felt as confident as I could given the circumstances. The problem was, I came in $116,000 lower than the cost of construction (and that didn’t include the value of the land which they already own). I braced myself for the phone calls which I knew would surely come.

Sure enough, a few days later my business line was ringing loudly. Now, this is a good client. I have done a lot of work for them, and did not wish to make them upset, but I also am not one to sell my integrity in order to keep working for a particular lender. As you can imagine, the lender is now in the unenviable position of explaining to their client that they will need to bring over $100 Grand to the closing table if they want to proceed. I understand their concern for sure, but there was not much I could do and explained that to them. It was then that they informed me that mine was the second appraisal they had had done on the property. The first one was completed six months previously and the borrowers had to back out ‘because the timing was not right.’ Uh huh. Furthermore, just a few months ago, the value of the property was approximately $60,000 higher! It still did not cover their construction costs, but it was a lot less of a spread than my opinion of value.

As I talked to the nice lady on the phone, it became very obvious very quickly that the previous appraiser had done two things differently than I. First, he did not use any manufactured homes as comparables. Second, the stick-built homes that he used were custom-made, site-builds and no adjustment was made for the difference. I was asked, quite forcefully, to explain why two appraisers could come up with such dramatically different values. My answer was simply, “Well, I cannot speak of the other appraiser’s work as I have not reviewed their report, but I can tell you that an appraisal is an opinion of value. I can only speak to my work and I believe my opinion of value is well-supported.” And I did. My work file was full of statistical samplings between manufactured homes and modular homes as well as between modular homes and site-built, custom jobs. Granted, there were no comparisons in the subject’s neighborhood as there have been no recent modular home sales, but a fair amount of data from surrounding counties showed a definitive percentage difference between manufactured and modular and modular and site-builds. In other words, in my opinion, the original appraiser was just wrong. He or she either did not know enough to look at the differences or they chose to ignore the data. Either way, we have two appraisal reports and two very different values.

This experience, once again, reminded me that appraisal is not an exact science. In many ways, what we do is scientific, but in many ways what we do is also an art. What is a deck really worth? Well, there are as many ways to answer that question as there are answers. Certainly we could agree that some ways of supporting adjustments are better than others, but in the end, the appraiser must make a decision. That decision is an opinion, but is not the only opinion. I do not know who this other appraiser is, but I do not have hard feelings toward them. They have their opinion, and I have mine. I just believe my opinion is the right one. We can use all of the fancy tools we can find to support our adjustments, but in the end, we are providing an opinion, nothing more.

Oh, and don’t forget to pick up your ego hat on the way out the door.

Have any comments or would you like to submit content of your own? Email


  1. But Dustin, it’s all so easy, all you need to do is use linear regression …didn’t you hear, it is the answer to all the appraisal questions. Your opinion is not going to be worth anything, the new wave of regression based modeling has ALL the correct answers. / sarcasm ON

    1. It may be that Big Data would show relationships that an individual appraiser would never see. Perhaps remote locations in Idaho, Montana, Oregon, and Wyoming have similar values. Perhaps there’s a good location function that could adjust for distance from the nearest city of 5,000 people. I don’t know, but I recall de Gaulle’s quote, “The cemeteries are full of indispensable men.”

  2. This is clearly a case of one appraiser who is competent and one who is not. If data is available, and even if its scarce, and if the appraiser is ethical and competent there is no reason why competing reports should not fall within 5% of each other. It is up to the lender to make an appropriate appraiser selection at the time the assignment was conveyed.

    Sadly, the two primary criteria or only for many entities that order appraisals are turn time and fee. When such criteria prevails events like those that Mr. Harris describes takes place.

    Now, should the 1st appraiser have accepted or completed the assignment?? NO. Should the client have inquired as the competency of the 1st appraiser??? ABSOLUTELY.

    I just had the rare opportunity to speak with a lender who is contemplating a loan on a property with unusual characteristics and wanted to know my experience level in general and with that property type. In this case they are compiling a panel so that compliance can make an informed yet unbiased decision.

    Communication is part of the management process. The more communication that takes place prior to the placement of an assignment the fewer incidents such as this would take place.

  3. Hey John,

    I don’t know why everyone is having such a hard time with regression. I did a sq. ft. regression with 20 recent sales, within 20% sq. footage, no brick, 3/2, no garage, exact same lot size in the same exact neighborhood. That should be a defensible calculation. Oops. Confidence Interval was .35. So I just did a Theorem Proof that Regression does not work in the real world.

    1. That is probably a superior confidence to what you can claim if regression were not used. If statistical analysis does not work in the real world neither does adjusting a hand full of pairs.

      1. The reason that the confidence level is so low is that the “unknown to the appraiser factors” ARE known to the buyer. So the 1200 sq ft home next door to the Sex Offender sold for less than the 1000 sq ft. one on the next block. The 1500 sq ft UGLY house on a downhill slope away from the road sold for 10% less than the 1300 sf one. Running regression on a local market basis will always have too few sales with too many reversed outcomes to give you a number that has an acceptable confidence score. So yes, adjusting matched pairs is just handpicking pairs to give you the outcome that you were looking for. That method is not statistically defensible either.

        1. So as for using either simple regression and matched pairs we don’t have methods that can be defended very well. That should caution any of us to be too sure about the opinions we arrive at. No wonder we appraisers so often use arrogance and intimidation in place of reason.

          1. As usual, you have a knack for seeing exactly what the point is. I am moving on to Cost Data adjusted for time with the mitigating effect of bracketed data to help eliminate the inherent error in the calculation. What do you think the effect of a 5% sf bracket is compared to a 10% bracket?

          2. Sorry. I have little (maybe no) idea what you are talking about with respect to “cost data” and the effect of bracketing it. I can imagine any such reliable factual market data would be far more difficult to find and analyze than are sales, but maybe that is just me.

            I am interested in reading your elaboration, but I have nothing to add. I’m just lost on this suggestion.

  4. I take it from your comments that the two opinions, yours and that of the other appraiser, are at this point to be given equal weight, notwithstanding your opinion of the market.

    1. In our neck of the woods modular and stick-built dwellings, often have better construction quality. Lines and saw-cuts are laser-measured, precision gluing and nailing, no high school kids hammering their first house together over the summer, etc. Modular homes often times get a bad rap, but the fit n finish is often times superior to a stick built home.

      1. Long ago when I still cared about understanding such things I ran a survey and study of HUD code manufactured housing, limited as it was by geography. I did consult studies from other communities.

        My conclusion overall was that lower sales prices are not due so much to the difference in quality of construction (the HUD code just isn’t materially different from the UBC or the IBC) as they are to the hold over stigma from the old mobile home parks. For what ever reasons, which are usually unstated, regular home mortgage lenders are almost universally reluctant to lend using solely the manufactured housing as collateral and often require the dealers to co-sign. And then of course there are the brokers and us appraisers who also compound the problem citing the market without citing its seemingly irrational reasoning.

        Modulars here seem to compare well with the common sticks, but not so for the manufactured. The market and the lenders say there is a big difference, but won’t say why.

        1. I should say my research also included engineering studies with respect to the structural soundness resulting from the various codes. The only issue the engineers found is that manufactured homes, being built on frame, have integrity related to the frame and cannot be modified without effecting structural soundness as can modular or site built homes.

  5. Justin, you are so right. As appraiser’s, I feel all kinds of relevant true facts, data, science, oh and whats that new ones “Big Data” and a “location functions’ could and should be used to support and help confirm the “true market value of a property’ when possible. However, the truth is no function, or regression analysis, no data sample can explain and define and account for all the individual impact of factors of a home such as curb appeal (two houses on the adjacent on the same side of the street can still have very different market appeal from the street or road (curb appeal), individual view (good or bad), or any deferred maintenance issue that would have occurred at the home since any last inspection. (the basement started leaking, hail damaged the roof siding and porch etc.) In 2008, when home prices declined so far, it was because the bad data fed upon itself. As the system is all based upon data, the bad data feeding the bubble happened with appraiser’s. in the system. Drywall was scarce and even cost approaches did not always show the sales data as skewed. As an appraiser if during the bubble, you did not have the experience of coming in below the contract price on a sale, and then having that contract come back to you in a few days or a week with a revision of the purchase price reduced to your contract price, then, I am sorry for you for missing that validation of your work, and as a ‘regulator’ you should question your role in the ‘bubble’.

    However, I remember orders coming in saying “no cost approach needed”. lol Many people making money off the over prices of houses did not want to see ‘that Big Data’. Lol An appraisal is a ‘micro analysis” using a limited but concentrated sample of very similar data. and approaches to value specific to the home. (Aka hopefully similar comparable sales and cost approach data if available) We all know: usually, there are factors that are barely bracketed or unique to the subject. This is why the micro concentrated experienced analysis is so important because on a project as complex and expensive as a home , it is worth a few hundred dollars to have someone with some experience, no interest in the project, and a professional standard to look at this investment in a micro statistical analysis with actual verification of conditions and market factors of the home and experience in the overall market.

    The point is this: the work is an art form and a valuable one. If the loan process goes automated without the opinion of good appraisers at least having some slow down and warning lights , the next bubble will be so much worse, and no one will see it coming; as, we all know; Goldman and Sacs and a few others did. Then the uproar and devastation will be so bad, Lehman’s will look like just the first little ducky that got pulled under never to be seen again by the monster (they know is there over leverage) in the river of history,

    The smarter ducky’s follow basic underwriting practices, push the appraisers to do better analysis, and respect the role of independent regulators in every financial process. The real lenders (not over risk takers with moral hazard is that as lender who you want to be?) will have no talk of replacing appraiser’s with a model, will respect appraisers, and know a model can not do our job!!!

    An opinion about or of anything is formed as an art form in experience, research, analysis, facts and procedure, (even if you like spinach or not) and everyone does it a little differently even under the same professional practices. CPA’s Appraisers,Lawyers, Doctors are all like this, not always the same results, not always the exact same way, The art is combining the science with experience, research, knowledge, and wisdom to get the best result no matter the profession . The ones who leave it all to numbers without judgement will reap what is sown, and you will know who did as the lack of wisdom usually leads to destruction in the business world as the “Great Recession” has shown by the names that are gone. (think of the banks in your area that have different names now if you question what I say oh great powers that be in the machine lol)

  6. All right, Dustin. You are right, it is only an opinion, your opinion. Based on years of experience and hard data(we hope). I’m impressed that you agreed that the other appraiser had an opinion. You. also, said that that appraiser had it all wrong without seeing that report. Your second opinion. The third opinion is that of the other appraiser. So, now we have three opinions. Which one is correct. We know that the unreviewed report is the least of the opinions since it is based on your bias of having done yours correctly, so the other report contains error of data, effort and opinion. But wait, dont you think that the other appraiser would feel the same about your report? I think that the attitude should be, He/she has their opinion and I have mine. In fact, neither may be the correct opinion. Yet, given the circumstances, both may be correct based on what type of opinion was being required by the lender at the time of the request. Given the scenario you have presented, I am suspicious as to why the lender did not return to the original appraiser for this report. Did they questioned the previous report? Apparently, they did not like the previous number opined. As far as John Appraiser’s admonishment regarding regression analysis, you do not have sufficient sales to plot a linear anything, so personal, unbiased experience comes into play as does you opinion. In my nearly 50 years in the business, I have based all of my opinions on what the market tells me, not the owner, lender, lawyer or anyone else. Markets are imperfect and ever changing which means my opinion of yesterday may be different today. So, yes, opinions are determined by the facts of the appraisal and the art of reading the market as of the effective date. If you are good, you’re employable, if not you should have already left the business AN THAT IS MY OPINION!

  7. I’ve always said that this is not based on science. If you stick ten of us in a room with the same data, we’ll produce ten different reports. The range of values might not be that far off, but the areas of adjustment and opinions of condition and quality most likely will be. The way it seems, the changes that the government has made over the course of the past 10 years or so has been to herd us toward a scientifically based report, verifiable by a computer program in some cold, well lit server room. This job will always include that opinion clause, and until they can get an opinion verification mode for the computer, I don’t know how they’re going to get us all on the same page. I guess with all of the “improvements” some of the software providers have introduced since the inception of the CU, the “opinion checker” might already be in the works.

  8. Shhhhhhhhh….don’t mention that obvious fact to those in charge of the Z13 project. They are convinced that if they just mine enough data from appraisers they can close in on the REAL VALUE of a property within 0.000001 cent with their mega computers. They only thing slowing them down now is the walk through. They can’t use drones (legally) and the arms and legs keep falling off of their almighty Z13 bots.


Need Help?

Our knowledge base will help you and find the answers you're looking for.