Blog #99: “In an Appraisal, Ranking Beats Bracketing All Over the Place”

There has been a lot written in the real estate appraisal press and blogosphere of late about an appraiser’s need to support his/her adjustments.  Such admonitions can be found in peer-reviewed journals, as well as in other publications more accessible to the practicing appraiser.  The premise of these admonitions is that market support exists (somewhere) for these adjustments.  Since the premise assumes such support exists, it presumes the appraiser merely must go out and find it. Generally, this is possible assuming the appraiser has the time, resources, and initiative to chase down the necessary support.  There, however, is an irony in this line of reasoning.  That irony is USPAP, the “bedrock” of all things pertaining to real estate appraisal, does not even mention the words “adjustment” or “adjustments” in Standards One or Two (but it clearly requires support for any and all adjustments and conclusions). 

Given this irony, one has reason to wonder just what is the premise on which the need for adjustments has its base?  In logic, one reasons to a conclusion from a premise.  Therefore, for a conclusion to be worthy of belief, its premise, its foundation, has to be built on the bedrock of proof (or, at least, support), not the sands of conjecture, or the muck of mere assumption.

So, if the conclusion is that appraisers must adjust comparable sales (as just one example) up or down to equality with the subject, where is the premise, and what is the line of reasoning, that such equality is uniquely, exclusively, and to the rejection of all other methodologies and protocols, the result of dollar and/or percentage adjustments?  Please understand, this is not to advocate such adjustments are improper or unnecessary.  Rather, it is to advocate there are other appropriate ways to bring comparable sales into equality with the subject than solely via the methodologies of dollar and/or percentage adjustments.  Further, via these other methods, it is possible to form a credible value conclusion.  Indeed the value indication(s) from these other methods is (are) just as likely to model the subject’s market value as is the process of adjustments. 

One of those other valuation methodologies is ranking, which is a form of bracketing (which most lenders and their AMCs – and way too many appraisers – don’t understand).  Ranking can be a valid alternative valuation protocol whether it is applied with or without adjustments.  The remainder of this monograph assumes the appraiser uses ranking post-adjustments.  The reason for this is that most residential appraisal assignments, since they are for secondary lenders such as Fannie Mae, et al, require dollar adjustments.

It is clear from the first three steps in the valuation process[1] that much of an appraiser’s preliminary efforts in an appraisal assignment go toward understanding the subject and its physical, legal, and economic characteristics[2].  The reasoning for this is clear:  before the appraiser can attempt to bring the comparables into equality with the subject (thus to synthesize an unknown market value from known market prices), the appraiser must first understand the subject well enough to accomplish this. More or less simultaneously, the appraiser also collects, verifies, and analyzes[3] general market data as well as the comparable property data.  Only after taking these steps is the appraiser properly prepared to analyze the demand and supply dynamic as the market currently interprets it, and how that dynamic affects the subject, its market value, and its marketability in that market.

At this point in the process, it is acceptable for the appraiser to form a preliminary opinion of the subject’s highest and best use as if vacant (if that is both appropriate and necessary), as well as its highest and best use as improved.  It is equally appropriate to ask the question:  if the appraiser’s opinion(s) of the subject’s highest and best use is (are) a conclusion, what are the premises and line of reasoning behind it (them)? In other words, as CU, USPAP, clients, and state appraisal boards all want to know, what is the appraiser’s support for this (these) conclusion(s)?  Or, stated differently, from what credible[4]^,[5] premise, and via what credible line of reasoning, did the appraiser reach the equally credible highest and best use conclusion(s), as well as the credible value conclusion(s) that result from a credible highest and best use analysis?  These two separate acts of (A) having a credible premise and credible line of reasoning, and then (B) explaining why these lead to a credible highest and best use and value conclusions, is what constitutes the support for those conclusions (since reasoning requires support).

Now let’s take this model of

 “credible premise à credible line of reasoning à credible conclusion(s)”

and use it in a specific example. The following is from an article in WorkingRE Magazine (citation by reference) by USPAP instructors Rachel Massey, SRA and Timothy C. Andersen, MAI, relative to the use of ranking in a residential appraisal and report.  First comes the premise:

“The appraisal has provided three closed sales within close proximity to the subject and within a recent period, including an inferior sale, superior sale, and one that is generally similar”.

The purpose of the premise is to “set the stage”.  It is the base or foundation of the line of reasoning from which the conclusions eventually flow.  In this premise, the appraiser makes it clear that the three sales s/he chose from all of those available are meant to bracket the subject.  Note that even though these sales “bracket” it, they are nonetheless comparable sales.  The inference is that, despite one superior sale and one inferior sale, the three sales are competitive with each other, as well as the subject; are viable analogues for the subject; and have the same (or highly similar) highest and best use as does the subject.  If these are not the case, then it is necessary to choose three (or more) other comps.  The credibility of this premise has already been established when the appraiser analyzed/described each of these comps earlier in the report and, as part of that analysis and description, demonstrated how/why these three were the best comparable sales there were.

After the credible premise follows the credible line of reasoning from that premise (credible premise à credible line of reasoning).  Please consider this line of reasoning:

“The inferior sale closed for $350,000, [therefore] providing the logical lower end of the value range, whereas the superior sale closed for $450,000, [therefore] providing a logical benchmark for the high end of the range.  The most similar sale closed for $400,000 and is also the most recent of the comparable sales.  This indicates that the value of the subject should be within a close range of that similar property.  After applying adjustments to the units of comparison that the appraiser considers most relevant, namely site size, condition of each property in comparison to the subject, gross living area differences, basement finish and other amenities, the adjusted sales price range has narrowed from a low of $380,000 to a high of $400,000”. 

This line of reasoning takes the market data (i.e., the support) and draws a preliminary value conclusion.  Note in this preliminary conclusion the appraiser concludes as to a value range rather than to value point.  From the Fannie Mae Selling Guide, as well as from USPAP, it is clear there is absolutely no requirement such a preliminary value conclusion be a point or a single dollar figure. Therefore, at this point in the line of reasoning, a value range is acceptable.

This line of reasoning is credible since (a) it introduces no new evidence from what the appraiser has already introduced up to this point in the appraisal report, (b) the support comes from market data, the veracity of which the appraiser has already demonstrated, and (c) the comparable sales demonstrate clearly the subject has a value above $380,000 (since this is an inferior sale), yet less than $400,000 (since this is a superior sale).

Now, from the credible line of reasoning, comes the credible conclusion [credible line of reasoning à credible conclusion(s)].  Notice the value conclusion flows smoothly and credibly from the line of reasoning which, in turn, flows smoothly from the premise.  There are no leaps of faith necessary to understand it; there is no opaque “black box” at work here; there are no unreasonable assumptions.  Up to now, everything the appraiser has done to arrive at a value conclusion has been clear, well-explained, and based on market data.  So, consider the conclusion: 

“Sale 1, which is overall most similar to the subject, has an adjusted sales price of $395,000, and [the appraiser has] weighted the value most heavily on this sale, which had the strongest indicator of value and required the least amount of adjustments. [The] opinion of market value as of the effective date is $395,000 based on both the adjusted and unadjusted sales price ranges.

To review then, to arrive at a credible value there is first a credible premise, then there is a credible line of reasoning, the unavoidable, logical, and linear results of which are a credible conclusion.  The model is “credible premise à credible line of reasoning à credible conclusion(s)”. 

A credible premise has its foundation in the appraiser finding, verifying, and then analyzing all of the data necessary for credible assignment results.  Then, a credible line of reasoning has its foundation in the credible analysis of the data the appraiser has amassed and verified. There are many ways to analyze data credibly, one of which is ranking, which is the center of this article, although by no means the only credible[6] manner to analyze real estate sales data.  Finally, a credible premise, followed by a credible line of reasoning, leads inescapably to a credible conclusion. 

Consider this credible premise, credible line of reasoning, and inevitably credible  conclusion:  “My spouse loves getting roses on special occasions. Our anniversary, a special occasion, is next week.  Therefore, I will purchase and then give to my spouse roses for that special occasion”.  Orchids are indeed beautiful flowers, but the purchase and presentation of orchids does not flow inexorably from the above premise and line of reasoning (and toproffer orchids rather than roses could result in the necessity to purchase and then present gold, diamonds, automobiles, month-long cruises, and/or real estate to ameliorate the ineffably stupid error, in this instance, of presenting to the wife orchids in lieu of roses).

It is true the conclusion that ranking is the only way in which real estate market participants decide on purchase/sales prices as part of the buy/sell decision is open to question.  However, that it is one way of many they reach that decision is essentially beyond debate. 

In the Massey & Andersen monograph, the premise (see above) is that the comparable sales before adjustment should bracket[7] the subject, thus even before any adjustments should be able to indicate a reasonable (albeit wide) value range.  In other words, the subject has a value higher than the highest inferior sale, yet a value lower than the lowest superior value.  However, this unadjusted range will be rather wide.  Thus merely choosing a number randomly from somewhere in the middle of that unadjusted range is not an acceptable protocol in the formation of a credible value opinion. 

In the line of reasoning (see above) Massey and Andersen demonstrate this is the place to explain why (a) of all the sales that could have been used, the appraiser used the specific combination of sales featured in the report; (b) why these sales are competitive analogues for the subject, (c) what adjustments it was necessary to make to these sales and (more importantly, why it was necessary to make them); and (d) the why the appraiser weighted[8] these sales the way s/he did.  From this array of adjusted sales, the appraiser will have narrowed the absolute spread of the range of sales.  Again, the implication is that the final value opinion will have come from this narrowed range.  

Finally, as a direct result of the premise and the line of reasoning, the appraiser winds up with a narrow range of sales prices.  The subject’s “…most probable [sales] price[9]…” should, therefore, be somewhere in this range since the premise and the line of reasoning should not lead anywhere else.  From this narrowed range, the appraiser then chooses a final value opinion and, more importantly, explains why this single number is more applicable to the subject than any of the other numbers within that range.

So, the take-away here is that while ranking the sales (both pre- and post-adjustment) has some subjectivity to it, the markets from which we garner the data are also subjective, thus subjectivity in forming a credible value opinion is inevitable.  This is why an appraiser’s value conclusions are opinions whose purpose is to conclude the most probable sales price of a property (not its exact price), under a given set of circumstances, at a specific moment in time. Such a credible conclusion is the result of a credible line of reasoning, all of which stems from the foundation of a credible premise[10].

Ranking the comparables, as described here, is another tool for the appraiser to use to support the formation of a credible value opinion. And, finally, the appraiser maintains all of that support in the workfile as USPAP requires.

Have more questions?  I’m here to help  you!  Contact me, Tim Andersen, the Appraiser’s Advocate, at tim@theappraisersadvocate.com, or 561/635-5265.

(The author gratefully acknowledges the contributions of Rachel Massey, SRA, in the preparation and architecture of this monograph.  The author, however, takes full responsibility for the contents.)

---

[1] See The Appraisal of Real Estate, 14^th ed., Figure 4.1, p. 37 for a flow-chart showing this process.

[2] 2020-2021

 USPAP, SR1-2(e)(i) in the context of identifying the characteristics of the [subject] property that are relevant (see generally e.g. SR1-2(e)(i-iv),   See AO-2 and AO-25 for more on this.

[3] See SR1-4

[4] Too many appraisers, and way too many reviewers, are concerned with an appraisal’s “accuracy”.  This is poppycock and it reflects not only a deep and inexplicable misunderstanding of the appraisal process by both parties, but a fundamental flaw in appraisal qualifying education and continuing education (especially the quality of the latter) since in them there is no refutation of this silly concept.  Since an appraisal is by definition an “opinion”, it cannot be “accurate” or “inaccurate” since an opinion is neither a fact to be proven, nor an object to be measured.  An appraiser’s opinion is either credible or it lacks credibility.  When there is a credible premise, accompanied by a credible line of reasoning, the conclusion(s) from them cannot be but equally credible.  With a false or unproven premise, or a faulty line of reasoning, however, the conclusions from them cannot but lack credibility. 

[5] USPAP defines credible as “worthy of belief”.  While there are some holes in this definition (do you believe what I believe?  Why?), there is every reason to use it here.

[6] Ranking, paired-sales analysis, regression analysis, GRMs, and so forth are subject to various biases, either those inherent in the method itself, or those imposed (inadvertently?) by the appraiser.  Therefore, at best, any method of forming a value opinion results from a bias/imperfection which, in turn, results in an approximation of value. This is why any value conclusion is an opinion, not a fact to be found.  We appraisers are imperfect, the market data on which we depend are imperfect, markets are imperfect since their participants are imperfect, and so forth.  Therefore, the best we can do is to collect, verify, and then analyze enough data so the obvious crap stands out (thus can be eliminated), and then the really prime data balance out/compensate for the mediocre data. Remember that opinions are either validly and credibly formed, or they are not.  They are neither “accurate” nor “inaccurate”.    

[7] Some appraisers have problems with the concept of bracketing (but that is the subject of another monograph).  However, when the appraiser credibly verifies the bracketing sales, and those sales are indeed comparable, competitive, and analogous sales, bracketing serves as another tool in the appraiser’s tool box.  There is no reason, however, it should serve as the only tool in that box.

[8] Yes, this weighting is subjective (and maybe even arbitrary).  But remember, CU and USPAP et al don’t require proof so much as they require support.  Therefore, in the narrative, admit the weighting is subjective, but that it is part of the appraiser’s scope of work to assign weight to the sales as part of the process of forming a credible value opinion.  The explanation (not a description!) of why the appraiser gave this weight is what CU and USPAP require and is evidence of the steps the appraiser took to form a credible value conclusion.

[9] See the definition of Market Value most appraisers use when appraising residences.

[10] A monograph this long needs 10 footnotes.