Repurchase Demands and Unacceptable Appraisal Practices - Part 2


Why Not Earn And Learn
> How to Support and Prove Your Adjustments
(Earn 7 Hours Online CE)


Repurchase Demands and Unacceptable Appraisal Practices – Part 2
By Rachel Massey, SRA

Lenders are rightly concerned about repurchase demands because they are time consuming, costly, and take away from their ability to free up money to originate other loans. Here’s how avoid problems altogether.

Repurchase demands occur not only on old non-performing loans but also on performing loans that are recently originated. There are different issues that could trigger an audit of a performing loan, one of them being simple random audits. Other factors could relate to anomalies in the credit file, or even issues with the appraisal that relates to the Unacceptable Appraisal Practices, or other issues such as non-permitted accessory dwelling units, leased solar panels treated as real property, etc. The takeaway is that these demands are not just on defaulted loans, but occur on loans that are current and can be completely random in nature.

Appraisers who primarily do mortgage-related work often find that review stipulations are endless and often seem meaningless. But if the client’s needs are better understood it helps alleviate some of these concerns. Repurchase demands are prevalent in the mortgage industry and reviewers often are the first line of defense. Reviewers try to make sure the appraisals they accept do not contain “unacceptable appraisal practices” or UAPs that can trigger a repurchase.

This is the second in a series of three articles that touch on some of the unacceptable appraisal practices outlined in Fannie Mae’s Selling Guide. In this article we are focusing on the following;

1. “Failure to adequately analyze and report any current contract of sale, option, offering, or listing of the subject property and the prior sales of the subject property and the comparable sales.”
2. “Use of adjustments to comparable sales that do not reflect market reaction to the differences between the subject property and the comparable sales.”
3. “Not supporting adjustments in the sales comparison approach.”

“The failure to adequately analyze and report any current contract of sale, option, offering, or listing of the subject property and the prior sales of the subject property and the comparable sales.”

This may seem self-evident but sometimes the search parameters for the sales and listings of our properties and comparable sales may simply miss data. For instance, if we rely on the Multiple Listing Service (MLS) history function, one misplaced dash in the assessor’s property identification number or a missing direction, can cause the MLS data to not associate with prior listings and sales. This can cause us to miss a listing that is relevant.

(story continues below)

(story continues)

Sometimes there is a misperception of what analysis is; analysis is more than stating when something sold and for how much it sold for. It is about taking a deeper dive into the market data to find the differences between the current appraised value and the prior sale or listing of the subject. The comparable sales need to be discussed as well. For example, a house that was purchased two and a half years ago out of foreclosure may have sold at a discount at the time due to the condition of the property as well as the terms of the sale. Since that time, the house may have undergone a complete renovation and the market may have increased significantly. Any large increase (or decline) in price is a trigger for further explanation. Explanation should be correct and verifiable, such as what we see in the following example.

“When the subject property last sold on 2/19/15 it was in poor condition, needing extensive work to the drywall, cabinetry, flooring and mechanical systems. Since the date of sale, the new owners have gutted the drywall, re-drywalled the house, installed new hardwood floors throughout, and also installed new Wood-Mode cabinetry with Quartz counter tops in both the kitchen and the bathrooms. In addition, the siding, windows, roof, furnace and A/C were replaced. Due to this significant remodeling, the appraised value is substantially higher than the prior sales price. Not only did the remodeling contribute to this increase in value, but the market has also increased in the area 6.5% since the prior sale (see market statistics that follow).”

This explains the “why” of the increase in value which helps the client and intended users understand why the sales price in February 2015 for $100,000 does not have direct correlation to the current appraised value of $200,000. If photographs support your analysis they can be included. At the very least, they should be retained in the workfile in case of a complaint filed at the state level.

“Use of adjustments to comparable sales that do not reflect market reaction to the differences between the subject property and the comparable sales.”

This unacceptable appraisal practice and the one that follows have had appraisers concerned over the last few years with the advent of the Collateral Underwriter program (CU). As appraisers, we have always needed to have some basis of support for our adjustments but CU has focused mortgage appraisers on being able to support the adjustments we use. There is a myriad of ways to support adjustments but in the end it is the individual appraiser’s analysis of what they find in the market that needs to take precedence. We have all seen results from paired sales, grouped data, depreciated cost, regression, and so forth that simply do not make sense. This is why it is helpful to use more than one method on some major units of comparison and to synthesize the results based on experience and common sense.

For example, if we look at a garage adjustment for a house with a two-car garage, it is very possible that paired sale data might tell us that the difference between a 2-car attached garage and a 1-car attached garage in a certain market segment is between $10,000 and $12,000. If we then look to cost, we might find the actual difference in cost is only between $6,000 and $8,000. After applying the adjustments between $6,000 and $12,000 to our sales comparison grid, the adjustments may end up significantly widening the range at $12,000 but at $6,000, it may widen the range but much less so. Perhaps the answer is that the value of that garage in this particular marketplace is less than $6,000. If we poll market participants as to what they would pay to have the extra garage, we may find the buyers tell us that they would pay $2,000 to $3,000 more, and the agents may tell us that they typically see buyers paying between $3,000 and $5,000 more for the extra stall. If we try the $2,000 from the low end of the buyer poll, and the range widens again, then we know that this is too light an adjustment, but if we try between $3,000 and $5,000, we might find the sweet spot where adjustments narrow the most, to be $4,000. This is an example of sensitivity analysis, which can be a useful tool once we have other adjustments isolated.

(story continues below)

(story continues)

The Collateral Underwriter system has “model” adjustments that are extracted from data based on what other appraisers in the area use, as well as their own algorithms. Our adjustments may be higher or lower than the model adjustment but if we can explain our thought process and how we arrived at this adjustment, a human being can read our comments even if the computer system cannot. If the CU indicates our appraisal contains inconsistent models from what is expected, in all likelihood it will be kicked to review. It makes much more sense to explain the thought process that went into an adjustment up front, than in response to stipulations from review later. The comment that “all adjustments are self-explanatory” or that “all adjustments are based on paired sales” may sound good in the comment section but if this isn’t actually how the adjustments are extracted from the market, then the appraiser could find themselves trying to dig out of a hole they needn’t have dug in the first place. Simply stating that the garage adjustment was based on a combination of paired sales analysis, cost and market participant polling, plus appraisal synthesis and common sense, should be enough to alleviate most lender/underwriter concern. But only if the comments are actually true and can be backed up with support if under scrutiny.

Not every adjustment can have market support every time. Sometimes there are adjustments that need to be based on the appraiser’s market expertise. An example might be a fireplace. The appraiser knows that the market typically desires a fireplace (in states like Michigan but not so much in states like Arizona). However, what if traditional approaches of paired sales, depreciated cost, regression, and so forth show inconclusive data and polling agents results in mixed information? Looking at cost data, the fireplace is $3,000 to install new or $6,000 to install in an existing house. The appraiser could logically conclude to it having less value than the install on an existing house, and possibly less than installing new, as long as a comment is made as to why the adjustment was made or not made. If there is support in the appraisal, does anyone really think that CU is going to focus on a $2,000 fireplace adjustment? Maybe, maybe not, but in the end the appraiser can explain it is their market expertise, as long as this is not the rationale used for the majority of adjustments.

It’s important to remember that a solid reconciliation can go a long way in supporting the overall adjustment process. The logic of providing a sale that is superior and one that is inferior for “benchmarks” or “bracketing” to highs and lows is sound, although it cannot always be achieved at the same time as providing comparable properties. Many appraisers complain about stipulations for bracketing sales but the rationale behind bracketing is the logical conclusion of a benchmark to the high end and the low end of the value range. If the appraisal only provides overall superior sales or overall inferior sales, then all the appraisal has done is proven what a superior or inferior property will fetch in the market, not how the subject is positioned. A good, solid reconciliation will take care of much of the concern but the reconciliation goes far beyond saying that the three most comparable sales in the market were used in the report. When faced with adjustments that are weakly supported, reconciliation is the perfect place to tie together the value conclusion based on common sense.

“Not supporting adjustments in the sales comparison approach.”

The above about a weakly supported adjustment ties into the third UAP covered, which is “not supporting adjustments in the sales comparison approach.” Collateral Underwriter may show a different model of adjustments but as long as appraisers can back up most adjustments, this “not supporting adjustments” UAP should not be relevant. Unfortunately, oftentimes there simply is no support offered in the appraisal because the appraiser has not gone through the process of trying to isolate adjustments and is instead relying on an old “list” that was handed down to them from their mentor 15-years ago. That list was handed down to their mentor from their mentor’s mentor all the way back to Moses in some cases, and may not be based on any reality in the current marketplace. While this is written largely in jest, unfortunately sometimes it is also reality.

There is a preference in the secondary market for “paired sales” data but there are other methods of extracting an adjustment. In imperfect markets, which most of us operate in, paired sales data can be difficult to come by. Being able to explain the adjustment process is critical in situations where the appraisal lands in the cross-hairs of the CU and/or an in-depth review. If the appraiser has gone to lengths to develop support for some of their adjustments, how difficult is it to write a sentence or two about what was done.

Take the example of a 15-year old house with a cost new of $250 per sqft to build. The house is approximately 10% depreciated, meaning that the improvements are $225 per sqft as depreciated. Each of the sales used in the appraisal report sold between $300 and $350 per sqft. including the site value, and have a residual between $200 and $250 per sqft. after site value is removed from the sales. The above grade living area contributes around 65% to 75% of the total cost to build, indicating that the depreciated cost per sqft is between $146 to $169 for the subject’s above grade space and between $130 and $188 for the comparable sales. While adjustments of $130 per sqft. may not make sense, neither would $20 per sqft. from the standpoint of pure logic. The appraiser could lay this out for sensitivity analysis to see where the range narrows and where it widens if paired sales data was not available but grouped data could be reasonably applied. Since there is enough sample size for analysis they could state that whatever adjustment they used made the most sense based on this information. While CU could show something different than the appraisal, at least the appraisal provided support.

Fannie Mae Selling Guide offers a lot of guidance for appraisers as to how appraisals should be supported, and section B4-1.3-09, Adjustments to Comparable Sales (12/16/2014) specifically states:

Fannie Mae does not have specific limitations or guidelines associated with net or gross adjustments. The number and/or amount of the dollar adjustments must not be the sole determinant in the acceptability of a comparable. Ideally, the best and most appropriate comparable would require no adjustment; however this is rarely the case as typically no two properties or transaction details are identical. The appraiser’s adjustments must reflect the market’s reaction (that is, market based adjustments) to the difference in the properties. For example, it would be inappropriate for an appraiser to provide a $20 per square foot adjustment for the difference in the gross living area based on a rule-of-thumb when market analysis indicates the adjustment should be $100 per square foot. The expectation is for the appraiser to analyze the market for competitive properties and provide appropriate market based adjustments without regard to arbitrary limits on the size of the adjustment.

It is useful to read relevant sections of the Selling Guide and to re-read the certifications in the Fannie Mae forms, specifically Point 9 in our certification as it relates to two of the UAPs addressed above.

“I have reported adjustments to the comparable sales that reflect the market’s reaction to the differences between the subject property and the comparable sales.”

While none of the issues addressed above are new to any of us, the UAPs section of the Selling Guide deserves a repeated read and further consideration. Reading them over can help all of us doing mortgage related work reduce some of the stipulations we receive, because in the end, many are related to the reviewer trying to help their employer avoid a repurchase scenario. As appraisers, if we are aware of the needs of our clients it can help us address potential issues in advance, saving valuable time and frustration after the delivery of our appraisal report.


New Online CE – 7 Hours
How To Support and Prove Your Adjustments
Presented by:
Richard Hagar, SRA

Do you have the proper support for your adjustments? Stop taking the same old CE courses and learn proven adjustment methods with instructor Richard Hagar, SRA. Fannie Mae states that the number one reason appraisals are flagged is the “use of adjustments that do not reflect market reaction.” Stay out of trouble with Fannie Mae, your state board and your AMC/lender clients with solid, supportable adjustments.  Up your game, avoid time-consuming callbacks and earn approved CE today!

“Why wasn’t this taught years ago?” – Jackie Henry

How to Support and Prove Your Adjustments
Sign Up Now!  $119  – 7 Hrs. Approved CE
(OREP Member Price: $99)


About the Author
Rachel Massey, SRA, AI-RRS has been in the real estate field in the Ann Arbor area since 1984, first in sales, and then as a full time appraiser since 1989. She has a Bachelor’s degree from Siena Heights University with a real estate concentration, and is an AQB Certified USPAP instructor. Rachel was one of the original members of the Michigan Council of Real Estate Appraisers and has a passion for helping other appraisers through writing, teaching and with peer review. She has expertise in lake appraisal, Relocation appraisal work and other residential work in Washtenaw County and surrounding communities. When not appraising or thinking about appraisal, she can be found enjoying sunsets, walking, and the occasional toss about the mat in aikido. Rachel can be reached at or through her website,


Send your story submission/idea to the Editor:

Tags: ,

Comments (8)

  1. Wow, what an excellent article. I like to refer to those methods as distributed allotment or proportional distribution. Whatever you like to call them though, sensitivity analysis, juggling numbers to find reasonable streamlined adjusted value, whatever, it’s all just logical human analysis. And the 2 hour appraisers continue to get all the love, and the 30 pieces of silver. The home is the pie, the grid the sections of the pie and every pair you need, is present in the 6 comps which have the most logical and realistic similarities and relationships to the subject, in one form or another. I say release the 3 comp rule personally, and stick it with 6 every time. The defensibility points are in there, even if you don’t necessarily take the time to measure them exactly. Meanwhile, the consumer base is watching real estate shows and ask silly questions like if I invest a dollar into my kitchen, will I get a dollar and ten cents back? The response is that I don’t know, because a kitchen improvement alone is not that substantial, it’s the track record of maintenance and whole home improvement that matters or not. I like to start off categorically high/mid/low, and narrow down from there. Logical analysis is where I hang my hat. Boilerplate is for the birds and pencil pushers. Most volume appraisers who may try to swing a hammer may very well risk their own lives in the process. They know math, but don’t necessarily know real estate. Double dare you to try and explain depreciated analysis in standard grid work to a regular outsourced underwriter. I’ve given up on that, and instead just chunk a big adjustment in the C line for materials, and explain it that way instead. Cheers!

    - Reply
  2. no other profession has to explain every minutia in detail as to why, or why not. i agree with everyone above. they also do not wish to pay for all these written prove-ables. i do like it now that this is a ce course, better to have some idea of a defense.

    - Reply
  3. by Detroit area appraiser

    If one wants to keep appraising until the day THEY choose to stop … as opposed to the day when some other entity forces you to stop by relieving you of your licensure … then the long addendums and reconciliations are a fact of life now. Hence, the volume of work potential has been decreased. Therefore, what is one to do? (this is a rhetorical question folks). The answer is simple … raise your fees folks! Know your value as a professional.

    - Reply
  4. As appraisers we all run the extremely high daily risk of having our work unfairly assaulted by an incompetent fannie mae reviewer. I sincerely hope incompetent reviewers are the exception not the rule, but I know for a fact some incompetent ones are out there: appraisers beware. The fannie mae reviewer is a bureaucrat paid by the hour and is at liberty to waste an umlimited amount of the appraisers time. Sadly my experience is having had an incompetent fannie mae reviewer disagree with my opinion of value on a complex appraisal of a multi-unit residential rental property. The reviewer failed to capitalize a $600/mo income stream and proposed a value too low by precisely this capitalized amount (this is for a purchase mind you). This nonsense caused my client to be unjustly threatened with a repurchase. This nonsense also cost me thirty hours to rebut and lost me much work from my best client. I no longer accept complex work because I don’t trust fannie mae to review it competently should it pop up as a random audit. Incompetent fannie mae reviewers are a pestilence on the industry. Please fannie mae fix this!

    - Reply
  5. Finally, some one agrees with what I have been practicing (and preaching) for 43 years as an appraiser! When extracting an adjustment factor from data when matched pairs are not available (never!), applying a test adjustments to comps (hopefully after adjustments for other variances have been applied and assuming comp data in question does bracket the subject amenity), as the range narrows when different factors applied, it would logically confirm the market reaction amount for the factor in question. It is not rocket science! Only convincing an AMC ‘reviewer’ (not all, but we know the ones I am referring to) is the challenge.

    - Reply
  6. No one can do more than one appraisal a day and do it in the manner the author suggests. All appraisals are reviewed for problems and must be near perfect.Why would anyone want to be in the appraisal business?

    - Reply
  7. Just think how long it took this gal to write this class. Just think how long it REALLY takes to do all this stuff on every report! Who can make a real living at this? I know I cannot. I do 3 appraisals a week if they are reasonable. I’ve done these for 29 years now. How does one person to 10-12? Many say they can!

    - Reply
    • I have done 10 full ones a week before – work 10-12 hours a day , 7 days a week. – 3 a week? I try to do one every working day – so at least 20 a month – I agree with you Debbie, if you write the long addendum the author states- you will be getting 5 calls a day and emails galore wanted to know where my stinking report is.

      - Reply

Leave a Reply

Your email address will not be published. Required fields are marked *