From AppraisalPort’s monthly newsletter

Author: Steve Costello, who attended the recent Valuation Expo

“Fannie Mae’s Murphy stated that over the past year, the GSE had been focusing on “quality” and “condition” ratings of comps used in multiple appraisals by the same appraiser and found many cases where the appraiser has changed the quality and/or condition ratings on the same comparable from appraisal to appraisal.  Now, based on the examination of the Uniform Appraisal Dataset (UAD) data, Fannie Mae’s focus for the next 12 months will be on adjustments.  The data indicates that many appraisers are not using proper methodology to make their adjustments.  Murphy stated that some appraisers are still using the old standard $20-$40 per square foot adjustment on properties that are easily valued at $500-$650 per square foot.”

“Murphy explained that Fannie Mae is planning to re-evaluate appraisers based on their adjustments and the GSE will expect appraisers to comment on all adjustments if necessary. And, ‘it will be necessary,’ he said, adding that Fannie has seen a lot of under adjusting. To be safe, appraisers should document their logic and reasoning for making any specific adjustments.”

My comment: The easiest adjustment is time. Fannie got that done by requiring 1004MC. The next easiest adjustment is sq.ft. – very easy and reliable using statistics. Of course, as we all know, unless you are appraising a conforming tract home, it is very, very difficult to “prove” all your adjustments. If you know the local market makes adjustments, they should to considered in your appraisal. State regulators are looking for support for adjustments. I am seriously thinking about not using dollar adjustments for 1-4 unit appraisals. Many years ago there was a Fannie form that just required plus and minus adjustments.

I seldom make any dollar adjustments on my apartment and commercial appraisals except for time adjustments, which are easy to support. I find it very strange that residential appraisals have such a high standard. I guess it is due to the lenders telling appraisers what they have to do. I am so glad I don’t do any residential lender appraisals any more. I never like them telling me how to do my appraisals.

I don’t know how Fannie will evaluate adjustments. I make many of my adjustments on a qualitative basis as I work in an area where most homes were built prior to 1920 and are very dissimilar. I know what my market wants, and doesn’t want. If I am not sure, I ask local real estate agents. Of course, they seldom know the dollar amount.

I wonder how well “bracketing” will work for adjustment support?

click here to read the full newsletter

Appraisal Today newsletter

  1. When UAD first appeared, I heard Mr. Murphy talk about his new system of standardization of appraisal data for “Risk Analysis”. While not all of the new appraisal standardized format (UAD) are awful, I believe my appraisal reports are less accurate using the UAD format vs. having the freedom to use other general purpose appraisal forms. The UAD format is similar to using just one type of shoe for several sporting events. What may work in downhill skiing, doesn’t work well in basketball or golf. The UAD data collectors for attempting to force square data into round holes. Thank you, but don’t pee on my leg and tell me its raining. I didn’t just fall off a Turnip truck. FNMA is ignoring the most important section of USPAP, in Scope of Work, “the appraiser relies on peer-reviewed methodology to formulate an acceptable work plan”

  2. The concept of adjusting per square foot is great…except in my area almost every mls listing has the incorrect square footage as the agents in my area “guess” or go off the “owner/s’ guess. And no my board really does not care, we complain and it does not change, after all the board was developed for realtors not appraisers.

  3. When did FNMA become geographically competent in my area to make such a decision for me?

  4. In situations where adjustments are not clear cut such as exact or near model matches in cookie cutter developments, the best way I know of to bulletproof your adjustments is by regression. It isn’t hard to do if your MLS contains an adequate amount of data. There is software out there with a low learning curve and if you have Excel on your computer you already have the tools you need provided you have a good working knowledge of spreadsheets.

    BUT… the problem is the push back you get from lenders, underwriters, reviewers & government staff appraisers who are used to seeing “traditional” adjustments applied the “traditional” way. They typically do not understand alternatives like regression and usually want no part of it. If a regression based on 100 sales tells you that the GLA adjustment rate is $58 but the lender is used to seeing, say, $30 guess what happens? You either use what they demand you use or you lose your client and probably your fee too. Even when you know for certain & have proven that your analysis is correct! Style over substance rules the roost.

    In residential mortgage appraisal, we are often not allowed to solve the problem using a scope of work & methodology that we deem accurate & appropriate to the situation, yet are probably going to be judged by Fannie using the very same statistical technology that our clients (who are the ones paying us) will not accept. Until Fannie rectifies this gap, they will continue to see adjustments that don’t jibe with their statistically based expectations. In my opinion, Fannie, FHA, VA and the rest need to encourage lenders to accept methods like regression or at the very least, encourage them to accept from us whatever other means Fannie chooses to use to evaluate our adjustments.

  5. Mr. Murphy is simply parroting the FNMA company line. It is evident FNMA and their portfolio analysts are fixated with the “data” they cull from their vast database. Of course, this is populated from properties across the country, and allows their programs to crunch elements derived from broad areas, or down to locales, cities and likely zip codes and census tracts. What these worthies fail to take into account is that appraisers DO NOT HAVE ACCESS to all this wonderful data. FNMA does. FNMA got it from the different, separate appraisers. FNMA got the data over time, and it was funneled into their database, accumulating from disparate suppliers. FNMA can afford sophisticated programs to crunch the data, extract the data, zero in on particular items with decent levels of confidence. Appraisers DO NOT have such benefits during THEIR analysis. All of which takes place under pressure from the client and under time duress. And all for a miserable diminishing appraisal fee. Mr. Murphy and his colleagues at FNMA, I submit, are under no such ‘promptings’. And from their lofty ivory like venues, they can interpret, evaluate and conclude their resounding “voila” judgments. All after the fact, of course. Easy enough to do. If the disconnect between FNMA and the current state of the appraisal function wasn’t so vast, the staffers and managers at FNMA ought to get out and about, and prepare and complete a few appraisals themselves so they could get in touch with reality. And do so under their own silly requirements as well as dealing incognito with the clowns appraisers are now forced to deal with in the mortgage lending food chain. So don’t hold your breath with this one.

    No wonder they keep coming up with their foolish UAD, 1004MC claptrap and such, as some type of panaceas. They have the luxury and time to indulge their farcical exercises and consequent appraisal requirements. It’s what they do best. They should just be upfront about it and simply eliminate the requirement for an appraisal, which is no doubt a guiding behest from their benefactors.

    And Murphy expects or says FNMA is looking for feedback from appraisers? Please, enough with the self-serving platitudes.

  6. Agree with the previous responses. It seems apparent that FNMA is leaning heavily toward the regression type of “appraisal” that many software developers are pushing…as well as at least one company consisting of several members/founders of a state appraisal group. What a house sells for per square foot used to be almost meaningless. The appraisal profession is systematically being dismantled in favor of AVMs and statistically regression. Mr. Murphy’s comment, unless it was somehow taken out of context, is very telling…and patently incorrect.

  7. This type of nonsense is really astounding. It is so obvious that FNMA is clueless to what a paired sales market analysis is about. They are going to drive every appraiser right out of this business. Why don’t they just use some type of AVM and get rid of appraising property. The UAD was the end of field appraising with all the data collection. As someone previously stated we are being marginalized. I wish them a lot of luck with data collection as a method to appraise property without a physical inspection. You think fraud was rampant during 2001-2007 once they go to using automated valuation data base collection via UAD the fraud will be ten times what is was during those years. Any time government get involved it becomes the ruination of that industry. I have 29 years in this business and you can barely make a living at it any more. It is really disheartening how foolish these government agency have become.

  8. I agree. It is Murphy at Fannie Mae who is in serious error. It is obvious he is looking at what homes sell for per sq. ft. and not what the market actually recognizes for differences in GLA. I would be curious to see his examples of homes where $500-$650 per sq. ft. are warranted, and his support for those adjustments. Not just “well, this sold for this per sq. ft. and so that’s what the adjustment should be”. I’ve had realtors make the same flawed argument when setting the proper asking price.
    The smaller the house in the same neighborhood, the more per square foot the sale price of the comps. They generally have an inverse relationship. Take 2 similar homes, same neighborhood with similar utility. One just happens to be smaller but with equal appeal. I think there is way too much emphasis on GLA adjustments. A smaller house many times commands the same or even more in some instances, but we make the adjustment anyway. And our adjusted prices actually move further apart when ideally they should be tightening. Not making the adjustment is many times unacceptable to the client. How much sense does that make? Sometimes it is obvious the market couldn’t care less about the GLA. One house had greater appeal for other reasons. The GLA had no bearing whatsoever. Where is the line item for “I just like that house” that reflects the market reaction for overall attractiveness/appeal? Obviously, it does not exist. That is why I am fairly conservative with my adjustments. It is not an exact science. Some homes are in good shape, modernized and updated, but just plain ugly. There’s no UGLY adjustment. Buyers make decisions for all sorts of reasons, many of which are not quantifiable on a grid. Murphy’s statements should have valuation professionals calling for his resignation. His comments are a window into just one of his flawed opinions on how appraisals are developed from market data. I wonder how he thinks about other areas of the appraisal. (one size fits all?) To have an individual like him making policy decisions at Fannie is a big problem that ought to concern everyone in the industry.
    But really folks, at the end of the day, this is all about data collection. Every report has data on at least 4 properties, and most of the time, 6 properties ( the subject + 3 closed and 2 actives). By the time 10 years will have gone by since UAD, they will likely have physical data on about 80% of the housing stock in the US. They will have this because of the 7 year turnover rate as well as refi appraisals.
    We are being marginalized as we faithfully execute their plan.

  9. If fannie Mae knows how to make adjustments, why do they need appraisers. The client can have no influence on the value determination. Try proving that time adjustments are required when there are only 4 or 5 comparable properties all year using the MC form, lol. This needs to be determined on a much wider look at the market area a lot of times.

  10. It seems to me that FNMA is calculating appropriate GLA adjustments by analyzing the gross sale price per square foot calculation and claiming that the appraiser’s methodology is wrong. Any comparison of the gross per square foot calculation or even construction cost is flawed. Gross per square foot numbers includes the land and all improvements. An adjustment for per square foot should be based upon paired sales analysis for that specific line item. This is a market derived per square foot adjustment. As in, not how much it costs to build but how much the market rewards for additional square footage alone. That means exclusive of land, baths, basement, pool or any other improvements. If a house sells for $500 psf, this number includes all improvements. Any attempt to compare the two numbers is flawed, not the other way around.

We want to know what you think!! Please leave a comment.