GLA Adjustments & Pizza Slices – They Do Correlate!
37 33 19
Determining the GLA Adjustment in Appraisal Reports
I promised in a previous post I’d send out my method for determining the GLA adjustment in appraisal reports.
Goodness! Appraisers actually found that statement buried in the prior post, and have ‘rung my chimes’ requesting this info!
OK, but first, the pizza analogy. I often tell clients and other appraisers that appraising (residential) real estate is far more complex than grabbing a store-bought pre-boxed pizza (that has exactly the same ingredients from box to box no matter which store you shop in) at a grocery and then applying 425 degrees of love for 18 minutes in your home oven to achieve the tasty “opinion” desired. For my example here, I’ll use pizza to help you “visualize” the concept for determining the GLA adjustment. This process/method works for every property type.
The second thing you need to know is this: the method I use was something I learned about 10 +/- years ago from another California based appraiser, in a small book discussing adjustments written by that appraiser. Since then, I’ve learned that other appraisers across the US use this or a very similar GLA adjustment method, which I’ve used for a decade. So this method is not some kind of weird science or hocus pocus!
The other benefit is it satisfies the wonderful FNMA Collateral Underwriter review robot focus because the GLA adjustment amounts in my reports from property to property are always different dollar amounts. They are based on plainly visible report data.
Third, there presently are multiple software-based regression programs being sold, both independently and via the forms providers. These are designed to have you put your properties into the software, and they magically spit out adjustment figures – alleged to be highly accurate – based on their on-board algorithms. Unfortunately, you probably don’t know anything about how they work internally. That’s a problem!
In actuality, regression programs focus first on GLA, then when the program can’t determine proper adjustments for additional amenities, they add to… you guessed it… the GLA adjustment amount and don’t make the others! So regression software is an extra step beyond what I do, and in most cases, it costs extra buckaroos to acquire the software. But some do say that they can provide ‘other’ adjustments also. You might find that important.
My GLA adjustment system is FREE. And the components are right in front of your nose! Once you learn this, doing the other adjustments is relatively easy.
Fourth, several printed or on-line residential construction cost guides are available. Many are used by appraisers to help determine adjustments. That’s perfectly fine, and does provide a degree of accuracy that some might consider the absolute precise way to figure out the GLA adjustment. I’m OK with that if that process floats your boat, in terms of time spent! Actually, it’s a good way to double check my method.
Fifth, there are approximately 81,000 independently licensed appraisers in the US as of July 2016. A number of ‘adjustment deriving’ classes are being taught by some highly experienced appraisers, live or on-line, whom I greatly admire for their willingness to teach. But, truth be told, there are somewhere between 1 and 81,000 ways to calculate the GLA and other adjustments! So which one is the better one to use? Which one has an accuracy you are comfy with? Is one better (i.e., more accurate) than another, or several others?
Is my method worse or better, or easier, than what you are presently doing? You will have to decide for yourself what is best for your appraisal business based on what you can support – if you are asked to do so. That’s the kicker. You need to be able to explain the mathematical method and reasoning. Most residential appraisers are basically in a ‘production mode’ 100% of the time. Meaning, we accept assignments and need to get them churned and burned and out the door relatively rapidly. That does not mean sloppily, or inaccurately, because now with all the various ‘review’ exams employed, our reports are being given a proctologic exam within 30 seconds of submittal. The process I’m about to explain will help you produce a credible, and mathematically supported, GLA adjustment easily.
By the way, this document is not about trying to “sell” you anything for money by leading you up to the ‘closing question’ I used to do in my prior 30 years of sales experience. I provide this info freely, because my goal is to help appraisers do better, more supported reports. Those who care to may distribute this to other peers freely, without restriction, with my permission given here.
OK, I’m getting hungry. Let’s examine the pizza analogy relative to the GLA adjustment.
When we are slaving away on the computer, a few inches off the tip of our nose is an appraisal report staring back at us – on the monitor. What’s at the top of the grid area in those reports – just below the Proximity to Subject line?
The Cost per Sq. ft is a mathematical function, dividing the Sale Price by the GLA.
So, the Sale Price/Gross Liv. Area = Cost per Sq. ft is what is staring back at you. This
is what we’ll be dealing with for the rest of this discussion.
The Cost per Sq. ft is the ‘whole’ pizza analogy that I have referred to. The Cost per Sq. ft represents the entire pie, of whatever tasty variety you prefer, thin crust or deep dish. It doesn’t matter.
The important item to remember is the Cost per Sq. ft represents 100% of all the property characteristics and amenities, such as GLA, view, location, quality, condition, basement, garage, site size, etc. As such, we never use the ‘total’ amount shown in the Cost per Sq. ft for the GLA adjustment. This is not totally understood by many people who come in contact with our work, such as real estate agents, underwriters, etc. And even some appraisers who don’t do residential reports regularly.
Let’s digress for a moment. On the 1004 Form, there are a total of 24 fields per comp where adjustments can be made, 21 pre-printed and 3 extras you can insert. GLA is only one of those. Not all fields need to be used. In some cases, line item fields can be combined to make one adjustment. And the GLA adjustment can include difficult-to-
adjust minor items not adjusted individually – usually the ones below the basement line.
OK, wander back into your pizza joint. Next time you order one fresh out of the brick oven, ask the cook to divide the pizza into different sized slices? WHAT??? The cook will have a conniption fit. But it’s what we appraisers need to do as we determine what size ‘slice’ the GLA represents. It’s not the same size as the other slices (the other adjustments). More than likely, the GLA adjustment will be the largest size. Sometimes not.
This, by the way, also points out what the CU robot is looking for, and is seeing too often. Appraisers are using ‘the same size slice’ (i.e. the same GLA adjustment amount) for all the reports they do – regardless of property characteristics. FNMA does not like that, nor should anyone!
“One size fits all” is improper methodology. And appraisers are getting their knuckles wacked by the FNMA CU reviewer robot due to this. And by warm body reviewers also.
We’re getting close to the ‘meat’ of this topic… pepperoni or ham? With oozing cheese and sauce… all the other aspects to be considered.
To be honest, when I have discussed this method I use with other appraisers, I often get a deer in the headlights look back, because they’ve often never considered or thought about this relatively simple, yet supportable, process. Maybe the simplicity is confusing??!
But I have continued to use this method, and have never had a problem from any report reviewer. In over a decade, remember?
The GLA ‘size slice’ is a PERCENTAGE of the Cost per Sq. ft. Again, it’s never 100%. The trick, and the stickler about this, is to determine the appropriate percentage to use. That’s the push back I normally get. “Well, I can’t just pick a number.” Actually, yes you can, once you understand the dynamics.
So where are we headed with this pizza slice analysis analogy?
The goal is to figure out what PERCENTAGE to use to apply against the Cost per Sq ft. This percentage figure will yield the $ cost per sf of adjustment between the Subject and the Comp(s) GLA difference.
You can also double check, and possibly modify, the percentage based on your own experience in your market and assignments for different properties you’ve completed. Do that by looking at old reports, if you feel your adjustment in those is OK.
Take your report individual Comps Cost per Sq. ft $ amount (at top of grid), and multiply that by 20%. Check that $ adjustment amount by what you used. Is it lower than your adjustment per sf, close or even higher? Then do it again, this time using 65%, and check how that relates to your report.
If your comp Cost per Sq. ft is $150, then 20% will yield $30; 65% will yield $97.50.
The Fannie Mae CU process is finding far too many reports with low percentage / low dollar amounts for the GLA adjustment. Being too low is really not a valid amount if one believes most people use GLA (living space) as a key purchase decision, except in Pittsburgh, PA where I’m told they don’t!
I’d like to suggest that you start with at least 40%. Using our $150 comp, that would be $60 per sf for the GLA adjustment. This means that the significant slice of the pizza, the GLA, represents a high ‘weighting’ in the eyes of the property buyer, or owner if a refi. For perspective, I currently start at 45% for most assignments.
Fannie is fond of explaining that the ‘rating numbers’ assigned to properties are Absolute, and not Relative. Well, in this case, the GLA adjustment percentage and amount is Relative! Relative to characteristics, features and amenities with each property. Meaning, it is changeable.
It’s also Relative to the measured GLA size. In most cases, the Cost per Sq. ft changes with the house size; smaller homes generally have a higher Cost per Sq. ft than larger homes. But regardless of that, the adjustment method explained here works across all types and sizes of properties.
So this ‘relativism’ is also related to the appraiser’s perspective and opinion on how much weight to place on GLA. This means you can ‘adjust’ the PERCENTAGE figure you use to determine the GLA adjustment. This is key in this discussion. The appraiser’s opinion is an important factor in this method. Other factors with the property will affect the PERCENTAGE used.
In my opinion, and actual use, using a PERCENTAGE of the Cost per Sq. ft at 40% to 50% is a good and safe place to start. You will not be criticized using that figure. Appraisers who drag out a cost manual, factor in depreciation and/or rely on regression software are probably close to that percentage. This is why this ‘pizza slice’ method works.
At this stage of pizza ‘consumption’ I’m guessing you’re questioning the rest of the Cost per Sq. ft percentage from the 100% total. What’s left over represents the other 23 items that can be adjusted. Each has a slice of their own, or maybe they are combined. You as the appraiser have to use your ‘relative’ thinking cap to decide how much each one gets, in terms of their own percentage.
If your property is in a typical built-out neighborhood of relatively uniform homes, then the 45% or so PERCENTAGE you choose is probably good for the GLA adjustment.
On the other hand, if your property has amenities that contribute to its overall value, such as acreage, area location, view, waterfront, basement, multi-car garage, etc., then you know that it or those slice(s) contribute a higher percentage to the overall Cost per Sq. ft figure. You can then reduce the GLA % amount because more of the overall value is tied up in the other amenities.
The point of this is that if you only make a GLA adjustment, and none other, the GLA % can be higher (but never 100%!). But if there are several other adjustments made up or down the grid, especially high dollar ones, then use a PERCENTAGE amount somewhere between 28% to 45%, or so. This is what I meant by ‘the dynamics’ mentioned above.
Now that you understand the basics of the ‘dynamics’, let’s discuss the “math mechanics” of how to calculate the adjustment dollar amount. Trust me, you won’t need to use a HP12C calculator, a fancy software spreadsheet program, hire a college level math professor, or your 12 year old neighbor. Just a trusty simple desk calculator with a percentage (%) key is enough!
“But first, there’s more” is the famous infomercial line. Here, I have to reveal that there are two basic schools of thought that apply to GLA adjustments. One says that adjustment is to be different (% or $ amount) for each comp. The other, more frequently used, is each comp has the same $ per s/f difference applied. I’m in the latter camp, and it is not incorrect to do so. At least one of the report software providers allows it to be done either way easily, so pick your poison!
Then there are folks who make every adjustment on the grid EXCEPT for GLA a ‘rounded’ dollar amount, but the GLA is shown down to the nearest 1 dollar. C’mon folks, and even some of you reviewers, what is the logic with that? Most value opinions are reported to a ‘rounded’ dollar figure, so why are GLA adjustment amounts done to the $1 figure? To me, that doesn’t make any sense. So round off GLA also! “We aren’t that good to conclude any value down to a dollar!” You can quote me!
In order to arrive at a dollar amount for the GLA adjustment, we have to first determine the PERCENTAGE to use, then multiply that against the appropriate Cost per Sq. ft figure. If you are in the camp that figures each comp separately, it’s easy. But if you prefer making the GLA adjustment a common amount for each comp, it takes a few pecks on the calculator. That’s what I will demonstrate next.
The Sale Price/Gross Liv. Area = Cost per Sq. ft is $165.68, $142.44 and $168.94.
When I do my reports, I analyze these figures and then decide whether or not to use all comps, two comps, or just one comp as the basis to derive the $ figure. Because most homes have differences, the Cost per Sq. ft between them often will be different.
My goal is to find a ‘common’ starting point, or Cost per Sq. ft to multiply the % figure by. It could be done by ‘averaging’ the numbers, but that takes more key strokes on the calculator.
What I do is find the ‘mid point’ value between the numbers (if I’m using 2 or 3 comps). Sometimes the ‘mid point’ is actually the median when calculated, but since I don’t know for sure, I call it ‘mid point.’ Some math whizzes will probably challenge this – oh well!
In the example above, I decided that comps 1 & 3 were most similar due to their Cost per Sq. ft values. The math works like this:
High Number [minus] Low Number [divide by 2], then add back Low Number = mid point
Mid point x % = $ amount for the GLA adjustment (rounded).
$168.94 – $165.68 / 2 = $1.63 + $165.68 = $167.31
$167.31 x 45% = $75.28 Round to $75, applied to all comp GLA difference. Aand then let the report software ROUND the resulting dollar adjustment in the grid!
In this example, if the actual ‘low’ number is used, it looks like this:
$168.94 – $142.44 / 2 = $13.25 + $142.44 = $155.69
$155.69 x 45% = $70.06 Round to $70
Picking the numbers is part of the appraiser’s ‘relativism’ I mentioned above. You just need to be able to explain what you did, and why you did it – disclosed or if asked.
The advantage of this method is the appraiser is in total control of the process. It’s easy to comprehend and to do the simple math. Anyone can replicate what has been done. A spreadsheet or regression program is not needed to arrive at a supportable GLA adjustment.
If you’ve read this far, you know basements have not been mentioned, yet. You’re probably asking “why is the basement living space excluded from the Cost per Sq. ft.” To be honest, I don’t know. I didn’t design the form, or make the rules!
But, in many MLS systems, the basement living space is added to the Above Grade living area (by the agent) so that the listing can reflect the total furniture placement and usable space. Remember, agents are ‘selling space’ while appraisers are quantifying individual areas. Please don’t criticize agents for this; it’s how they are trained, it’s how buyers think, but it’s different from what appraisers do.
When the basement living space is added to the Above Grade area, the Cost per Sq. ft figure will become smaller than the number you calculate using this Above Grade GLA adjustment method.
In your reports when you have basements, you can separately calculate (off report) the Cost per Sq. ft by combining Above and Below grade areas so that the building functions similar to a dwelling built all above grade. Then apply your PERCENTAGE to that figure to determine the GLA adjustment dollar amount. When doing this, you probably should reduce the GLA PERCENTAGE because you’ll also be making a basement adjustment on one or both basement lines. Remember, the basement adjustment is one of the other ‘slices’ of the pizza.
This brings up another anomaly I see in reports. Appraisers adjust basements differently from one another. Remember – 1 to 81,000 ways!
I’ve seen reports where “the basement” is adjusted on the first basement adjustment line, then the “rooms” are adjusted on the second line. Talk about confusing and lots of extra work! I advocate keeping the basement adjustment much simpler.
Most reviewers, underwriters, appraisers, FHA, VA and even the GSE’s have the impression that basements are ‘worth less’ than the above grade areas. Yes, challenge-able and debatable, but let’s just assume that to be acceptable. Then use KISS!
The KISS method I use for basement adjustments is again a PERCENTAGE, but this time, based on the calculated GLA dollar amount. In my previous example, the GLA $ adjustment was calculated at either $70 or $75. I figure a Finished Basement is ‘worth’ at least 80% of the GLA (due to fewer walls, plumbing, electric, etc.), a Partially Finished at about 60% (which varies by the finished space), and an Unfinished Basement at about 20%. Use your ‘relative thinking cap’ to calculate these.
Finished Basement = $56 to $60 (80%)
Partially Finished = $42 to $45 (60%)
Unfinished Basement = $14 to $15 (20%)
Make these basement adjustments on the first basement adjustment line only, and don’t make an adjustment on the second line. You will insert a “0” instead to get it through UAD.
To keep the report simple to read, round the basement adjustments the same way you do all other adjustments.
I round to the nearest $100 because most homes are sold in $100 increments. And I don’t make adjustments for less than $499. I also adjust “all” the GLA and basement areas. It just keeps everything easier to calculate and to read.
Fine tuning … or ‘sensitivity analysis.’
In most appraisal literature, text books, etc., the ideal goal is to have each of the Adjusted Sale Price of Comparables be exactly the same. In the real world, that’s almost never achieved. So the next best outcome is to have the ‘range’ between high and low be as close as possible. Many reviewers and underwriters get knickers twiddled if the percentage or dollar difference between the range is too high. Far too many think real estate is absolutely perfect. Experienced appraisers know it is not.
When using this PERCENTAGE method, you can first use a figure that you believe is acceptable to ‘run the numbers.’ Write down the Adjusted Sale Price numbers. Then modify the percentage a bit higher and a bit lower to get a different dollar amount for the $ per Sq Ft difference. Or in the report software, just change the $ figure for the GLA adjustment you use. Write down the results from the higher and lower $ figure. Calculate the difference of all these samples, and use the GLA adjustment $ figure that results in a tighter range.
Describe the process.
Now that you’ve gotten to the zen moment in your report with the GLA $ figure, the next item is to explain the process. Some people believe it is absolutely necessary to include every single calculation you do to arrive at the math conclusion -> in the report. Well, USPAP does not say that. However, the GSE’s and others instructions are a little different, in that they want to see explanations. An ‘explanation’ can be short and sweet, or a full blown master’s thesis. It’s up to you to decide what to do.
I tend to hang with the ‘short and sweet’ crowd, over here in the corner, or dancing in the center. My explanations look like this:
GLA at $XX s/f (factored from comps XXX price per s/f). I include which comps were used. This is on page 2 in the report in the Summary of Sales Comparison Approach comment section.
In the Addendum, I have extended comments about adjustments made. This is how I explain the GLA adjustment:
GLA is factored as a percentage from the ‘range middle value’ cost per square foot as shown at the top of the comparable sale grid columns for indicated comps, and shown in the comment section below the grid.
The word ‘factor’ is used due to its definition: Noun – One of several elements or causes that produce a result. Technically, ‘factored’ is a Verb because it describes an action or occurrence.
Note that I don’t show the actual ‘math’ calculations, but you certainly can if you think it is necessary. But no one, not even the CU robot, or any other reviewer, has ever challenged me on this method, or the $ per Sq. ft used in my reports. And as I’ve mentioned before, every single report I’ve done (since adopting this method) has a different $ per Sq. ft adjustment from one report to the next. And also from one property type to the next! This PERCENTAGE method is property dependent.
The 10% – 15% – 25% guidelines:
If you’ve paid attention to the revised FNMA Selling Guide and various news releases, blogs, etc. which mention these, you will know that they have been dropped by FNMA. However, many other report users still believe they are vital indicators of report quality.
The reason why these guidelines came into existence was the belief that exceeding them caused the resulting Opinion of Value to be suspect, or questionable. Improper comps tended to increase values due to inflated adjustments. After the UAD mandate was introduced in 2011, FNMA began to recognize (using their Collateral Underwriter robot) that appraisers were purposely ‘fudging the numbers’ just to keep adjustments within those guidelines. So they dropped the guidelines. FNMA wants reports to report reality, not bogus information. You CAN exceed those guidelines, if warranted.
FNMA has finally realized that real estate is imperfect. The square box cannot be pounded into the triangle shaped hole and have the ‘smushed data’ emerge as believable. Other users, however, think it’s possible. They still clutch the 10% line, 15% net, and 25% gross adjustment tight to their chest and expect appraisers to be mindful of this ridged, potentially flawed, perspective.
I bring this up now as a cautionary note, more than anything. And also to re-emphasize that this PERCENTAGE method to derive the GLA adjustment is not based on multiplying a percentage against the property Sale Price. No, it’s a percentage against the Cost per Sq. ft.
However, the resulting actual adjustment $ amount, the pizza slice wedge, for the difference in GLA does have a correlation with the line item percentage. I’d like to suggest that your actual dollar adjustment be kept to 10% or less of the Sale Price whenever possible. Using a PERCENTAGE of the Cost per Sq ft, starting at about 45%, will yield a line item percentage of less than 10% in most cases.
You can achieve this by keeping the comps GLA between 75% and 125% of the subject’s GLA size, or ideally tighter if possible. In other words, search for comps using GLA, not sale price. That way, the GLA adjustment $ amount is relatively low.
The other slices:
Every adjustment line in the grid (up to 24) consumes a hunk of the pizza. They are organized on the form in relative importance. Some of those are ‘stand alone’ like the GLA, but others can be combined into one entry. The first group that can be combined are Location, Site and View. The reason for this is they relate to the allocation of the ‘dirt’ to the actual dwelling, in terms of value.
Allocation is related to mass appraisal, i.e., how the local assessor divides up the contribution of the ‘site’ value to the contribution of the dwelling value. Generally, the allocated ‘site’ value (including Location and View) is in the range of 15% to 50% or possibly more of the total assessed value. Homes in subdivisions tend to have about a 20% to 35% ‘site’ contribution to the total assessed value. Homes on sites with stellar views in “beneficial” locations tend to have higher ‘site’ values. Similarly, homes at or near the end of their economic life tend to contribute less to the overall value. That’s when they become tear-down homes and the site is redeveloped.
The point here is the appraiser needs to employ some ‘relativity’ thinking to determine what, if any, adjustment should apply to these three categories. And the appraiser can make ONE adjustment on one of the lines to incorporate all three factors. The appraiser just needs to explain the process in the report.
Next on the grid for combining are Actual Age and Condition. Both are related. They both incorporate depreciation for short and long term components. Some appraisers like to make an adjustment on both lines. Others, like me, only make an Age adjustment in the report, when possible, while keeping the Condition rating the same across the grid. This avoids actual, or the appearance of, double-dipping, and having to explain both adjustments. Explaining one is simpler!
Related to this is Effective Age, which per FNMA guidelines, can be made on a separate line, not on the Age line. Some appraisers believe their crystal balls or Ouija boards and do make this adjustment. Others, like me, don’t, due to the inherent subjectiveness of the separate depreciable components not actually observed. Effective Age for comparables involves too much guesswork, unless you’ve actually inspected the home recently.
Items below the basement, except for the garage/carport, generally are smaller slices of the ‘pie’ and command a much lower percentage of the total. Making a specific adjustment of 1% – 2% of the sale price for minor items is really questionable, in my view, and may be difficult to quantify with market data. These type items include the heating/cooling, sheds, patios & decks, fences, etc., because direct comparison/observation between the subject and comparables may not be possible.
One other combinable group adjustment is one you may not be aware of. But this one IS allowed by FNMA in their Selling Guide. It can be used when situations involving imperfect real estate are encountered and comparables are lacking. Specifically, it has to do with combining below grade square footage and room count with the above grade GLA/room count. If you encounter situations where the same type building design are extremely limited, but you feel compelled to use one or more ‘imperfect’ comp, you can do this. This happens when design elements are not exactly similar between the homes. Examples: subject is a two story, but comps have finished basements with internal stairs. Or the subject has a finished basement, but comps are multi-story.
Granted, the above is not something done very often. And some reviewers will get their knickers twiddled. But you can use this technique as a last resort in specific circumstances. Just explain what you have done.
Homes with daylight basements:
In this area, homes with daylight basements are marketed in the MLS as having combined living space with the main level(s) of the dwelling. These homes compete with 1.5, 2 story, and Bi/Tri level designs due to internal stairs between levels. There does not seem to be any adverse buyer reaction to daylight basement style homes, especially when the finish in the basement is similar to the main level. Therefore, for the purposes of this report, the main and the basement level living spaces (for the subject or Comps XXX) are combined to report the GLA and room
counts. This is an acceptable appraisal method in this area, and does not adversely affect the OMV as reported.
I’ve gone off on this tangent to help an understanding of how the other adjustments relate to the GLA, and how all the adjustments relate in percentage to the Cost per Sq. ft and to the property sale price. Both are correlated.
Obviously, we don’t want the adjustment percentages relating to the Cost per Sq. ft to be more than 100%. Similarly, the actual dollar amount of the line item adjustments ideally will be as low as possible.
Enjoying the pizza:
The take-aways from this presentation are these:
- Calculating the GLA adjustment using this PERCENTAGE method is simple
- An appraiser does not need a 2 or 4 year college degree to figure it out
- There is no need for an HP12C calculator to be used; a simple calculator works
- This PERCENTAGE method is free to use; no additional software is required
- This method is supportable, logical and easy to replicate
- The appraiser is in full control of the ‘relative’ percentages or actual dollar amounts used which are based on local market evidence
- This method is similar to how regression works to determine adjustments
- This method works for all property types
- This method satisfies the CU robot in how it examines individual appraiser reports for inconsistencies in how the GLA adjustment is indicated
There are up to 81,000 ways to calculate adjustments. Mine is one of those. Use it if you like. Share it with peers if you want to. But if you decide your method is far superior, I don’t need to have spears, bricks or flames thrown my way.