I was hired last week to perform an appraisal by a homeowner who is getting ready to put their home on the market. As you may already suspect, he was disappointed in the outcome and said his realtor, who sent him my way in the first place (likely to interject some logic into an emotional situation), felt that a price per square foot analysis placed the property at a higher value.
Almost nothing makes an appraiser cringe harder than a discussion that includes price per square foot analysis. While it can be reasonably accurate in an area that has a high degree of conformity, it isn’t as applicable when an area has a broad range of home styles, ages, designs and utility.
The subject property I was asked to appraise is an older home in Downtown Orlando, an area where vacant land is almost non-existent and home prices range from $195,000-$1,500,000. Upon submitting the appraisal, the gentleman requested some clarity regarding the $50 per square foot adjustment made to account for the differences in living area between the comparables and his property. His request for clarification is as follows:
“By using such a low value and applying it to all 5 inferior sized comps means that there is no balance in the adjusted metric. The value for size is drawn down and thus giving your suggested values for our property the lowest price per sqft.
I would be curious to see your adjustments with a larger sqft. comparable property such as 1630 E. Livingston St. (2/2, sgl fam, sold 5/16/2019 for $377,000). ”
That all seems very reasonable and I completely understand his confusion. It would seem that while on the surface, this logic was not applied in my appraisal, however there is context missing, and as you likely already know: I’m all about context.
Adjustments within an appraisal report are market extracted using sensitivity analysis. To better understand how that might work, picture making the adjustments in the same way you would add stones to a double pan balance scale. If you are too heavy handed on one side, you lose balance on the other. Let’s say we changed the adjustment from the supported $50 per square foot, to $100 per square foot. Below, you’ll see a comparison of the original adjustments for price per square foot at $50 and the second at $100.
Original appraisal with $50 adjustment for differences in square footage:
Test for reasonableness with a $100 adjustment for differences in price per square foot:
You can see that not only did raising the adjustment to $100 not get us in the upper $300,000 range, which is where I suspect this seller wanted to be, but the figures are less reliable and unsupported. The gross adjustment totals are now higher and Comparable #2 appears to be an outlier. When I see a significant outlier among my range of comparables, those generally get removed. An appraisal is intended to demonstrate support among several comparables, not just one, that a certain value estimate is reasonable. With the new adjustments of $100 per square foot, these three comparables now range between $329,500 to $353,000, a difference of $23,500. This is far less narrow, and reliable, a range of indicated value than the report I submitted which ranged from $326,500 to $337,500, a difference of $11,000. The more broad the range of indicated value, the more room for interpretation and therefore more room for bias and error.
The owner provided me with the perfect example of the overarching concepts behind sensitivity analysis and conformity with the sale on E. Livingston. I did, in fact, analyze that sale on Livingston and the adjusted sales price not only lacked correlation, but was significantly lower than all the others and so I had omitted it.
The original comparable sales had adjusted sales prices of $326,500, $336,500, $337,500, $331,000 and $329,000, respectively. This range in adjusted sales prices is very narrow, with minimal adjustments required, and provides a reliable range of indicated value for the subject property of between $326,500-$337,500. At the owner’s request, I added the sale on E Livingston to the software grid as new Comparable #6 (see below) and made the adjustments it in conformity with the other 5 comparables.
In spite of it’s proximity (mere blocks) and similar bed/bath count, it fails to support the other data in the report, having a grossly lower adjusted sales price of less than $300,000. Here it is added as Comparable #6 so you can see what that comparable does to the schematic…
Using the same adjustment values that were used for all the comparables in my original appraisal report brings that comparable down to under $300,000; and that’s being VERY gentle on the adjustment for condition which should be significantly higher. That home had recently been completely remodeled, with modern finishes, and people are willing to pay a major premium for modern design. Our subject property is in good condition, but it’s materials and color scheme are reflective of what was in fashion approximately 10 years ago, not today.
The reason it adjusted so significantly lower is that…
a: it was considerably larger
b: it was in superior condition with modern updating
c: it’s site size was almost three times that of our subject.
When a property is located in an area that has a vacant land shortage, paired with strict zoning and building requirements, site size is a major factor in value. This is a great example of how a price per square foot analogy would fail given that no consideration in that basic formula is given to anything other than living area. Price per square foot analogy = price/living area.
Because of this property’s downtown location, the site contributes to at least 50% of that of $247.74, likely a little more. The remainder of that $123.87 has to be shared by not only square footage, but amenities like the garage, fireplace and porches, appliances, condition factors and quality factors. The adjustments of $50 for square foot are made to consider what the market will pay for DIFFERENCES in size between nominally smaller or larger homes. It is not intended to consider cost and it does reflect the concept of a diminishing return. The market gives us these figures; they are not arbitrary. Using homes that are significantly larger or smaller results in a skew that becomes less and less reliable.
When using this “sensitivity” method of adjusting comparable sales within a Sales Comparison Approach, it is critical that the sales selected for comparison be similar to the subject, and represent both smaller, larger or the same sized units of comparison. In this case, Comparable #3 was essentially the same size as the subject property, being 3 feet smaller. 3 feet is 2 one hundredths of a percent difference, so it would not be recognized by anyone as being “smaller”; it’s as close to identical as one can possibly expect. For that matter, many appraisers would not even have made an adjustment for the nominal difference in square footage for Comparable #1 given that it’s only 53 sq/ft smaller. Many appraisers don’t address anything under 100 sq/ft difference. I happen to feel that in smaller spaces, it is important to reflect at least a nominal affect on marketability.
The application of Price Per Square Foot analogies is far too basic, inept and broad to ever be considered reliable in this, and most, Orlando markets. At the market value estimate of say $330,000 for this particular house, that provides a price per square foot of $247.74. Keep in mind, the smaller the house, the higher the price per square foot. It is highly likely that the agent that is representing this seller is using smaller homes, with higher prices per square foot, and then applying those figures to the subject. The rub is this, very few agents actually understand how to adjust and reconcile prices per square foot when presented with multiple examples. When seeking representation, make sure to find an agent who thoroughly understands the more relevant principals of “substitution” and “supply and demand” that are at work in your marketplace.