Excessive GLA and improvements

[Workshop]

I have an appraisal on a home that is overly improved for the area. The GLA is greater than 8,000 square feet than any other home in the area. This has become a bear. Any opinions?

 

[Metamorphic]

I have an appraisal on a home that is overly improved for the area. The GLA is greater than 8,000 square feet than any other home in the area. This has become a bear. Any opinions?

Measure the market. Develop a basis for discounting the value of the over-improved footage.

Dont forget to look for the last sale of the subject. Comparison of that sales price with the market at the time may give you a clue as to how it would relate today.

Also, make sure you've throughly explored HBU. Something that size may have comps and a use in the commercial genre.

 

[Workshop]

I have ana djustment value to work with. This is not a property for commercial purposes. Three counties bordering the subject have been considered. There simply is nothing of this size. I am considering placing a cap on the adjustment, i.e. the GLA is 17,000 and the largest comp is 12,000. I would apply dollar adjustment up to the 12,000 mark. All other comps max at 7,500. I also am considering capping the bath count to 6 and half vs. the 8.5.5.5.5.

 

[Metamorphic]

I have ana djustment value to work with. This is not a property for commercial purposes. Three counties bordering the subject have been considered. There simply is nothing of this size. I am considering placing a cap on the adjustment, i.e. the GLA is 17,000 and the largest comp is 12,000. I would apply dollar adjustment up to the 12,000 mark. All other comps max at 7,500. I also am considering capping the bath count to 6 and half vs. the 8.5.5.5.5.

Having an adjustment is not the issue. The issue is how that adjustment decreases as you look at increasingly larger footages. A 100/sf might be a great rate for the difference between a 3000 sf house and a 3500 sf house. But if you apply that rate between a 5000 sf house and a 15,000 sf house you'll likely over-appraise the 15,000 ft house all other things being equal. As you look at increasingly larger houses you will reach a point of diminishing returns. Its even reasonable to expect that there is a point at which your GLA adjustment number would be negative; which is to say the "typical buyer" would not tolerate that large a house because of taxes, cleaning, heating, maintenance, etc), and would actually pay less for a smaller house to save those costs. That's a big part of the reason all the mansions on the Gold Coast were torn down in favor of smaller mini-mansions; the negative affect on value from the over improved robber-barron style mansions shifted the HBU from mansion to something else.

Call it a sliding adjustment scale, or a cap or whatever you want, you'd be advised to have some market based evidence for where you set it.

Here's some analysis I put together for a overbuilt-for-market property. The red polynomial shows a decreasing value for additional square footage. As you can see my comp was larger than anything that had sold in the market in the 3 years preceding the valuation.

That curved polynomial is basically the graphical representation of an over-improvement phenomenon that you're trying to describe. You could illustrate it for your report like I did. Or you could do it by looking at historic sales of over improved properties (with the subject being the one that's most germane), and try to extract a read on how value diminishes at extreme size, or you go where ever you need to to find a comp and then make a very difficult location adjustment.

 

[hastalavista]

What a great demonstration (thanks, Meta).

Just as a visual, it would seem to me that the two upper-priced sales creating the upper value range have influences affecting their value that are not related to GLA; likewise, the lower set of sales have similar influences (not related to GLA) that creates the lower boundary.

With the exception of the last data point on the linear regression line, it seems that most of the values for upper-end GLA homes are clustered around the $775k to $850k range. That seems like a pretty "tight" range for large-sized homes (all other things being equal).

Given this data set, and assuming there isn't anything else embedded in the data points, I'd be inclined to view my subject's value in the $800k to $850k range; the single high-sale point on the linear regression line would be an outlier to me.

In my opinion, $850k appears to be this neighborhood's top-end; anything more than that, the likely buyer would opt-out and go to another neighborhood with higher overall values.

Meta-

How did you interpret the data?

 

[Metamorphic]

What a great demonstration (thanks, Meta).


In my opinion, $850k appears to be this neighborhood's top-end; anything more than that, the likely buyer would opt-out and go to another neighborhood with higher overall values.

This was actually done for a field review. The OAR put it at $1.25mil. I appraised it at $875k.

One of the things I really like about the indexed price analysis is that it provides some sturdy guardrails to the valuation problem. Even if a person doesn't really understand appraising, or the process that you went through to make the chart, they are usually comfortable with the idea of a top, bottom, and middle of the market. Within that frame work they readily accept that some specific factors about the property will push it towards one end of the range or the other. Net result is even a noob can quickly play a "that pushes it up but this pulls it down" game and satisfy themselves that the overall magnitude of the valuation is correct. At that point your OMV becomes just a fine tuning of the value they've already bought into.

Meta-

How did you interpret the data?

INDEXED SALES PRICE CHART EXPLANATION AND ANALYSIS:
THIS CHART COLLECTS ALL SALES FROM THE NORTH LAKE TAHOE, AND TAHOE CITY MLS AREAS, SINCE 1/1/06. ONLY PROPERTIES WHICH LACK SUBSTANTIAL VIEWS OR LAKE FRONT AMENITIES, BUILT BETWEEN 1970 AND 1990 (AS INDICATED IN THE MLS) WERE INCLUDED IN THE ANALYSIS; THIS WAS DONE TO TALOR THE ANALYSIS TO THE SUBJECT'S SUB MARKET. A LINEAR REGRESSION OF THE SALES PRICE PER SQUARE FOOT OF THESE PROPERTIES OVER TIME WAS PERFORMED TO DERIVE A MARKET CHANGE ADJUSTMENT VALUE. USING THIS ADJUSTMENT VALUE THE SALES PRICES WERE ADJUSTED TO INDEX THE SALES PRICES OF THE PROPERTIES TO THE DATE OF VALUATION IN THE OAR (MAY 09). THE RESULTING INDEXED SALES PRICES ARE PLOTTED AGAINST THE SQUARE FEET OF GLA OF EACH SALE. THROUGH THIS PROCESS A LARGE SET OF TIME ADJUSTED SALES DATA IS GATHERED THAT COLLECTIVELY DESCRIBE THE RELATIONSHIP BETWEEN THE SIZE OF A PROPERTY (IN TERMS OF GLA) AND ITS MARKET VALUE.
FOR A PROPERTY OF A GIVEN SIZE, MARKET VALUE CAN BE ESTIMATED BY DRAWING A VERTICAL STRAIGHT LINE UP FROM THE DESIRED GLA, AND THEN A HORIZONTAL LINE LEFT ACROSS TO THE EXPECTED SALE PRICE. HOW FAR UP ONE GOES BETWEEN THE TWO LIMIT LINES (ORANGE) DEPENDS ON THE RELATIVE LEVEL OF QUALITY, CONDITION, UPDATING, AND AMENITIES RELATE TO THE REST OF THE MARKET. AN ADDITIONAL VERTICAL LINE REPRESENTING THE SQUARE FOOTAGE OF THE SUBJECT (BLUE) HAS BEEN PLACED ON CHART 4. THE INTERSECTION OF THE BLUE AND ORANGE LINES REPRESENT THE UPPER AND LOWER LIMITS OF VALUE FOR THE SUBJECT BASED ON THE 103 CLOSED SALES INCLUDED IN THE ANALYSIS.
SEVERAL IMPORTANT ELEMENTS OF PROPERTY VALUE CAN BE DERIVED FROM THIS CHART. FIRST, IT IS OBVIOUS THAT THERE ARE NO DIRECT SIZE COMPARISONS FOR THE SUBJECT; STRONGLY SUGGESTING THE SUBJECT IS OVER-BUILT FOR THIS MARKET. SECOND, THE SUPPORTED VALUE RANGE FOR THE SUBJECT IS BROAD, RANGING FROM ABOUT $750K TO OVER $1.5 MIL. THE MIDDLE (DEFINED BY THE PROJECTED LINER REGRESSION OF THE DATA - GREEN LINE) SHOWS A MEDIAN (AVERAGE) VALUE OF JUST UNDER $1 MIL. HOWEVER THE RED, POLYNOMIAL REGRESSION CURVE, SHOWS A TREND OF DECREASING VALUES FOR OVERSIZED PROPERTIES AND SUGGESTS A MEDIAN PRICE OF ABOUT $800K. THE RED REGRESSION CURVE MORE ACCURATELY REFLECTS THE MARKET AS WITHIN ANY MARKET THERE IS A POINT OF DIMINISHING VALUE WITH ADDITIONAL SIZE. ALL MARKET INDICATIONS SUGGEST THIS PROPERTY HAS REACHED THAT POINT.
GIVEN THAT THE SUBJECT IS IN AVERAGE CONDITION (AT BEST) RELATIVE TO COMPARABLE PROPERTIES IN THIS MARKET, AND HAS SIGNIFICANT FUNCTIONAL UTILITY ISSUES RELATED TO ACCESS EASEMENTS AND COVERED PARKING, A PRICE BETWEEN THE $750K (THE LOWER LIMIT OF VALUE- LOWER ORANGE LINE) AND $950K (THE MOST OPTIMISTIC INDICATION OF VALUE FOR AN "AVERAGE" CONDITION PROPERTY WITH BELOW AVERAGE UTILITY AND MARKETABILITY FACTORS - INDICATED BY THE GREEN LINE) IS SUGGESTED BY THIS ANALYSIS.

Sorry about the all caps.....

 

[Captain Common Sense]

That's pretty awesome Meta.

 

[stangtime]

Oh, Meta, I meant to ask you the last time I saw you were doing some linear regression; do you just use the graphs in Excel with your own designed spread sheets or have you found some cool software for your graphs?

Thanks,
Dan

 

[Workshop]

GLA Market Analysis

My word! That is some of the most impressive work I've in some time. Thanks for the demonstration.

I'll do my best to follow your lead.

Thanks.

 

[Metamorphic]

Oh, Meta, I meant to ask you the last time I saw you were doing some linear regression; do you just use the graphs in Excel with your own designed spread sheets or have you found some cool software for your graphs?

Thanks,
Dan

That's just straight up excel. The only real trick to it is to turn the equation from the regression curve you put on the Sales-Price vs. Time analysis into an adjustment to the sales price of the properties used in the analysis.