The Infamous Price-Per-Square-Foot Valuation Method

The gold standard for many buyers and Sellers is to try to determine the value of a home in terms of price-per-square-foot.  Price-per-square-foot is calculated by taking the anticipated sales price of a home and dividing it by the livable square footage.  It is then common practice to compare that value to the same value of a group of comparable homes that sold to determine if the subject house is a good buy.  It is human nature to seek out this type of simplicity and hang onto a single quantitative figure of merit.  For example, we have the controversial Body Mass Index to tell us we are overweight.  We have a credit score to tell us we are credit worthy.  We have a GPA in school to tell us how smart we are.  So it is natural to seek a figure of merit to easily tell us we are getting a good deal on a home.  But like all Figures of Merit, they aren't very accurate and are subject to much misinterpretation.

On the surface it may seem like everyone in Real Estate uses price-per-square-foot.  The Realtors have it in their statistics, it's on appraisal sheets, builder's quote teaser rates based on price-per-square-foot, and online sites like Zillow and Realtor.com throw out that statistic.  But does anyone really rely on that metric, and how reliable is it?  Well I can answer that pretty quickly.  A lot of people use it, and it is NOT accurate at all in determining if a house is priced correctly.

Why Using Price-Per-Square-Foot Fails

There are plenty of reasons why price-per-square-foot is a poor metric, but here is what you really need to know.  First, not everything that contributes to the value of the house has anything to do with the livable square footage of the home.  For example, the land cost has nothing to do with livable square footage.  Doubling the square footage of the house doesn't double the cost you pay for the land, but the land price goes into the final sales price.  So doubling the square footage of a home would typically lower price-per-square-foot, rather than keep it  the same, as Price-Per-Square-Foot would imply.  3-car vs 2-car garage, finished basement vs unfinished basement, vs poorly finished basement are all factors.  Second, you never hear a Buyer say "Find me a house in the \$120 per square foot range and I will buy it".  They shop around and compare homes.  Third, if you look at bank owned homes, the best flippers can double the resale value of homes they purchase just by addressing condition.  Condition isn't linked to square footage, but it can account for a large portion of the value, and it can contribute to over half of the price in many cases.  A house can be updated and done poorly.  Curb appeal, lot placement, interior lot vs backing up to a road all contribute to the value differences from one house to the other.  I could go on, but you can see where this is going.

For price-per-square-foot to work, you need compare nearly identical homes.  If homes are not identical, it is not just because of square footage differences, there are lots of reasons.  If homes were identical, there is no need to take the extra step and convert the sales price to price-per-square-foot, you could simply use the average sales price of the comps to determine the value of the home.  That is unless you found that special case where the comps and the subject homes are identical in every way except in terms of square footage.  What a needle in the haystack that would be? When they are not identical, which is the case most of the time, you are far better off using the method that appraisers use the most, which is the Sales Comparison Approach.   That approach adjusts for location, various options, conditions, lot situation, superiority, etc., to normalize the homes, and avoid price-per-square-foot.  Appraisers DO adjust for square footage differences when it is significant, such as greater than 200 square feet on a 3000 sqft home, but they don't adjust using price-per-square-foot.   They use a value such as \$25-\$35 per square foot for Main and Upper square footage in the Midwest, and even less for basement square footage differences.

Demonstrating the Weakness of Price-Per-Square Foot

If Price-Per-Square-Foot was a legitimate method to use to estimate the price of a home based on 5 or 6 sold comps, then you should be able to turn around and apply it to those same comps, for which you have the actual sales price number, and test the results.  The accuracy of which it can be used to predict the sold prices of the comps where you know the sold price, is probably similar to the accuracy you would get applying it to the subject home.  Keep in mind that Sellers usually think their home is better than it really is and Buyers tend to think Seller's are always over priced, so they are each a little biased when picking comps.

For example, below is a list homes a Realtor gave to us recently that they had used as comps against a home we had listed.  The agent was indicating that our listing was \$10K overpriced compared to other similar homes in the neighborhood.  They determined that by calculating an average price-per-square-foot and then applying it to our house.  The homes the Realtor provided as comps are all from the same builder, were all sold new, and were sold within the same year.  The market was stable to maybe slightly upward moving due to short supply.   The homes all had 5 bedrooms, 3 bathrooms, 3 car garage, partially finished basements, and were in the same neighborhood on comparable lots.  I used a spreadsheet to evaluate these comps to see how close the average price-per-square-foot value they derived from them came to predicting their known sales prices.  Table 1 holds the data and the results.

Table 1 – Accuracy Assessment of Price Per Sq Ft Calculation vs Actual Sold Homes

 5 Bedroom, 3 Bath, 3 Car Garage in Same Neighborhood Adjusted Price Using \$95.87 Adjusted Price Using \$78.53 Address SFM/U SFtotal Price \$/sqft (M&U) \$/sqft (Total) Method 1 Error Method 2 Error 14481 Livingston LN 3236 4148 \$   351,500.00 \$     108.62 \$        84.74 \$  312,129.22 \$     39,370.78 \$  319,476.97 \$     32,023.03 14518 Camden Ln 3478 4298 \$   337,875.00 \$       97.15 \$        78.61 \$  335,471.39 \$       2,403.61 \$  331,029.90 \$      6,845.10 2652 Newington Ln 3478 4298 \$   331,370.00 \$       95.28 \$        77.10 \$  335,471.39 \$      (4,101.39) \$  331,029.90 \$         340.10 14481 Livingston Ln 3236 4148 \$   325,000.00 \$     100.43 \$        78.35 \$  312,129.22 \$     12,870.78 \$  319,476.97 \$      5,523.03 14460 Livingston Ln 3478 4298 \$   324,695.00 \$       93.36 \$        75.55 \$  335,471.39 \$    (10,776.39) \$  331,029.90 \$     (6,334.90) 14521 Camden Ln 3464 4378 \$   318,890.00 \$       92.06 \$        72.84 \$  334,121.01 \$    (15,231.01) \$  337,191.46 \$    (18,301.46) 14480 Timmis Circle 3478 4298 \$   310,935.00 \$       89.40 \$        72.34 \$  335,471.39 \$    (24,536.39) \$  331,029.90 \$    (20,094.90) Totals: 23848 29866 \$2,300,265.00 \$       96.46 \$        77.02 Min \$    (24,536.39) Min \$    (20,094.90) Max \$     39,370.78 Max \$     32,023.03 Error Range \$     63,907.17 Error Range \$     52,117.93

The spreadsheet in Table 1 sums together the sales prices and sums together the square footages, and calculates two price-per-square-foot values.  The first is the price-per-square-foot average using Sales Price summation and Main and Upper square footage summation.  The second is the price-per-square-foot average using the Sales Price summation and Total square footage summation.  I calculated both because I find people use both when deriving their price-per-square-foot.  The two columns labeled Method 1 and Method 2 hold the results calculated from taking the two average price-per-square-foot values and multiplying them by their corresponding square foot measurements.  Those results represent the estimated sales values of the homes for which we know the actual sales value.  The two Error columns show the pricing error between the known sales price and the estimated sales price.  The spreadsheet also calculates a min/max error value from each of the two error columns so that you can determine the range of errors each method produced.  The red cells hold the error ranges.  The error range for Method 1 was approximately \$64K and the error range was \$52K for Method 2.  For Method 1 that is close to a 20% error range.

So in this particular example, the Agent and/or their Buyer were trying to convince me my house was over priced by \$10K using a set of comps they assumed were sufficiently identical, and a pricing method that yielded a \$50K-\$60K error range.  That doesn't make for a convincing argument does it?  I've used this method to test the accuracy of PPSF many times and the results are similar each time, sometimes I get a range of \$200K on \$500K homes.  What this means to me is that in a world where the top Realtors can predict sales price to within 0%-3%, they can't be using price-per-square-foot as their tool no matter how prevalent the term may be.  It is certainly not something we ever use.

Origin of this Technique

So if price-per-square-foot is that inaccurate, how did it become so widely used by Buyers?  It is hard to say exactly why the price-per-square-foot standard became so popular among buyers, but it probably has something to do with a combination of creative advertising from the big box builders who advertise more square footage for the price, the dominance of production builders in general who build similar houses within a neighborhood, a lack of formal training in proper methods for home valuations within the industry, including for lenders, and its historical use by custom builders, insurance adjusters, and appraisers for rough estimating new construction costs.  A lack of available data within the public domain probably drives some of this "go with what you know" approach as well.  It may also be further propagated by new and inexperienced real estate agents who have not yet figured out the correct way to evaluate comps, and perhaps some who never will, but who somehow find a way to survive in spite of it (most likely as Buyer Agents rather than List Agents).  Perhaps in a lot of ways, plus or minus 10 percent is reasonably close estimate for most people?  But when you are dealing with big numbers such as with a home purchase, +-10% or worse is a big range.

Errors in Square Footage Quotes Don't Help

Quoted square footage values on listings aren't even that accurate.  As Realtors we have seen cases where the same exact floor plan within the same neighborhood varied 200-400 square feet.  At \$100 price-per-square-foot that would result in a \$20,000 to \$40,000 price difference.  Realtors and assessors often copy those misquoted numbers across many homes rather than measure on their own, or obtain the correct info from the builders.  In some cases it may be a mistake (such as using the square footage from another listing with the same floor plan that had a bonus room, while the current home does not, or because the homeowner made modifications without obtaining a permit to alert the Assessor), but in other cases there seems to be a pattern to suggest some fluffing may be occurring to make the Buyer feel like they are getting a better deal.  Sellers often overestimate their square footage when providing it to the agent. It is best to verify consistency where identical floor plans are available.  In most cases the square footage used by Realtors comes from the Assessor.  On new construction, the builder data is used.  The Assessor estimates the square footage rather than actually measures it, or in the case of New Construction, he obtains the info from the Builder Permit.  On reassessments, the Assessor rarely goes into the home.  They often ask the homeowner to verify information.  Regardless, don't put too much weight on reported square footage.  You rarely see a Buyer come in with a tape measure to verify square footage before making an offer.  That should tell you a little about the weighting factor.  They will come back two or three times to look the house over, but usually not to verify the exact size.

Statistical Errors

Without delving into statistics too much, be aware that if you take the highest and lowest sales prices within a given neighborhood and convert that to price-per-square-foot, the variation can be great.  Averaging several homes smoothes things over, but the data can be skewed to one side or the other depending on the types of homes that have sold recently.  Also, you can have an average price-per-square-foot and there be no homes that sold at or significantly close to that average price-per-square-foot.  Think about the height of the starters for a basketball team.  If you had two 7' players and three 6' players, the average height would be 6’4".  There are no players on the team at 6’4".  Is it fair to say that the 7' players are too tall or that the 6' players are too short?  The same applies for price-per-square-foot when using limited data.  So even though lender's prefer the appraiser to use comparable homes sold in the last 6 months, you have to be aware that there may be more direct comparable homes in the neighborhood that have not sold recently.

For the type of stratified sampling needed for price-per-square-foot to work, the population of homes you choose for your evaluation must be direct comps.  All homes in the neighborhood are not direct comps.  Think of a Buyer looking at homes on a lake.  If the homes he likes are waterfront properties, then the homes across the street that are not water front properties are not direct comps, even if the houses otherwise are identical.  As another example, in a production neighborhood if you are comparing a premier series home on an interior lot, to a lower series home backing up to a road, that is not a direct comp either, even if they have the same same builder, same square footage, and are both in pristine condition.  Look for other differences as well that might affect building costs, and also trigger buyer interest, such as two story family rooms, 9' basements, interior lots, trim packages, etc.  These issues and the errors mentioned earlier are just part of many reasons price-per-square-foot is a bad technique for valuing a home.

What Then is the Proper Use of Price-Per-Square-Foot?

Price-per-square-foot does have its use in real estate, but maybe not so much in the determination of re-sale value of a detached residential property, or even it's rental value.  It probably works reasonably well on condominiums for example, and is certainly used in commercial space applications.   Average sales price-per-square-foot taken over time can be a useful tool in determining if the prices of homes in the local area are rising or falling and by how much.  The median price-per-square-foot can be useful in determining if the affordability of existing homes or new construction is rising or falling.  But there are other methods for this as well.

Price-per-square-foot is used by insurance adjusters to roughly determine the replacement cost of homes in case of a fire or tornado.  You can always over insure to make up for any slop.  It can be useful when evaluating builders in estimating construction costs for specific sized homes and with a standard set of options (i.e. when you are providing the lot).  However, there can be a wide variation in construction costs, particularly between production builders and custom builders.  Most custom neighborhoods command a higher resale price, thus justifying the higher construction price-per-square-foot initially.  Their costs are reflective of several factors including differences in quality, uniqueness in design, additional gables or more complex blue prints.  Custom builders don't enjoy the economy of scale savings that production builders have, so custom homes cost more just in materials.  Even in production neighborhoods there are wide variations in price from floor plan to floor plan and between different front elevations, as well as trim options or upgrades within that same floor plan.  These are typically due to layout and material quality.  When considering custom builders, watch out for signals of potential estimation errors, such as demanding a cost plus contract vs a fixed price contract.  In any case, most builders won't quote you a price-per-square-foot value, they will quote you an estimated contract price with limitations or budgets on options.  The price-per-square-foot exists as a statistic, but not a particularly useful one when pricing out your custom home.  Finish level and quality are big factors.

Other Factors to Consider

When considering the subject home or its comps, looking at initial sales price of the property can be an important factor in determining value.  For example, if a comp in the same price range looks dramatically better than the subject property, see if there is a past history that shows it once selling at a higher price.  This could be a sign that the condition of the home was not as good as the pictures indicate, and could indicate that to a shrewd buyer, the home was a steal at that price.  Couple that with a comparison of the average price-per-square-foot in the neighborhood to determine direction and magnitude of price movement.  I once sold a house that was the highest priced home in the neighborhood, which are always tough.  A potential Buyer tried to convince me, based on price-per-square-foot, that the house was \$150K higher than it should be.  The history on the house showed that it had three times before sold in the \$900K range, which is where I had it listed.  The median neighborhood price had been about the same through all three of those sales, and the condition of the home was in line with the neighborhood.  To me that was a much better indicator of final sales price than some hokey price-per-square-foot comparison.  In the end, that home sold more in line with my expectations than it did with the price-per-square-footer's expectations.

In an increasing or decreasing market, you should have an understanding of what is driving the increase or decrease in average and median price-per-square-foot and how that might affect future pricing.  Is it due to material cost, labor, building restrictions, inflation, deflation, economic collapse, financial collapse, or some other important factor?  Just because the price seems in line with today's market, doesn't mean today's market is sustainable.  Anytime housing inflation is growing faster than wage inflation, there will be an adjustment coming in the future.   Low interest rates can fuel housing inflation because it lowers payments.  If interest rates are starting to rise, it could be a problem for homes at the higher end of the price range.  A tight money market can artificially lower the price of homes, making bank owned homes very lucrative.  There is no substitute for years of experience working within the market to be able to recognize which homes will command a premium vs those that will not.  Buyers should always work with an experienced agent who services their area of interest if they truly want to find a good deal.  Some homes push the market higher; others do not.

Remember Price-per-square-foot Ignores Features that Affect Price the Most

Below are a list of items that significantly affect price, but ARE NOT tied to square footage.  Unless the buyer or Realtor has visited the comps personally, the pictures or a drive by only tell part of the story, so leave some padding in your estimate.  Photographers are clever and don't usually focus on the bad points of a home.  Consider the following:

• Elevation of the home (even production builders have significant cost variation)
• Floor plan, such as an open floor plan vs closed
• Two-Story rooms and vaulted ceilings vs All Single Story flat ceilings
• Interior and Exterior Condition (i.e. shows like a model vs looks worn and needs paint and carpet)
• Three-car vs two-car or attached vs detached garage vs no garage (some condos)
• Professionally finished basement vs owner finished basement, vs unfinished basement
• Walkout basement vs Daylight basement vs Egress window basement vs No Window basement, vs Partial Basement, vs No Basement
• Lot condition such as preferred interior lot on cul-de-sac, vs corner lot, or backing up to road, vs backing up to pond, vs mature trees, etc..  This variation can is amplified  on lakefront or beachfront properties vs those on the opposite side of the road.  The identical house on the non-preferred side of the road can cost a fraction of those on the water.
• Lot Size or acreage 1 acre vs ˝ acre vs 1/3 acre (lot premium can significantly affect price)
• Home placement (nice backyard vs no backyard)
• Number of gables vs big box home (affects actual construction price-per-square-foot)
• Overhead power lines or intrusive easements
• Vinyl Siding vs Hardi-Plank vs all Brick
• Pool vs No Pool
• Hardwoods and Tile vs Vinyl
• Granite Countertops vs Solid Surface vs Formica
• Limited counter space or cabinets
• 8' ceilings vs 9' ceilings
• Functional obsolescence (limited number of bathrooms, small closets, little or no insulation, etc.)
• 3 bedrooms vs 4 bedrooms
• Downstairs Master vs Upstairs Master in certain neighborhoods
• Neighborhood to Neighborhood Variations
• Well manicured lawns with sprinkler system, fence, and privacy from other homes or public access will greatly be appreciated by buyers over unkempt or hard to maintain lawns backing up to other homes and without any privacy.
• Landscaping
• Builder brass vs updated lighting and fixtures
• Proximity to a busy road, railroad, or public buildings vs quite interior neighborhood setting.
• Financial predicaments such as divorce or foreclosure.
• Builder still in neighborhood vs closeout phase, vs builder out
• Paint and Carpet Colors or wallpaper
• Mold
• Age of Home
• Property Taxes and Schools
• Stigmatized Property
• Poorly maintained property next door

As you can see in the list above, there are numerous factors that should be adjusted for that are not considered in the majority of straight price-per-square-foot comparisons.  Buyers will often pick the home within their price range that is in the best condition, or has the best lot, best curb appeal, more functional floor plan, or most desirable options overall, and then they want to compare that home to lower priced homes that they really wouldn't have been interested in otherwise.  Seasoned Realtors know that most Buyers in this market are reluctant to take on a "Project” house.  They also know Buyers tend to overestimate the cost of repairs in general when buying, but underestimate those same costs when choosing homes to compare price to.  You can bet though that it won't be price-per-square-foot driving a Buyer's initial selection of a home, but it could be driving their rejection of the home.  As I mentioned earlier, you never have a Buyer say "find me a house in the \$120 per square foot range and I will buy it", but you occasionally hear that the Buyer is walking away because he feels the house is over priced based on price-per-square-foot.

Price-per-square-foot implies that a group of houses are identical except that they differ in square footage.  It also implies that each square foot added or subtracted proportionately changes the features about the house that are not tied to square footage, such as the lot cost, the cost of the three car garage, the cost of the landscaping, etc.  Understanding the error in price-per-square-foot valuations can sure help save a deal when a deal starts to fall apart because of it.

Remember that no matter what pricing model you use, a house is worth what a Buyer is willing to pay and what a Seller is willing to take, unless, of course, there is third party involved.  That third party is usually the bank.  Banks can have a dramatic affect on the price of homes either by requiring more money down, commanding higher interest rates, and placing limitations on buyer qualifications and appraisals.  When the government gets involved, such as with mortgage backed securities, they can tie the hands of all banks, through Fannie Mae guidelines for example, so that they behave in a similar way.

Jeff Wilson

The Wilson Team

F.C. Tucker Company