Pricing homes is often more art than science. Some of the factors that play into a recommended pricing range are nearly impossible to measure or quantify. For example, how do you put a price on location? Everyone knows the old adage that the three most important factors in real estate are location, location and location. But how do you systematically assign value to such a complex quality?
The practical answer is that we look for similar homes that have either recently sold or are currently on the market. We then make adjustments to account for differences between the properties. One day, there will be databases that can quantify what we currently consider impossible to measure, but we’re not there yet. In the meantime we visit as many properties as possible so that we can rely on basic metrics and human judgement.
Price per Square Foot ($/SqFt) is one of the most common basic metrics used to quantify the intangibles. The $/SqFt of homes often differs greatly between towns as a result of the desirability of individual communities. Perceptions about crime and schools play into the values, as do the physical characteristics of the housing stock. The $/SqFt can even vary widely within towns if certain neighborhoods or streets are highly desirable addresses.
I decided to take a closer look at how the $/SqFt varies within one particular school district in one particular town. The main question I wanted to explore was, “Could I calculate the list price of a home if all I knew was how big it was?” Here’s what I discovered about the 28 homes currently on the market in the Norfeldt school district in West Hartford.
As one would expect, there is much less dollar variability in price for smaller homes as there is for the larger properties. Interestingly, there are two homes that are both about 5,500 square feet that are priced about $800,000 apart. There must be a number of factors at work to produce such a large price difference.
Knowing the size of a home is a surprisingly good predictor of its list price in this particular school district. Excel reports that our very simple linear model for pricing homes accounts for over 87% of the variability. So if I had a 2,000 square foot house to sell, the model suggests pricing it at $352,570.
Now let’s calculate the $/SqFt for all these properties and see what that tells us. First we’ll look at it as a function of the home’s price.
The best fit line once again slopes up and to the right – more expensive homes tend to cost more on a $/SqFt basis. I suppose this shouldn’t be surprising since high-end properties are more likely to have luxurious extras.
The most interesting result of this chart is that there is much more variation at the lower price points. Perhaps this is best explained by my friend Alan’s observation that here in Connecticut (as opposed to The South) two adjacent homes, even though they look exactly the same on the outside, can be completely different inside due to their level of updating. His previous experience had been that all houses in the neighborhood were very nearly identical both inside and outside since they were built as part of a large, recent development.
Out of curiosity I also plotted the $/SqFt as a function of the home’s size. In this case there is very little relationship and lots of variation at all sizes. Anecdotally, it is common to see smaller homes that are impeccibly upgraded and maintained with the highest $/SqFt.
Although it is by no means a perfect metric, understanding $/SqFt is important for both buyers and sellers as a proxy for the many intangibles of real property. Buyers should think critically about whether a property in which they are interested warrants a premium price. And sellers need to consider $/SqFt relative to their competition when deciding what to ask for their home. $/SqFt will likely be one of the many supprting data points raised during negotiations – but only after buyers fall in love.