The valbots are coming
Via the Washington Post’s technology columnist comes an update on an apparently-long-overdue innovation that, like brokerage’s baby bang, might revolutionize another part of the home-selling services market:
When I saw a demo of the Zillow.com real estate service last month, it struck me as so obvious I wondered why no one had done it before.
It’s hard to quibble with the company’s goal — “free, instant valuations and data for 60,000,000+ homes.” You type in any address, and in most cases Zillow will spit out a free estimate of the property’s market value.
Then when Zillow launched on the Web last week, I realized why.
Offering automated property valuations via the Internet turns out to be much harder than it seems — especially if you expect them to be accurate.
At this point the consulting real estate professional murmurs, “What does ‘accurate’ mean?”
“The proper price,” Watson replied.
Watson wondered how the internet could work before the advent of electricity.
But, Holmes might have asked, is there such a thing? Does ‘the proper price’ ever exist?
Is there such a thing as ‘the proper price’?
Any given transaction is a trade — good for good (barter), or good for money (sale) — and as we all know, each individual trade is a highly personalized event. A market is a place — physical, electronic, or otherwise — where many trades occur in a continuing sequence. Thus an item’s market price is the expected outcome of the median possible trade of that item.
There’s one and only one way to find out for sure: sell the object. The only definitive proof is a trade, and that is irreversible. So we have the classic quantum-mechanics problem, where we do not know with certainty where the electron is at any given moment.
In Bohr’s model, electrons were planets orbiting a star.
In quantum mechanics, an electron is a cloud of probability density.
In the same way, from an appraisal perspective, we cannot know with 100% certainty what an object’s value is until we lose the object by selling it.
We had to sell the property to value it properly.
Since the appraiser cannot know the property’s value for absolute certainty, we base our decisions on estimates of value. In real estate terminology under the Uniform Standards of Professional Appraisal Practice (USPAP), it’s typically defined as Fair Market Value, with a definition something like this (quoting Fannie Mae residential appraisal form 1025):
DEFINITION OF MARKET VALUE: The most probable price which a property should bring in a competitive and open market under all conditions requisite to a fair sale, the buyer and seller, each acting prudently, knowledgeably and assuming the price is not affected by undue stimulus.
Implicit in this definition is the consummation of a sale as of a specified date and the passing of title from seller to buyer under conditions whereby:
1. Buyer and seller are typically motivated
2. Both parties are well informed or well advised, and each acting in what he or she considers his or her own best interest
3. A reasonable time is allowed for exposure in the open market
4. Payment is made in terms of cash in U. S. dollars or in terms of financial arrangements comparable thereto
5. The price represents the normal consideration for the property sold unaffected by special or creative financing or sales concessions granted by anyone associated with the sale.
Most importantly, any appraisal is an opinion. Can a bot perform one as well as a human?
“In my opinion, yes.”
Appraiser’s knowledge and insight: the ‘expert system’ approach. What distinguishes analysis from intuition, physics from alchemy?
“Irreproducibility.”
Replicability and algorithms. If a process is analytical, it can be replicated algorithmically. Indeed, fifty years ago the great Alan Turing demonstrated that any analytical process can in fact be replicated by a sufficiently expert system.
But then he ate the poisoned apple ….
Today we have expert systems that play world-class chess, diagnose illness, pick stocks to buy and sell, and guide callers through help-desk questions.
Maybe that last one’s not such a good example.
If we can do all that, why can’t we get there with automated appraisals, especially of the single real estate asset, the single-family home?
“And in time, you’ll be able to balance your checkbook on it!”
Even allowing for the uncertainty inherent in any valuation, single-family homes are a financially simple asset, for which there is an enormously thick market nationwide and a fairly thick market in every major city or metropolitan area. Thus, under the Law of One Price, there ought to be a formula, encompassing relevant variables, that uses linear weights on relevant attributes to derive a realistic value for the subject property.
Everyone thinks such a formula must exist, with weights that vary from market to market:
The Philosopher’s Stone of Valuation
Value = f(v1, v2, v3, v4 … vN)
Where each variable’s weight depends on its market
but no one has truly discovered it.
“We’ve unearthed the formula.”
As a result, everyone’s ‘comparable sales’ approach to value (there are two others, the cost approach and income approach) seeks to infer the weights by listing comparables and then reconciling them back to the subject. (See, for instance, pages 1 and 3 of the Fannie Mae form cited above.) What Zillow does is similar:
How does Zillow work? It starts by buying massive amounts of real estate information from commercial data collectors — home addresses, tax assessments, square footage, lot sizes, number of bedrooms, prior sales prices and the like. Then its computers identify similar homes that recently sold in the same neighborhoods and use mathematical models to compare their traits and create a “Zestimate,” or “estimated market value.”
Naturally, Garbage In, Garbage Out:
Trouble is, because Zillow is loaded with dirty data in some places and missing key factoids in others, its Zestimates often miss the mark — sometimes so widely that I fear that anyone trying to buy or sell a home could get burned by relying on Zillow.
Anyone who relies on free information has no one else to blame.
“What scares me is the consumer who goes out there and makes a decision based on that data,” said Richard Powers, president of the Appraisal Institute, the nation’s largest appraiser association with 21,000 members. “Consumers really have no way to judge the accuracy of the estimate — that really is the problem.”
I have the greatest respect for the Appraisal Institute (besides, twenty-some years ago they gave me a nice award for an article (pdf link to shorter version) presenting a theoretical advance in appraisal methodology) and they are right; binding decisions need much better information than just a catchy Zestimate. Indeed, the site’s proprietors make this and other points in a very useful ‘how to’ background section.
Zillow president and co-founder Lloyd Frink said the free, advertising-supported site doesn’t aim to replace home appraisers or real estate agents. “It is meant as something to help buyers and sellers start a conversation.”
Still, it’s free, the information it uses is compactly compiled, and it invites users to improve the estimates.
With just how big a grain of salt must it be taken?
In my own random tests of dozens of local properties, I found about half of the estimates to be sharply off — more than 10% off the actual recent sales price or what someone knowledgeable about the property deemed its market value to be. Many were off by 20% or more.
Markets are about information: the more information, the better resolution.
More data, better resolution
Zillow offers a tweaking tool called My Zestimator that allows you to fine-tune estimates by adding data about a home, such as the size and cost of renovation work — and then view, but not save, the modified value.
Zillow probably wants to prevent its data from being ‘corrupted’ — although it might well do better to allow an open-source database that lets users update information provided that the results are accepted as provisional, not verified.
Below its estimate, Zillow shows a mesmerizing array of other data, including features describing the house, a chart of estimated value over the past year and list of similar houses sold recently. You can call up scrollable neighborhood maps with images of the homes, with their values superimposed.
The site’s sponsors seem determined to keep collecting information:
So far, Zillow has identified about 60 million homes and collected enough data to offer estimates on slightly more than 40 million, with more being added daily. Coverage is spotty to nonexistent for the District,
What the doubters miss is that data-oriented networks only become more effective over time: just as we eventually focused the Hubble telescope, the quality of computer-model weather forecasting has improved dramatically, with better data and more computing power:
Those estimates were likely due to a glitch in Zillow’s valuation formula, which has been overriding actual sales data too frequently, according to Stan Humphries, Zillow’s director of advanced analytics. He said the formula is being adjusted to give greater weight to sale prices.
Frink predicted that Zillow, which received plenty of media hoopla when it debuted, will improve gradually as the company adds housing data and tweaks its algorithm for automated value calls.
Observe how a self-organizing network creates business opportunities:
· GPS satellites enable us to locate property.
· Google offers mapping software.
· High-speed computing offers essentially infinitely fast algorithmic calculation.
· For-sale-by-owner cooperatives spring up.
· Now comes a do-it-yourself value estimator.
How does Zillow address the data paradox? In two ways:
· Selling eyeballs. By selling advertising.
· Premium features. Once the site is established, I strongly suspect it will premium features — like owner-to-owner sales.
Worried yet, brokers?
