Introduction
Often, I have dreamed of retracing and revisiting the islands described in “Travels into Several Remote Nations of the World” by Jonathan Swift, first published in 1726. In the book, Lemuel Gulliver travels by ship to various lands, including Lilliput, Brobdingnag, and the land of the Houyhnhnms. The most famous of these, known to many children, is Lilliput.
However, this is not a children’s story it is much more. At the time of its publication, the book was considered scandalous, as it mocked the political order of the day, with each island portraying a different critique. The author, like a true artist, challenged perceptions of what is considered the natural order of things.
Likewise, in real estate valuation, there is also an “order.” In this tale, we revisit a fictitious island where data scientists hold the sole key to determining automated valuation models (AVMs), and explore what happens when, on the horizon, another ship of values arrives.
The Kingdom of Homogeneous
Once upon a time, there was a prosperous kingdom named Homogeneous. The king, queen, and royal assembly lived contentedly until one morning, a sailing ship from the far reaches of the world arrived. The inhabitants of the island gathered at the harbor to welcome these travelers.
The visitors introduced themselves as “data scientists” and claimed expertise in all things mathematical. The king welcomed them warmly, and over the course of a meal, it was determined that they could be useful in assessing the value of all properties on the island of Homogeneous.
The data scientists took residence in the king’s castle and were given access to the kingdom’s entire real estate transaction database. They worked laboriously and eventually produced an all-knowing AVM. As all properties were considered homogeneous, a single price per square meter was determined, adjusted quarterly for market movements.
Their computer system churned, and out came a total value for all properties in the land. Delighted, the king’s treasurer applied a tax rate and levied taxes accordingly. Stability followed until several years later, when another ship appeared on the horizon.
A Second Opinion
As before, the islanders gathered at the docks to meet the newcomers. These travelers were real estate valuers and appraisers, trained in the intricate discipline of property valuation. The king, being a prudent man, asked them to review the island’s property models and values.
Unlike their predecessors, the valuers declined the king’s invitation to stay in the castle. Instead, they requested transport to survey the island firsthand. Traveling by donkey to its farthest reaches, they spent weeks examining properties.
Upon reconvening, they confirmed that the properties were largely homogeneous, as described. However, they also observed critical differences. The valuers developed an adjustment grid accounting for physical deterioration, functional obsolescence, and external obsolescence. Armed with this framework and detailed property inspections, they valued each property individually using pen, paper, and calculation tools. It was a monumental task, but eventually, a total valuation was produced.
The Reckoning
On the day of the findings, the king assembled both the data scientists and the valuers in the city square before the citizens. It was a day meant to reaffirm the value of the kingdom. The data scientists, with great ceremony, presented their updated index. The valuers, however, appeared uneasy. When it was their turn to present, the results caused an uproar. Their conclusion: the total value of the land was significantly lower.
In essence, they argued that once properly adjusted, the kingdom was operating under a fiscal illusion its tax revenues were overstated, masking a structural deficit. The implications were profound. For years, property values had been artificially inflated, leading to unsustainable taxation and financial mismanagement.
As the gravity of the situation became clear, unrest spread among the citizens. Led by the king, the crowd turned against the data scientists, who fled to their ship and were exiled from the island.
Conclusion
In the saga of Homogeneous, several key lessons emerge. Foremost: intelligent and capable data scientists who build AVMs are not necessarily trained valuers. This is the first red flag.
Whenever an AVM is used, the first question a client should ask is: “What are your valuation credentials?” Too often, there are none. This is akin to the Dutch children’s game “Ezeltje prik” (Pin the Tail on the Donkey), where success depends more on chance than skill.
Valuation is a profession, much like that of a doctor, surgeon, dentist, or pilot. Society does not permit unqualified individuals to operate in these fields and rightly so. Yet, paradoxically, this is increasingly tolerated in real estate valuation.
The axiom “If it seems too good to be true, it probably is” is often conveniently ignored in what can only be described as a spiral of silence.
AVMs are useful tools efficient, cost-effective, and often directionally accurate. However, when were their disclaimers and limitations last truly examined? Many fail to adequately account for depreciation or include alternative data. For an aspiring valuer, this would be a fundamental error one that would not pass an introductory exam. And yet, the market continues forward using uninformed AVM, not unlike the tale of the emperor with no clothes.
In real estate, as in life: let the buyer beware. This is, ultimately, a cautionary tale.
