The Resurgent Importance of AVMs

by Romi Mahajan, Quantarium Board Advisor
and Malcolm Cannon, COO Quantarium 

August 2, 2019
All industries have standards and solutions that generate debate about their relative importance. Sixteen years ago, Nicholas Carr wrote an article suggesting that IT had become a commodity and as such was not a source of comparative advantage for particular companies as measured against others. The article turned a bold and growing IT sector on its head, with defensive marketing emanating from all corners and panel discussions dominated by paid pundits who railed against Carr for what they considered myopia. I We wrote extensively suggesting that while Carr had a variety of incredibly important arguments, that in fact technology in general and IT specifically can certainly be part of any organization’s re-invention. No doubt, the debate continues even in the face of a technology industry that has, in the paraphrased words of one prominent Venture-Capitalism, “eaten the world.”

In Real Estate, we have our own ideas that generate heated debate. One of these relates to the relative importance of Automated Valuation Models (AVMs). Interestingly, while AVMs have been around for decades, they’ve resurged in importance and, as such, deserve increased analysis in the press and amongst all members of the real estate ecosystem.

For some, AVMs are a commodity that are used just because “we have to.” In that conception, as long as AVMs meet the minimum bar, they are good enough. Whether votaries of this belief share their “good enough” view with the customers and partners that depend on them is unlikely, but indeed all industries have some groups that work in this middle-of-the-road space. I do not believe this view represents either the potential of the real estate industry or its future.

We subscribe to the opposite view- that AVMs are a high-value add area that are essential to the new world of Real Estate. We believe that industry trends, the tea-leaves so to speak, read in favor of this point of view.

To be fair, it is worth giving the opposite view some credence. After all, AVMs have largely been defined by an inherent antagonism- between precision and breadth of coverage. In direct opposition to the adage about “having one’s cake and eating it too,” most purveyors of AVMs have asked customers to choose which of these- precision or coverage- is more important. In philosophy, this is called “Hobson’s Choice” which is to say you’re asked to choose between two insufficient or untenable choices. “I want both!” is the sensible customer response. If this were a persistent and immutable fact of AVMs, then one could understand the reticence amongst industry pundits of elevating AVMs to the status of high value-add.

Luckily, this is not the case. Advances in applied Artificial Intelligence (AI) and Data Science have allowed for the creation of AVMs that are not in the sacrificial game. Machine-learning, a process in which suo moto, the engine learns and improves, when wielded with the wisdom and dexterity necessary in the world’s largest asset class, has allowed us to leapfrog the AVMs that are built on old models. Great AVMs must be AI-native and not after the fact “bolt-on” technology.

Their importance, especially in light of these advances, has resurged. As large companies move into the “iBuying” space, and as billions of dollars of VC and PE money flow into residential real estate, AVMs move into the core of business. Good enough will soon be bad business. AVMs, hardly a commodity, will now be a source of comparative advantage and will soon be the differentiating factor between success and failure, between winners and losers in the grand arena of Real Estate.

These economic trends, coupled with a loosening regulatory environment, and the enormous interest and focus of well-capitalized and technology-forward companies bode well for AVMs.

So pick carefully and never accept any rhetoric about trade-offs. Expect and demand a great AVM and you’ll find yourself set up for enormous success in this market.