How Will Commercial Building Managers Budget for the Unknowable? | TAKE THE SURVEY→

Understanding the Effects of the Pandemic on Real Estate Requires Property Level Data


Long before any of us had heard of the virus we are calling COVID-19, the commercial real estate industry was already undergoing a big change in the way it understands property value. We were already seeing the limitations of the commonly used technique of using two or three comparable properties to inform an assessment of value. Appraisers, brokers, and investors were beginning to understand that they needed more sophisticated techniques to help them determine the value of a completely unique product like a building or piece of land. 

Now that we are starting to see the impact of the virus on our cities and our economy, the need for a shift towards advanced modeling has become even more evident. Trying to predict the future value of a building is hard under normal circumstances. Models must take into consideration economic outlooks, regulatory risks, supply and demand curves, and cultural shifts. But now they are finding themselves unequipped to understand and predict an event as unprecedented as the COVID-19 pandemic. 

The blanket effects of the stay at home orders for entire states and countries require analysts to rethink how they calibrate their models. Unlike natural disasters, which affect certain geographies more than others, the impact won’t be concentrated to specific regions. Unlike economic downturns, which affect certain industries more than others, the impact won’t be concentrated in specific cities. Unlike financial crises, which affect certain lenders more than others, the impact won’t be concentrated on specific financial products. And unlike consumer trends, which affect certain businesses more than others, the impact won’t be concentrated on specific property uses.

Previously, the industry didn’t look at this much at the individual property level because it wasn’t feasible that tenants would be distressed in general like this, but I think now that is changing.

Ron Bekkerman, CTO of Cherre

Understanding what will happen to property prices over the next few months will require analyzing each property down to the occupant level. Ron Bekkerman, CTO of real estate data platform Cherre, told me how this is a big step for commercial real estate analysis, “Previously the industry would be able to understand connections between certain events and certain neighborhoods, but this is different,” he said. The same neighborhood could have some buildings that are impacted and others that are not. Bekkerman continued, “Previously, the industry didn’t look at this much at the individual property level because it wasn’t feasible that tenants would be distressed in general like this, but I think now that is changing.”

Examining each parcel isn’t just about categorizing every building. It means creating a knowledge graph of the entire property landscape. Every data point of every building must be mapped so that correlations between every point can be analyzed. In this process, each data point is called a node and each correlation is called an edge. “We can tie a person to the deed of a property, for example, making an ownership edge,” Bekkerman explained. “Then we can compare it to other changes in other edges like financing or tenancy. If we put a timestamp, we can know how it changed from start to finish.”

This ends up being an enormous task. Even information as simple as a property address can result in multiple data points. “You can have multiple addresses for one building and multiple residents of one address,” Bekkerman said. Ron says that his team is working with around a billion nodes and over a billion and a half edges. This kind of analysis requires a lot of help from computational techniques like artificial intelligence, but Ron thinks that that term has gotten a bit overused. “So many things have gotten categorized as AI these days that I try to stay away from the term,” he said.

Knowledge graphs as big as the one that Cherre is creating needs advanced computing techniques like machine learning to help make sense of it. But they also need the expertise of computer scientists and the intuition of a property industry veteran to turn the analysis into opportunity. The stress that the global shutdown inflicts on the property industry will reveal a lot about how resilient property prices are to this kind of event. But the insights that will be gleaned will only be seen by those with the capacity to understand changes property by property and tenant by tenant.

Your email address will not be published. Required fields are marked *

Propmodo is a global multimedia effort to explore how emerging technologies affect our built environment.

More Stories
ibreakup: Keller Williams, Offerpad Part Ways on Home-Buying Partnership