We try to compare real estate to other asset classes like stocks and bonds but the analogy has some flaws. The biggest of which is that a business or a financial instrument exists mostly on a ledger. Properties, on the other hand, have an actual physical presence — exposed to the wear of use and the elements. The value of most properties is much more than just its replacement or repair cost, of course, but the condition of a building can impact its future profitability and therefore its value.
Understanding historical data can provide insight into future property risk. While the process to mine and process data for insights is cumbersome, the rewards might outweigh the costs. “Knowing about changes in the overall condition of properties, from the individual address to an aggregate, national level, can give an idea of what people are spending to maintain these expensive assets. This creates insights into micro- and macroeconomic trends affecting the U.S. housing market,” according to Jonathan Kanarek, COO of BuildFax.
Take consumer confidence, for instance. BuildFax’s latest Housing Health Report revealed a 6.47 percent decrease in year-over-year maintenance volume. This activity is highly correlated with consumer confidence because homeowners are less likely to invest in sizable housing projects during times of economic instability.
As one of the largest providers of property condition and history data, BuildFax has spent the last ten years refining its data solutions for the insurance and financial industries. Kanarek explained, “We collect from tens of thousands of public data sources and put it into a format that is usable for our customers. It starts by ensuring the data ties back to a single identifier, typically an address, and then put it into a standard format in our data lake.”
A lot has changed since BuildFax started. “We used to get annual cycles of property data, now it is weekly and sometimes daily,” said Kanarek. Having more data points has helped them create more complicated training models, something that he sees as being one of his company’s unique advantages.
According to Kanarek, one of the biggest changes to the data landscape, particularly in insurance, is the use of aerial imagery. So much high-quality image data is available from a number of sources, including big tech companies like Google, that he thinks it will start to become a larger part of property underwriting. “While aerial imagery has opened the door for new possibilities within insurance, the view is limited. Additional data sets are needed to provide interior risk conditions,” Kanarek said.
Aerial Imagery has already changed the way a lot of industries understand the world. Finance pros can look at parking lot occupancy as a proxy for the activity in a store. The shipping industry is constantly taking stock on how many shipping containers are piling up at which docks. Now property images might help alleviate a lot of uncertainty in insurance. For example, Kanarek says that many insurance companies are moving to a “no-questions-asked” system by leveraging big data and analytics to automate the quoting process.
There are typically security concerns with data processing and collection, however, property data is a different animal. “We don’t collect personally identifiable information about the homeowner or property owner,” Kanarek said. “Unlike a credit agency, which is focused on the who, we are focused on the what and how the property has changed over time.”