How Automation and Predictive Analytics Can Help Real Estate Analysts Underwrite More Effectively
In American folklore, John Henry was a steel-driving man — a man tasked with hammering a large steel drill into rock to help construct railroad tunnels. According to legend, John Henry’s prowess as steel driver was measured in a race against a steam-powered hammer. He gave the race everything he had for more than a day, ultimately beating the steam-powered hammer- but only to die in victory, with hammer in hand, as his heart gave out from stress.
A skilled analyst may see themselves as the John Henry of underwriting, but everyone has limits. Automation and predictive analytics are the “steam-hammers” of real estate analysis and, when used correctly, they can help analysts push beyond their limits to underwrite more deals with better results.
Analyze More Data, Faster
Automated data collection systems can collect and analyze more data than a skilled analyst. An analyst may be able to visit dozens of property websites and, using a seemingly limitless repertoire of keyboard shortcuts, collect data at lightning speed. However, an algorithm can do this for thousands of properties instantly. When technology exists to instantly collect real estate data, it is inefficient for an analyst to spend time on these activities. Knowledgeable analysts should let technology do the grunt work of data aggregation, and focus their time on delivering insight instead.
Remove Outliers Quickly and Effectively
Algorithms are better at removing outliers than your analyst. For a skilled analyst, it is easy to spot rents that are too far above or below market or property averages to be used in their analysis. Perhaps these rents represent short-term leases posted automatically from revenue management software to a property website, or maybe they were posted at an artificially low price just to get potential tenants in the door. Both situations would be clear to a competent analyst, but they are even more easily picked up by an algorithm. Additionally, machine learning algorithms can perform this analysis instantaneously for hundreds of properties. Instead of meticulously trying to spot outliers, analysts should leverage the technology available to remove them before they begin their analysis.
Eliminate Human Bias
Machine learning algorithms will not skew results based on human bias, as there is no incentive for an algorithm to overstate or understate the expected performance of a deal. In some situations, an acquisitions analyst may underwrite more aggressively, as they are incentivized to close more deals. Conversely, an asset manager may underwrite more conservatively and close fewer deals, as they are more incentivized to boost performance in an existing portfolio. These motivations can influence both investment decisions and results, resulting in spurious assumptions and incorrect valuations. Algorithms follow a simple and consistent set of rules and work the same way no matter what the situation. For this reason, they can often be used as an objective, third-party source to confirm an analysis is unbiased.
These are just a few of the many ways technology can be used to improve underwriting results. The important thing is to view technology as a tool rather than a competitor. Analysts shouldn’t see themselves as John Henry fighting the steam engine — they should let the steam engine help them do more.