Aggregations, algorithms, AI and judgment - by David Kirk

October 08, 2021 - Appraisal & Consulting
David Kirk

Judgement continues to be the discerning characteristic in commercial property investment and development. And in the analysis supporting underwriting and appraising commercial property. Not necessarily in defiance of the ubiquitous data and traditional methodologies, and the artificial intelligence that forecast outcomes. In every step of thereof. The Federal Reserve Bank exemplifies the process and the challenges and abuses that go with the territory of reconciliation of the accessible data, trends and forecasts that result in judgments that relate to the process. The transparency of our FED activity and process is instructive for best practices and missteps. Commercial property is still unique investment class among many options or alternatives. However, the FED Is a major impactful force on the economy and commercial property is an economic derivative. Gobbledygook, circularity, not really. Simply a start to understand the pervasiveness and determinism of sound judgment.

Data overload is the bugle call of the naysayers of the accessibility and transparency of volumes, rivers and piles of data extracted from our economy, markets and collateral sources. Instantaneously sliced and diced and multiplied. Aggregations and anecdotes. Too much data? The collectors and purveyors of data provide consistent and reasonable push-back on anyone denying the importance of data, saying the abuse occurs in the selection, purpose, use, manipulation and promotion of data which is easy peasy once you have got it - the data.

Data is ubiquitous. Even single points of data as well as near infinite aggregations of data points, collected, generated, or imagined, require massaging. Platforms for ordering and recording data are now readily available and accessible to the professionals in the real estate sector. Data selection is judgmental. And a massage is optional.

For commercial property, slicing and dicing to produce commonality or comparability or adjusted indicators is logical and essential. That process is produced by an algorithm. The data firms sell data and platforms. Methodologies for analyzing raw data abound and platforms continuously evolve to suit the sector and the user. These algorithms are generally available and accessible. Sometimes not so. These algorithms are often set, fixed and immutable! Judgment is required to select the appropriate methodology or algorithm that is representative, current and suited for the mission. Clunky stuff does not lend itself to modification based on judgment.

Lastly, the intended mortar to solidify the data and the algorithms - artificial intelligence. Not needed for quick and dirty, windshield, ballpark or rough estimates, judgment and reconciliation still is the linch pin or lunch pail. Reconciliation is a final nominal step in the academic appraisal. Little or no text attempts to articulate the reconciliation or judgment process. Artificial intelligence does not have its toes wet. Data will continue to spill upon the process. Algorithms will evolve to manage the data. And judgment of the practitioner, as well as of FED’s FOMC, will dominate each step along the way. With these self truths in mind, proceed with confidence and diligence and the judgment that distinguishes the accomplished practitioner.

David Kirk is founder, chief executive officer of Kirk & Company, Real Estate Counselors of Boston, Mass.



Add Comment