Bits Meets Bricks: Leveraging AI in Real Estate
SCHACK PROFESSOR INSIGHT - PROFESSOR SCOTT DUNPHY
We are currently living through the Fourth Industrial Revolution. Only this Industrial Revolution is happening over the span of years, rather than decades or centuries. The driver of this revolution is Artificial Intelligence (AI).
Machine Learning and Generative AI (GenAI) are two branches of AI that are rapidly changing the way people work in the commercial real estate industry.
Machine Learning is primarily used for predictive analytics and portfolio optimization. Many firms are beginning to hire data engineers and data scientists who can source and model vast amounts of data to make predictions about markets, property sectors, and properties themselves. How will rents change? Where will occupancy go? Where will people move?
More than this, machine learning can help simulate various economic scenarios and the impact on real estate. This allows real estate investment managers to better understand and manage risk well beyond the traditional ‘best’, ‘base’, ‘worst’ case scenarios that often rely on simple adjustments to rent growth or cap rates. Better forecasting and risk management means better investments and better risk-adjusted returns.
Real estate analysts aren’t expected to know how to build machine learning models but they should be knowledgeable about them to understand the outputs, model limitations, and the risks associated with them. AI is powerful but humans must always remain in the loop to assess and challenge the output.
ChatGPT was publicly released less than a year and a half ago but already many real estate firms are looking at ways to implement GenAI in their businesses. GenAI is an artificial intelligence model that is trained on vast amounts of data and afterwards can create or generate new and original content. This could be images, video, audio, text, code, etc.
Commercial real estate is a document rich industry. In it we have leases, purchase and sale agreements, joint venture agreements, environmental reports, property condition assessments, loan agreements, and rent rolls, among many others. All of these need to be reviewed and analyzed during a transaction and in the normal course of business. GenAI has the potential to make the analysis of these documents faster, more efficient, and more accurate. Acquisition teams can use it to quickly compare rent rolls to leases and identify risks in environmental reports and property condition assessments. Lawyers will be able to compare newly received contracts against precedent contracts or templates and quickly draft an initial redline or markup.
And while GenAI can’t do math directly, the models are capable of writing code that can do math. ChatGPT and Google Gemini both have the ability to do complex data analysis by generating and executing python code.
These capabilities are democratizing advanced data analysis. Savvy users of GenAI tools are able to do in minutes what would take an analyst hours to do in Excel. If you have a ChatGPT Plus or Gemini subscription, you can test these features for yourself. Early in my career I had to create all sorts of complex formulas and pivot tables for simple metrics like weighted average cap rate. Now an analyst only has to ask an AI chat bot to “Calculate the weighted average cap rate by property type” and they are instantly presented with a well-formatted answer.
In my opinion, GenAI technologies will become as common in real estate as Excel and Argus. It has always been the case that analysts with the best knowledge of Excel and Argus were given the most interesting and valuable work. So for students entering the workforce, mastering these new technologies will give them a competitive advantage versus those that don’t. Not only that, it will give them a competitive advantage against those analysts who have been in the workforce for several years already but haven’t bothered to focus on ChatGPT or Gemini or any of the other models.
The best way to get started is to play with these tools. Find publicly available real estate research from the large asset management firms and use ChatGPT to summarize the documents or ask questions about sections you don’t understand. Ask it what terminology you should know going into a real estate internship and to define those terms. If you're Excel and macro savvy, have it write code to automate your spreadsheets. And if you’re willing to spend the twenty dollars per month for ChatGPT Plus, find or create datasets and test the data analysis capabilities of ChatGPT’s Advanced Data Analysis tool.
The commercial real estate industry has historically been slow to adopt new technology. Artificial Intelligence is a revolutionary technology though and is actively being explored and adopted by firms large and small. AI skills are becoming increasingly in demand and eventually will be expected the way Excel skills are expected.