REIT investor relations (IR) teams that have dipped a toe in the water with AI tools are finding a rapidly expanding list of practical use cases. While the technology has the potential to boost output and efficiency, it also creates new challenges in the form of maintaining the security of information and remaining mindful about how it can be interpreted, among other issues.
AI has come a long way in a short period of time. IR teams are now leveraging AI for tasks such as gathering, summarizing, and analyzing information, preparing earnings call transcripts, and anticipating analyst questions. “A year ago, I would have told you AI is a great intern. Today, I would say it's a great IR manager,” says Scott McLaughlin, senior vice president, investor relations and tax, at Invitation Homes Inc. (NYSE: INVH).
For example, Invitation Homes uses AI tools to help prepare remarks for earnings call scripts. “I consider myself a good writer, but AI often will have minor tweaks here and there that take it from great to excellent,” McLaughlin says. He also can use AI to download a presentation into a PDF, analyze the content, and recommend improvements, such as different headlines or talking points. “That is not something an intern could do,” he adds.
Other REITs are taking a similar approach. Extra Space Storage Inc. (NYSE: EXR) has always prided itself on its innovation, first with machine learning and now with AI. “When this started to develop and evolve a couple years ago, we assigned a task force with people from our data science group, marketing, and operations teams to figure out how to best utilize AI in various processes,” says Jared Conley, vice president, investor relations at Extra Space.
Conley now uses AI tools in a variety of areas, such as prepping for earnings calls, identifying and preparing for potential analyst questions, analyzing analyst reports, and monitoring what peers are doing.
“It’s an effective tool to be more efficient and be able to do things quicker, but you also have to be wary of the output,” Conley says. For example, he has found that he can alter the AI response by changing how the prompt is worded. “It's a tool that can get you 80% to 90% of the way, but you've got to be aware of its limitations and understand that it is a tool,” he adds.
As IR teams continue to look for ways to incorporate AI into processes and workflows, they also are paying close attention to how AI is transforming the investment landscape. “The buy side is using AI in many of the same ways we are—to grab data and analyze data on a much faster and streamlined basis than we could in the old days,” McLaughlin says. As such, IR teams are thinking about how to use AI more strategically to communicate and engage with their investor base.
Changes in Buy Side Processes
The buy side has long been at the forefront of leveraging technology, big data analytics, and AI to get an edge on where to find new opportunities in the investment universe. “Where we see the biggest impact of AI on how the buy side is analyzing companies is that they can use these tools to look at language, such as in earnings calls,” says Chris Blake, an executive director of advisory innovation & technology within the investor relations solutions team at S&P Global Market Intelligence.
Earnings call transcripts now can be fed through different types of algorithms to score sentiment. The buy side is using that sentiment analysis to quickly evaluate what are positive or negative earnings calls. Where are the trends that they need to be aware of in terms of increasing or decreasing language complexity that management teams are using?
“What that means for IR teams is that they need to be able to take into account how those algorithms are going to score what they're going to put out in the market,” Blake says. “Essentially, AI and technology become kind of a new persona that they have to be catering to when they start writing what they're going to include in earnings call scripts.”
Adapting to AI Scoring Algorithms
In the new AI era, IR teams need to be more mindful of the content they’re putting out, what they’re saying, and how they’re saying it, because the meaning could be interpreted differently by an AI scoring algorithm. For example, a member of a company’s management team might say, “If you recall, we discussed XYZ on our last call.” However, the word “recall” can be flagged as a negative by an AI algorithm because it has been trained to view “recall” as a negative related to finance.
That's one area where understanding how the buy side is leveraging this technology, and how they're analyzing transcripts, becomes really important for IR teams to think strategically about the impact of their language and word choice. “It’s important for IR teams to identify those areas of disconnect when they think about how they approach drafting, scripting, and practicing earning calls,” Blake says.
The buy side also is beginning to lean more on AI for efficiency. For example, a portfolio manager may not have listened to or read the entire transcript of an earnings call and instead is relying on an AI summary of the event. “While that creates a lot of efficiency for the buy side, it's important to be aware of that happening, because sometimes there might be specific nuance or context that could be missed in those situations,” Blake notes. That could result in situations where the buy side investors or managers are not getting the full picture or story that an IR team is trying to convey, he adds.
There are different ways that IR teams can account for that shift. Foremost, they need to be aware of what is being included in AI summaries. Using tools that keep non-public information secure, they can vet an earnings call script through an AI solution to get insight into how it will be interpreted. IR teams also can be proactive and create their own summary to control the narrative.
Another proactive step IR teams can take is to use generally available AI tools, such as ChatGPT, Grok, Perplexity or Google Gemini, to ask for summaries and thoughts on their company in terms of the publicly available information. That allows IR teams to see how their company is being portrayed through those types of tools. “That's one of the easiest places to start in terms of leveraging AI, because you want to know what these tools will say in the same way that you'd want to know how high you would show up in a Google search,” Blake says.
Buy Side Perspective
Buy side investors and analysts are embracing AI to varying degrees. There also is going to be differences in how groups with a dedicated team of analysts covering the REIT sector incorporate AI into their processes as compared to generalist investors.
“So far, we have largely been using AI to enhance our research in several different ways, and it doesn't necessarily have to do with the information that's coming from the REITs themselves,” says Uma Moriarty, senior investment strategist at CenterSquare Investment Management.
For example, in the apartment sector, CenterSquare’s research team is using AI to pull available data on new supply and mapping that against multifamily REIT portfolios to show how they might be impacted. CenterSquare also is exploring AI tools that can assist with secondary coverage, such as providing key takeaways on earnings for major corporate tenants that might impact a REIT.
REITs tend to be a niche part of the broader equity market, and generalist investors often try to look at REITs in the same way they would for broader equities, and that doesn't always work, Moriarty says. “Similar to what we're doing to enhance our secondary coverage, I could see more people using AI to get smarter about REITs,” she adds.
Navigating Downside Risks
IR teams are understandably cautious about the downside risks. The biggest concern is unintentionally leaking non-public information, so it’s critical to use secure and safe platforms. Inaccurate information and biases also remain on the list of top risks.
“It isn't the perfect tool for everything. There are limitations in the responses and outcomes, so we have to be careful and not be too reliant on it,” Conley says. “AI has its own biases or hallucinations and other things that can creep in to create confirmation bias based on information I've given and how I've asked the questions.”
Another big risk is leaning too heavily on AI to the point where people are losing some of the knowledge gained from manual tasks. Taking the time to read and digest an earnings call or presentation has a knowledge and learning component that can be lost by just reading a short summary. Additionally, if you don’t have a baseline knowledge of a topic, and are just reusing an AI summary, it’s difficult to gauge if the information you have is correct.
Getting results relies on good prompts, which takes a certain level of insight and understanding of the issues. So, IR teams still need to have a human doing the prompts and reviewing the results. “I think it will always need the human that’s helping guide it, but I think AI will continue to get better and better,” McLaughlin says. “It’s really a force multiplier, not a threat.”
Ongoing Evolution
Despite rapid advances in AI, it’s still early days when it comes to how IR teams are implementing AI into their processes. To use a baseball analogy, it’s only the bottom of the second or top of the third inning, McLaughlin notes.
Given the fast-paced evolution of AI, forward-thinking IR teams are continuing to look ahead to future applications. For example, there is more experimentation around using AI tools to analyze a speaker’s tone. What’s the message behind the message based on the speaker's tone and inflection? “I think that's very early days, but I can certainly see AI taking us there at some point,” McLaughlin says.
People also are anticipating advances in the ability to combine agentic AI, which requires little human intervention, with other AI tools to move from task-oriented functions to an ability to transform an entire end-to-end process.
For example, IR teams often spend a lot of time targeting investors. In the future, an AI tool could help to identify new targets, find contact information, and draft an outreach email so that the IR team only has to proofread it before it is sent out. “I think that level of putting all the pieces together and truly becoming an assistant, from an AI perspective, has a lot of exciting potential for the world of IR,” Blake says.