Artificial intelligence – the stuff of sci-fi legend – is arguably coming into its own. Those IROs who grasp the power of AI paint a picture of a future in which data-mining and predictive analytics will reshape the traditional IR role in dramatic ways.
“AI engines will be able to look at past behaviour of shareholders and at massive amounts of data that the human mind would be unable to recognize patterns in,” says Greg Secord, Vice President, Investor Relations, at Open Text Corporation in Waterloo, Ontario. “This could provide the IR professional with a better picture, not just of the shareholder base and their investment style, but how individually and collectively shareholders may react.”
As a seasoned IR professional at OpenText, the largest artificial intelligence focused company in Canada, Secord has a unique perspective on how AI might reshape the IR function. He points out that AI is spawning new tools that will help IROs deliver clearer messages and even help influence their companies’ strategic direction. These tools could, for instance, model the Street’s reaction to a corporate event, such as a possible acquisition, and predict which new institutional investors might buy shares should a company raise its dividend.
Although some IR officers may feel threatened by the incursion of AI into their relationship-based domain, Secord thinks that the emerging tools will make the tech-savvy IRO more valuable than ever. “The reality,” he says, “is that IR is going to have a much more strategic role once we have these data sets.” He suggests that engaged IROs who embrace technology will have more impact when they sit down with management because they will be offering “clear, concise, measured, and backed-up advice.”
Many of the applications for AI are still years in the future, but AI itself is now perceived as a serious undertaking. In a November 2016 report, Allied Market Research estimates that the market for AI will reach $19.48 billion by 2022 with a compound annual growth rate of 45.4% in the six years between 2016 and 2022.
While many experts foresee machine-learning, big data, and predictive analytics reshaping investor relations, no one anticipates the function disappearing. Instead, some see technological advances as an answer to how to accomplish more, given steady or decreasing IR headcounts.
Christine Marchuska, Director of Investor Relations for Nuance Communications, a natural language and systems integration company in Burlington, Massachusetts, says “IR groups today are slim.” She explains that a company the size of Nuance, which has $5 billion in market cap, would once have had an IR team with two-to-three employees.
“Today,” she says, “the group has just one person, and this is a trend that is expected to continue.” For Marchuska, AI could be the breakthrough enabling smaller teams to have the firepower to continue practicing cutting-edge IR.
Computer Generated Reports: A New Reality
In an October 26, 2017, research report AI: The Coming Disruption on Wall Street, Greenwich Associates found that by next year, roughly three-quarters of banks and financial firms will be actively exploring or implementing AI technologies, using these offerings to generate insights from massive amounts of data.
One of the current applications for AI is the writing of routine earnings and other reports.
Laura Pressman, Marketing Manager for Automated Insights, a Durham, North Carolina-based technology company, notes that the Associated Press has been using her company’s natural language generation (NLG) platform – Wordsmith – to write articles on earnings reports. “We can look to companies in any industry, and can take their data, either internal or client facing, and transform it into a written narrative,” she says. “Human writers alone are just not capable of this speed and scale.”
NLG relies on the creation of a template with built-in conditional logic around data, explains Pressman. She notes that at its core, Wordsmith uses “if/then conditional logic statements so that if spending is greater this month than last, the computer will say: ‘The amount of spending rose compared to the same time last month.’”
While AI-generated reports may hold enormous potential appeal for IROs who might consider using the technology themselves, there is a downside. When there is a mistake inside a report generated by NLG, there is no one to contact for a correction. Computers and robots do not explain their logic – and do not have the ability to rectify problems that they have created.
Worse still, in a world in which ‘journalists’ may be robot engines reporting on research reports also generated by robots, an initial inaccuracy may be replicated many times over like a reflection in a hall of funhouse mirrors.
That’s why for IROs, automated research can be an unsettling prospect. Carol Hansell, Senior Partner at Hansell LLP, an independent legal and governance firm based in Toronto, says that this is an area that IROs should devote time to understanding. “You need to worry about the quality of the information that’s being generated,” she says.
Surveillance, Targeting, and Better Strategic Decisions
Generating written reports relies on a fairly straightforward AI application. Many believe that the more powerful applications come from the ability of computer programs to identify patterns and detect trends.
The good news is that while computers have combed social media to gauge investor sentiment for years now, they are finally getting quite good at doing so. Jim Hendler, an AI researcher and Director of the Institute for Data Exploration and Applications at Rensselaer Polytechnic Institute in Troy, New York, says that AI programs can interpret sarcasm and other subtleties that famously stymied computers until quite recently.
Hendler points out that one of the game changers for investor relations is the fact that more IR data is communicated in machine-readable formats, such as XBRL. Information about earnings and income “is now more extractable,” he says, “and more electronic disclosure means that IR officers have more to get their hands on at the data level.”
Stock surveillance and market intelligence systems are other areas that have made quantum leaps, thanks to far more sophisticated data mining.
As part of a September 6, 2017, panel discussion on capital markets intelligence, Mike Coffey, Vice President of Business Development and Intelligence for Toronto-based Q4, said that his company can break down factors impacting stock price on a 15-minute basis. These near real-time insights are very valuable for IROs. “Going on the road, telling your message and being able to see if your message is resonating is key,” according to Coffey.
Secord offers a similar view, noting that both better market intelligence and more accurate targeting are goals within reach of many IROs given the anticipated emergence of AI-enabled tools.
“There will be more data than one person – or even a group of people – can use to make efficient summaries of data patterns. And that is where the only tool that will truly track, manage, and give insights into this data will be a machine,” says Secord. “These machines will learn as they go. And we’ll have smarter tools, and better IR.”
Arguably, the arrival of such tools matters now more than ever because passive investing is on the rise and senior managers feel pressed for time.
In a world in which the pool of active investors is shrinking, IROs may also feel that their persuasive skills are less important. However, AI can let IROs tap into the most promising shareholders, delivering a message that is carefully calibrated to their needs.
“At the end of the day, strategic targeting is becoming more and more important to what an IRO does,” says Secord. Meeting with investors takes management away from running the business, and so an effective IR officer will want every tool he or she can utilize to schedule coveted one-on-ones with the most qualified prospects.
The Changing IRO Role
Savvy IROs will soon venture beyond the confines of their own departments to tap into data elsewhere in their organizations, says Hendler. He points out that while traditionally, an IRO might not have sought input from the web group about advertisements on social media, today “there’s a much wider range of information that could be integrated into the investor relations task.”
Hendler is convinced that the arrival of AI tools might turn IR, which was “once more of a public relations effort,” into “a data-intensive, decision-making position.” This, he believes, is a very important development for the profession and its prestige.
Secord also sees the IR role evolving because of AI and other computer-enabled capabilities, but he defends the IRO’s relationship-building skills and intuition as invaluable.
“Nobody is going to make an artificial intelligence engine to do the relationship work that investor relations officers do,” emphasizes Secord. “You need that human element, and that won’t go away. But imagine going into a meeting with senior management and offering a strategic opinion based not only on your experience and your history, but also on data sets that use AI to support the predictions or recommendations you make. These will be incredibly powerful tools.”