Investor Relations Meets The Matrix: Is AI the Future of Earnings Analysis?
Anyone who has read the news, bought a product, or looked at a billboard in the last five years knows that artificial intelligence (AI) is, without a doubt, the newest technology of interest in all facets of life. It’s literally being applied to everything. And earnings calls are no exception.
You may have heard stories over the years about hedge funds and quant funds using AI technology to analyze their earnings calls. Like many investor relations officers, you probably wrote it off as a fringe practice. And, for many years, it was. But today, a tremendous number of capital markets participants have access to advanced technology to analyze earnings calls—and it’s not just the tech-inspired hedge funds in Silicon Valley anymore. Nowadays, your standard investment managers and even retail investors are using tools and data providers that offer AI technology.
Will Companies’ Fates Soon Be at the Mercy of Machines?
Before you panic that the bots are taking over, it’s worth gaining a better understanding of AI technology and what the trend is really all about. There are four very important questions to consider:
- How does the technology work?
- How are investors accessing it?
- How are they using it?
- Perhaps most importantly, what, if anything, you should be doing differently as you build your next earnings script?
We’ll dive into each of these questions in order to help you properly gauge and respond to investors’ use of AI in earnings analysis.
How Does the Technology Work?
As you would expect, the full details on how this technology works could get pretty longwinded. The short story is this: typically, investors are using data platforms that leverage two technologies: natural language processing (NLP) and/or machine learning (ML). Both technologies fall under the umbrella of AI, though they’re far from the machines you’re used to seeing in movies like The Matrix.
NLP uses a series of algorithms to allow software to interpret and analyze language, similar to the way our brains do. It can understand sentence structure, word choice, and speech complexity to, for instance, evaluate if the person speaking is optimistic, pessimistic, or neutral. And it can often do so with greater subtlety and awareness than humans can. This is often referred to as sentiment analysis, and it is an important way this technology is used on earnings calls.
ML is similar to NLP, and the technologies are typically used in tandem. ML also uses algorithms to learn from data, identify patterns, and anticipate future outcomes. What makes ML technology unique is that it can quickly analyze massive quantities of data and make sense of it in short order. For instance, if your CEO is prone to begin good news with “so” and bad news with “uh,” you can rest assured ML will pick up on that—even if you don’t.
When used together, these two technologies allow software to digest a speaker’s points, interpret their delivery, and understand norms and standards of speech for individuals. More importantly, they allow software to detect if something is, perhaps, a little off.
How Are Investors Accessing this Technology?
In many cases, investors are using the same platforms to access NLP or ML data on earnings calls that they use to access standard financial data. A few commonly used platforms include: S&P Global’s Textual Data Analytics, FactSet’s Thematic Sentiment, and Sentieo. All three providers offer a sentiment analysis output using their own (or a partner’s) technology to describe the sentiment(s) of a transcript, an individual speaker, or a section of the script. And, in some cases, this data is capable of being integrated with other investment, CRM, or reporting software tools to allow for seamless analysis and decision making. This is especially appealing to investors who want to get a leg up, and do so rapidly.
How Are Investors Using this Technology?
The short answer is that most active investors are not using this technology. Further, most investors likely are not even looking at the output of this technology. For better or for worse, the typical institutional investor is still a long way from leveraging NLP or ML for sentiment analysis. And while quant or algorithmic investors may be using this type of analysis, their decision making is beyond our influence anyhow.
So for the group of active investors who are using this technology, they are certainly not using it in a vacuum. The application comes alongside a myriad of other considerations. In other words, virtually no one is making final buy or sell decisions based on the positive or negative sentiment of a transcript according to NLP/ML software. Rather, the analysis serves as one tool among many to aid certain portfolio managers in their attempt to understand what’s going on behind the curtain.
For an investor who is skeptical of a company’s decision making and displeased with its performance, negative sentiment on an earnings call may be the catalyst for reaching out to management and scheduling an in-depth discussion about what’s been going on. For an investor who has been researching a stock from the sidelines and is interested in getting in, an overtly positive read on the company’s recent call could provide the nudge needed to finally purchase shares.
But, for most, AI analysis is a small drop in a big pond, unlikely to independently push a portfolio manager in one direction or another.
How Should the AI Trend Affect Your Next Earnings Call Script?
It may be tempting to tear down and rebuild your entire earnings script in an effort to convey more positive sentiment. But we’d advise against overthinking this—or putting too much effort in trying to “sway” the machines. It’s likely that by using your standard approach to script writing, you’re not setting off any false positives (or negatives).
Rather, take some time to consider if what you are saying in your earnings call is actually what you believe and feel.
Rather think about whether your language on the earnings call is actually reflective of your true confidence in the business. And, beyond that, consider if in a quarter’s time (barring exceptional circumstances) your statements will remain justified. If the answer is no, then the disconnect between your confidence and your words may get picked up.
Now, this doesn’t mean that you should not or cannot keep some things close to the vest. It just means you should be more careful and thoughtful when saying one thing and thinking another, as tech-minded investors may be more aware of this than you’d expect.
And keep in mind that there is no universal good or bad outcome tied to specific words, as they are all interpreted in context. Further, many leading edge hedge funds are, in fact, more concerned with body language than diction. But possessing an awareness of this technology and its presence in the market is important for CEOs, CFOs, IROs, and others who are regularly speaking to the investment community.
At the End of the Day, Keep It Real.
In most cases, it’s okay to keep doing what you’ve been doing as long as you’re using your earnings calls for authentic and meaningful communications with your investors. Yes, it’s best practice to be aware of new technologies and how investors are utilizing them. But rest assured, the fundamentals of effective, strategic investor communications aren’t going by the wayside anytime soon.
If you could use some help with your script writing or general investor communications, reach out to us.Back To Blog