You ask your new IR assistant – let’s call her Irma – to set up a meeting with an investor to fill that hole in the London roadshow schedule. Irma scans her database to find you the most suitable target to meet with and gets to work scheduling a one-on-one meeting. Nothing so strange there – except Irma isn’t a junior in the IR team: Irma is a piece of software.
This might be the future, where instead of asking Siri to set an egg timer or Alexa to read the news headlines, you ask your virtual IR assistant to find new investor targets, analyze the shareholder base or fill a gap in the roadshow schedule. That’s just where Adam Frederick, senior vice president of intelligence at Q4, says IR tools are heading. In fact, Q4 has already come up with a name for what it hopes your newest – and possibly cheapest – member of the IR team will be called: iris.
The first job for this ‘AI engine’, which was launched in January this year, is to tackle real-time shareholder analysis. Q4 is so confident about iris that it made a public commitment to keep clients up to date on accuracy rates and even offers a money-back guarantee if shareholder ID accuracy falls below 80 percent. But this is just the beginning, says Frederick. Other services already offered by the firm are backed by artificial intelligence (AI), and Q4 is launching its own AI targeting tool at NIRI’s annual conference in June. It certainly isn’t the only company offering such services. For example, Intro-act’s AI-powered targeting tool moved to the beta stage at the start of the year and the firm is pushing into the corporate access space.
Q4 is betting hard on the technology, however. ‘AI is core to us and we are building all of our platforms around this technology,’ says Frederick. ‘Everything we are developing has attributes of AI, machine learning and big data as driving forces.’
Behind the buzz
For Barry Star, founder and CEO of Wall Street Horizon, the term AI is a bit of a bugbear. ‘I think AI is over- hyped – it’s just a buzzword,’ he says. ‘But technology – and getting smarter and better, and using technology to achieve that goal – that’s absolutely happening.
Star’s firm, which specializes in earnings calendar and events data, relies on ‘a half-million lines of proprietary code that we use to make our analysts quicker and smarter – and we use technology to get rid of all the drudgery so that they can focus on the interesting stuff. Do we call that AI? There are plenty of industry people who would, but it really isn’t AI at all. If you want to get technical about it, it’s really about making pareto-optimal decisions in your development queue to figure out how and when to automate the things you should automate.’
In other words, Star says, ‘it’s the 80-20 rule. If we can automate 80 percent of a task so we can have the analyst focus on the other, really complex 20 percent, that’s a great use of automation and a great use of technology. Would I call that artificial intelligence? Maybe as a marketing term, but not really. It’s just being smart about how one automates a process.’
Taking away the tedium
What Star and Frederick agree on, however, is the purpose of this technology: to take the tedious tasks away from humans, freeing them up to do the stuff machines can’t do. ‘The point is to make you as an IRO faster, better and able to handle more information (in both directions), because as information keeps increasing you need to really use these technologies to leverage your intelligence,’ says Star.
For him, these technologies are evolutionary rather than revolutionary but for Frederick – perhaps unsurprisingly – AI has the potential to be a game changer for investor relations. Many IROs are already using this sort of technology, he says – often without realizing it. But that’s the point.
‘For IROs, it should be seamless,’ points out Gregg Lampf, vice president of IR at network provider Ciena. ‘What I’d be looking for from AI is how to better digest and act upon the information I already get. If there’s additional information that can help me further, that’s fine, but if it’s just giving me another dose of data, that’s not going to be that helpful. I’m looking for something that makes data more actionable, helps create insights and can elevate the importance of information.’
Much of what Lampf says would benefit his three-strong IR team in terms of new technology reflects the challenges around some of his more work-intensive tasks – and that’s likely to be the case for most time-strapped IROs. ‘Targeting is a good start, at least from a strategic point of view,’ he says. ‘But I suppose from a more tactical point of view, with all the trading activity in any given stock (and a significant amount of it is machine-driven), if AI can bring to my attention the more important drivers of the stock and help me understand why it’s going up or down from a machine perspective, that would be great.’
And, like Star, Lampf believes such tools should be designed to enhance his own insights, adding that it’s unlikely he’d completely drop some of his human activities even where AI was helping.
‘Management sometimes asks me why the stock is behaving in a certain way; if it is behaving a little out of trend, having a tool that adds to my answer is something I would appreciate,’ he explains. ‘Right now, what I have to do is stay on top of it through all the emails I receive and through relationships with trading desks – and frankly I would still probably do both those things – but if AI maturity gets to a point where I can lean on it more, that would be great.’
The role of regulation
The maturing of the technology is only one element in the future of AI, however. There’s also the issue of uptake, and Frederick says part of this drive has to come from service providers. ‘Historically speaking, the IR space has been a very stagnant industry: it’s slow to move, and the same players have been around for the past couple of decades,’ he explains. ‘Usually – and in any industry – innovation comes from smaller or less mature companies that have a different perspective.
‘With any new technology adoption is often slow at first, and we see that today in the IR space. But the biggest reason for this is not necessarily that the end- user – the IRO – has been slow to adopt it. Rather, it’s that many vendors have been slow to implement it.’
Another driver is regulation. Mifid II may be a European directive, but it has a wide reach and one of its generally accepted effects is going to be an increasing workload for IROs – resulting in a need for greater resources and smarter ways of working. This will push the industry to seek out better IR tools, says Frederick. But when you’re arguing for a bigger budget, you also need to show the return on your investment.
‘With Mifid II, IROs will be increasingly pressed to have quantitative analytics they can show senior management and say, This is why we attended X conference, this is why we paid for access to be in a meeting with Y investors,’ he explains.
And that brings the conversation back to Frederick’s idea of the centralized, automated AI assistant of the future. ‘Envision an IR workflow solution that basically automates almost every single aspect of your day,’ he says. ‘From targeting and prospecting new investors, tracking your activities, notes and meetings, reporting on meeting success and the return on investment of your marketing efforts to monitoring real-time institutional ownership changes and market-structure analytics. Using a centralized intelligent platform binds it all together. We’ve only scratched the surface of what AI can do in this industry.’
Now, if you could just get Irma to make your morning coffee...
Machine to machine: your future earnings call
When it comes to artificial intelligence (AI), IR is pretty late to the game. The buy side is where you find the pros, putting machines to work on everything from exchange-traded funds to computer-driven hedge funds.
One area where the technology is being employed by the buy side – and where IR professionals could also put machines to work – is in the analysis of transcripts, earnings releases and executive quotes that examine tone or look for any mismatch of sentiment. This is a relatively new issue but it is something that Gregg Lampf, vice president of investor relations at network provider Ciena, says he’s looking to ‘be smarter about’.
‘What those machines are looking at now – to my understanding at least – is the earnings release, for example, and the various numbers,’ he explains. ‘They will look at the executive quotes and see whether there’s a mismatch between what the numbers are compared with consensus, and the content or perhaps tone of the quotes to make an assessment as to how the stock may behave and, presumably, drive its own trading as a result.’
Managing the tone of your call or the words used in your messaging is nothing new, of course. But when you script the earnings call of the future you might be thinking of the robots on the line as well as the analysts and investors. That would be an area machine learning could potentially help with.
‘I imagine AI would be much better at understanding the algorithms we need to consider – and staying on top of them as they change over time,’ says Lampf.
This article appeared in the summer 2018 issue of IR Magazine