Big data and IR: Turning the tables on investor targeting
As in so many professions and industries, technology continues to transform the face of the IR world. And with investors sparing no expense when it comes to collecting data on the corporates they back – or are planning to – it can seem sometimes that IR is at a disadvantage when it comes to collecting intelligence about the decisions investors make, and predicting the moves they may be about to make.
This was the focus of conversation at the second breakfast briefing that IR Magazine held atop London’s famous Gherkin building in association with Q4. It was an examination about how big data had already – and would continue to – impact the capital markets, and the implications this held for the role of the modern IRO.
A quick poll of the audience found the majority did not already use big data analytics to guide their IR programs or investor-targeting strategies, with some adding that the costs seemed to outweigh the potential benefits. The majority also agreed, however, that big data and associated technologies were not a passing fad – an opinion that each of the day’s panelists returned to.
Julian Schwarzenbach, chief data evangelist at DPA, was the first to address some of the terminology that would be used throughout the day. He pointed out that the term ‘big data analytics’ was one he preferred, as it reflected the tools at use in the digital world rather than the daunting prospect of reams and reams of information with little recourse to understand it.
Schwarzenbach, something of an expert on data quality, often made reference to the idea of being ‘skeptical’ about the sources of your data, and not trusting it just because of its volume alone. An important distinction he made was that these analytics, artificial intelligence algorithms or machine-learning capabilities rely on the inputs being robust to start with. ‘What it really does is codify the intelligence you already have,’ he explained. ‘People sometimes lose sight of the fact that the business logic they’re trying to codify is actually very complex.’
This was a recurring theme of the morning: while big data techniques can save IROs a lot of time and mental load, the intelligence provided still requires a skilled IRO to make sense of and turn it into action. With that in mind, Amit Sanghvi, Q4’s managing director for Europe, explained how his firm’s intelligence gathering and processing can give IROs a detailed look at the profiles of investors that could be shifting into their company’s stock.
Q4’s AI Targeting tool takes in more than 5 mn data points on each company it covers. These include fundamental and economic data on an issuer, its dividend yield and earnings per share, how it has performed against various industry benchmarks – all backdated for the past 10 years. When combined with publicly available data about each fund or institution’s holdings, Q4 can process this information to begin to get an idea of the decisions money managers are making.
‘The purpose and goal of this system is to take the data and look at a given fund manager’s portfolio and what he or she bought and sold at different points in time,’ said Sangvhi. ‘It then combines that with information about the company and the historical macroeconomic environment and tries to unpick the underlying decision-making process for that fund manager when he or she bought or sold a particular stock.’
At present, Q4 can give users a detailed list of potential investor targets rated between 1 and 100 along with the five factors that drive their investment decisions, according to its analysis. It’s a powerful tool in formulating investor targets, and reflects something of a redress of balance in an age where investors will exploit non-financial disclosures and information you may not be aware your company is even disclosing to fuel their decisions. For example, Sanghvi explained how satellite photos of supermarket parking lots can give a clear indication of how business is going and how likely a company is to meet its guidance.
As Gunhild Grieve, head of IR at RWE, put it, being aware of these information sources is crucial to understanding how investors perceive your company. ‘There might even be other data points you’re not aware of that analysts and investors will [know about],’ she explained. ‘Other people will draw their own conclusions and even publish on it.’
Grieve illustrated this with an example of a recent incident when, ahead of some strong Q3 results, RWE’s shares were trading at a premium of up to 40 percent – good news for her and her team. On the day the results were published, however, the firm’s stock price nosedived. ‘Talking to investors we found out that although we had performed well, it was mid-November and they had year-end in mind,’ she explained. ‘They weren’t interested in profits. They thought it better to exit the stock and come back into it after year-end.’
It will take machine learning and AI tools a while to account for these sorts of situations, the panelists agreed, though they are becoming increasingly sophisticated in their ability to adapt. But tools such as Q4’s fill a notable gap in IR’s reach, particularly in a post-Mifid II environment where the sell side is less able to give corporates a good idea of investor targets thanks to it having fewer direct conversations with investors.
Turning up to meetings armed with this intelligence can also be a powerful tactic. Grieve outlined how she often confirms the findings of Q4’s analysis with investors in person when on roadshows or other similar opportunities. Spending the last 10 minutes turning the tables on investors can yield yet more insight into how your stock is held.
‘You can turn to investors and ask, Is this correct? How much do you own? Why have you held, bought or sold?’ Grieve explained. ‘You’ll be surprised how much information you may get.’
In the meantime, she added, modern IROs still need to ‘combine the intelligence with your own insight’ and sometimes your own input to achieve the strongest results. Though the buy side may have developed robo-advisers, the robo-IRO is still something of a pipe dream.
Schwarzenbach summed it up well. ‘Data is a very powerful enabler of what you want to achieve, but it’s not the whole answer,’ he said, with an IRO’s skill set, experience and expertise still required to turn these powerful sources of intelligence into effective action.
‘The key is to give you back time and free up your mental load for the rest of job,’ Sanghvi concluded.