Recent research from the University at Buffalo highlighted how earnings reports can be enough to get a CEO fired.  The key is something known as corporate earnings persistence, which is the degree to which the earnings of the company are expected to recur or not.

“Firms with high earnings persistence understand that the performance in the current period is likely to carry forward with the incumbent CEO, so they’re more likely to fire a CEO who yields poor earnings,” the researchers explain. “In contrast, boards of firms with low earnings persistence are less likely to fire a CEO with a poor performance because it’s likely temporary.”

Pleasing analysts is therefore something that most CEOs will be only too aware of.  How does that situation change with the rise of robotraders?  That was the question posed by new research from Georgia State University, which examined whether executives are modifying their behavior to better appeal to the machines that read financial statements and monitor earnings calls.

Tailoring the message

The research finds that executives are adapting their statements to suit the algorithms, with an emphasis on positive language over anything that might be perceived by the technology as negative in any way.

“Increasing AI readership motivates firms to prepare filings that are more friendly to machine parsing and processing,” the researchers explain. “Technological progress and the sheer volume of disclosure make the trend inevitable.”

Of course, this kind of customization of communications is nothing new, and we’ve been doing it for a long time with our human audience.  That companies and executives are doing likewise for an artificial audience is an interesting development, however.

The researchers believe that the trend began around 2011 when technology to gauge sentiment in financial reports began to reach the market at scale.  They reveal that companies who were seeing a large number of machine downloads of filings began to modify the language in those filings.

While the technology has been growing in popularity, its relative predictability has enabled executives to tailor their communications to suitably please and impress the algorithms.

“Because such rules are transparent, observable, or reverse-engineerable to at least some degree, agents who are impacted by the decisions have the incentive to manipulate the inputs to machine learning in order to game at a more desirable outcome,” the authors conclude.

So if automated technologies are increasingly assessing financial statements, it may be the uninitiated that fall into hot water as a result of their statements, leaving those able to game the rules free from trouble.