‘We’ve missed earnings forecasts lol’

Stay informed with free updates

Simply sign up to the Corporate earnings and results myFT Digest — delivered directly to your inbox.

Alphaville has belatedly been made aware of some research on a subject close to our heart — bad jokes about bad news.

From a paper published in the Review of Accounting Studies earlier this year:

When managers use humor on an earnings call, stock market returns and analyst forecast revisions following the call are more positive, primarily because of a muted response to negative earnings news. Consistent with managers’ successful use of humor being a favorable signal of future firm performance, we find no evidence of a return reversal over the subsequent quarter, and managers’ use of humor predicts more favorable news at the subsequent quarter’s earnings announcement.

There’s a growing body of work on the subject of linguistic analysis and finance as economists use tools like natural language processing to systematically scour central bank speeches, newspaper headlines, earnings calls and even recondite bond docs for valuable nuggets.

This paper — by Andrew Hall, Rachel Flam, Joshua Lee and Nathan Sharp — uses machine learning to look for instances of laughter on almost 12,000 earnings calls, who triggered it, and what effect it seemed to have.

As you’d expect, most CEOs are serious people. Just 11.9 per cent of earnings calls feature levity of some kind. Nor are we talking Friday Night at the Comedy Store. Here’s an example of something that triggered some hollering from Honeywell’s Q3 2016 earnings call:

Steve Tusa, JPMorgan: “Who’s making the call on the buyback at this stage?”

Dave Cote, chair and CEO of Honeywell: “Well, consistent with our policy over the last 15 years, I try to make all of these decisions with no input from anybody.”

Mwa wa wa. Still, even the lamest of jokes can get a laugh out of access-thirsty analysts. And it seems to work, judging by the market reaction.

The researchers examined the two-day market reaction after earnings calls, and found it tended to be more positive when it had been mirthful. But the results were mostly driven by a more muted response when the earnings were disappointing. Joking around somewhat blunted the impact of bad news, while joking about good results generally didn’t have much impact.

A follow-up analysis 30 and 60 days after the call indicated no subsequent reversal either. Maybe, rather than jokes being a means for obfuscation, levity is a hint that management remains confident about the outlook:

In summary, after controlling for the earnings surprise, the tone of analysts’ questions and managers’ responses, and other analyst and firm characteristics, we show that humor plays a meaningful role in public earnings conference calls. Further, humor has a significant association with immediate stock market reaction, subsequent analyst behavior, and future firm performance. Taken together, our evidence suggests that humor can soften the disclosure of negative news and signal relatively stronger future firm performance.

By the way, the paper also indicates that about 20 per cent of analysts managed to trigger some laughter at least once in the 11,721 earnings call sample, but more should probably do it more often:

We find that analysts who use humor successfully are allowed to ask longer questions, receive longer responses, and have more opportunities to interact with management later in the Q&A session than other analysts on the call — collective evidence that humor is associated with immediate benefits on the call.

We’re looking forward to monitoring the impact of this research in upcoming earnings calls. Suggestions on the funniest CEOs in the comments below (we already know who is the sweariest).