If Elon Musk Tries to Rattle Tesla Earnings Call Again, a Robot Will Know

An A.I. startup is making tools that can detect whether a CEO is lying on investor calls.

Elon Musk
Elon Musk has a habit of surprising Wall Street analysts on Tesla earnings calls. FREDERIC J. BROWN/AFP/Getty Images

Elon Musk’s embattled electric carmaker Tesla is set to report second-quarter earnings after the bell closes on Wednesday. And Musk’s subsequent investor call will once again be a nerve-wracking experience for Wall Street analysts. As Tesla (TSLA)’s past calls have shown, these events are often filled with unpleasant surprises that can throw Tesla’s stock price into a tailspin.

Sign Up For Our Daily Newsletter

By clicking submit, you agree to our <a href="http://observermedia.com/terms">terms of service</a> and acknowledge we may use your information to send you emails, product samples, and promotions on this website and other properties. You can opt out anytime.

See all of our newsletters

On a quarterly call last May, for instance, Musk publicly shamed an analyst for asking “boring, bonehead questions.” On another call this January, Musk unexpectedly announced that Tesla would replace its longtime chief financial officer Deepak Ahuja with a guy no one had heard of.

SEE ALSO: Is Elon Musk a Broke Billionaire?

But this kind of problem may soon be solved by a tool using one of Musk’s favorite innovations: artificial intelligence.

Amenity Analytics, a software company founded by a team of entrepreneurs and scientists in the U.S. and Israel, aims to use natural language processing, a subset of artificial intelligence, to parse a company’s current and past earnings call transcripts and detect warning signs, such as when a CEO lies or attempts to skirt around important questions.

“We have identified approximately 50 different ways a CEO can avoid answering a question. One of our customers calls it the ‘bullshit detector,'” Roy Penn, Amenity’s head for engineering, told Israeli newspaper Haaretz in a January interview.

An important factor is historical context, an area often overlooked by human analysts and particularly challenging for run-of-the-mill artificial intelligence tools.

“For example, Nvidia spoke over the course of several quarters about the gaming market and then suddenly stopped. The analysts might not notice that the company had stopped talking about it, but our system could, by analyzing previous quarters and generating insights about something suspicious,” Penn explained. “If a fairly honest executive, who only avoids answers 10 percent of the time, is suddenly lying four times as much, I would want to look at the questions that he didn’t answer. That’s critical.”

To make the results useful to analysts, Amenity’s software assigns each transcript a numeric score measuring how positive or negative the information is to investors, as it’s not always clear whether what a CEO says means good news.

“In a sentence like ‘Toyota declares a recall of its most successful and best-selling care,’ most NLP engines will analyze it positively, [because] it includes the words ‘successful’ and ‘best-selling.’ But we would be able to understand that it’s a negative event for investors because it’s talking about a recall,” Amenity’s chief product officer Mati Cohen explained to Haaretz.

When a Bloomberg reporter recently tested the software on the transcript of Musk’s infamous “bonehead questions” call last May, it scored the call as “off-the-charts for deception.”

Amenity is currently used by Intel, ratings firm Moody’s, Nasdaq and Time Warner, among other large institutions.

If Elon Musk Tries to Rattle Tesla Earnings Call Again, a Robot Will Know