Detecting Deception in Conference Calls

Research paper: Detecting Deceptive Discussions in Conference Calls, by David F. Larcker and Anastasia A. Zakolyukina.

Abstract: We estimate classification models of deceptive discussions during quarterly earnings conference calls. Using data on subsequent financial restatements (and a set of criteria to identify especially serious accounting problems), we label the Question and Answer section of each call as “truthful” or “deceptive”. Our models are developed with the word categories that have been shown by previous psychological and linguistic research to be related to deception. Using conservative statistical tests, we find that the out-of-sample performance of the models that are based on CEO or CFO narratives is significantly better than random by 4% – 6% (with 50% – 65% accuracy) and provides a significant improvement to a model based on discretionary accruals and traditional controls. We find that answers of deceptive executives have more references to general knowledge, fewer non-extreme positive emotions, and fewer references to shareholders value and value creation. In addition, deceptive CEOs use significantly fewer self-references, more third person plural and impersonal pronouns, more extreme positive emotions, fewer extreme negative emotions, and fewer certainty and hesitation words.

Posted on August 26, 2010 at 6:15 AM23 Comments


Glenn Willen August 26, 2010 7:24 AM

“In addition, deceptive CEOs use significantly fewer self-references, more third person plural and impersonal pronouns, more extreme positive emotions, fewer extreme negative emotions, and fewer certainty and hesitation words.”

Not anymore.

Grant Gould August 26, 2010 7:34 AM

That pretty well describes the CEOs of the last three failed companies I worked for.

If nothing else, it’s a pretty good when-to-polish-up-the-resume metric.

Anon. August 26, 2010 8:10 AM

So they do a bunch of analysis and get 4-6% better than flipping a coin? Wow! 🙂 OK, guess that’s better than not doing any objective analysis at all.

Unfortunately, I suspect that (some) CEOs would make deceptive statements even if they knew their lies would be detected. As long as it happened after their quarterly bonus.

kingsnake August 26, 2010 8:43 AM

I simply assume that anyone with any power is lying all the time.

I am rarely disappointed …

Piper August 26, 2010 9:06 AM

I have an easier and more accurate algorithm for this.

bool IsCeoDeceptive()
return true;

kangaroo August 26, 2010 9:15 AM

@Anon: you’ve got it backwards — the point is that looking at balance sheets and other such process oriented methods for identifying whether execs are lying about the company going forward are basically random.

So personal credibility — reading the exec — is better than an army of accountants and laws to keep things on the up-and-up. Of course, the reduction to “impersonal pronouns,” etc is trivial.

But goes to show once again that there is no objective PROCESS for running a human system well — there is just good and bad judgment that has to fall on a person. The further we go down the path of trying to automate & bind by rules the making of judgments, the more corrupt the system will become.

It also shows the vast advantage that a small elite have in investments over the common stockholder — they not only have multiple face-to-face opportunities to identify lying, they have a hell of a lot of personal information on the exec to help parse out his lying. They know him from the hamptons, they know whether he picks up his dogs poop, how he acts when driving around town, whether he double-dips at the chip tray.

All the accounting rules aren’t as good seeing how a man treats his lawn guys.

Neil Smith August 26, 2010 10:11 AM

“Our primary assumption is that CEOs and CFOs know whether Financial state-
ments have been manipulated, and their spontaneous and (hopefully) unrehearsed narratives
provide cues that can be used to identify lying or deceitful behavior.”

Not knowing your Y makes any analysis on your Xs kind of pointless!

k August 26, 2010 10:20 AM

I wonder how this would work combined with accruals (classic quant signal for aggressive accounting).

HJohn August 26, 2010 11:07 AM

@Anon: “Unfortunately, I suspect that (some) CEOs would make deceptive statements even if they knew their lies would be detected. As long as it happened after their quarterly bonus.”

Part of it is the nature of power, or more specifically the lack of scrutiny that comes with enormous power. Since there are rarely consequences, they will say what they want and what works.

A CEO that my wife used to work for did something unbelievably unethical if you ask me. He took some employees, laid them off, and rehired them under contract… he did this so he’d retain their work, but they would no longer have sick time and the office would no longer match their pension contributions. Because he regards himself as so valuable, he upped his salary to $250,000 a year–meanwhile, he cheated people making about $30,000 out of their pension and benefits.

Most of his 50 employees and contractual employees are furious at him, but no one tells him because they dont’ trust him not to rip them off too (or rip them off again).

Lack of accountability is bad for anyone. I think that explains why power may corrupt.

Jon August 26, 2010 11:09 AM

Oddly enough, the sociopathic CEO is the successful CEO.

If you really want to analyze the treatment of the lawn guys, look for a CEO who promises his lawn guys the world, screws them over repeatedly, and they still kinda sorta respect him or her.

That’s the guy you want to improve your stock holding valuation.

But you might want to sell in a couple of years, before the bubble pops.


BF Skinner August 26, 2010 11:43 AM

@Jon “Oddly enough, the sociopathic CEO is the successful CEO. ”

Nothing odd about it. The successful performance of the sociopath within corporate hierarchy has been noted since the 60’s in the previous century.

A person without conscience, moral sense, whatever, has a greater range of alternatives to choose from. Once they make the jump that they can treat people as things any atrocity becomes possible for them.

It says a lot more about corporate culture that, if not actually encouraged, it’s tolerated and within norms. Perhaps it’s a consequence from separting management from firm owners. I know AT&T fought our stockholder resolutions tooth and nail to keep from divesting in South Africa during apartheid.

Remember 1 in 24 people in the US now are qualifying sociopaths.

kangaroo August 26, 2010 12:02 PM

@BF Skinner:

Modern society has been based on a sociopathic elite all the way to its founding. The treatment of human beings as economic resources can be seen with the Irish famine in the 19th century, the commodification of slavery in the 18th century and before — the underlying model of modern civilization.

Note Sitting Bulls response when he did his tour of major American cities. Here was a man who was quite unsentimental, more than willing to go into a low-probability fight and lose the bulk of his fighting population — but he found the objectification and discarding of human beings by the “moderns” as beneath contempt.

Not only is it not odd, it’s odd that anyone would think it odd that sociopathy is rewarded by modern societies. Put it in reverse — the bulk of us who aren’t sociopaths are actually mad.

BF Skinner August 26, 2010 12:31 PM

@kangaroo “…us who aren’t sociopaths are actually mad.”

Somthing I think I often suspect. But I wouldn’t go cynical.

I have a theory (it could be bunnies…no wait something isn’t right there) that in the same way cultures maybe more evolved than others (for a given value of evolution) our minds and beliefs evolve too. But not everyone is at the same level at the same time. So when someone says that guy is a real cave man he or shee can be speaking the literal truth. Our biology is still calibrated to the savanah (Africa not Georgia) So we’ve got, in the U.S., a lot of people still living in the old west 19th c. and they (with their emphasis on rugged individualism) come into conflict with minds calibrated for the industrial age and their focus on cooperative large scale enterprise (who are in conflict with those calibrating for the information age with our emphasis on distributed networking and reuseable objects and Halo night).

It could be back in the Ogg days a far larger percentage of human population was sociopathic and it’s been through millenium of civilization we’ve developed ever widing empathy for those who are not just ourselves, our families our neighbors, but even (god help us) people who don’t look, talk, or pray like us. So the percentage of sociopaths is actually a positive sign of progress (with more to do).

What is worisome is the behavior of our corporate citizens. The US is giving ever greater ‘person’ rights to organizations that have no moral sense. If a ‘person’ can’t make a moral choice how can they be law-abiding or held to be acceptable in society? The choice should always come down on ‘profit to shareholders, screw the stakeholders’. We’ve cited, here time and again how an organization faced with a profitable or a legal course of action has chosen the profitable fully intending to pay the fines.

BF Skinner August 26, 2010 12:34 PM

The concept i was groping for was ‘accountable’. A person can’t be held accountable for their actions under some circumstances.

David Thornley August 26, 2010 2:24 PM

@kingsnake, Piper: Much like my terrorist detection device, which functions at least 99.9999% accurately in normal use, which has yet to be properly evaluated by the TSA or DHS.

Ravan Asteris August 26, 2010 5:35 PM

About 6 months before Global Crossing went south, I was listening to a conference call where they discussed their quarterly results. My BS detector was going into overdrive, because of the sheer number of accounting buzzwords and obfuscation that I heard. Yet everyone afterward was talking about how much stock they would get. Me? I started circulating my resume.

If it is not straightforward, plain English and simple income, expenses and profit, but relies on funky accounting acronyms and arcane tricks to explain the results? Then it’s probably BS.

Sasha van den Heetkamp August 26, 2010 10:42 PM

I wonder how an elevated state of perception and skepticism can influences someone else to the point that the other person feels the need to convince you by utilizing social lies which are important for daily social functioning, or distort your perception thereof. It’s my adage; I am skeptical of skepticism.

Jon August 27, 2010 2:10 AM

@BF Skinner, @kangaroo:

What’s odd is that deliberately antisocial behavior is rewarded, both financially and powerfully, by the society it is actively working against.

Given that, why should they change their ways?


Winter August 27, 2010 3:55 AM

Sociopaths rewarded? Of course.

If you send in the troops to loot, pillage, and massacrer, you need a homicidal maniac or sociopath mercenary. You will not put a renowned altruist in charge.

CEOs are put in place to extract as much money out of a company as is possible. Preferably leaving nothing of value standing. If they also rip off the shareholders who hired them this is the “risques du metier” of hiring mercenaries (not likely, it is mostly other shareholders that are ripped off).

It is not so much that society rewards sociopaths. It is that some quarters of society put sociopath mercenaries to good use ripping of their fellow citizens.

paul August 27, 2010 9:23 AM

You don’t even need the CEO to be aware of balance-sheet manipulation for this to work (statistically speaking). The kind of CEO who uses this kind of language will also generally be the kind who pressures subordinates to report positive results under all conditions, and the kind who does not inquire closely into how those reported results are achieved.

Leave a comment


Allowed HTML <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre> Markdown Extra syntax via

Sidebar photo of Bruce Schneier by Joe MacInnis.