Table of Contents
Every quarter, a ritual plays out across corporate America. CEOs step up to the microphone, clear their throats, and begin speaking to analysts, investors, and the broader market. They talk about “robust pipelines” and “strategic positioning” and “exciting opportunities ahead.” They sound confident. They sound assured. They sound, frankly, like they have rehearsed this forty times in front of a mirror. Because they have.
Here is the thing most investors never stop to consider: these words are free. A CEO pays no penalty for saying “we are incredibly optimistic about the second half.” There is no fine for misplaced enthusiasm. No regulatory body will knock on their door because their “transformational initiative” turned out to be a mild process improvement. And when words are free, economists have a precise term for what follows.
They call it cheap talk.
A Quick Detour Through Game Theory
Cheap talk is not an insult. It is a formal concept in game theory, a branch of microeconomics that studies how rational actors make decisions when their outcomes depend on what others do. The idea was rigorously developed by economists Vincent Crawford and Joel Sobel in 1982, and it describes a situation where one party can send messages to another, but the messages cost nothing to send and nothing to verify.
Think of it like poker. A player who goes all in might have a royal flush or might have nothing at all. The bet itself carries a cost, which makes it a meaningful signal. But imagine a version of poker where players could simply announce their hand before betting, with no obligation to tell the truth. That announcement would be cheap talk. It carries no inherent credibility because there is no penalty for lying.
Now replace the poker player with a CEO and the poker table with a quarterly earnings call. The structure is almost identical. The CEO has private information about the company. Investors want that information. And the CEO has every incentive to present it in the most favorable light possible.
Why CEOs Are Not Exactly Neutral Narrators
This is where the incentive structure matters, and where most investors stop thinking too early.
A CEO’s compensation is typically tied to stock price performance. Their reputation, their job security, their legacy narrative all hinge on whether the market perceives them as competent stewards of capital. So when a CEO tells you that things are going well, you have to ask a simple but uncomfortable question: would they say anything different if things were going poorly?
In many cases, the answer is no. And that is the core problem with cheap talk. If a sender would send the same message regardless of their private information, then the message contains no actual information at all. It is noise dressed up as signal.
This does not mean every CEO is lying. That would be too simple and too cynical. What it means is that the structure of the situation makes it rational for them to be optimistic regardless of reality. The game theory is not about morality. It is about incentives.
The Crawford Sobel Model, Without the Math
Crawford and Sobel showed something surprising in their original paper. Even in a world of cheap talk, some information can still get through. But only under specific conditions.
The key variable is what they called “bias.” If the sender’s interests are perfectly aligned with the receiver’s interests, cheap talk becomes fully informative. The sender has no reason to distort. But as the bias between sender and receiver grows, the amount of real information that can be transmitted shrinks. Messages become vaguer, more pooled, less useful.
Apply this to earnings calls and the insight is immediate. A CEO whose compensation is entirely in long term restricted stock, whose interests are closely aligned with long term shareholders, is more likely to be informative. A CEO sitting on a pile of options that vest next quarter has a much larger bias. Their words should be discounted accordingly.
This is not a radical idea. But it is remarkable how few investors actually perform this calibration. They listen to the words and forget to examine the mouth they are coming from.
The Language of Saying Nothing
If you start listening to earnings calls with a game theoretic ear, patterns emerge quickly. There is an entire vocabulary that has evolved specifically to sound meaningful while committing to nothing.
Phrases like “we remain confident in our long term strategy” are the corporate equivalent of a weather forecast that predicts “temperatures.” Yes, obviously. There will be temperatures. But which direction?
“We are seeing green shoots” is another classic. It borrows the language of organic growth, of nature, of inevitability, to describe something that may or may not be happening. Green shoots are conveniently impossible to quantify. You cannot ask a follow up question like “how many green shoots, exactly?” without sounding absurd.
Then there is the masterclass move: strategic ambiguity. “We are exploring several options to unlock shareholder value.” This sentence contains no information whatsoever. It does not tell you what the options are, when they might be pursued, or what “unlock” means in any operational sense. But it sounds proactive. It sounds like leadership. And that is precisely the point.
The Poker Tell: When Cheap Talk Becomes Expensive
Here is where it gets interesting for investors who actually want to extract signal from noise.
Game theory predicts that cheap talk becomes more credible when it is costly. Not costly in dollars, necessarily, but costly in terms of reputation, future flexibility, or strategic positioning. When a CEO says something specific enough that they will look foolish if they are wrong, the cheap talk framework starts to break down. The talk is no longer cheap.
Consider the difference between these two statements:
“We expect strong performance in the coming quarter.”
“We expect revenue between $4.2 billion and $4.4 billion, driven primarily by our North American enterprise segment.”
The first statement costs nothing. The second costs a great deal. If that revenue number comes in at $3.8 billion, there will be consequences. Analysts will replay the tape. The financial press will have a field day. The CEO’s credibility takes a measurable hit.
This is why specific, falsifiable guidance is so much more valuable than vague optimism. It is a signal precisely because it is costly to fake.
What the Analysts Are Really Doing
Watch the Q&A portion of any earnings call and you will see another game being played in real time. Analysts are not just asking questions. They are designing strategic probes intended to force the CEO out of cheap talk territory.
A good analyst question does not ask “how do you feel about the competitive landscape?” That invites a scripted non answer. A good analyst question asks something like “can you quantify the impact of the pricing change on your gross margins in Q3?” This is a question that demands specificity. It is harder to answer with vague optimism. It forces the CEO into costly talk territory, where the information content goes up.
The best analysts understand this intuitively. They are not looking for the CEO to reveal secrets. They are constructing conversational traps that make it expensive to be vague.
The Silence Signal
There is one more dimension to this that most commentary overlooks, and it might be the most valuable of all.
In game theory, what you do not say can be just as informative as what you do say. This is sometimes called the “unraveling” result. If a company has good news, they have every incentive to share it. If they stay silent on a topic, rational observers should infer that the news is probably not good.
Pay attention to what gets omitted from earnings calls. If a CEO spends twenty minutes discussing three business segments and barely mentions the fourth, that fourth segment is probably struggling. If last quarter’s call featured extensive commentary on a new product launch and this quarter it vanishes from the script, that launch is likely not going well.
Silence is not neutral. In a world of cheap talk, silence is the most expensive signal of all, because it forfeits the opportunity to spin.
The Counterintuitive Takeaway
Here is the part that feels backwards but is supported by the theory: you should sometimes trust a pessimistic CEO more than an optimistic one.
Think about it through the incentive lens. A CEO who admits to challenges, who acknowledges headwinds, who says “this quarter was disappointing and here is why” is engaging in costly talk. They are paying a reputational price for honesty. They are doing the equivalent of showing you a bad hand at the poker table voluntarily. Why would they do that unless it were true?
Optimism is cheap. Pessimism is expensive. And expensive signals are credible signals.
This does not mean you should invest exclusively in companies run by pessimists. That would be absurd. But it does mean you should weight negative disclosures more heavily than positive ones when forming your view of a company’s true position.
Playing the Game Better
Earnings calls are like games, and like any game, you play it better when you understand the rules.
The rules are straightforward. Words are cheap. Specificity is expensive. Incentives shape messages. Silence speaks. And the structure of who is talking, why they are talking, and what it costs them to be wrong matters far more than the words themselves.
Next time you listen to an earnings call, do not just listen to what the CEO says. Listen for what it costs them to say it. That is where the real information lives.
The bluff only works on players who do not know the game is being played.
