Dear clients, friends, and others we hope find this interesting,
Is there a difference between a CEO who communicates well and one who doesn't? A financial difference. Like in dollars and cents.
Is it even possible to put a number on it?
This report is, as far as we can tell, the first to answer that question. And the answer is yes—we can put a number on it.
One of those numbers is $367 million.
This study measures thought leadership: "owned content" in which corporate leaders share their perspective on their business, their industry, or the world.
It finds that when the CEO of a publicly traded company releases an op-ed, delivers a keynote, or publishes an annual letter, the difference between that communication being high-quality and low-quality—we'll get to how we define those in a moment—is associated with a 0.9% difference in the company's stock price over a single week.
For the median S&P 500 company, that's $367 million in shareholder value.
For the average Magnificent Seven company, it's $25 billion—like adding Zoom's entire market capitalization between Monday's opening bell and Friday's closing.
CEO thought leadership closer to best-in-class writing was associated with stronger stock performance the following week.
Let's pause here before we tip over. Because it would be very easy to take these findings too literally, making it seem like a CEO who simply writes their annual letter like Warren Buffett will, by default, generate the shareholder value equivalent of the GDP of a small island nation. And they might. But that's not what this study is claiming. It doesn't prove causation. Nor are we claiming those $367 million and $25 billion figures are exact. Measuring something like this is, by nature, imprecise.
The dollar values are not this study’s most important findings. What matters is that even in their imprecision, those numbers are telling a pretty remarkable story. When a CEO publishes a piece of thought leadership, the quality of those thoughts and the way they're articulated create value independent of any material information.
Put another way: A CEO's words aren't a vessel for value created somewhere else.
The words are the value.
This is something communications professionals have always intuited. I'm fairly certain I once crashed the Bitcoin market by writing a dumb sentence for Janet Yellen at the U.S. Treasury. (If you're going to accidentally crash a market, I recommend Bitcoin: give it a few minutes and it's right back where it was.)
But there have never been numbers to show it. Now there are.
Why is that?
Conducting this kind of measurement has only been possible in the last few years. The math and economic theory have been around since the 1970s, but the compute power required to do this kind of research wasn't.
There has also always been a more basic problem: To measure the value of good communications, you have to define what “good” is.
But good is subjective, right? Who's to say what’s good thought leadership? And what’s bad?
As it turns out: Someone has to say it. Someone has to be willing to look at a piece of writing and say, "This one has it, and this one doesn't." Otherwise, conducting a study like this is, as far as we know, impossible.
We were resistant to this at first. We were hoping there'd be a formula—some rough checklist that could predict value. Our initial hypothesis was that ninth-grade English teacher rules could explain the difference between bottom-tier and top-tier thought leadership.
While starting Cardinal40, my co-founder Jake and I commissioned a national poll (really, Jake had a friend at a polling company who let us throw in a question). And it found that a majority of investors said a CEO's ability to communicate clearly factored into their investment decisions. So we thought we might be able to do this with a basic rubric: Does the CEO include details and data? Good. Do they use a lot of jargon? Bad.
We tested 63 text features along these lines: word count, use of data, sentiment. None predicted abnormal returns.
We tried other, increasingly advanced approaches, including whether the model itself could give us an intelligible understanding of what made high-return documents good and low-return ones bad. Nothing.
Eventually, after several months, we stopped searching for a formula that would define quality. There probably wasn't one.
Instead, we decided to run a taste test. Our idea was to curate a library of the best executive communications ever produced — or at least the ones we thought were the best. Yes, it would include Warren Buffett's shareholder letters. Also Steve Jobs's Stanford commencement, Andrew Carnegie's "The Gospel of Wealth," Marc Andreessen's "Why Software Is Eating the World," and dozens of other pieces, including some that weren't business writing at all but belong in any canon of persuasive first-person prose.
We began obnoxiously calling this library, “the canon.” And we wondered: If we benchmarked every other CEO communication in our study against it, would the ones most similar to the canon produce the strongest returns?
This, in retrospect, was something of a professional bet: If CEO communications that resembled this canon didn't generate stronger stock returns, it would mean one of two things:
(1) there was no measurable link between communication quality and shareholder value, or
(2) my taste just sucked.
Either way, I'd have just founded a consulting firm on a thesis that was wrong.
I'd love to tell you I clocked this risk. That I looked at the possibility of being publicly wrong about the one thing I'm supposed to be right about—and still summoned the courage to say: "Screw it. Let's run the numbers anyway!"
That would’ve been a great story.
But the truth is my wife and I just welcomed a baby boy, and for the past four months, the only risks I’ve focused on are the ones related to keeping him alive. My brain isn’t heroic. It’s just tired.
Anyway, good news: Julian is now 17 pounds and sleeping (mostly) through the night. And our taste test came back positive. The CEO thought leadership closer to the canon generated stronger returns.
There was also deeper satisfaction in the methodology itself. The language model we used doesn't reduce a piece of writing to a single quality score. It places every document at a precise location in a universe of 384 dimensions — a space no human mind can visualize — and it can see that great writing occupies a particular region of that space. But it can't explain what put it there.
The quality that separates valuable thought leadership from worthless (or value-destroying) thought leadership doesn't live in any structural formula — no series of rhetorical Xs and Os. Sure, you can spot the obvious notes across the good ones. But no combination of ingredients explains the result.
In other words, you can make the case that thought leadership is something of an art form.
And this study finds it is an art form the market appears to judge as particularly valuable.
One final note: Cardinal40's Thought Leadership Alpha report is, by design, the first of many. We plan to publish one in 2027, 2028, and all the years we are in business.
With more compute, a larger model, and more time, we could move from 384 dimensions to thousands—a finer-grained universe in which the distance between good and great becomes even more legible.
We could expand the corpus to include podcast transcripts, social media, and long-form interviews.
We could study whether a newly hired CEO who communicates like the best in our canon is the early signal that a value stock should become a growth one.
We could move beyond public markets entirely and study venture pitches—data that might produce an even cleaner signal, since a founder walks into a room and either gets the check or doesn't.
We are just now arriving at the borderlands between words and numbers, and in every direction, the territory is unmapped.
If you've been studying something adjacent—in finance, natural language processing, or anywhere else where language meets value—we'd love to hear from you. Please reach out.
Founder & CEO, Cardinal40
P.S. We didn't run these particular numbers, but if we did, my expectation is that I, personally, accounted for no more than a small fraction of this report's value or completion. There are many others to thank:
Shivram Viswanathan, Harvard PhD candidate and Cardinal40's senior economist, led the analysis and unknowingly lent his name to the core finding—the "Slackshiv." We considered naming the 0.9% finding the "Thought Leadership Alpha Index," but we agreed that sounded pretentious. Plus, every time I knew a new batch of analysis was incoming, I'd obsessively Slack Shiv for the latest results. So in my head, the cumulative abnormal return numbers in this report became the Slackshiv — and for as long as we produce this report, that's what it will be. Sorry, Shiv.
Adam Jorring, Cardinal40's chief economist and Assistant Professor of Finance at the University of Massachusetts Amherst, helped design and validate this study. Adam is the rare economist who believes rigorous analysis and great storytelling aren't opposing forces — and everything about this report reflects that belief. I feel very lucky he's part of Cardinal40.
Benjamin Sore, our data analyst, crunched the numbers and, in the process, became a veritable expert on the history of economic studies on communication. He's one of the hardest workers I've ever met.
Many people helped us source the corpus of documents at the heart of this report, including dozens of my friends, all excellent speechwriters, who suggested examples to measure the rest of CEO communications against. I especially want to thank Gideon Leek and Serhii Tupikin who compiled a significant portion of the documents; and Kevin Gee, who archives great corporate thought leadership via his Substack and had excellent sources and suggestions.
This study took eight months. It would've taken much longer without Doris Telford, Cardinal40's head of business operations, and her remarkable ability to coordinate meetings across 10 time zones. The fact that she showed up to those meetings because she thought the study was genuinely interesting makes me feel like people who aren't strangely obsessed with speechwriting might read and enjoy the following report.
And finally: Jake Leffew, my partner in both business and ideas. His editorial judgment has sharpened every piece of work Cardinal40 has produced, and every decision we've made. Neither this report—nor our firm—would exist without Jake. And even if it could, I wouldn't want to work with anybody else.
