Updating Your Priors [Edition 2]: The Narrative Machine ft. Ben Hunt, founder of Epsilon Theory and co-founder of Second Foundation Partners
How Stories Shape Markets and Influence Behavior. Reading time 10 minutes.
How do companies build and propagate narratives?
That’s been a question on my mind for the last few weeks, for work and non-work related reasons. Then I saw a podcast recommendation in my colleague Sajith Pai’s 2024 wrapped featuring Ben Hunt on the Capital Allocators podcast by Ted Seides. Ben is the creator of Epsilon Theory and co-founder of Second Foundation Partners.
On further research, I observed that he writes and invests through the lens of narratives, which created enough intrigue to fall through the ‘Ben Hunt’ rabbit hole. Thanks to my podcast app Snipd (which lets me see all episodes featuring any guests across different shows), I came across 2 more excellent Ben Hunt episodes.
These notes about narratives, story-telling, semantics, and more are from three podcast episodes:
Ben Hunt - The Stories That Drive Markets, on the Capital Allocators Podcast
Ben Hunt, Author of Epsilon Theory & Co-Founder/Cio, Second Foundation Partners, on the Inside the Coffeehouse podcast
Ben Hunt — Preserving the Physical Metaverse on the Infinite Loops podcast
Let’s dive in.
Stories are used to fill in the information vacuum created by an event
The media or experts cannot simply report price movements - they are compelled to create explanatory narratives, even when movements could be completely random. These reactive stories spread like viruses through financial media, creating a self-reinforcing ecosystem of explanation-seeking.
From the Capital Allocators podcast:
…stories come about I'll say in two ways. stories will be created in reaction to price. so earning season I don't know who's first up with financials is JP Morgan. for whatever idiosyncratic reasons they have a good report and the stock jumps three, four percent that day. The job of our news media is to tell you why. You can't just say, oh, yeah, I was, I guess it was a good quarter or what often happens when stocks move up and down. You can't just say, that was variance. Right? doesn't get eyeballs. Variance isn't a good story. You have to tell your reader, your listener, you have to tell them why. You have to tell a story. So that evening Kramer will come on and he'll say, yeah, I'm bullish on the banks because XYZ. And he'll tell a story. That's not a knock on Kramer or anyone else who writes a story for the Y, that's the job. So these stories of why we're now bullish on financials or why we now think that financials are cheap, they have words and phrases and an arc to them, and you can track density of that language and the way it spreads virus-like. It's the same math that you use to track viruses.
Ben double-clicks on this in the Infinite Loops podcast:
…one of the things emerges from this research is that narrative follows price, not the other way around. So price moves for at the time, often reasons that no one knows about or understands... And so in the early two thousands, our most aggressive growth strategy, which was a small cap, but also had a momentum factor, all of a sudden started buying a bunch of tiny steel companies... Well six months later, write as one of the most prominent stories in the wall street journal, china building the equivalent of one boston every month. And what were they using to build those cities? Steel. So my point is we were buying those steel stocks because of price movement, basically, and that they were cheap. And the story didn't catch up for many, many months.
Narratives can also trigger events
Beyond reactive storytelling, there's a deliberate narrative machine at work in markets, where stories are strategically crafted to influence investor behavior. Understanding why one is being told a particular story at a particular time is crucial for market participants.
From the Capital Allocators podcast:
The other kind of story is when somebody wants to change price or behavior. So this is, the is hawkish, the fed is dovish. These are stories that are constructed in order to change behavior... you have to understand this is what's happening.
What the crowd thinks the crowd thinks
This one absolutely blew my mind and shifted how I think about investing.
Ben says investing isn't about picking what you think is valuable or even what you think others think is valuable. The real game is figuring out what the crowd thinks the crowd thinks - a third-level mind game that most investors fail to grasp.
From the Capital Allocators podcast:
This is the newspaper beauty contest where a newspaper would print the pictures of 15 pretty girls. What Keynes said is the first level of decision making is you look at the pictures and say, okay, I think that's the prettiest girl. And that's like you're a discretionary portfolio manager like I was and say, oh, I think that's the prettiest stock. I'm going to short that or I'm going to go long that. Kane says, all right, that's not it. That's not how you play the game. But it's also not how you play the game to say, oh, I talked to a lot of my friends, I get a sense of the crowd. I think the consensus is that it's this other girl over here. Kane said, that is also not how you play the game. What Kane said is that you need to play it third level, meaning what does the crowd think that the crowd thinks? It's never about you. It's never what you think, and it's never what you think the consensus is always what does the crowd think. And the way that gets determined, and this is called the larger category here in Game Theory, it's called the common knowledge games.
Markets, folklore, Hollywood movies—They all follow a few essential story arcs.
Like Hollywood has a finite number of basic plots, market narratives follow predictable structural patterns that repeat themselves across different contexts and times.
From the Capital Allocators podcast:
Story is a script. Story is a story arc. Story has a beginning, a middle, and an end. Hollywood has figured this out for a lot of years. There are famously five scripts. What's true for Hollywood and what's true for the scripts of Hollywood, and you can type in the name of your favorite movie plus three act structure, and you can get a chart of what's called The rising action and the declining action, the three act structure. It's all the same. That's also true in markets. That's also true in what we read about why this analyst is bullish on a stock or why this commentator is bearish on this sector. We tell the same stories with the same structures over and over again. There are more than five in markets, but it's a finite number. And once you start looking for those structures, once you start thinking in those terms, it really changes everything about how you see what we do for a living.
In the Inside the Coffeehouse podcast, Ben adds more details about this, citing the Hero’s journey as an example and explaining how humans are hardwired to respond only to certain types of story arcs. Thus, while the story creator may change characters, places, time periods, and context, the arc remains the same because they want the audience to accept it.
…if I'm talking to somebody from Hollywood, I say, well, an example of semantic structure is the hero's journey, right? That old story, it's a story arc, it's a script, it's a narrative archetype... And the kicker is that, I really believe this, is that we humans are hardwired to respond to that story... Now, in Wall Street, in investing, an example of semantic structure might be, well, we're bullish on the new thing. Again, can take a thousand different forms, not dependent on what the new thing is. It's dependent on telling a story in a certain way that we humans are also hardwired to respond to."
Universal Story Archetypes Ancient storytelling patterns keep reappearing in modern contexts, suggesting these narrative structures are fundamental to human communication - whether in folk tales or financial markets.
Here’s an example Ben shared in the Inside the Coffeehouse podcast of how even thousands of-year-old folktales share the same handful of story arcs.
I bought a big, thick book recently that collected Indo-European folk tales. These go back thousands of years. The story of Rumpelstiltskin, that story is way older than Christianity, right? That story goes back to like 3,500 before the Common Era... But once I started reading like three or four following one of these thematic areas, it kind of clicked to me that, oh, you know what? There's a motif in a lot of sci-fi horror movies I've seen that have this in it. I don't think it was intentional on the part of the authors. I think it's just, we are hardwired to tell these stories and we get a receptive audience. Or when an author stumbles on one of these old stories and starts to craft it and mold it for a modern age, I think they get a good response to it... There are thousands of these different thematic structures, most of which are stories that I've never heard of or read, but stories which a large language can recognize immediately.
Maintain Critical Distance from Narratives to Understand them Better
The most crucial skill in today's market isn't just understanding the narratives being presented but maintaining critical distance to see what stories aren't being told and why.
From the Capital Allocators podcast:
…you have to understand this is what's happening. And you have to ask yourself the question, I think all the time, why am I reading this now? Why is this being presented to me? Why am I being told this story? And what stories are not being told to me? The focus I'm talking about provides critical distance and requires you to go actively search for the stories that are not being presented to you. And those stories that I think are so important are the stories that other countries and their governments and their commercial entities are presenting to their people. It's always the domestic media that I find gives the most signaling power. That's true for our domestic media, signaling to us. But it's also true for understand what's happening in China or in. I want to look at what their governments are telling their own people.
Semantic Structures vs Surface Understanding
This was the biggest takeaway from the Inside the Coffeehouse podcast—Markets, like humans, tend to focus on surface-level topics and keywords, like what are the words being used - are these negative or positive/are these happy or sad.
But the deeper patterns lie in semantic structures - the underlying ways we tell and understand stories. This gap between surface and semantic understanding is crucial for market analysis.
You read the white papers from the model builders and the like. They talk a lot about semantics, the semantic layer... But in our own everyday world of investing or making movies or writing about sports teams or playing a fantasy sports game, right? We don't think in terms of semantics. We tend to think on kind of what's on the surface of our language. And on the surface are specific topics. I'm going to search for XYZ. Tell me what's being said about, in other words, what's the new thing? To use that Wall Street example I was using, let me do a search on the AI or whatever you think is the new thing. But that's not what this non-human intelligence of a large language model, that's not how it thinks. It thinks in semantics. It thinks in story. You're right, it's predicting the next word. What predicts the next word? Well, it's the underlying structure of human communication. And the underlying structure of human communication is not sentiment, it's not using mean words or nice words, it's not topic, it's semantics.
I hope you enjoyed reading this edition of the Updating Your Priors.
Rohit
Links to the three episodes covered in this edition:
https://www.capitalallocators.com/podcast/the-stories-that-drive-markets/