The Epsilon Strategy
Epsilon Theory
March 25, 2019·11 comments·Money
The entire investment industry is built on fundamental analysis, yet the most successful investors ignore it. They're watching something else entirely. Wall Street creates narrative pressure to move money into and out of sectors, and this pressure follows patterns that can be measured. What if the real skill in investing isn't understanding companies, but understanding who's trying to get you to notice them and why?
- The gap between theory and practice runs deep. A brilliant $4 billion portfolio manager ignored his analyst's hundred-hour DCF models entirely. He made decisions based on sector rotation and money flow. Yet the entire investment profession is structured around doing the opposite.
- Wall Street knows something we don't, or at least acts like it does. Before the 1950s, intrinsic value analysis didn't exist on Wall Street. Traders read the tape and made money. Then we decided we were smarter and built an entire religion around fundamental analysis. The people who still get rich seem to be playing a different game.
- The game has a structure that can be measured. Financial media creates waves of attention around sectors through repeated narrative effort. That effort has quantifiable properties: how much attention it commands, how focused the messaging is, and what emotional slant it carries. These can be tracked mathematically using natural language processing on media texts.
- Different narrative patterns predict different money flows. When Wall Street creates a subtle message with positive sentiment (Early Drumbeats), money flows in months later. When a narrative gets too much attention (Overbought), money reverses. When focus disappears while positivity remains (Wavering Bull), slow selling begins. The Street's own signal-sending reveals what's coming next.
- The question isn't whether this works, but whether we can measure it fast enough to trade on it. A research program using computational analysis has generated 24 months of sector-level data. Preliminary results exist but remain unreleased. The real test is whether narrative patterns can be turned into a systematic edge before the patterns themselves change.
The Why of Epsilon Theory
- Direct access to leading narrative-tracking technology across global news.
- Deep analysis of how narratives shape markets, politics, and society.
- An active online community of independent voters, investors and thinkers.
Subscribe to Premium
Already a member? Log in
Looking for Deeper Insights?
Unlock exclusive market intelligence, trade ideas, and member-only events tailored for investment professionals and active investors with Perscient Pro.
VISIT PRO
Money
Money
Comments
This reminds me of a story an old London currency trader friend told me about working in the 80s and 90s. There were fewer computers thus less ‘anlysis’ and more gut feeling. One guy, the most successful trader on the floor, ignored virtually all the financial media but always carried a copy of the Sun with him. It ‘told’ him what the mood was that day and he’d trade on that. A bit like reading the tape but at a more meta level.
I should also ask if you think there’s anything that Foucault or those old Euro Marxists could tell us about markets and narrative. My gut feeling - and as I get older I realise that gut sometimes needs to be believed - tells me there’s something in there that’s worth studying. Certainly better value I’m sure than the CFA textbook. I say this as a former fundamentally-trained equity analyst at a long-only value fund who hardly ever used value indicators or models in his personal account.
I just saw this video today of neurons making connections, and I’m struck by the visual and metaphorical similarity to your narrative maps. https://www.youtube.com/watch?v=hb7tjqhfDus
It’s EXACTLY the same thing.
I am Baudrillard. Just in convenient English language form.
I’m guessing you can do full backwards-testing with Quid, where you can, say, examine the narrative that existed before significant market moves in a stock? That sounds pretty useful…
What exactly is the “causal” latency between the attention/cohesion/sentiment and actual money flows? I’m not sure how one would determine what the “+1” in time “T+1”. Some work with PCA maybe? I suppose that it could vary in relation to volatility and trading time.
It seems like it would be something hard to do without trying to extrapolate what is the shape of the straw that will break the camel’s back.
The Narrative Machine is a low-resolution camera right now. More like the early iPhone than a Galaxy S10. So we’re taking very simple implementations here on the sector rotation strategies, like monthly snapshots and rebalancing.
Just a little note - I tried to pick up my two red aces right handed to reproduce the Malkovich photo, and it’s possible but relatively awkward. Is there a special poker code being displayed?
Continue the discussion at the Epsilon Theory Forum...