Epsilon Theory In Full
The soul of Epsilon Theory is our long-form content, a library of hundreds of pieces written by Ben, Rusty and others over the course of the last 5+ years. These are the print-and-take-home-for the weekend notes that made Epsilon Theory what it is today.
Programmable money, ImageNet: the data that changed AI research, Auto Public Offerings and the paradox of historical knowledge.
Back by popular demand, it’s the Epsilon Theory Mailbag! Today’s edition covers notes from the past two months including “Tell My Horse”, “Post-Fed Follow-Up”, “Notes…
Part 1 of this note highlighted the supremacy of the risk decision in portfolio construction. In this follow-up, Rusty observes that many investors may be assuming that the natural risk of asset classes is “right” for them.
Massively complex complexes of algorithms, AI vs. human performance, the Alpha male brain switch and explaining vs. understanding.
On episode 23 of the Epsilon Theory podcast, we’ve assembled the all-star team — Jeremy Radcliffe (Salient’s President), Rusty Guinn (Salient’s EVP of Asset Management), Neville Crawley (Founder & CEO of Engram Labs) and of course, Dr. Ben Hunt — to discuss whether we are at the inflection point when the proverbial punch bowl is taken away, and, as investors, what we do now.
The question isn’t whether the barge of monetary policy has turned around and embarked on a tightening course — it has — the question is how fast that barge is going to move AND whether or not the market pays more attention to the actual barge movements than what the barge captain says.
One model to learn them all, semantic scholars, AI is/isn’t taking over the world, and autonomous learning investment strategies.
Part 2 of Ben’s Notes from the Field series, in which he considers the question: what can a bird teach us about value investing? To everything there is a season.
What does farming have to do with investing? Quite a lot, actually. In this first of a series that takes on a life of its own, Ben discusses bees and bonds, eggs and ETFs, and more.
Of all the decisions you make as an investor, how much risk you take outweighs all of them. It is more important than costs, more important than diversification, more important than picking the right stock / fund / investment.
A quick post-Fed follow-up to “Tell My Horse”, the best-received Epsilon Theory note to date (thank you!). I’ll jump right into what I’ve got to say, without the usual 20 pages of movie quotes and the like. Well, I’ve got one quote above, because I can’t help myself. They’re the lyrics to the best break-up song ever, and they’re what Janet Yellen was singing to the market on Wednesday.
AI and video games, tricky chatbots, the quantum age has officially arrived, and your high dimensional brain.
On episode 22 of the Epsilon Theory podcast, we’re in Las Vegas at the 2017 EQDerivatives conference. Both Dr. Ben Hunt and our guest, Devin Anderson, managing director in equity derivative sales at Deutsche Bank, were speakers at the event this year. In a nod to David McCullough’s 2015 book, The Wright Brothers, this episode explores whether the ubiquitous ideas floating around finance today actually have wings and can fly.
DARPA funds a graph analytics processor, exploring long short-term memory networks, auditing black box predictive models, fast iteration and language from police body cameras.
In which Jeremy Radcliffe recommends Bob Lefsetz, Scott Galloway, Scott Belsky,Tim Urban and the gang at Hoisington.
So yeah, I’m overweight and I need to get more sleep. I’m not happy about the market, and I’m anxious about living up to my obligations to my partners and clients. But I wake up every morning thinking independent thoughts about idiosyncratic risks. I’ve got a Tribe. I’m nobody’s horse. And that’s about as good as it gets here in the Hollow Market.
The second moral license from a wise emphasis on passive investing is spending inordinate amounts of time on tilts, trades and tactical ideas that will never influence our portfolio results.
What web searches correlate to unemployment, verbal and nonverbal behaviors, and methodologies with a fragility problem.
Complex systems, machine learning software creating machine learning software, one-shot imitation and the power of the platform.
Proximity of verbs to gender, wiki-memory, fool me once (and twice), and a veritable zoo of machine learning techniques.