Thesis
Almost everyone believes AI is going to keep getting better. Almost no one has a way to invest in that. Absolute Value allows you to invest in AI progress as a whole. It uses frontier AI models to find investment signals that feed into a quant algorithm, improving as AI improves. Most AI-oriented funds buy the chip makers and the model companies, but that's a bet on a few names, not on AI itself, and they are hard to access. This is a bet on AI as a whole, structured so the general public can make money from AI development.
In Absolute Value, AI ingests nearly any data stream in the world and turns it into signals a machine-learning quant model can reason over, running over 100 million calculations a day to decide which stocks to hold. This isn't ChatGPT picking stocks. The AI turns the raw world into numbers, and the quant model does the deciding, like a calculator that can't make things up. Each day it reviews its own process and adjusts how it learns, teaching itself to get better over time.
Long term, this is a bet on AI's ability to find investment signals, not on whether the market goes up. Investing always carries risk, and this can lose money. But if you believe AI is going to keep getting better, there's almost no other way to put money directly behind that. This is a fund built to harness the power of AGI.
The idea
It's hard to know why a stock moves. It could be because of earnings, the news, a product launch, or a shift in the economy. Instead of guessing which of those matter most, I built a system that tracks all of them and measures, at a very large scale, which ones have really been correlated to price changes.
Every day it measures each U.S. stock, from company data like revenue and earnings to economy-wide signals like inflation.
Like a weather forecast, it reads today's conditions and ranks which stocks are most likely to rise or fall over the next week.
It buys the stocks ranked most likely to rise and shorts the ones ranked most likely to fall.
The system won't be right every time, but it doesn't need to be. Roger Federer won only 54% of the points he played but 82% of his matches. A small edge, repeated many times, adds up.
How the portfolio is built
See it live
Progress timeline
About me
I'm a student at Stanford University where I study Artificial General Intelligence. In high school, I went to St. Mark's in Dallas and was a U.S. Presidential Scholar. This summer I worked at BlackRock as an AI implementation analyst on a new hedge fund launch called TriaXial, and last summer I worked in Private Wealth Management at Morgan Stanley.
When I was 16, I got Long Covid, and it combined with earlier concussion symptoms, leaving me with severe brain fog and memory loss. I took a year of medical leave from school and learned to rebuild how I think as a system: I took photos, kept notes, and lived by my calendar so I could organize far more than I could ever hold in my head. I learned to outsource my own intelligence. That recovery process is the inspiration behind Absolute Value. The system measures and remembers more about every stock than any person ever could.
Get in touch: patflan14@absolute-value.ai
Absolute Value n. — the magnitude of a number, regardless of sign. Events have a size. You decide the sign.