When Jim Simons started Renaissance Technologies, most of Wall Street thought he was solving the wrong problem.
Hedge funds in the late 1970s and 80s were built around people who could read balance sheets, interview CEOs, and forecast economies. Simons built something almost perverse by comparison — a firm that avoided hiring economists and traders altogether. Instead, Renaissance filled its ranks with mathematicians, physicists, astronomers, and cryptographers. People who had spent their careers finding faint signals inside massive, noisy datasets — cosmic radiation, encrypted messages, particle collisions — not people who had spent careers explaining markets.
The bet was simple to state and brutally hard to execute: if markets have any repeatable structure at all, it will show up in the data long before it shows up in a story anyone can tell about "why." So instead of theories, Renaissance built models. Instead of predictions about the economy, it looked for statistical relationships — patterns that recurred often enough, and robustly enough, to bet real money on, even if nobody could fully explain why they existed.
This was not a smooth start. In the early-to-mid 1980s, Renaissance's models produced middling, sometimes losing, results. Simons himself stepped back from direct trading and, crucially, made the decision to trust the system-building over his own market instincts — even though he had run a successful discretionary macro fund before this.
In his own words: "I don't have to worry about the human tendency to get scared out of things at the wrong time or get excited about things at the wrong time — because we've programmed the system to do this."
That single decision — to trust a model over a feeling — is what eventually gave rise to the Medallion Fund: the vehicle that would go on to post the best long-term returns of any investment fund in history.
Next: inside Medallion — the fund almost nobody outside Renaissance was ever allowed to invest in.