The Codebreaker Goes to Wall Street

Inteliview Guru Story — Jim Simons EP.1

In 1964, a 26-year-old mathematician sat in an underground bunker at the U.S. National Security Agency (NSA). His job was to decrypt Soviet codes. The front line of the Cold War. Yet this mathematician had published an article in the New York Times opposing the Vietnam War and was fired by the NSA. Twenty-six years later, he was running the highest-returning hedge fund in Wall Street history. This is the story of Jim Simons' transformation from codebreaking to market decryption.

1. The Math Boy from Brookline

James Harris Simons. Born in Brookline, Massachusetts in 1938. His father, Matthew Simons, was a manager at a shoe factory. An upper-middle-class Jewish household.

Simons was fascinated by numbers from childhood. Family testimony suggests he taught himself to count at age three. In a later interview, he recalled:

"I became fixated on one idea very early on. If a car's fuel tank is halved repeatedly, does it never reach zero? Half of a half of a half… couldn't it run forever? Later, I learned this was called Zeno's Paradox. I was thinking mathematically before I even learned mathematics."

He attended MIT, majoring in mathematics. He graduated in just three years. At 22, he earned his Ph.D. in mathematics from UC Berkeley. His dissertation focused on differential geometry—specifically, research on holonomy groups on manifolds.

What does this have to do with investing? No direct connection. But there is an indirect one. Differential geometry studies patterns in high-dimensional spaces. It's training in mathematically capturing invisible structures. This is exactly what Simons later did in financial markets—mathematically capturing invisible price patterns.

2. NSA — A Mathematician in the Cold War

In 1964, the 26-year-old Simons joined the Institute for Defense Analyses (IDA), a research institution under the U.S. Department of Defense that supported NSA cryptanalysis operations.

It was the height of the Cold War. The Soviet Union and America had nuclear weapons aimed at each other. Military communications on both sides were encrypted. Decrypting these codes was critical to national security.

Simons' work was to mathematically analyze Soviet encryption systems and develop decryption algorithms. Specifics remain classified.

He learned two key lessons from this work.

First, patterns exist everywhere. No matter how complex a code, given sufficient data and mathematics, patterns emerge. Both the patterns designers intentionally hide and those they don't consciously notice. This was later applied identically to financial markets. No matter how random price movements appear, with enough data and mathematics, patterns reveal themselves.

Second, the power of computers. Cryptanalysis in the 1960s already relied on computers. Machines processed vast amounts of data human brains could not handle. Simons gained experience in large-scale data analysis using computers.

But his NSA career ended abruptly.

3. Dismissal

In 1967, the Vietnam War was escalating. Simons opposed it. He submitted an op-ed to the New York Times publicly expressing his opposition to the war.

The NSA was under the Department of Defense. The department could not tolerate an employee publicly opposing its policies. Simons was fired at age 29.

"I thought I could do two things simultaneously—work for the government while opposing its policies. I was naive."

Yet this dismissal opened his next path. Just as Thiel moved to Silicon Valley after losing a Supreme Court clerkship, and Munger started investing after growing dissatisfied with law practice.

4. Department Chair at Stony Brook

In 1968, at age 30, Simons became chair of the mathematics department at Stony Brook University, a public university in Long Island, New York. A department chair at 30—extraordinarily young. Upon taking the position, he did something remarkable: he recruited world-class mathematicians to Stony Brook. Using his connections and scholarly reputation, he brought top-tier mathematicians to the university. Within years, Stony Brook's mathematics program became one of the best in America.

During this period, Simons made his most important mathematical contribution: the Chern-Simons form. Developed jointly with the great Chinese mathematician Shiing-Shen Chern, this theory connected differential geometry and topology through invariants. It became foundational to theoretical physics—particularly quantum field theory and string theory.

The Chern-Simons theory became one of the most-cited mathematical papers in physics. Simons could have secured his place in mathematics history through this work alone.

Yet he was not satisfied with mathematics alone.

"I love mathematics. But mathematics has one problem—it doesn't make money."

As department chair, he began investing as a side business. Initially, he traded commodity futures with personal funds—sugar, soybeans, currencies. Trading based on instinct.

Results were not bad. He made decent returns over several years. But Simons was dissatisfied.

"When trading on instinct, you're sometimes right, but you don't know why. And you don't know why you're wrong either. For a mathematician, this is unbearable."

He decided to replace intuition with mathematics.

5. "Let's Build a Model"

In 1978, at age 40, Simons founded the investment firm Monemetrics while retaining his department chair position. This company later changed its name to Renaissance Technologies.

Simons' approach was unlike anything on Wall Street at the time.

In 1978, Wall Street investing worked like this: Fund managers analyzed companies. They met with executives. They researched industries. They forecasted macroeconomics. Then they bought and sold based on their judgment. Buffett, Soros, Lynch, Druckenmiller—all used this approach. Human judgment was central to investing.

Simons sought to eliminate human judgment.

"Humans have biases. Munger created a list of 25 biases, right? I wanted to 'remove' those 25 biases. Mathematical models have no bias. No emotion. No fear, no greed. If the model says 'buy,' we buy. If it says 'sell,' we sell. Humans don't get involved."

His plan was straightforward:

• Step 1: Collect vast amounts of financial market price data.
• Step 2: Find statistically significant patterns in that data.
• Step 3: Transform those patterns into mathematical models.
• Step 4: Let the model trade automatically. Humans only build the model; computers execute trades.

This was the birth of quantitative investing.

6. Assembling Mathematicians

Simons did not recruit from Wall Street.

The people he hired: mathematicians, physicists, astronomers, computer scientists, statisticians, cryptographers.

MBAs? None. Finance veterans? Almost none. What Simons wanted was not financial knowledge but pattern recognition ability. His hypothesis was that anyone who could decrypt codes could also decrypt market patterns.

Early key figures:

Leonard Baum: Mathematician and co-developer of the Baum-Welch algorithm, a method for estimating Hidden Markov Model parameters that became foundational to speech recognition technology. Simons applied it to financial market analysis—using it to estimate market "hidden states."

James Ax: Mathematician and Cole Prize recipient (mathematics' equivalent to the Fields Medal). Algebraist. He developed early trading models with Simons.

Elwyn Berlekamp: Mathematics professor at UC Berkeley, master of information and coding theory. After his involvement, model performance improved dramatically.

Later, physicist Robert Mercer and computer scientist Peter Brown joined. Both had researched machine translation at IBM, bringing statistical natural language analysis methodology that was also applied to market data analysis.

Simons' team-building approach resembled PayPal's Peter Thiel's: hire only extraordinarily talented people—then don't teach them finance, but rather apply their tools (mathematics, physics, computer science) to finance.

7. Early Struggles

Renaissance Technologies' early years were not smooth.

In the early 1980s, Simons' model traded commodity markets: currencies, bonds, futures. Results were mixed. Some years were good, others bad.

Two problems emerged:

First, insufficient data. In the 1980s, digitized financial data was scarce. Simply obtaining computer-analyzable price data was difficult. Simons' team had to manually digitize old newspapers, exchange records, and government statistics.

Second, model limitations. Early models used simple trend-following strategies. If prices rose, bet they'd rise more; if they fell, bet they'd fall further. This worked in trending markets but lost money in sideways markets.

Simons continuously improved the model. He recruited more mathematicians and scientists, collected more data, and refined algorithms.

Then, in 1988, a turning point came.

8. The Birth of the Medallion Fund

In 1988, Simons launched the Medallion Fund. Named after the mathematics medal (Oswald Veblen Prize) that Simons and Ax had received.

Medallion was Renaissance's flagship fund. It employed the most sophisticated models and was managed by the most brilliant team.

• 1988 (launch year): 16%—decent, but not extraordinary.
• 1989: -4%—a loss. Simons extensively revised the model.
• 1990: 55%—explosive.
• 1991: 39%
• 1993: 39%
• 1994: 71%

The model was working. And once it started working, it never stopped. From the late 1990s through the 2000s and 2010s. Thirty years.

That 30-year record became the most remarkable figure in human investment history. That story comes next.

9. Three Lessons from This Story

First, revolutionary innovation comes from complete outsiders.
Simons was a mathematician with zero Wall Street experience. The people he hired were physicists, cryptographers, and astronomers—not finance experts. Yet these "outsiders" beat every Wall Street "insider." This mirrors Lynch finding stocks at shopping malls, outperforming Wall Street analysts. People immersed in existing methods struggle to escape them. Those with completely different tools create new possibilities.

Second, eliminating human judgment is itself a strategy.
Munger taught recognizing and avoiding 25 human biases. Simons went further—not avoiding biases but completely removing humans from the trading process itself. The model decides; the computer executes. Humans only improve the model. This isn't applicable to every investor. But everyone can ask: "Am I trading emotionally?" If yes, adopting automated rules (stop-loss orders, automatic rebalancing) is the individual investor version of Simons' approach.

Third, early failure doesn't guarantee eventual success, but eventual success is impossible without early failure.
Medallion's first year returned 16%, the second year -4%. Nothing extraordinary. Had Simons quit here, humanity's greatest fund would never have existed. Just as Livermore returned to New York after his first bankruptcy, Burry weathered two years of insurance premiums, and Keith Gill posted videos from a basement for 18 months. Enduring early failure is a prerequisite for success.

10. End of 1988

1988: Jim Simons at age 50. Twenty-one years since his NSA dismissal. His first year away from Stony Brook. Medallion Fund's launch year.

He had 36 years ahead. During those 36 years, Medallion would achieve returns no investment vehicle in human history had ever approached. And the secret of those returns has never been fully disclosed.

What Simons created was not just a hedge fund. He created a machine for decrypting markets. The mathematician who decrypted Soviet codes built a machine to decrypt market codes.

That machine's performance is revealed next.

Next Episode Preview

EP.2 — Medallion: Humanity's Greatest Returns
1988-2020. Thirty-two years. 66% annualized pre-fee; 39% annualized after fees. A $1,000 investment becomes approximately $400 million in 32 years. Buffett, Soros, and Lynch never approached this record. And even Simons himself never fully explained why this machine worked.