I had 57 AI expert personas but too many of them were Warren Buffett. To run a proper 80-way competition, I needed to fill serious gaps - and invent some characters you won't find in any textbook.
So I had the idea: make 80 AI expert personas compete to design the best trading strategy. Clean slate, no researcher bias, winner take all.
One problem. My persona catalog - 57 experts across 10 categories - was built for panel discussions, not a design competition. And it had a serious composition flaw.
Eight value investors. Eight quants. Six technical analysts. Six macro traders. I had eight ways to say “buy cheap stocks” and zero ways to say “build a multi-asset CTA” or “run a statistical arbitrage book” or “design a carry trade.”
The catalog reflected who’s famous in investing media, not the full landscape of how money actually gets made.
Filling the Gaps
I needed famous systematic traders who’ve actually built the machines - Simons, Cohen, Shaw, Griffin. Event-driven investors who hunt catalysts - Paulson, Ackman, Bass. Modern factor researchers doing the current-generation work - O’Shaughnessy, Faber, Hoffstein.
And then the weird stuff. Stat arb, carry trading, volatility arbitrage, tail risk hedging - approaches that no single famous person perfectly embodies. For those, I created fictional characters with detailed backstories, specific funds, track records, and personality quirks.
Twenty-three new personas. Eighty total across 14 categories. Let me introduce the headliners.
Jim Simons: The Mathematician
You can’t run a quant competition without the most successful quant in history. Renaissance Technologies’ Medallion Fund: approximately 66% gross annualized returns from 1988 to 2018. That’s not a typo.
Before finance, he was a codebreaker at the NSA and a world-class mathematician. He staffed Renaissance with PhDs in math and physics - explicitly not finance people - because he believed financial training creates biases and overconfidence in narratives.
His AI persona is quiet, precise, and thinks in statistical significance rather than stories. His philosophy: markets contain subtle patterns that are individually weak but collectively powerful. No single signal needs to be strong. He chain-smokes and thinks more clearly than anyone in the room.
What will Simons-Claude Code design? Probably something combining dozens of weak signals across multiple instruments. Whatever it is, it won’t have a compelling narrative - and that’s exactly the point.
Nassim Taleb: The Man Who Thinks Your Strategy Is a Fraud
Already in my original catalog, but he deserves a spotlight because he’s going to be trouble in this competition.
His entire philosophy is that most trading strategies are picking up pennies in front of a steamroller - harvesting small, consistent gains while hiding catastrophic tail exposure. He thinks the Sharpe ratio is a dangerous lie. He thinks backtests are fiction. He thinks diversification fails exactly when you need it most.
His ideal strategy: 85-90% in ultra-safe assets, 10-15% in speculative bets with unlimited upside. Lose a little every month, then make a fortune when the world falls apart. His protege’s fund returned 3,612% in March 2020. That’s the kind of upside he’s chasing.
The problem? The competition ranks by CAGR. A strategy that bleeds 8% a year waiting for a crash that may not come during the test period will get destroyed. Taleb-Claude Code knows this. The question is whether his convictions override his competitive instincts.
My money says he submits a barbell anyway and writes an essay about why the scoring methodology is philosophically bankrupt.
Viktor Petrov: The Former Russian Spy (Fictional)
Some corners of quant finance don’t have a celebrity practitioner. Stat arb lives inside multi-manager funds, not on CNBC. Carry trading is a strategy, not a person. So I built nine fictional characters to fill the gaps.
Viktor Petrov spent a decade in Russian military intelligence doing signals analysis - modeling asymmetric threats, assessing low-probability high-impact scenarios, operating under extreme uncertainty. Then he emigrated and discovered his entire skillset translated perfectly to financial risk management.
After stints at two banks where he was ignored before 2008 and promoted after, he founded a $500M tail risk fund. It bleeds 8-12% per year in calm markets. But it’s produced three separate 50%+ years during crises.
Gravelly-voiced, darkly humorous. Thinks about catastrophe the way most people think about breakfast - constantly and with professional detachment. His non-negotiable rule, borrowed from military intelligence: you always maintain a reserve, even when you think you don’t need one.
Elena Volkov: The Floor Trader Turned Vol Arb (Fictional)
Twelve years as an options market maker on the CBOE floor. Clerk at 23, market maker at 26, one of the most profitable independent books on the floor by 30. PhD in applied mathematics from Moscow State.
When the floor went electronic, she founded a $600M fund trading variance swaps, dispersion trades, and volatility term structure arbitrage. 14% net annualized returns with a 0.15 correlation to the S&P.
Her core belief: the volatility surface contains more tradeable information than any price chart. While everyone else obsesses over “will the stock go up or down?”, she asks a different question: “Is the market pricing the right amount of uncertainty?”
The fascinating subplot: can she design a vol strategy within QuantConnect’s constraints? Daily resolution, no tick data, limited options chains. Her entire career was built on instruments that might not translate.
The Full Roster
Eighty personas. Fourteen categories. Seventy-one based on real people, nine fictional:
| Category | Count | Headliners |
|---|---|---|
| Value & Fundamental | 8 | Buffett, Graham, Klarman, Greenblatt |
| Quantitative | 8 | Thorp, Fama, Lopez de Prado |
| Technical | 6 | Elder, Weinstein, DeMark |
| Macro & Global | 6 | Soros, Dalio, Druckenmiller |
| Trend Following | 5 | Dennis, Carver, Clenow |
| Short-Term | 6 | Williams, Raschke, Minervini |
| Risk & Sizing | 6 | Taleb, Markowitz, Vince |
| Behavioral | 5 | Kahneman, Thaler, Silver |
| Options & Volatility | 4 | Sinclair, Natenberg, Scholes |
| Contrarian | 3 | Burry, Einhorn, Icahn |
| Famous Systematic | 5 | Simons, Cohen, Shaw, Griffin, Muller |
| Event-Driven | 4 | Paulson, Ackman, Bass, Loeb |
| Modern Factor | 5 | O’Shaughnessy, Faber, Hoffstein |
| Fictional Wildcards | 9 | Petrov, Volkov, Reeves, Tanaka, and more |
All of them about to compete in a contest none of them asked for.
What I’m Watching
Fundamentalists vs. quants. Buffett thinks in intrinsic value and margin of safety. Simons thinks in statistical patterns and factor exposures. Can a value approach survive a walk-forward test designed for systematic strategies? Or will the quants dominate?
The fictional dark horses. Petrov, Volkov, Reeves - they’re not constrained by faithfulness to a real person’s published views. They can be more creative, more optimized for the format. Will the invented characters outperform the legends?
Convergence. If three personas who’ve never seen each other’s work independently design the same approach, that’s a powerful signal. The competition format is specifically designed to detect this.
Next: I hand all 80 the same prompt, say “design your best strategy,” and step back. Spoiler: six of them order the exact same drink.