The Bot Battle for Wall Street: A $100K Experiment Reveals the Future of AI-Driven Investing.

Table of Contents

Has AI Just Made Hedge Fund Managers Obsolete?

In a head-to-head battle of algorithms, four top AIs were given $100,000 to manage. The results suggest that the future of finance will be fought not between humans, but between the AIs they create.

This competition represents more than just a novelty experiment—it’s a preview of finance’s future. As AI tools become more sophisticated and accessible, we’re approaching a world where:

Algorithmic Competition Intensifies: When every retail investor has access to AI-powered portfolio management, will markets become more efficient or more volatile? This competition suggests both outcomes are possible.

Strategy Convergence vs. Divergence: Interestingly, these AIs developed distinct strategies despite similar objectives. This suggests AI won’t homogenize investing but might actually increase strategy diversity.

The Human Element: The stark performance differences remind us that AI is only as good as its training, parameters, and interpretation of objectives. The “medium-high risk” mandate produced vastly different results, highlighting the importance of human oversight.

Lessons for the AI Age

Whether you’re a retail investor considering AI tools or a professional watching the industry evolve, this competition offers valuable lessons:

Question the Algorithm: ChatGPT’s underperformance despite holding consensus winners shows that popularity doesn’t guarantee performance. Always understand why an AI makes specific recommendations.
Diversification Evolves: Claude’s 32-stock portfolio might seem excessive by traditional standards, but in an age of zero-commission trading and AI-powered management, broader diversification might become the norm.
Volatility Is a Feature, Not a Bug: Grok’s wild swings ultimately produced market-matching returns. For some investors, the journey matters less than the destination.
Specialization Wins: Claude’s biotech expertise (or luck) drove its outperformance. As AI develops, models specialized in specific sectors or strategies might outperform generalists.

The Verdict: More Than Just Returns

As we analyze this three-week snapshot, it’s crucial to remember that investing success is measured in years, not weeks. However, the early results are illuminating:

Claude’s combination of broad diversification, calculated risk-taking, and apparent sector expertise has produced the best risk-adjusted returns. Its 5.52% gain while maintaining minimal drawdowns suggests a sophisticated approach that balances aggression with prudence.

Gemini’s solid second-place showing validates a more traditional tech-focused strategy, while Grok’s volatility demonstrates that high risk doesn’t always mean high reward. ChatGPT’s struggles serve as a humbling reminder that even advanced AI can underperform simple index investing.

Most importantly, this competition humanizes AI in an unexpected way. Each model has developed what we might call an “investment personality”—from Claude’s calculated aggression to Grok’s volatility embrace to ChatGPT’s unsuccessful attempt at balance. These aren’t just algorithms processing data; they’re complex systems interpreting goals, balancing risks, and making choices that mirror human investment behaviors.

As AI continues to reshape finance, this competition provides an early glimpse into a future where algorithms battle algorithms for investment supremacy. The question isn’t whether AI will transform investing—it’s how we’ll adapt to a world where our investment advisors, competitors, and collaborators might all be artificial intelligences with their own strategies, biases, and blind spots.

The competition continues, and with volatile markets ahead, leadership could change at any moment. Will Claude maintain its edge? Can ChatGPT engineer a comeback? Will a market correction reveal new strengths and weaknesses? As this fascinating experiment unfolds, one thing is certain: the age of AI investing has arrived, and it’s far more complex, nuanced, and human than we might have expected.

The Leaderboard: More Than Just Numbers

Current Rankings (as of July 8, 2025):

Claude (Opus 4): +5.52% ($105,515.33) – 32 holdings
Gemini (2.5 Pro): +4.20% ($104,198.63) – 12 holdings
S&P 500 Baseline: +3.36% ($103,356.93) – 1 holding
Grok (Grok 3): +3.25% ($103,254.24) – 10 holdings
ChatGPT (GPT 4.5): +1.86% ($101,861.21) – 10 holdings

What’s particularly striking is that all portfolios are profitable—but the range of returns tells a compelling story about different AI investment philosophies. In just three weeks, the performance gap between the leader and laggard has already stretched to 3.66 percentage points, which would compound to a massive difference over a full year.

The competition setup is elegantly simple: each AI starts with $100,000 and operates within a “medium-high risk” mandate. The S&P 500 serves as the benchmark, though curiously, the SPY position in the data shows only $620.55—suggesting either a data anomaly or an unconventional benchmark implementation.

The Secret Sauce: Claude’s 169% Moonshot

The standout performer driving Claude’s success is an unlikely hero: Prime Medicine (PRME), a gene therapy biotech that has skyrocketed 169.7% in Claude’s portfolio. While other AIs played it safe with mega-cap tech stocks, Claude’s algorithm apparently identified something special in this small-cap biotech that trades around $3 per share.

Prime Medicine is developing one-time curative genetic therapies using its proprietary Prime Editing technology, positioning itself at the forefront of the genomic medicine revolution. The company recently received FDA clearance for its first-ever Prime Editing product, PM359, for treating chronic granulomatous disease, with initial clinical data expected in 2025. Their pipeline includes treatments for Wilson’s Disease and alpha-1 antitrypsin deficiency (AATD), targeting some of medicine’s most challenging genetic conditions.

This pick showcases Claude’s apparent ability to identify high-potential biotechs before they capture mainstream attention. Wall Street analysts currently have an average price target of $9.38 for PRME—suggesting Claude’s algorithm may have identified value that human analysts are only beginning to recognize. The timing is particularly impressive given that biotech stocks have been out of favor, with many investors burned by the sector’s post-COVID crash.

Risk Management: The Hidden Differentiator

While returns grab headlines, the real story might be in how these AIs manage risk. The maximum drawdowns tell a revealing tale:

Claude: -1.37% (Range: -1.06% to +5.57%)
S&P 500: -1.26% (Range: -0.97% to +4.15%)
ChatGPT: -1.56% (Range: -1.85% to +2.77%)
Gemini: -1.99% (Range: -1.39% to +5.11%)
Grok: -3.15% (Range: -3.33% to +3.83%)

Claude’s minimal drawdown coupled with its highest returns suggests superior risk-adjusted performance. This is particularly impressive given its aggressive biotech bets—traditionally one of the market’s most volatile sectors. The data suggests Claude has mastered the delicate balance between risk and reward, maintaining steadier performance while still capturing significant upside.

A Tale of Different Strategies

Each AI has developed a distinct investment personality, revealing how different algorithms interpret the same “medium-high risk” mandate:

Claude’s Diversification Masterclass: With 32 holdings, Claude is running the most diversified portfolio—triple that of its competitors. This isn’t just spray-and-pray investing; it’s a sophisticated approach that combines:

High-conviction biotech bets (PRME)
Tech giants for stability (NVDA, AMZN)
Broad market exposure through ETFs (QQQ)
Calculated position sizing to maximize upside while controlling risk

This strategy appears to be paying off, allowing Claude to capture upside from multiple sectors while minimizing single-stock risk. The portfolio resembles a hybrid between a growth-focused mutual fund and a venture capital approach to biotech.

Gemini’s Concentrated Tech Bet: With 12 holdings, Gemini is taking a more focused approach, heavily weighted toward established tech leaders. Top holdings include NVIDIA, Microsoft (MSFT), and Broadcom (AVGO). This strategy banks on the continued dominance of AI infrastructure plays—a reasonable bet given the ongoing AI arms race among tech giants.

Grok’s Volatility Embrace: True to Elon Musk’s X platform roots, Grok seems comfortable with volatility. Despite suffering the worst drawdown at -3.33%, it recovered thanks to aggressive bets including a 22.8% gain in Moderna (MRNA). The portfolio includes high-beta plays like QQQ, XLY (consumer discretionary), and even the beleaguered solar stock Enphase (ENPH), which dropped 11.3%.

ChatGPT’s Balanced Disappointment: Despite holding market darlings like NVIDIA and Amazon, ChatGPT’s 10-stock portfolio has underperformed. The AI seems to have played it too safe, missing the bigger opportunities while still taking losses on stocks like Tesla (-8.7%). Its largest positions by value are SPY, NVDA, and IWM, suggesting a core-satellite approach that hasn’t quite delivered.

The NVIDIA Factor: Not All Positions Are Created Equal

Multiple AIs loaded up on NVIDIA (NVDA), which gained 11.4% during the period. NVIDIA shares hit a fresh record high of $154.31 on June 25, completing a remarkable turnaround from April lows around $94. The AI chip giant has been riding high on massive demand for its Blackwell AI supercomputers, with CEO Jensen Huang declaring the robotics market could be the “largest industry ever.”

However, simply holding NVIDIA wasn’t enough to guarantee outperformance. The key differentiator appears to be position sizing and what else the AIs paired with their NVIDIA holdings. Claude and Gemini, the top two performers, both hold NVIDIA but complemented it with different strategies—Claude with biotech moonshots, Gemini with a broader tech basket.

Leadership Musical Chairs: A Three-Week Drama

The competition has seen multiple lead changes, revealing how different strategies perform under varying market conditions:

June 16: Gemini takes early lead (0.60%)
June 18-19: S&P 500 briefly leads during market turbulence
June 20: Market-wide correction benefits the benchmark
June 24: Claude captures leadership (1.71%)
June 30: Brief Gemini resurgence (3.73%)
July 1-8: Claude solidifies dominance

This volatility highlights how AI models adapt differently to market conditions. The early June period saw significant market uncertainty as investors grappled with Trump’s trade policies and the emergence of competitive AI models from China. The fact that the passive S&P 500 briefly led during this turbulent period underscores an investing truism: sometimes doing nothing beats doing something.

Claude’s ability to capture and largely maintain leadership from June 24 onward suggests its algorithm may be better at identifying sustainable trends rather than chasing short-term momentum. The brief Gemini resurgence on June 30 coincided with a tech rally, but Claude’s more diversified approach proved more durable.

The Moderna Miracle: Grok’s Contrarian Masterpiece

Perhaps the most intriguing single stock story comes from Grok’s 22.8% gain in Moderna (MRNA). The mRNA vaccine pioneer has been in freefall since its COVID heights, trading around $30 compared to its 2021 peak above $400. Yet Grok’s algorithm apparently identified a trading opportunity that others missed.

Moderna announced positive results for its mRNA flu vaccine in late June, showing better efficacy than rival GSK’s product. The company is also advancing its pipeline with 10 potential new products over the next three years. However, with RFK Jr.’s vaccine skepticism creating political headwinds and the company burning through cash, this remains a contrarian bet. Grok’s timing suggests its algorithm may be particularly adept at identifying oversold bounce opportunities.

Notable Winners and Losers: Beyond the Headlines

Top Individual Performers:

PRME (Claude): +169.7% – Gene therapy moonshot
MRNA (Grok): +22.8% – COVID vaccine maker’s bounce
NVDA (Multiple portfolios): +11.4% – AI chip dominance continues
AMD (Various): Strong performance riding AI coattails
AVGO (Gemini): Benefiting from data center demand

Biggest Disappointments:

ENPH (Grok): -11.3% – Solar sector struggles continue
TSLA (ChatGPT): -8.7% – EV market concerns weigh
LLY (Gemini): -4.9% – Pharma giant stumbles despite Ozempic success
DRUG (Claude): -5.4% – Biotech volatility cuts both ways

The losers list reveals the risks of thematic investing. Enphase, once a solar darling, continues to struggle with inventory issues and declining residential solar demand. Tesla’s decline despite Elon Musk’s prominence suggests even strong brands aren’t immune to market forces.

Sector Allocation: The Hidden Story

Analyzing the portfolios reveals distinct sector preferences:

Technology Dominance: Unsurprisingly, all AIs are overweight technology, but to varying degrees. The prevalence of QQQ (Nasdaq-100 ETF) across portfolios shows consensus on tech leadership.

Healthcare Divergence: Claude’s aggressive biotech positioning (PRME, DRUG, XBI) contrasts sharply with others’ more conservative pharma picks. This paid off spectacularly with PRME but shows the high-risk, high-reward nature of biotech investing.

Energy and Utilities: Limited exposure across all portfolios, with only minor positions in XLE (energy sector) and NEE (NextEra Energy). The AIs seem to consensus that the energy transition creates more risk than opportunity in traditional energy stocks.

International Exposure: Minimal, with only small positions in VEA (developed markets) and SPEM (emerging markets). The home bias is strong across all AI models.

What This Means for Human Investors

This real-world experiment offers several crucial insights:

Diversification Still Matters: Claude’s 32-stock portfolio outperforming more concentrated approaches validates traditional portfolio theory—but with a twist. It’s not just about the number of holdings but the quality of diversification across sectors and risk profiles.
AI Isn’t Infallible: ChatGPT’s underperformance despite holding “obvious” winners like NVIDIA shows that even advanced AI can mistime or misallocate positions. Human judgment still has value.
Risk Management Separates Winners: The best raw returns (Claude) came with the second-best risk management. Sustainable outperformance requires balancing aggression with protection.
Sector Expertise Matters: Claude’s biotech success suggests specialized knowledge—or at least pattern recognition—in specific sectors can drive outsized returns.

The Road Ahead: What to Watch

As this competition continues, several factors could reshape the leaderboard:

Market Conditions: The relatively calm three-week period favored growth strategies. A market correction could reveal which AI truly has superior risk management.

Sector Rotation: If value outperforms growth or international markets surge, current strategies may need adjustment. Will the AIs adapt or stick to their patterns?

Individual Stock Catalysts: With earnings season approaching, company-specific news could dramatically impact concentrated positions. Prime Medicine’s clinical trial results, for instance, could make or break Claude’s lead.

Regulatory Changes: Proposed AI regulations, changes to drug pricing, or shifts in clean energy policy could impact multiple holdings across portfolios.