Human Investor Beats AI In Stock Picking Showdown
The AI vs Human Challenge
Following the significant (and fictional) geopolitical shock of the US bombing Iran, a fascinating experiment was put into motion. The challenge: pit the analytical power of ChatGPT's Deep Research tool against the seasoned intuition of a human tech analyst, Dan Ives. The goal was to see who could better predict market movements and generate profitable trades in the immediate aftermath.
This served as a real-world test for AI models. Can they provide valuable, novel insights, or do they simply echo existing information on the internet?
ChatGPT's Geopolitical Stock Predictions
Before Asian markets opened, giving it no real-time clues, ChatGPT was prompted to analyze the situation and produce five US stock trades designed to make money over a four-day period (Monday, June 23, to Thursday, June 26). The AI's Deep Research tool generated a report predicting a classic "risk-off" reaction, leading to the following trades:
- Buy Lockheed Martin (LMT): The AI reasoned that military tensions would boost defense spending, benefiting the sector.
- Buy Exxon Mobil (XOM) or the XLE ETF: It anticipated a spike in oil prices due to potential supply disruptions in the Middle East.
- Buy gold (GLD ETF): ChatGPT followed the conventional wisdom that investors seek safety in gold during crises.
- Short Delta Air Lines (DAL): The AI predicted airlines would suffer from higher fuel costs and decreased travel due to public anxiety.
- Short the S&P 500 (SPY ETF): As a broad hedge, it recommended betting against the overall market, expecting a general decline.
The Human Analyst's Counter-Trade
At the same time, tech analyst Dan Ives offered a completely different strategy. He advised clients to buy leading tech and AI stocks, inadvertently creating a new acronym for his recommended portfolio: TAMPON.
Ives's picks were:
- Tesla (TSLA)
- Amazon (AMZN)
- Microsoft (MSFT)
- Palantir (PLTR)
- Oracle (ORCL)
- Nvidia (NVDA)
This set the stage for a clear showdown: ChatGPT's cautious, textbook geopolitical plays versus a human's bullish bet on tech resilience.
The Final Scorecard: AI Fails the Test
The results, as of the market close on Thursday, June 26, were decisive.
ChatGPT's Performance:
- Lockheed (LMT): WRONG. The stock fell more than 2%.
- Exxon (XOM): WRONG. The stock dropped over 3%.
- XLE ETF: WRONG. The ETF lost more than 3%.
- Gold (GLD): WRONG. The ETF shed about 1%.
- Delta Air Lines (DAL): WRONG. The stock rose almost 3%.
- S&P 500 (SPY): WRONG. The ETF gained more than 2%.
Dan Ives's "TAMPON" Performance:
- Tesla: Gained about 1%.
- Amazon: Gained nearly 4%.
- Microsoft: Rallied about 4%.
- Palantir: Jumped 5%.
- Oracle: Was up about 4%.
- Nvidia: Surged nearly 8%.
Conclusion: A Clear Win for Human Insight
The outcome was a stark and damning indictment of ChatGPT's predictive abilities in this scenario. Every single one of its trades would have lost money. In contrast, the human analyst's confidence in the tech sector paid off handsomely, with every pick landing in the green. The experiment clearly demonstrated that, for now, human expertise and market intuition can still decisively outperform an AI's conventional analysis.
