AI Stock Picking: Can Machines Outperform Human Investors?

Artificial intelligence (AI) is emerging as a potential alternative to traditional stock picking, with preliminary studies suggesting it may outperform human investors. Recent analyses, including Dalbar’s 2025 Quantitative Analysis of Investor Behavior (QAIB), indicate that many individuals struggle to achieve consistent returns through stock selection. The findings raise questions about whether AI could offer a more reliable approach to investing.

Dalbar has documented investor behavior since 1994, revealing that individual decisions around buying and selling stocks often lead to disappointing outcomes. According to their latest report, the average equity fund investor earned 8.5 percent less than the S&P 500 in 2024. This gap is largely attributed to psychological biases, such as loss aversion and herd behavior, which frequently undermine investor returns.

In a stark comparison, research conducted by Hendrik Bessembinder at Arizona State University examined nearly a century of stock market data. His findings show that over half of the 29,078 publicly listed common stocks had negative cumulative returns from 1926 to 2023. While a few stocks delivered extraordinary returns—like Altria Group, which saw a staggering 265 million percent cumulative return—most stocks fared poorly, with a median return of -7.41 percent.

Despite the grim statistics for human investors, AI-driven stock selection has shown promise in certain experiments. For instance, a project by Finder.com earlier this year tested ChatGPT‘s stock-picking abilities. The AI portfolio gained 4.9 percent during a period when the average human-managed fund lost 0.8 percent. Additionally, a high school student’s experiment with ChatGPT yielded a return of 23 percent while the Russell 2000 index increased by just 3.9 percent.

AI’s Role in Stock Selection

The allure of stock picking stems from its excitement and the belief that individuals can outperform the market. Yet, research confirms that this endeavor is significantly more complicated than many realize. AI strategies often follow established market patterns, akin to momentum investing. For example, the Comet Portfolio, created by AI, featured major tech stocks like Amazon, Nvidia, and Microsoft. While this approach reflects market trends, it raises questions about whether AI is truly generating innovative insights.

Dalbar’s QAIB report emphasizes how emotional and psychological factors hinder investor success. For instance, the report notes that many investors tend to pull out of the market during downturns, further damaging their potential returns. The study identifies behavioral pitfalls, including loss aversion and excessive optimism, which consistently lead to subpar performance.

Although AI offers a potential solution to some of these emotional biases, it is not without limitations. AI is only as effective as the instructions it receives. If those instructions reflect human biases, the AI may not outperform traditional market strategies over the long term. Moreover, even the best AI traders could face termination by human investors during market corrections, which are inevitable.

The Future of AI in Investing

While AI has demonstrated impressive capabilities in stock trading, its successes so far are largely tied to existing market wisdom rather than groundbreaking strategies. Recent studies indicate that timing-based investment strategies utilizing large language models often fall short of passive benchmarks. These models may act too conservatively in bull markets and excessively aggressively in bear markets, leading to significant losses.

Currently, AI has not proven its ability to consistently outperform a strategy of investing in broad market indices. Professional stock pickers and AI alike have not established superiority over simple index strategies. Therefore, investors are encouraged to approach claims of individual stock selection with caution, particularly when high fees accompany such services.

As AI technology continues to evolve, it may eventually provide investors with tools to avoid losses and navigate market fluctuations more effectively. Until then, the wisdom of investing in diversified index funds remains a prudent strategy. Investors should maintain a healthy skepticism toward AI investment advice until it demonstrates consistent, long-term market-beating capabilities.

For those interested in the data behind these insights, Dalbar’s 2025 QAIB report offers valuable statistics. The average asset allocation was 74.45 percent equities and 25.25 percent bonds by the end of 2024, indicating a slight shift towards risk assets. Additionally, investors guessed market direction correctly only 25 percent of the time in 2024, tying the lowest success rate recorded since 2019. Historically, 73 percent of negative years for the S&P 500 have been followed by positive returns the subsequent year.

In conclusion, while AI presents exciting possibilities for stock picking, its current performance suggests that it may not yet be a replacement for traditional investment strategies. As technology advances, the investment landscape may shift, but until then, a focus on index funds remains a sound approach for most investors.