AI Exposes the Myth of Human Fund Managers with Sixfold Returns

Stanford researchers thought something must be wrong with their AI-driven investment bot. In reality, it was just six times better than most professional fund managers. That’s not a metaphor, it’s a hard fact. The AI outperformed 93 percent of fund managers by an average of 600 percent.

For a year, researchers combed through the model, convinced there had to be a flaw. Professor Ed de Haan admitted they simply couldn’t believe the outcome. The bot had absorbed in a matter of days what human fund managers have struggled to master over three decades.

To train it, the AI was fed public market data from 1980 to 1990. It then synthesized 170 different economic and semantic indicators, from interest rates to natural-language analysis of corporate financial reports. After that, it went to work, testing its self-built algorithm on 3,300 US investment funds across three decades, from 1990 to 2020. The result was a quarterly rebalancing strategy that multiplied fund performance.

The AI ranked investments by expected return, swapped out underperformers for better alternatives, and when an asset proved truly worthless, it sold it off and redirected the capital to index funds. All of it perfectly rational, completely logical, and—perhaps most damningly, exactly what human fund managers should have been doing all along.

Just two years ago, the idea of an AI trader was laughed out of the room. Today, that room is quiet. Some fund managers are likely bracing for the axe, while others are busy coding up models of their own.