A number of investing platforms are jumping on AI agentic trading. My hot take: it is a bad idea for long-term investing.
The pitch is compelling: give an AI agent your brokerage account, let it read the news and earnings calls, and ask it to trade and find alpha.
But professional active managers already have dedicated investment teams, deep research budgets, and proprietary data. Even so, 89.93% of active U.S. large-cap funds underperformed the S&P 500 over the past 15 years.
This should make us skeptical that a retail agent, acting on generic prompts like “buy the dip” or “analyze this earnings call,” will reliably beat the market.
The more promising use of AI is helping an investor or researcher test a specific thesis across far more information than a human could process alone.
A few areas that seem promising:
1. OTC biotech and pharma research
Use AI to synthesize molecular targets, trial design, readout dates, competitive landscapes, and prior clinical evidence. The goal is to identify cases where the market may be underpricing the probability of success.
2. Small-cap and international research
Many companies have thin analyst coverage and fragmented disclosures across local filings, investor presentations, and non-English sources. Translation and structured extraction can make that information far easier to analyze.
3. Regulatory and policy exposure
AI can help map an FDA decision, tariff proposal, Medicare reimbursement change, defense contract, energy permit, or state-level rule to the companies and revenue streams that may be affected.
My concern is that many retail investors who are seeking alpha will treat AI agentic trading as simple prompts into a chatbot. In reality, using AI requires a well-defined thesis, deep domain knowledge, a thorough AI workflow, and real work to validate the output.