Everyday investors are handing stock portfolios to AI agents at brokerages including Robinhood and Moomoo, a trend the latter projects could account for 20 percent of its trading volume by year-end.
Quay lại
AI agents run retail portfolios as brokerages target 20% automated volume
Brokerages including Robinhood, Webull and Moomoo now let retail investors hand their portfolios to AI agents that buy and sell assets autonomously, with Moomoo projecting 20 percent of its trading volume could be agent-executed by year-end.
"These people are becoming mini hedge funds," Neil McDonald, U.S. chief executive at Moomoo, said. The platform made agentic trading available in April.
Colin Edsman, a hairstylist and stay-at-home dad, runs three Claude agents — Alex, Sarah and Elena — from his kitchen table laptop at Robinhood, each assigned distinct tasks from scanning for promising stocks to generating weekly performance reports. Dean Ahrens, a 19-year-old content creator, grew his Public account from $3,000 to $8,000 in a few months using a Codex agent that scores options flow data, including one Micron Technology trade that returned more than 500 percent.
The shift extends the set-it-and-forget-it approach from index funds to advanced day-trading strategies, but researchers warn that AI models trained on similar public data could crowd into identical positions, echoing the 2007 quant meltdown when funds rushed to sell simultaneously.
From prompts to positions
Agents can be instructed in plain language to buy energy stocks as oil prices rise, sell an options contract once it generates a target return, or monitor the president's social-media posts for market signals. Angel Gutierrez, a full-time options trader, built an agent that scores contracts to determine when to sell, describing it as "a software version of me with no emotions."
The technology closes a gap that has separated professional quants from retail traders for decades. Institutional investors have long used algorithmic tools to sift data and execute trades in milliseconds; AI assistants now let everyday investors build similar automated strategies from written prompts. Ahrens said his Codex agent sifts through options flow data to identify large institutional trades, scores them on multiple factors, and piles in when it finds a "high-confidence trade."
The retail investing boom of the past decade — from commission-free trading at Robinhood to the meme-stock era — has already expanded the role of everyday investors in U.S. markets. Just over a decade ago, Americans were starting to buy stocks on their smartphones. Now they can run what McDonald calls "mini hedge funds" from their laptops, with agents handling everything from options trading to tax-loss harvesting on autopilot.
The herding risk
A working paper distributed by the National Bureau of Economic Research found that AI models asked to build a general investment strategy recommended concentrated portfolios, stocks with high valuations and companies with the most frequent media coverage. "AI takes risk, recommends a narrow set of assets, focuses on specific industries, and does not appear to exhibit better performance" than passive benchmarks, researchers wrote.
Irene Aldridge, an engineer and former quant trader who researches AI and financial markets, said widespread use of agents scouring the same publicly available data could create volatility as they crowd into similar positions. The 2007 quant meltdown, when a number of quant funds rushed to sell simultaneously and triggered deep losses, offers a precedent.
Brokerage executives said their customers are implementing the agents thoughtfully, and their platforms have guardrails to manage risk. At Robinhood, AI-managed portfolios are kept in a dedicated separate account and send notifications for every trade. Edsman, Ahrens and Gutierrez said they had not experienced major losses or instances of their agent going rogue.
For Edsman, the promise of the technology is worth experimenting with something untested. He hopes agentic trading becomes easy enough that his 57-year-old mother can play the market in retirement. "This could change the way people make money," he said.
The stakes extend beyond individual returns. If Moomoo's 20 percent projection holds and competitors follow, agent-executed retail volume could reach a scale where herding behavior moves prices — a structural shift in how retail money participates in U.S. equity markets.
This article is for informational purposes only and does not constitute investment advice.
