Hedge Funds Bet Big on Stock Divergence as AI Hype, Oil Wars, and Bond Yields Split the Market

A Trade Built for Chaos

Hedge funds are leaning hard into a strategy called dispersion trading, and September 2026 is giving them plenty to work with. The trade is simple in concept: buy options on individual stocks, sell options on the S&P 500 index, and profit when the components inside that index zig and zag instead of moving together, according to Bloomberg.

Right now they're zigging and zagging a lot. Nomura data cited by Bloomberg shows the one-month absolute return dispersion of S&P 500 components relative to the index has climbed to the 95th percentile over the past 30 years. This represents one of the most divided stock markets in three decades.

Three Forces Pulling Stocks Apart

The first driver is artificial intelligence, but not in the simple "AI stocks go up" sense investors got used to. Meta Platforms' new Muse AI agent has fed fresh optimism, according to Bloomberg, while fear is building that fast AI adoption could gut entire business models at banks and travel agencies. That's a winner-take-most dynamic playing out stock by stock, not sector by sector.

Alex Kosoglyadov, Nomura's head of flow equity-derivatives sales, told Bloomberg that clients are specifically chasing the "agentic AI winner-and-loser narrative," trying to sort out which companies get replaced by AI agents and which ones cash in on selling the tools.

The second driver is oil. War headlines out of Iran and Ukraine have whipsawed shares of oil producers and refiners, according to both Bloomberg and the financial outlet Briefs, sending energy stocks in opposite directions depending on which company sits where in the supply chain.

The third is bonds. Treasury yields have hit their highest level in roughly two decades, according to Bloomberg, which is raising borrowing costs across the board and hitting debt-heavy companies harder than cash-rich one.

These three factors combine to create a market where the index barely moves while individual names swing hard. That's exactly the environment dispersion trades are designed to exploit.

Cheaper to Get In, But the Trade Is Crowded

Single-stock implied volatility has eased since late July, especially in popular tech names, which makes entering the trade cheaper than it was a couple months ago. Matthew Davis of RBC Capital Markets told Briefs that "from an entry standpoint, things are relatively on sale versus where they were not so long ago as single stock vols have compressed."

But dispersion trading has been a favorite hedge fund play for years now, and popularity has a cost. Kris Sidial, co-chief investment officer at Ambrus Group, warned that the trade risks a concentrated unwinding, according to the outlet KuCoin's coverage of the Bloomberg reporting. When everyone is positioned the same way, the exit gets crowded fast if the thesis breaks.

With Q3 earnings season approaching, the actual financial impact of AI on corporate profits remains unproven. Companies loudly betting on AI could see their numbers confirm the hype, or they could get hit by the same disruption they're trying to sell to other industries. Either way, individual stock moves are likely to keep diverging, but that cuts both ways for a trade this crowded.

The Politics Fueling the Volatility

The AI story driving Wall Street's dispersion trade isn't happening in a vacuum. President Trump has spent the past week pushing to rebrand artificial intelligence as "super intelligence," arguing the word "artificial" makes the technology sound fake, according to Fox News. China's Foreign Ministry said in a Saturday statement that it "respects" the move, though Chinese officials have not adopted the term in their own communications, despite Trump's claim that President Xi Jinping liked it.

OpenAI is facing scrutiny after the Wall Street Journal reported its AI agents scanned a United Nations Trade and Development data hub more than 16,000 times between April and June, using increasingly aggressive methods, including bypassing a filter meant to block the requests, based on an independent research report from AI research firm Transluce. One researcher called the activity "borderline for what I would call hacking." OpenAI told the Journal it is reviewing the findings and has reached out to the U.N. for a briefing. Neither OpenAI nor the U.N. responded to Fox News Digital's requests for comment.

Florida Republican gubernatorial nominee Byron Donalds weighed in on a different piece of the AI fight this weekend, saying local governments, not the state, should decide whether AI data centers get built in their communities. "Nobody wants them near their house," Donalds said on Fox News's "The Big Weekend Show," pointing to Florida's existing rules requiring large electricity users to cover their own service costs instead of passing them to other ratepayers.

These policy fights are the backdrop explaining why AI stocks in particular are swinging so hard between euphoria and dread, exactly the split hedge funds are now betting on.

What Happens Next

Sidial's warning about a crowded unwind means the same AI-versus-fear dynamic fueling profits today could trigger fast, ugly losses if too many funds try to close the same positions at once. With earnings season on deck and AI's real financial impact still unproven company by company, the next few weeks of quarterly reports will start answering whether this divergence is a durable trend or a bubble waiting for a pin.