Hedge funds shift from AI infrastructure to monetisation plays, says Hazeltree

During July, global hedge funds adjusted their exposure within the artificial intelligence sector, focusing on companies demonstrating clearer links between heavy AI investment and revenue or earnings growth, according to Hazeltree's latest Data Insights Crowding Report. This shift indicates a refinement of risk management rather than a complete exit from AI, with the findings derived from anonymised securities-finance data from over 700 hedge funds across approximately 16,000 securities.

In the Magnificent Seven, Tesla and Amazon saw modest increases in long positions, while Apple, Meta, and Nvidia experienced notable declines in long holders and increased short positions. This realignment reflects growing skepticism about whether the substantial capital expenditure on AI infrastructure will yield sufficient commercial returns.

Tim Smith, managing director at Data Insights, noted that investors are placing greater emphasis on fundamentals such as earnings, revenue growth, and the ability of companies to commercialize AI. Semiconductor stocks saw stable hedge fund interest, with around 70% of companies in the PHLX Semiconductor Sector Index maintaining net long positioning in July.

However, individual stock performance varied significantly, with Intel shifting from short to long bias and Marvell moving from long to short bias. SpaceX's share price declined by 31% in July, coinciding with a 55% decrease in the ratio of long to short positions. The report also highlighted increased short interest in mid-cap companies such as Whirlpool, Core Scientific, and Repligen, while long participation grew in companies like Darling Ingredients, Monday.com, and EXL Service.

Overall, hedge funds appear to be reallocating capital towards companies with stronger potential for monetizing AI investments, rather than abandoning the sector entirely.