Greed and Fear report: Christopher Wood sees 'AI fatigue' setting in. Why Jefferies is turning bullish on India

Greed and Fear Report: The spectacular rally in artificial intelligence (AI)-linked stocks may be losing momentum, according to Christopher Wood, Global Head of Equity Strategy at Jefferies. In his latest GREED & Fear newsletter, Wood said investors are increasingly talking about "AI fatigue" as the market begins rotating away from expensive momentum stocks toward cheaper value opportunities, particularly in Asia.

Against this backdrop, Jefferies has increased its preference for markets such as India and China, which have largely missed the AI-driven rally and now offer relatively attractive valuations. While Wood remains constructive on the long-term AI investment cycle, he believes investors should increasingly focus on the beneficiaries supplying the infrastructure behind AI rather than the technology giants spending billions of dollars building it.

"The new quarter has opened with much talk of 'AI fatigue' as investors look out for a peaking out of momentum and rotation into cheaper 'value' names which have not been part of the AI trade," Wood wrote in the latest edition of GREED & Fear.

According to Wood, the recent correction in South Korea's AI-heavy stocks should not be viewed negatively. He described the decline as "both natural and healthy" after what he called "hyperbolic moves." The Kospi has fallen 22% from its June 19 peak, while leveraged ETFs tracking SK Hynix and Samsung Electronics have dropped around 30% from their asset highs.

The rally in AI-related semiconductor companies has also been extraordinary. Since the beginning of 2023, a market-cap weighted basket comprising Micron, SK Hynix and Samsung Electronics has surged around 760%, compared with a 180% gain in a basket of Alphabet, Amazon, Meta and Microsoft.

Despite expecting a rotation within AI beneficiaries, Wood remains bullish on memory chip manufacturers.

"Long-term GREED & Fear would still rather own the DRAM makers. The demand for compute can keep growing even as the cost of tokens collapse," he said, adding that he has “no idea which of the hyperscalers, if any, are going to be successfully monetising their AI capex.”

Wood believes the AI investment cycle remains intact, but questions are beginning to emerge around whether the companies making massive investments will ultimately generate adequate returns.

Jefferies estimates that the four major US hyperscalers will spend around $700 billion on capital expenditure this year, increasing to more than $800 billion next year. Including companies such as Oracle, Anthropic, OpenAI and neo-cloud providers, total AI-related capital expenditure is expected to exceed $1 trillion in 2027.

According to the brokerage, that figure would equal roughly 3% of US GDP, around 22% of US non-residential fixed investment and nearly 33% of the total pre-tax profits of all US non-financial companies, highlighting the unprecedented scale of the AI investment cycle.

Wood also pointed to growing financial pressures behind this spending. The four hyperscalers have increased projected capital expenditure to 92% of projected operating cash flow, issued $169 billion worth of bonds so far this year and accumulated $662 billion in future data-centre lease commitments that remain off the balance sheet. Their total undiscounted lease obligations now stand at nearly $969 billion.

Against this backdrop, Jefferies has shifted its Asia Pacific ex-Japan allocation in favour of India. The brokerage currently assigns India a recommended weight of 12%, compared with 10.9% in the MSCI AC Asia Pacific ex-Japan Index, implying an overweight position of 1.1 percentage points.

Although memory stocks have corrected, Wood continues to remain Underweight Taiwan and Neutral on South Korea after reducing Korea's allocation from 24.6% to 20.8% since late June. The firm has also retained exposure to smaller ASEAN markets primarily to maintain a presence in the region.

China is another market Jefferies expects to benefit from any susta...