Summary of this article
Wood sees India benefiting if investors rotate out of AI momentum stocks
He prefers AI infrastructure firms over hyperscalers amid monetisation concerns
Rising AI spending and complex financing warrant greater investor caution
Christopher Wood, the global head of equity strategy at Jefferies and author of the brokerage's widely followed GREED & Fear report, has said that India could emerge as one of the markets that benefit if investors start shifting money away from expensive artificial intelligence (AI)-linked stocks. His view comes at a time when global investors are increasingly debating whether the AI rally has run too far and whether it is time to rotate into relatively cheaper markets and sectors that have lagged the AI boom.
Wood has said that such a shift, if it gathers pace, could favour Asian markets outside the AI winners.
“MSCI China is one obvious area of mean reversion rotating out of AI plays in Asia, as is India…,” he wrote in his report, discussing a broader move away from momentum-driven AI trades rather than presenting a standalone bullish call on India.
According to the report, conversations around “AI fatigue” have become more prominent as the new quarter begins, with investors looking beyond the biggest beneficiaries of the AI trade. Wood said that market participants are considering whether gains in AI-linked companies have become stretched and whether capital could start flowing into cheaper “value” stocks that were largely left behind during the rally. He cited Tencent as an example in Asia and said Jefferies’ quantitative research has also examined this theme.
According to Wood, some correction in AI-related stocks is both expected and healthy after the sharp run-up seen in recent months. He gave the example of South Korea, where AI-linked shares rallied exorbitantly before coming down sharply, saying the market had experienced “hyperbolic moves”.
While acknowledging that momentum stocks can continue to correct, he said he did not believe that the broader AI investment cycle was over.
He said his preference, however, remained firmly with companies supplying AI infrastructure rather than those spending aggressively to build AI capabilities. Wood said the biggest beneficiaries continued to be the “picks and shovels” businesses that supplied memory chips and other hardware needed for AI computing. He said these companies profited from AI spending regardless of which platform ultimately succeeded.
He was, however, sceptical about the largest cloud companies' ability to generate adequate returns from their enormous AI investments.
Referring to Alphabet, Amazon, Meta and Microsoft, Wood wrote in his report that the demand for computing power may continue to rise even as AI usage became cheaper, adding that he has “no idea which of the hyperscalers, if any, are going to be successfully monetising their AI capex. There is a real possibility that none of them could be.”
That uncertainty, he said, makes the upcoming earnings season particularly important.
“Quarterly results from the major US hyperscalers, starting later this month, will provide the next major test of whether investors remain convinced that record AI spending will eventually translate into meaningful profits,” he said, adding that investors previously welcomed higher capital expenditure guidance even as the largest cloud companies projected spending equivalent to nearly all of their operating cash flow.
The scale of AI investment itself is another reason for caution, headed. Wood said that the four largest US hyperscalers are expected to spend around US$700 billion on capital expenditure this year. Including Oracle, Anthropic, OpenAI and emerging cloud providers, AI-related spending could exceed US$1 trillion in 2027. He added that this would amount to roughly 3 per cent of the gross domestic product (GDP) of the US and about one-third of total pre-tax profits earned by all US non-financial companies.
He also flagged rising financial risks accompanying this investment boom. He said the biggest AI companies are increasingly funding expansion through debt, with Alphabet, Amazon and Meta collectively issuing tens of billions of dollars of bonds this year. Beyond borrowings, Wood also highlighted Moody's estimate that the leading hyperscalers have accumulated $662 billion of future data-centre lease commitments that have yet to appear on their balance sheets. He said these hidden obligations deserved closer scrutiny.
Wood also highlighted a warning from the Bank for International Settlements, which says AI companies are becoming increasingly financially linked to one another. He said cross-holdings, long-term supply agreements and data-centre financing structures have created a web of “circular financing” that could make risks harder for investors to assess.
He also cited recent moves by Elon Musk’s xAI and the Meta to monetise computing capacity as signs that even major AI players were looking for ways to improve returns on their huge infrastructure investments.
Against that backdrop, Wood is of the opinion that investors could start looking at markets that did not benefit as much from the AI rally. He said China and India could attract some of this money if investors move away from expensive AI stocks.
At the same time, he has cautioned that investors should keep an eye on China’s economic challenges, including weak household borrowing, slower consumer spending and rising bad loans in the retail segment, before making investment decisions.
