Goldman Sachs says Asia is having a very good AI year. The bank points to Japan, India and China as the places where the momentum is strongest.
The driver is money spent on AI infrastructure: chips, power and data centers. Goldman thinks that spending wave is far from over, and it has raised its forecasts accordingly.
Kevin Sneader, Goldman’s president for Asia-Pacific excluding Japan, made the case during a Bloomberg TV interview on October 14, 2026. He singled out North Asia as a particular winner from AI-related capital expenditure.
The numbers behind the optimism
Goldman now projects a 26% return in US dollar terms for the MSCI AC Asia Pacific ex-Japan index over the next 12 months. The bank’s target for the index is 1,120.
The earnings path the bank sketches is steep. Goldman expects regional earnings growth of 75% in 2026, followed by 25% in 2027 and 14% in 2028.
Put differently, the firm is anticipating earnings growth averaging 34% a year through 2028. It also lifted its 10-year annual earnings growth forecast for the region to 10%.
Much of that profit is expected to come from one sector. Goldman forecasts that IT will account for 47% of Asian regional earnings in 2026.
Within that slice, memory chips are expected to make up about 70%. So a large share of the region’s projected profits ultimately rides on a single product category.
Goldman expects hyperscaler AI capital expenditure to top $6 trillion between 2026 and 2030. At the same time, the bank flags a record undersupply of memory that could persist through 2030.
Goldman maintains overweight recommendations on Korea, Taiwan, Japan and China A-shares, citing their significant technology exposure.
India’s quiet AI trade
Goldman’s research identified 42 AI-enabling companies listed on Indian exchanges, concentrated in power, data centers and semiconductors.
Together, those companies carry a combined market capitalization of around $670 billion. They gained roughly 60% in 2026.
Over the same period, the broader Nifty fell 12%.
China and a split AI world
Goldman also sees an emerging bifurcation in global AI investment ecosystems. The bank highlighted China’s progress within that split, noting those advances come despite China lagging in leading-edge semiconductor technology.
Goldman’s overweight on China A-shares fits that read.
What this means for investors
If IT generates 47% of regional earnings and memory accounts for about 70% of that, Asia ex-Japan equities become heavily leveraged to one segment of the chip cycle.
The front-loaded earnings profile also matters. Growth of 75% in 2026 decelerating to 14% by 2028 means the steepest gains are expected early.
The gap between the 42 AI-enabling Indian names and the falling Nifty shows that broad index exposure can miss a powerful theme entirely.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.