Author: Li Jia
The AI rally has entered a phase of high volatility. Can tech stocks still be bought in the second half of the year? Goldman Sachs' answer remains: Stay bullish, but shift from 'buying the sector' to 'picking the company.'
In its latest report, Goldman Sachs points out that there are currently no signs that the AI-driven tech cycle has peaked, as signals of supply exceeding demand and slowing technological progress have not yet appeared. Analysts at Goldman Sachs believe this cycle could become one of the largest and longest-lasting technology upturns in history. Following profit-taking in related stocks after July, the report characterizes this as a healthy correction after a rapid price increase, not a trend reversal.
Regarding stock selection strategy, the report proposes three core themes: First, maintain a bullish view on AI server and data center-related hardware stocks. Second, in subsectors where supply and demand are already tight, place greater emphasis on fine-grained assessment of individual stock risk-reward. Third, when market risk appetite declines, focus on software and IT services stocks that are leveraging the AI disruption wave to explore new business opportunities as a defensive allocation.
AI Cycle Has Not Peaked; The Correction Is a Healthy Pullback
Goldman Sachs maintains an overall bullish view on the Asian AI supply chain.
The report notes that determining whether a tech cycle is nearing its end primarily involves observing two signals: first, whether semiconductors and electronic components shift from undersupply to oversupply; second, whether technological innovation slows, and industry competition reverts to being price-driven rather than performance-driven. Currently, neither of these signals is present.
Goldman Sachs believes that AI infrastructure investment is still in an expansion phase, and future new applications such as physical AI and edge AI will take over from AI servers and data center construction, further extending this tech cycle. Therefore, recent profit-taking in related stocks should be seen as a healthy adjustment after rapid gains, not a fundamental reversal.
At the same time, tight supply and demand are gradually spreading from hot areas like memory and optical communication to more semiconductor sub-sectors, indicating the scope of industry prosperity is still expanding.
Focus for the Second Half: Shift from Sector Selection to Stock Picking
Following significant gains in many AI beneficiary sectors, Goldman Sachs believes the investment logic for the second half will gradually shift from "getting the industry right" to "selecting the right company."
The report suggests that companies worth attention typically share several common characteristics: they can directly benefit from product price increases; possess strong capacity expansion capabilities to capture profit opportunities arising from tight supply and demand; have AI business growth potential not yet fully reflected in valuations by the market; or possess unique catalysts not fully priced in by the market.
In other words, after an overall valuation uplift, future excess returns will increasingly come from a company's own competitiveness, rather than industry beta.
Defensive Strategy Shifts to AI Applications, Not Traditional Defensive Sectors
Beyond continuing to allocate to AI hardware, Goldman Sachs also proposes a new defensive approach.
The report argues that when market risk appetite declines, instead of avoiding the tech sector, investors should focus on software, IT services, and internet companies that are leveraging AI to create new business opportunities. Goldman Sachs points out that generative AI is spurring new enterprise service demand in areas like AI consulting, data infrastructure construction, and cybersecurity. Some software and IT services companies may actually see improved profitability from using AI tools to enhance development efficiency and reduce costs.
At the same time, previous market concerns about AI diminishing content value are easing. Goldman Sachs believes that AI is more likely to become a new tool for improving commercialization efficiency and operational efficiency, rather than simply replacing existing businesses. Therefore, the growth logic for some internet and digital content companies is improving.
Overall, Goldman Sachs believes that second-half tech investment in Asia should still adhere to the AI theme, but the allocation approach needs to be more balanced: on the offensive side, continue to focus on AI infrastructure and hardware industry chains with sustained improving fundamentals; on the defensive side, pay attention to software and IT services companies that can leverage AI to create new demand and improve efficiency, balancing growth and defensiveness in a more volatile market environment.
