Chen Sijing of Guotai Haitong Asset Management: The tech sector is undergoing a "20/80 shift," with excess opportunities moving beyond computing power

Article body

English text is being prepared.

Chen Sijing, deputy general manager of the Equity Investment Department of Guotai Haitong Asset Management, said in an interview with the reporter that the technology sector is undergoing a round of "80/20 rotation." "Compared with the computing power sector, the risk-return ratio of some other sectors within technology is improving. The excess returns of technology investment are hidden in the 80% outside computing power."

"80/20 rotation": crowded computing power and the other 80%

The latest disclosed public fund third-quarter reports show that "reducing computing power" was the mainstream adjustment direction in the third quarter: optical module leader Zhongji Innolight was reduced by more than 12 million shares in the third quarter, with a reduced market value of about 70.3 billion yuan; Eoptolink was reduced by more than 40 billion yuan in market value during the same period.

At present, the valuation of the CSI TMT Index has returned to near its historical average, and the overall crowding of the sector has declined accordingly. According to analysis by the Huachuang Securities strategy team, the share of TMT sector holdings by market value fell from a historical high of more than 61% at the end of the second quarter to 53.3%, a single-quarter reduction of about 8.4 percentage points, with electronics and communications becoming the main targets of reduction.

Even after this decline, the overweight ratio of the technology sector remains at a historical high in the 98th percentile since 2013. The allocation inertia of "equating technology with computing power" is loosening, and funds leaving the most crowded position are looking for the next landing spot within the industrial chain.

Chen Sijing believes that computing power accounts for only 10%–20% of the number of stocks in the entire technology sector, and another 80% of sub-sectors have not risen in the past two years, or have even fallen, which may be exactly where the next round of excess returns is breeding.

Tick-Tock strategy: writing discipline into trading

The Tick-Tock strategy is the most recognizable part of Chen Sijing's strategy. It draws on the "Tick-Tock" rhythm of chip giant Intel—alternating upgrades of process technology and microarchitecture. Mapped to investment, he divides the cycle into two stages: "generating returns" and "controlling drawdowns": establishing a pricing model for each stock so that trading is more grounded.

Chen Sijing introduced that the essence of this system is trading discipline: giving up the last segment of returns during a bubble period in exchange for drawdown control in a volatile market. The other side of discipline is an instinctive vigilance against "perfect logic." Chen Sijing explained that investors think linearly: the more stock prices rise, the more the logic is reinforced, eventually becoming a belief; once the market concludes that the logic of a certain field is perfect, that is precisely when the risk is greatest.

Beyond stock selection, Chen Sijing also has clear standards for corporate governance: only management that truly treats secondary-market shareholders as partners deserves long-term trust. "With management whose values are problematic, even if you can make money at a stage, the money will be taken away at some point in the future."

Bottom-up: understand the company first, then talk about the industry

In technology investment, many people focus their efforts on "picking the right industry"—betting on tracks and doing rotation. Chen Sijing's path is somewhat different: he is more willing to place the foothold of research on specific companies one by one, rather than first judging which industry will win. In his view, the "meso trend" at the industry level is more like a secondary confirmation of changes in individual stocks, and excess returns mainly come from long-term tracking of specific companies.

Reviewing recent years, this method has worked in different market stages. In 2024, when a certain IoT chip leader was "ignored," he took a heavy position and earned gains from the rise. After DeepSeek was released in 2025, he again reduced positions at high prices when domestic computing power chain stocks surged, and instead laid out early in relatively unpopular areas such as on-device AI and complex software.

Chen Sijing mentioned that these operations are not predictions of industry rotation, but a natural extension of bottom-up stock selection. In his view, continuous tracking of individual stocks and industries is constantly supplementing "input" for the system; the longer the time, the more reliable the judgment. He admitted: "After seeing many cycles, one's view of stocks becomes more objective, and it is less easy to get carried away." He summarized the core of mining excess returns in the technology industry as following the general trend of the industry, going against the small trend of the cycle, and going against the small trend of market sentiment.

Rebalancing after the ebb of computing power

Compared with computing power, Chen Sijing is more optimistic about opportunities in the other "80%." First, on-device AI: many companies already have good cost-effectiveness in growth and valuation even without considering the AI logic. Second, AI software: for specialized software deeply embedded in enterprise processes and with high reliability requirements, substitution remains very difficult, and future opportunities are worth watching.

On the broad direction, he is optimistic about the long-term value of semiconductors. "Globally, the major core technology companies may emerge from semiconductors, the internet, and software companies with very high barriers." He also cautioned that there are huge differences within semiconductors and that careful screening is needed.

For the fourth quarter, Chen Sijing's judgment is structural opportunities under a volatile market: earnings verification will become a key variable. "What truly widens the long-term return gap may not be how much was earned during rising markets, but losing less money in volatile markets." He hopes to bring investors a relatively better holding experience and to share in the growth of China's technology industry over a three-to-five-year horizon. It is understood that Guotai Haitong Growth Preferred, for which Chen Sijing is proposed as fund manager, is currently being issued; the fund adheres to a bottom-up stock selection method and selects subdivided opportunities in the technology industry.

(Source: Shanghai Securities News·China Securities Network) Middle of the body Embedded advertisement