Newsletter commentary Dec 2025

Year-End Review and Portfolio Rebalancing

In December, the CSI 500 index rose about 6%, closing the year on a strong note. However, some sectors, such as real estate, food & beverages, and pharmaceuticals, still posted declines, while sectors like non-ferrous metals, commercial aerospace, the robotics supply chain, communications, and insurance stood out with strong performance.

Looking back over the year, the clear winners were the AI industry chain, non-ferrous metals, and biopharmaceuticals, along with pockets of new consumer trends. For China, external demand emerged as a relatively common contributor, whereas domestic demand remained generally weak. Globally, traditional growth drivers are relatively weak, making the AI industry chain a key focus. The AI industry chain is the core focus, and supply shortages emerging at specific nodes along the chain present the greatest investment opportunities. This applies to the energy, optical communication, memory storage, and PCB sectors. At present, the economic development model seems unable to simultaneously drive broad-based growth across multiple sectors—this appears to be a widespread challenge.

In December, the market began to recognize that China’s interest rates may have bottomed, which is a strong support factor for insurance. Insurance also has significant potential to act as a channel for converting savings into investment, benefiting the liability side. Non-ferrous metals show some signs of a short squeeze, with certain marginal-demand commodities showing strong price resilience, though this will gradually be tempered by demand constraints. For example, parts of industrial demand for silver may face replacement risk from new technologies, and ultimately, changes in supply and demand will determine prices.

Commercial aerospace is undoubtedly a golden track with immense potential and inherent market appeal. Yet compared with the intelligent electric vehicle (EV) industry, breaking through and gaining a foothold in this sector proves far more challenging. In the intelligent EV space, Chinese manufacturers have leveraged differentiated innovation to catch up with and even surpass Tesla, forging unique competitive advantages. In stark contrast, SpaceX, spearheaded by Elon Musk, has refined the full-chain technology and operations of Starlink to the extreme based on first principles, setting an almost insurmountable benchmark in nearly every key link. What’s more, it has seized the lion’s share of high-value clients worldwide. While commercial aerospace is a strategic, future-oriented arena that we must enter and play a pivotal role in, there is no denying that it faces far greater hurdles than the intelligent EV industry in terms of commercialization and market penetration.

In biopharmaceuticals, the market has moved past the “everyone gets into college” stage; the focus now is on the “Olympiad,” where only the top three matter. Many currently licensed assets may ultimately hold little value if they do not rank among the top three, and the market is beginning to price in this reality.

The robotics supply chain has proven much more difficult than expected, likely an order of magnitude more challenging than large language models. As Elon Musk noted in early 2025, progress on robotics is significantly below expectations, yet stock prices are already elevated. Breakthroughs in physical-world perception and other areas remain elusive.

The biggest challenge for the AI industry chain going forward is the shift in downstream demand from model training to inference across thousands of industries and households. Cost-effective engineering solutions to ensure stability and reduce procurement expenses along the supply chain will be essential. High gross margins could become a constraint on industry growth.

Traditional sectors, such as real estate and traditional consumption, still require time. Some areas may be in a “waiting for Godot” stage, as many consumption scenarios have changed drastically compared with several years ago and are unlikely to repeat.

As we enter 2026, top-down perspectives suggest the presence of investment opportunities, while bottom-up stock selection remains challenging. From a macro perspective, considering asset allocation and global capital flows, Chinese assets still appear inexpensive. From the perspective of sectors that have already risen over a year and their historical valuations, they may seem less attractive. We tend to believe that the latter will ultimately conform to the former. Historical valuation references lose relevance as the environment changes, and what we are seeing now is not a continuation of past stories, but the emergence of a new narrative.