Recently, there has been good news every day in the chip and domestic computing power sectors, with third-quarter reports beating expectations one after another, and prices for domestic AI chips also being raised. Yet the STAR 50 Index has been moving downward since July, almost giving back all of this year's gains, and ETF products tracking the index, such as the STAR 50 ETF E Fund (588080), have come under pressure accordingly. On October 28, the STAR 50 rebounded again, rising more than 2% at one point during the session. With the sector clearly strengthening on one side and the index continuously fluctuating on the other, many people look at their accounts and worry: at this position, can I add more? First, a quick primer on what the STAR 50 is. You can think of it as a "report card" — a specialized compiling institution gathers the 50 largest and most representative tech companies on the STAR Market together and calculates an average value. If the report card rises, it means these 50 companies are doing well overall; if it falls, it means they are weakening. So you do not need to pick stocks one by one — a glance at this number lets you understand the temperature of the entire STAR Market. So why does the index keep fluctuating when the companies are clearly improving? There are three main reasons. First, U.S. Treasury yields rose, weighing on tech stocks. From July to October, the 10-year U.S. Treasury yield rose by about 100 basis points cumulatively, reaching 5.34% at one point, the highest since 2002. Companies in the STAR 50 have relatively high valuations and have not yet reached their profit peak, so when rates rise, this type of high-valuation company is suppressed first. Historically, once U.S. Treasury yields rise by more than 10 basis points in a single month, the STAR 50 falls by about 3.2% on average that month. The good news is that U.S. Treasury yields have recently retreated from their highs, and this pressure is easing. Second, trading has been sluggish and buying demand insufficient. Since July, the tech sector has consistently seen shrinking volume, and before the National Day holiday everyone held back and watched, while after the holiday overseas tech stocks surged and then fell back, which also weighed on A-share sentiment. The STAR 50's daily turnover fell from 200 billion yuan in early July to less than 80 billion yuan by the end of September. When the market is sluggish, fewer people dare to buy, and even if not many people are selling, prices can slowly be ground down. Once sentiment returns, buying demand will naturally come back. Third, uneasy news piled up after the holiday. U.S. inflation data beat expectations, and the market worried about another rate hike; revenue from AI leaders such as OpenAI fell short of expectations, making people question whether trillion-level computing power investment can actually earn back its cost through revenue; and the bond yields of computing power leasing company CoreWeave rose sharply, stirring concerns about AI infrastructure financing. Overseas tech stocks fell collectively, dragging down the A-share tech sector as well. However, looking back at the fundamentals, the STAR 50 has neither lost its profitability nor broken its industry trend. On the earnings side, market forecasts for the STAR 50's net profit over the next year have been raised by about 62.6% over the past year and are still being revised upward; the 6 companies that have already disclosed earnings forecasts in the third quarter have a combined weight of about 18.6%, and all of them issued positive guidance. On the industry trend side, overseas AI leaders' revenue falling short of expectations does not mean computing power demand has stopped. Nvidia itself has said that leading labs account for only one-quarter to one-third of its computing power demand, so judging all computing power investment solely by the revenue of OpenAI and Anthropic is inherently unfair. Conversely, TSMC and ASML both beat expectations in the third quarter, and their fourth-quarter guidance is also optimistic, showing that global AI is still genuinely putting money in its pocket. Domestic computing power is performing even more impressively. Recently, domestic accelerator cards have seen collective price increases, and profit expectations for four domestic AI chip companies have been revised upward for four consecutive weeks, with the upward revisions mainly driven by price increases, indicating that demand for domestic computing power has risen and companies now have the confidence to raise prices. Chart: Profit forecasts for four domestic chip companies over the next 12 months (100 million yuan). Data source: WIND, as of October 16, 2026. The above information is intended to introduce sector dynamics, is based on objective facts, and is for display only. It does not represent individual stock recommendations, and the same applies below. Therefore, this round of index adjustment is more about price fluctuation than a decline in value itself. On October 23, the STAR 50 rose more than 3% in a single day, and on October 28 it rose more than 2% intraday again, so the repair has already begun. From a valuation perspective, the STAR 50's current price-to-earnings ratio is about 45.4 times, about 40% lower than its July high, returning to the 55th percentile since 2020. In other words, more than half of the earlier high valuation has already been digested. From the perspective of holdings, this round has given back 93% of the previous gains, and only about 4% of holders still have floating profits. The current price is about 20% below the average holding cost. That is to say, most people who chased in are trapped, those who could sell have mostly already sold, and selling pressure has eased considerably. On the capital side, from October 12 to 15, ETFs tracking the STAR 50 saw net inflows of 11.69 billion yuan, indicating that medium- and long-term money has begun to take over at low levels. However, margin financing positions are still withdrawing, and public funds' tech allocations have not been fully digested, so short-term volatility may continue. Looking over a longer cycle, there is another signal: the growth-versus-value style divergence indicator has fallen to 0.01 and has been hovering at this level for two to three weeks. This means the style deviation has been suppressed for too long, and the probability of a rebound ahead is rising. The STAR 50 is one of the most elastic broad-based indices in the growth sector, and once style returns, its rebound magnitude often ranks near the top. Chart: Growth-value divergence. Data source: WIND, as of October 16, 2026. To sum up in one sentence: this round of STAR 50 adjustment was caused by a combination of investors waiting on the sidelines, external negative news, and interest rate pressure, rather than the sector itself getting worse. Profitability is accelerating, the industry trend is advancing, and both valuation and holdings have gone through a round of cleansing, so the cost-effectiveness right now is actually improving. If you are optimistic about China's core technology assets but do not have a stock account, you can pay attention to the E Fund SSE STAR 50 ETF feeder fund (A/C/Y: 011608/011609/022895), which can be purchased through bank apps, the fund company's official website, or internet platforms. Still have questions about index investing? Want to check more detailed data? Open the "Index Express" WeChat mini program and ask the AI directly for professional answers.
Comments