For stock trading, check the Jin Qilin analyst research report, authoritative, professional, timely and comprehensive, helping you tap potential thematic opportunities. Core conclusion: (1) Similar to the internet sector from June to September 2015 and the new energy sector from December 2021 to April 2022, this technology stock decline stems from high valuations and crowded trading. (2) Comparing the changes in allocation structure, decline magnitude, and shifts in market sentiment, the sufficiency of this correction is basically comparable to the previous two episodes. (3) After both previous corrections, rebounds emerged and were accompanied by new catalysts. This time, the new catalysts may be personal Agents, physical AI, and others. The Shanghai Composite Index has been correcting for five and a half months since its mid-May high, while the STAR 50 Index has been correcting for four months since its early July high. The decline since late September has been steeper, and investor sentiment has recently been sluggish. Has the market correction been sufficient? Where are the opportunities ahead? We draw on the markets of 2015 and 2021 for analysis, because both were typical bull markets led by technology growth stocks and are fairly similar to this one.
Drawing lessons from 2015 and 2021 to examine the rhythm and logic of declines
Looking back at history, the initial decline after a bull market peak stems from high valuations and crowded trading, while fundamentals remain sound. The bull market from December 2012 to June 2015 was dominated by the mobile internet industry, and the bull market from January 2019 to December 2021 was dominated by new energy. After the peaks of both bull markets, the trigger for the initial decline was high valuations and crowded trading, while fundamentals were healthy at the time. From June to September 2015, the ChiNext Index and the computer index fell at most 56% and 59% from their highs respectively. During December 2021 to April 2022, the ChiNext Index and the power equipment index fell 40% and 45% respectively. Because fundamental data was still relatively strong at that time and the ROE of leading industries was rising, after the first wave of declines, clear rebounds occurred from September to December 2015 and from April to July 2022. Once fundamentals showed a downward turning point, declines resumed after the rebound ended, such as computers in 2016 and new energy after July 2022. The background of the technology stock decline represented by the STAR 50 since July is similar to that of the ChiNext Index from June to September 2015 and December 2021 to April 2022. Since July, A-share technology stocks have clearly pulled back. As of October 29, the STAR 50, electronics, and communications had fallen at most 46%, 49%, and 48% from their highs respectively. Like in 2015 and 2021, this decline also stems from high valuations and crowded trading, while fundamental data has not deteriorated. The STAR 50 Index's attributable net profit in the first half of 2026 is expected to surge 166.4% year on year, and the consensus net profit growth forecast for 2027 has been raised from 48% to 50%. Moreover, overseas markets had already stabilized and rebounded earlier, with the Nasdaq 100 Index hitting a record high on October 6, mainly supported by strong fundamentals. During the same period, the Nasdaq Index's expected net profit growth for 2027 was revised up from 26% to 31%.
Comparing with 2015 and 2021 to assess the sufficiency of this correction
As mentioned earlier, the nature of the technology stock decline since early July is similar to mobile internet from June to September 2015 and new energy from December 2021 to April 2022. After four months of correction, has it already become relatively sufficient? We compare and analyze from the perspectives of institutional allocation, absolute and relative declines, and market sentiment. From the perspective of allocation structure, similar to 2015 and 2021, this time there has already been a clear rebalancing. Based on relatively transparent holdings data from the three major investor groups of margin financing balances, Stock Connect, and actively managed equity public funds, we analyze investors' allocation structure. To exclude the impact of stock price gains, we estimate the holdings of different investor groups in the leading industries of the bull market using a fixed-price weighted shareholding method, and measure the extent of reduction from their respective peaks. The detailed data is shown in Table 2. During the decline from June to September 2015, there was obvious deleveraging, so the investor group represented by margin financing balances reduced holdings relatively significantly during the computer sector decline, with shareholdings falling 28%. Apart from that, during the declines in leading industries in June to September 2015, December 2021 to April 2022, and since July 2026, investor groups reduced holdings by about 10%. From the perspective of correction magnitude, both the absolute and relative declines this time are close to those in 2015 and 2021. In terms of the absolute decline in stock prices, from June to September 2015, the ChiNext Index and computers fell at most 55% and 59% from their highs respectively, while in 2021 the ChiNext Index and the new energy rally fell 39% and 44% respectively. On the other hand, in terms of the degree to which gains since the bull market starting point were given back, when computers fell in 2015 and new energy fell in 2021, the ChiNext Index retraced 0.64 and 0.59 of previous gains respectively. Returning to this round of decline, electronics has fallen at most 46% since its early July high, retracing 0.61 of previous gains, both fairly close to 2015 and 2021. From the perspective of market sentiment, the cooling this round is basically comparable to 2015 and 2021. We observe changes in overall market sentiment during each decline, which can be described by the degree of contraction in turnover and turnover rate. During the market decline from June to September 2015, the weekly turnover rate fell from a high of 765% to a low of 421%, and weekly turnover fell from 9.8 trillion yuan to a low of 3.8 trillion yuan, contracting by about 50% and 60% respectively. During the market decline from December 2021 to April 2022, the weekly turnover rate and weekly turnover contracted by about 40% and 50% respectively. Returning to this round of decline, the two have contracted by 40% and 60% from their highs respectively, roughly comparable to the degree of sentiment cooling in 2015 and 2021.
Drawing on the previous two episodes, rebounds often come with new catalysts
Based on the analysis above, the sufficiency of this round of market correction is already close to that of 2015 and 2021, and the preconditions for a rebound are gradually in place. Drawing on the previous two episodes, ultimately emerging into a considerable rebound still requires new catalysts. The considerable rebounds from September to December 2015 and from May to July 2022 were both accompanied by new fundamental catalysts. In the first half of 2015, "Internet Plus" was in full swing, but after negative feedback in micro-level liquidity from June to September 2015, the market needed a new narrative to rebuild consensus. After September, favorable policies arrived frequently, and a new narrative of domestic informatization gradually emerged, driving a rebound. The domestic chip and operating system indices rose at most 127% and 118% respectively. From 2020 to 2021, the rapid increase in new energy penetration was the core logic behind stock price gains. The renewed rise in new energy in May 2022 stemmed from accelerated resumption of work and production in China, a sharp rebound in new energy vehicle production and sales growth, combined with high overseas demand for photovoltaics and price increases that enabled cost pass-through for silicon materials. The photovoltaic and battery sectors rebounded clearly, with the photovoltaic inverter and TOPcon battery indices rising at most 171% and 131% respectively. The new catalysts that may be needed for a rebound: personal Agents, physical AI, and others. In the first half of this year, capital expenditure by overseas cloud vendors continued to grow rapidly, and domestic substitution of computing power accelerated. However, as free cash flow of cloud vendors turned negative, a further rebound also needs support from a new narrative. Looking ahead, Anthropic recently publicly submitted an IPO application, effectively easing concerns about AI financing. At the same time, large models are accelerating their evolution from large language models (LLM) to Agents, and personal Agents represented by Muse are expected to continue penetrating multiple scenarios such as finance, media, and consumption. In addition, Tesla's third-quarter report on October 22 showed that Optimus has started mass production in Fremont, with weekly output expected to reach 20,000 units by the end of 2027. Many embodied intelligence-related products are gradually moving from the B-end to the C-end, and directions such as humanoid robots in physical AI may usher in an explosive growth period. These directions are expected to become new catalysts driving a rebound. Risk warning: technological industry development falling short of expectations; overseas monetary policy tightening more than expected.
Comments