The transaction accounted for more than 40% Can the AI sector still chase up?

The new high proportion of transactions means that short-term transactions are overheated and the probability of volatility increases, and the industry reminds investors to be cautious and chase higher

Since last week, the turnover of the computer, electronics, communications, and media industries extended by the concept of artificial intelligence has exceeded 40% of all A-share turnover many times - on March 3, the total transaction volume of TMT (technology, media, communications) and electronics reached 24% of the total A-share transaction. In the A-share market, there is a magic "digital curse", that is, a concept sector trading volume reaches 46% is a signal to peak. Can AI-related sectors triggered by ChatGPT break the "40%" curse of A-shares?

Artificial intelligence empowers new drug development?

Cambrian, known as the "first stock of AI chips", under the blessing of the concept of ChatGPT, the stock price soared from less than 55 yuan at the end of last year to a recent high of 188.88 yuan, up nearly 250% in less than a quarter. Cambrian's stock price change announcement a few days ago said that the company is not directly engaged in the development and sales of artificial intelligence final application products (such as ChatGPT-like applications), and the company's related products are in the adaptation stage and have not yet generated revenue; Whether customers carry out large-scale procurement is affected by factors such as product technical capabilities, delivery capabilities, product prices and customers' willingness to purchase, and there is uncertainty.

This week, artificial intelligence swept the field of new drug discovery. It is reported that HUAWEI CLOUD Pangu Drug Molecule Large Model is a large model jointly trained by HUAWEI CLOUD and Shanghai Institute of Materia Medica, Chinese Academy of Sciences, which can realize AI-assisted drug design for the whole process of small molecule drugs. Further manual experiments by researchers show that the accuracy of predicting druggability of Pangu drug molecule large model is 20% higher than that of traditional methods, thereby improving R&D efficiency, shortening the R&D cycle of lead drugs from several years to one month, and reducing R&D costs by 70%.

According to HUAWEI CLOUD's official account, the HUAWEI CLOUD Pangu Drug Molecule Model released in 2021 is a member of the Pangu Model Family, which is a pre-trained large model specially launched for the drug R&D field, aiming to help pharmaceutical companies open a new model of AI-assisted drug discovery. According to the analysis of Zheshang Securities Research Report, HUAWEI CLOUD Pangu drug molecule model is classified as AI-assisted drug design services. From the perspectives of AI+Pharma Cost Reduction and Efficiency IncreaseDrug X ushered in a major breakthrough, technology platforms and scientific research teams joined hands with AI and biomedical bidirectional empowerment, and full-process assisted drug design AI has great prospects in the future, the application examples of HUAWEI CLOUD's Pangu drug molecule model are listed.

It is understood that HUAWEI CLOUD Pangu pre-trained big model has completed the transformation from academic big model to industrial big model, forming a development path of "basic big model, industry big model, and subdivision scenario big model", and achieving cost reduction and efficiency improvement in energy, finance, healthcare, coal mining, agriculture, meteorology, and other fields.

Is the "40%" spell expected to be broken?

The A-share market has a magical "digital curse", that is, a concept sector trading volume reaches 40% is a signal to peak. Last week, the turnover of computer, electronics, communications, and media industries related to the concept of artificial intelligence accounted for 43% of all A-share turnover; On March 3, the turnover accounted for an unprecedented 24%, that is, nearly half of the funds in the market were concentrated in the AI-driven TMT+ electronics sector.

A reporter from Beijing Youth Daily combed through relevant information and found that in the past ten years, the concentration of transactions in a certain direction of A-shares exceeded 40% 7 times. TF Securities said that the high of 2019G in 5 is 40% after the peak, and similar situations are for automobiles, photovoltaics and new energy vehicles in June 2022. 6% means that the mood has reached a climax in a very short period of time, and the expectation is fully interpreted, which also means a sharp decline in the price-performance ratio, which may last for several weeks; Subsequently, under the influence of fundamental expected changes, style changes, external shocks, etc., the mood staged ebb, and the proportion of turnover began to digest and fall, and in this process, whether the sector fluctuated or fell sharply depends on whether there is strong fundamental support.

However, there are also different opinions in the market. Research reports published by a number of research institutions believe that the hype of the theme of artificial intelligence is not limited to short-term "hot money", and institutional funds are gradually running into the market from other fields (new energy, etc.), jointly promoting the current round of technology stocks. "In the process of deducing a round of technology stocks from thematic investment to the main line, there will be multiple rounds of institutional position increases, and the market's understanding of the technology sector will continue to deepen from initial attempts to consensus formation to cognitive maturity." CSC Securities wrote in the research report.

Industrial Securities believes that the current round of digital economy market can be compared to small and medium-sized enterprises from 2013 to 2014, core assets from 2016 to 2020, and new energy from 2020 to 2021, all of which confirm their mainline status in the process of continuously reaching new highs in the proportion of transactions. This round of digital economy market or similar to the core assets and new energy of the year. The dual drive of fundamentals and policy environment may support the "digital economy" through short-term trading congestion and continue to form the main focus direction of the market in the medium term.

Can the AI sector still chase up?

A number of industry insiders told the Beiqing News reporter that from the perspective of trading, the new high proportion of transactions means that short-term transactions are overheated and the probability of volatility increases. There is also the possibility of a short-term correction in some core high-end stocks. Investors are reminded to be cautious in chasing higher.

However, CITIC Securities believes that with the emergence of ChatGPT and its extremely fast technical iteration capabilities and strong industrial extension and penetration capabilities, the market's perception of it has also changed from purely thematic opportunities to potentially a large number of allocation-type opportunities, and over time, it may also evolve towards track-level opportunities with sufficient capital capacity.

CSC Securities believes that in the short term, the popularity of the sector is rising, the trading concentration is too high, and the unhealthy trading structure will indeed aggravate the risk of subsequent volatility. However, in the long run, under the premise of promoting industry trends and deepening market consensus, relevant sectors will still have considerable excess returns. Among them, the process of market pricing in the first stage or the first adjustment after the mood is overheated will be a valuable layout opportunity, and the odds for win rate will still be valid.

The hype of artificial intelligence has also attracted the attention of management. Regulators have maintained high attention to the speculation of ChatGPT concept stocks, and many listed companies such as Kunlun Wanwei, Tomcat, Inspur Information, and Tongyu Communication have received letters of concern or regulatory letters. Shenzhen Stock Exchange recently issued a regulatory letter to Kunlun Wanwei, saying, "Under the background that ChatGPT and related concepts have received widespread attention from investors, your company has failed to objectively and completely introduce and reflect the actual situation of the company's relevant business, the impact on the company's performance and fully remind the relevant risks when responding to investors' inquiries through interactive and easy response." ”

At the same time, during the period when the stock prices of relevant individual stocks rose sharply, they were reduced by important shareholders, such as Haitian AAC's recent announcement on shareholder reduction, China Mobile Investment plans to reduce its holdings of Haitian AAC shares through centralized bidding and block trading, a total of no more than 128.4 million shares; Hanwang Technology said in response to the Shenzhen Stock Exchange concern letter that three of the company's top ten shareholders reduced their holdings of the company's shares in early February; Tongyu Communication announced that the company's controlling shareholders, actual controllers, directors, Senior manager Shi Guiqing intends to reduce his holdings of the company's shares within 2 months, with a total of no more than 6,804,11 shares (accounting for 2% of the company's total share capital).

Text/Reporter Zhu Kaiyun

Co-ordinator/Yu Meiying

(Beijing Youth Daily)