news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

Securities Times reporter Wang Minghong

Since the beginning of this year, sectors such as dividends and artificial intelligence have been sought after by funds. The share prices of industry stocks represented by Nvidia have continued to rise, and the fund yields of the artificial intelligence sector have also outperformed other industries in the market.

In the first half of the year, Morgan Stanley Digital Economy, managed by Lei Zhiyong, deputy director of the equity investment department of Morgan Stanley Fund and manager of Morgan Stanley Digital Economy Fund, ranked at the forefront of equity-oriented mixed funds with an annual return rate of 25.46%. How did he capture the investment returns of artificial intelligence? Is the valuation of the artificial intelligence industry chain overvalued? What investment opportunities are there in this field in the future? In response to these questions, Lei Zhiyong recently accepted an interview with a reporter from Securities Times.

Finding Certainty in the Market

Lei Zhiyong graduated with a master's degree in computer software and theory from Peking University. He has served as a project manager at the headquarters of China Mobile Communications Group Co., Ltd. and an intermediate network operation support supervisor at China Mobile Communications Group Guangdong Co., Ltd. He joined Morgan Stanley Fund in October 2014 and is one of the few fund managers in the industry with work experience in artificial intelligence-related industries.

Lei Zhiyong believes that the reason why the artificial intelligence sector is favored by funds is that in the current uncertain market, macroeconomic and international environment, investors have placed high demands on certainty. As the market shifted to a defensive style this year, trading volume declined significantly. The two types of assets that outperformed the market in the first half of the year were high dividends and artificial intelligence computing power. In essence, they both reflect the same idea - "defense", that is, investors pursue certainty in an uncertain environment.

Lei Zhiyong recalled that during the period of 2011-2013, the domestic macro-economy and A-share market were also under great pressure, but the growth-oriented assets represented by smartphones had a good growth trend. In those years, China's industrial chain was accompanied by a large technology growth cycle, that is, the switch from feature phones to smartphones. The performance growth rate of the smartphone sector far outperformed other industries, and the industry performance also significantly outperformed the A-share market. The same thing happened this year.

"In fact, we have increased our AI computing power allocation since the second half of last year. At that time, our research found that the order increase of optical module manufacturers was very clear, and TSMC's order outlook also reflected that CoWoS packaging capacity was in short supply. These information all reflected that the AI ​​industry chain's performance growth was very certain."

However, there are many fund managers in the market who are planning AI industry chain opportunities. Why is the Morgan Stanley Digital Economy Fund ranked at the top? Lei Zhiyong said that on the one hand, the Morgan Stanley Digital Economy Fund did not invest in theme-oriented and concept-oriented computing companies due to excessive pursuit of flexibility. The holdings of the Morgan Stanley Digital Economy Fund focus more on performance realization and certainty. On the other hand, the Fund continues to optimize its holding structure based on closely tracking changes in the industry chain.

He revealed that in the process of tracking the industry, he found that in addition to making efforts in the computing cloud - deploying a large number of servers and training large models, giant companies at home and abroad have also begun to evolve towards the application side, especially smart terminals. In addition, giants represented by Apple have begun to deploy new products on the mobile terminal side. "After industry comparison, we found that directions such as smart terminals are better than simple large models. Large models have the problem of performance realization. In the smart terminal industry chain, domestic companies can serve as a supporting manufacturing industry chain with order support." Lei Zhiyong said that based on the above logic, he gradually added some positions on the AI ​​terminal side to the fund portfolio, and this judgment has also been verified by the market.

“Investment and Research Integration”

Improve your stock picking success rate

So, how did Lei Zhiyong capture the prosperity of the artificial intelligence industry and buy individual stocks in the industry chain at the right time?

"In recent years, with the continuous upgrading of the high-end manufacturing industry, a trend has emerged in industries such as automobiles, mobile phones, and national defense and military industries. A single industry requires multiple researchers to cover it together. For example, the automobile industry needs new energy, electronics, and computer researchers to participate together to compare the industry's prosperity and industrial cycle trends in the next two quarters." Lei Zhiyong introduced that Morgan Stanley Fund has established an "investment and research integration" system, dividing the equity team into growth groups, consumer groups, and cycle groups. Senior fund managers or researchers serve as team leaders and promote investment and research analysis, which more effectively improves the collaborative efficiency of integrated research and better opens up the link from research to investment implementation.

He said that after OpenAI was launched last year, the team initially thought it was a thematic investment, but members of the growth group continued to track and found that the prosperity of sectors such as PCB, servers, and optical modules was better than that of new energy and other industries. After selecting the industry based on the mid-term dimension, they compared the stock price and valuation to decide to buy.

How to evaluate whether the stock price of an individual stock matches its valuation and buy it at the right time, Lei Zhiyong introduced that it is mainly carried out through three dimensions: first, judging the position of the industrial chain, including paying attention to the evaluation from the giants of the industrial chain, as well as the evaluation of it by its upstream and downstream supply chain and competitors; second, interviewing the executives of listed companies and their business backbones, including order visibility, employee confidence, etc., and collecting feedback from employees; third, financial data, studying key indicators in the financial statements including accounts receivable, gross profit margin, etc.

The trend of artificial intelligence is still there

Lei Zhiyong, who seized the opportunity of artificial intelligence in the first half of the year, is optimistic about the development prospects of the industry. He believes that the greatest charm of the growth direction of science and technology lies in innovation and demand creation. "The iPhone is a typical example of innovation and demand creation. Before the iPhone came out, we used feature phones, but Apple created a brand new product and changed everyone's mobile phone usage habits. We see that ChatGPT has also created a brand new product and driven new demand." Lei Zhiyong said.

Lei Zhiyong believes that the charm of science and technology also lies in the fact that the iteration of technology is accompanied by industrial development and investment opportunities. By creating new products, a certain user experience is greatly improved, the needs of social life and work are met, and the industry itself also achieves substantial growth. In this process, the soil for the birth of bull stocks is laid.

However, he believes that the evolution of artificial intelligence is staged and may be lower than expected at a certain stage. For example, after the subsequent release of ChatGPT-5, everyone thought it was lower than expected, but this does not prove that artificial intelligence has failed. "The vision of artificial intelligence is the vast sea of ​​stars and general artificial intelligence. It is not just about writing an article or refining information on the Internet in a question-and-answer style. It must be able to replace humans in some difficult environments or some repetitive labor and help us complete these tasks. There is still a long way to go to achieve this goal." Lei Zhiyong said that from the industrial end, artificial intelligence is still on the left side, but the investment end is already very hot.

Recently, after the share prices of AI stocks represented by NVIDIA reached historical highs, they immediately began to fluctuate sideways at high levels. The market is worried whether the sharp rise in AI stocks has already overdrawn the valuation of the industry chain in advance.

Lei Zhiyong believes that the valuation of AI needs to be viewed from two dimensions. First, AI is in an upward cycle of industrial prosperity. Historically, the PE of the new energy industry in the current and second years of this stage can reach 30 to 40 times or even higher, but the current market generally expects that the PE of optical modules and PCBs will be less than 20 times in 2025.

"Secondly, from a historical perspective, during the innovation cycle when TMT companies switched from 3G to 4G, their PE ratios reached as high as 35 or even 40 times. Whether compared horizontally with other industries or vertically with the industrial chain of individual stocks, the current overall valuation of AI should be reasonable or even on the low side when it matches its performance," said Lei Zhiyong.

Lei Zhiyong added that whether the valuation is reasonable or not is based on the judgment of the trend of the artificial intelligence industry. With the expansion of computing power scale and the continued evolution of large models, the trend of industrial development has not ended and the rules have not yet been falsified. Therefore, the valuation is still in a reasonable range.