Zhitong Financial APP learned that Dongguan Securities released a research report saying that as the impact of the epidemic receded, the book retail market recovered this year, and the performance of the publishing industry picked up steadily. Looking forward to the follow-up, the changes of AI for the publishing industry are mainly reflected in three aspects. First, the publishing industry has the high-quality data resources needed for model training, and the acceleration of large model iteration is expected to fully release the value of industry data assets. Second, AI can enable the traditional publishing industry chain to reduce costs and increase efficiency. Third, AI education is the current hot development direction, local publishing state-owned enterprises are expected to rely on content channels and other advantages to take the lead in landing products and promotion, to create the second growth curve of "AI+ education". In the direction of investment, it is recommended to pay attention to the publishing companies with the advantages of content channels and active layout in the field of AIGC.

Related subject matter:Phoenix Media (601928.SH), Anhui New Media (601801.SH), Southern Media (601900.SH), Central South Media (601098.SH), Chinese online (300364.SZ), Century Tianhong (300654.SZ) and so on.

The main points of Dongguan Securities are as follows:

After the epidemic, the book retail market has been gradually repaired, and short video e-commerce channels have sprung up rapidly.

Affected by the epidemic, the book retail market has been under pressure in recent years. From a sub-channel point of view, due to social isolation restrictions on offline consumption, book sales accelerated the transfer process from offline to online during the epidemic, while online channel medium and short video e-commerce complied with the trend to become a rising star. In 2023, as the impact of the epidemic subsided, the decline in the size of the book retail market in the first three quarters continued to narrow compared with Q1 and the first half of the year, and the market recovery situation improved.

The performance is growing steadily, and the industry valuation is on the low side waiting to be repaired.

The leading enterprises in the publishing industry have the characteristics of sound performance, abundant cash flow, high dividend ratio and high dividend yield. According to iFind statistics, 2023H1, the publishing industry revenue and home net profit increased by 5.91% and 10.62% respectively compared with the same period last year, and the operation of enterprises in the industry continued to be sound. In terms of valuation, the publishing industry is currently in the 38.47% quartile of nearly five years of history. The overall valuation of the industry is on the low side and has a high margin of investment safety.

With the opening of the arms race of large models, the value of high-quality data in the publishing industry has been highlighted.

The emergence of ChatGPT leads the new wave of the development of generative AI technology, and then the domestic and foreign science and technology giants have invested in starting the AI large model arms race. As the training of large models needs the support of large amounts of high-quality data, the rich and high-quality Chinese data resources accumulated by the publishing industry are expected to be fully released, and the industry is expected to usher in valuation reshaping. With the steady introduction of training data copyright protection policy in China, the continuous improvement of infrastructure and trading system, and the gradual clarification of the charging rules for overseas data, it is expected to accelerate the pricing, circulation and realization of data resources in China, including the publishing industry.

AI empowers traditional publishing to build a new ecology of "AI+ Education".

In the aspect of traditional publishing, AI can be used in topic selection planning, content editing, marketing and other links to achieve cost reduction and efficiency. In terms of AI education, the integration of overseas AI+ education is accelerating, and the demand for related products is strong, while the landing of domestic AI education is still in the early stage. With the steady promotion of education information-related policies in China, the scale of state investment is increasing, and publishing enterprises, especially local state-owned enterprises, relying on the advantages of content, channels and funds, are expected to deeply participate in the reform of the "AI+ education" industry, thus opening the second growth curve.

Risk Tips:Policy regulatory risks; macroeconomic fluctuations affect book consumption demand; industry competition intensifies risks; the development and application of AI technology is not as expected; state support for educational informatization is not as strong as expected.