Investment advice
Industry strategy: the data center and cloud computing industry chain have a long-term prosperity, and the industry will maintain an overall CAGR of 20% for more than five years +: computing network infrastructure accounts for about 41% of the investment, and it is estimated that the market scale will be 878.5 billion yuan in 2025; It is expected that the data lake, artificial intelligence and other subdivided fields will double in three years; The final goal is to establish a national computing power network of "four high and three synergies" (high energy efficiency, high technology, high security, high computing power, digital cloud collaboration, cloud edge collaboration and digital network collaboration), which can drive an investment of 3 trillion yuan in the long term (National Development and Reform Commission). Incremental space is mainly concentrated in the construction of data center; ICT infrastructure is more a structural opportunity to evolve to high-end products. In the medium and short term, it is recommended to pay attention to the parts manufacturers with high investment, such as power supply and distribution system, cold source air conditioning and heating system; Recommended leading manufacturers of computing network infrastructure; The growth prospect of software services is considerable, and OPEX expenditure is more sustainable. IDC operators and application service software manufacturers deserve long-term layout.
Recommended combination: China Mobile, Zhongji Innolight Co.Ltd(300308) , Inspur Electronic Information Industry Co.Ltd(000977) , Kehua Data Co.Ltd(002335) , Yonyou Network Technology Co.Ltd(600588) .
Industry perspective
We believe that there are valuation repair opportunities in IDC sector, and adjust and supplement the mainstream research framework of the market. 1) Non bat cloud business is growing rapidly, the reference of capital expenditure growth of cloud giants is weakened, and the delivered data can best reflect the prosperity of the industry. Traditional enterprises need to gradually increase their penetration into the cloud, and the steady-state industry growth rate is expected to be 20% - 30% in the next five years. 2) In the short term, the supply and demand of eastern and Western data centers do not match, the supply side of 2021h2 has stabilized, and the share of head IDC manufacturers has increased. In the long term, China's super large data centers with high intensification and high computing power are insufficient. 3) The general view of the market is that IDC industry barriers are mainly reflected in scale, site selection and pue. Whether to obtain land and cabinet resources in the core area is the most important; We believe that industry barriers run through the whole life cycle of planning and site selection, construction speed, operation and maintenance capacity and financing capacity. The energy consumption policy has become stricter, the industry differentiation has intensified, and the share of leading manufacturers has increased. 4) IDC is a heavy asset industry, which is usually valued by EV / EBITDA; The historical valuation center is about 25X and the current value is less than 20x. There is room for repair. At the same time, the impact of depreciation accounting policy should be considered. In the long run, or switch to PE or Pb valuation.
The policy of "counting from the east to the west" is the starting point of the construction of computing power network. It is estimated that the market space during the 14th Five Year Plan period is about 2 trillion yuan. According to the disclosure of regional data center construction plans, it is estimated that about 2.42 million 6kW standard racks will be added to the national core computing hub nodes from 2021 to 2025. According to the cost of single cabinet of 200000 yuan, it is expected to directly drive the total investment of 484 billion yuan. As a systematic project, the length of the industrial chain involved in "computing from the east to the west" is not limited to IDC business, but also includes the comprehensive software and hardware investment of computing network infrastructure layer, arrangement management layer and operation service layer. Finally, it is necessary to realize a national computing network with "network everywhere, computing power everywhere and intelligence". Among them, the high proportion of total investment mainly includes power supply and distribution system (11%), server (17%), network security (8%) and comprehensive software and application services (20%) in IDC construction; In addition, there are data communication optical modules (20%), data Lake (31%) and artificial intelligence (30%) with fast compound growth rate.
The data center seems to be traditional, but technological innovation continues. It is suggested to pay attention to the investment opportunities brought by technological innovation. Large scale, intensive and green data centers are the basis for supporting the integration of computing networks across the country. In order to solve the problem of mismatch between land, hydropower and other resources and computing power needs, before the national intervention, all enterprise entities have invested in chip level cost reduction and efficiency enhancement technological innovation, and will migrate businesses with low delay requirements to the western region. All government departments carry out corresponding planning, not only to encourage the healthy and orderly development of the industry, but also to increase investment from the perspective of top-level design, and provide l0 level infrastructure services that enterprise entities cannot layout, such as hydraulic and power network deployment and corresponding planning that can match the data center. The main technological innovation directions are mainly reflected in chip energy saving, liquid cooling, heterogeneous computing, IP intelligence, etc.
Risk tips
The implementation of relevant policies of the data center is less than expected; The scale development of IDC is less than expected; The industrial chain coordination is not as expected; Technological innovation is less than expected; The progress of localization is less than expected.
