Baidu:relatively thin cloud coverage

类型：公司研究机构：德意志银行研究员：Alan Hellawell 日期：2017-08-15

Scrapping hard to build share; AI to provide the foothold.

    Baidu Cloud continues to hold a relatively small share of the domestic cloudmarket, though our 2017 CIO survey showed an impressive rise in userpreference for its services. Baidu has the potential to leverage its expertise anddominance in the country’s AI space to offer “AI + big data” focused cloudsolutions in the mid- to long-term. The company is also working on combiningits autonomous driving technology with cloud computing to develop cloud-tocarcomputing. We however do not believe that these initiatives are likely tocontribute significantly to the platform’s growth in the short- to mid-term.

    Maintain Hold as we wait for stronger signals of a recovery in core search.

    Users preference rising; AI, IoT possible areas to lift player out of relativeobscurity.

    Our centerpiece 2017 CIO survey showed impressive user preference for BaiduCloud, with a 33% user preference under the “multi-vendor” strategy.

    Respondents also ranked Baidu Cloud #2 for its solid performance and aftersalesservices, while ranking it #3 in terms of pricing, security,speed/stability/availability, and technological innovation. Nevertheless, manywith whom we spoke believe that the company generally continues to strugglein terms of market share as a result of its late entrance into the domestic cloudmarket.

    Cloud revenue to grow at a 4-year CAGR of 109% to reach RMB4bn by 2020.

    We expect Baidu Cloud to be on smaller scale vs. Alibaba and Tencent. Weexpect Baidu Cloud revenue to grow at a 4-year CAGR of 109% to reachRMB4bn in 2020, 10% of Alibaba Cloud and 14% of Tencent Cloud estimate.

    Maintaining Baidu TP at USD212.

    Our TP is based on SoTP: core search & newsfeed on 1.0X PEG 2018-20ECAGR; a share of iQiyi’s valuation; a Ctrip stake (1.1x PEG); a share of Baidu’sO2O valuation, and net cash. Risks: better/worse recovery in search,success/failure in newsfeed, faster/slower monetization of AI initiatives (p.4).