China Internet - Splash - BNP Paribas EQresearch
Transcription
China Internet - Splash - BNP Paribas EQresearch
EQUITIES RESEARCH CHINA INTERNET CONNECTING USERS TO ADVERTISING ■ This report focuses on the emerging business model in China of connecting more Internet users to more advertising through Ad Exchange platforms – a.k.a. Real-Time Bidding (RTB). ■ Ad Exchange platforms enable millions of websites to make their ad spaces available to advertisers through a real-time bidding process, optimizing pricing efficiency. ■ We believe large, established search platforms are in a good position to benefit as they already have large bases of diverse online advertising customers. ■ We initiate coverage of China’s two biggest search engine platforms: Baidu (BUY; TP USD113.00), and Qihoo 360 (HOLD; TP USD44.20). Alen Lin [email protected] +852 2825 1801 BNP Paribas Securities (Asia) Ltd research is available on Thomson One, Bloomberg, TheMarkets.com, Factset and on http://eqresearch.bnpparibas.com/index. Please contact your salesperson for authorisation. Please see the important notice on the inside back cover. PREPARED BY BNP PARIBAS SECURITIES ASIA THIS MATERIAL HAS BEEN APPROVED FOR U.S DISTRIBUTION. IMPORTANT DISCLOSURES CAN BE FOUND IN THE DISCLOSURES APPENDIX 27 JUNE 2013 27 JUNE 2013 SECTOR REPORT CHINA INTERNET POSITIVE Connecting users to advertising n We initiate coverage with a positive outlook and BUY rating on Baidu, HOLD on Qihoo 360 This report focuses on the emerging business model in China of connecting more Internet users to more advertising through Ad Exchange platforms. These platforms make the ad space on millions of websites available to advertisers through a real-time bidding (RTB) process, optimising price efficiency. We initiate coverage of China’s two biggest search engine platforms: Baidu (BUY; TP USD113.00), and Qihoo 360 (HOLD; TP USD44.20). n New business model getting off the ground in China Ad Exchange represented only 0.7% of total display ad spend in China in 2012, with IDC forecasting a 2012-16 CAGR of 145%, contributing 11.1% (USD627m) of total display ad spend in 2016. Platforms being established include Baidu Exchange Service, Tencent AdExchange, Sina, and Alibaba’s Alimama. Key success factors for Ad Exchange platforms include a large base of online advertisers and a network of Internet properties that covers a broad range of user segments. n Search engines in a good position to capitalise on the emerging opportunity We believe large, established search platforms are in a good position, as they have substantial existing bases of diverse online advertising customers. Baidu Union already includes hundreds of thousands of members that make their websites available for ad placement. We view Ad Exchange as a new product offering for search engine platforms that will incrementally drive their ad revenue growth as they expand from traditional search ads to include display ads on member websites. BNPP recommendations Company BBG code Rating Share price Target price Upside/downside Baidu Inc - ADR BIDU US BUY 91.09 113.00 +24.1% QIHOO 360 TECHNOLOGY QIHU US HOLD 44.14 44.20 +0.1% Prices at 25 June 2013 Source: Bloomberg; BNP Paribas Alen Lin [email protected] +852 2825 1801 2 Alen Lin China Internet Investment thesis Baidu – Initiate coverage at BUY with a TP of USD113.00 Ad Exchange, aka Real-Time Bidding (RBT), is starting to emerge as a new business model in China’s Internet industry; the platforms are well established in the US and it is a proven business model to connect advertisers with relevant Internet properties. Various platforms have been established including Baidu’s BES, Tencent’s AdExchange, Sina’s SAX, Alibaba’s Alimama, and Taobao’s tanx. We believe the business model favours platforms with strong established networks of online advertisers and Internet properties. We believe Baidu is in the process of diversifying its revenue beyond search advertising. While we expect search advertising to remain strong, with our forecast industry revenue CAGR of 31.9% between 2012 and 2015, we believe the company has yet to tap into new products such as Ad Exchange. We believe the exchange platform will enable large search engine platforms like Baidu to expand into new advertising segments to include the vast number of Internet properties available and mobile apps. We expect Baidu’s revenue growth to remain >30% y-y through 2015 and see margins stabilising in 2014, leading to a return to strong profit growth. We also estimate Baidu’s ROAE will remain above 30% through 2015. We view the search engine industry as a high-performance subsector of Ad Exchange that enables the serving of highly targeted ads to Internet users. Large-scale search engines have already built up significant online advertising customer bases for monetisation purposes – an important component for an Ad Exchange platform. In 2012, Google generated 28.5% of its advertising revenue from Google Network Member websites, which we believe is closely related to its DoubleClick Ad Exchange platform; we see significant advertising potential in China’s Internet industry, as the segment only contributed 0.7% of industry display advertising revenue in 2012. Qihoo 360 – Initiate coverage at HOLD; TP USD44.20 While Qihoo 360 has posed a challenge to rival search engine platforms in China, we believe organic market share gains will become more difficult. In our view, the company needs to maintain high levels of marketing and R&D spend, which could pose a downside risk to consensus expectations. In the near term, we expect key revenue contributors to remain navigation ads and online gaming. Given the high valuation (2013/14E PEs of 126x/43x), we believe potential upside to the share price is limited at current levels. Ad Exchange platform CONTENTS Executive summary 4 Connecting users to advertising 6 Internet search engine – a high-performance subset of Ad Exchange 14 Competitive landscape in the China search engine industry 22 Regulatory control; a significant force in the search engine 27 industry Ad Exchange favours participants with large user bases and 30 advertisers Company reports 31 Contribution to Ad Exchange display ad sales United States Western Europe Japan China Others 2013 2014 2015 2016 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2010 3 BNP PARIBAS 2011 2012 27 JUNE 2013 China Internet Alen Lin Executive summary Our report focuses on the emerging business model in China of connecting more Internet users to more advertising through Ad Exchange platforms – aka Real-Time Bidding (RTB). While various platforms have been established, including Baidu’s BES, Tencent’s AdExchange, Sina’s SAX, Alibaba’s Alimama, and Taobao’s tanx, the business model favours platforms with strong established networks of online advertisers and Internet properties. Regardless of the perceived utilities for Internet users for various Internet services, they are ultimately all in the business of connecting users to advertising. We believe multiple platforms can exist in China, targeting various industry verticals that differentiate through different segments of online advertisers and Internet properties. We place particular focus on the search engine industry as its core business has always been about connecting advertisers to the most relevant audience. We view search engines as a high-performance subset of the Ad Exchange structure. Search queries initiated by Internet users enable search platforms to serve highly targeted advertisements to users. By nature of the business model, large-scale search platforms such as Baidu have built up significant online advertising customer bases; we view this as a key component of an Ad Exchange. While this has enabled these platforms to monetise their respective search services, we believe they also lay the foundation for rapid expansion into the Ad Exchange platform. Google in 2012 generated 28.5% of its advertising revenue from Google Network Member websites, which we believe is largely attributable to its DoubleClick Ad Exchange platform. Ad Exchange as a new business model in China – Internet players already positioning for the opportunity Internet search engines in a unique position as online advertisements play to their core strengths On this theme, we initiate coverage on Baidu (BIDU US) with a BUY recommendation and TP of USD113.00. We also initiate coverage on Qihoo 360 (QIHU US) with a HOLD recommendation and TP of USD44.20. Baidu – Initiate coverage at BUY with a TP of USD113.00 We believe Baidu is in the process of expanding beyond its core PC search business to tap into new growth drivers. In our view, successful execution of its Baidu Exchange Service (BES) would enable the company to drastically increase its inventory of Internet properties for advertising, which could also include mobile applications. We assume the following in our investment case: § Baidu’s Internet search market share to remain dominant, given the positive feedback loop of the industry where a higher number of users drives a higher number of online advertising customers, leading to rising revenue/customer. This enables Baidu to maintain its ability to monetise its core search services. We estimate Baidu’s revenue growth for 2013-15 to be 37.4%/32.4%/32.7%. § Margin decline in 2013 (NM from 46.9% in 2012 to 33.6% in 2013), primarily because of product mix change including acquisitions of online video services and greater emphasis on BES. While BES is likely to have lower margins compared with the search service, the service is a new product offering that leverages Baidu’s existing customer base and creates a new revenue stream. § Baidu’s breadth of Internet services should help to maintain its strong user base. While the non-search services may not be profitable as standalone businesses, they are important in retaining Internet users on Baidu properties and serve as important traffic sources for Baidu’s advertising business. § We estimate net profit growth will return to 30%+ in 2014 as revenue growth remains robust on stable blended margins. This implies ROAE for 2013-2015E of 32.9%/31.4%/30.7%. Our target price implies 2013-15E P/E multiples of 23.8x/18.1x/13.5x. Initiate coverage of Baidu at BUY Qihoo 360 – Initiate coverage at HOLD with a TP of USD44.2 We believe the company will continue to depend on its navigation website and online gaming for revenue growth. While Qihoo 360 has made significant market share gains in Internet search in recent quarters (14.1% in May 2013), we believe organic 4 BNP PARIBAS Initiate coverage of Qihoo 360 at HOLD 27 JUNE 2013 China Internet Alen Lin growth in search market share will be challenging. In our view, other methods for market share gain would have a negative impact on margins or balance sheet. § Qihoo 360’s navigation site remains the key contributor for its online advertising revenue. As its search service mostly depends on its navigation site for traffic, organic growth of search traffic could be limited as its navigation site is a relatively mature platform with a stable user base. Nevertheless, we believe search revenue growth could be realised by focusing on an expanding number of search advertising customers. We estimate its online advertising revenue growth for 2013-15 to be +65%/+47.6%/20.8%. § Online games has seen strong growth (+72% y-y to 150 games) as Qihoo 360 added new games that drove up the number of paying users by 102% y-y in 1Q13 to 281k; we estimate the number will rise to 405k by the end of 2013, +70% y-y. We expect this to remain a key growth driver for the company. § We estimate net profit growth for 2013-15 of -10.3%/+191.2%/+77.7% on the effects of operating leverage; this implies ROAE of 7.9%/18.5%/24.6%. Our target price implies P/E multiples of 126x/43.2x/24.3x. 5 BNP PARIBAS 27 JUNE 2013 Alen Lin China Internet Connecting users to advertising The various services available online have different business models and offer different functions for users; however, they typically have a common theme – connecting users to advertising. Whether the perceived function is to search for information on the Internet, to consume media content, to socialise, or to shop for merchandise, advertising is a common revenue model across sites. Internet services are mostly about connecting users to advertising; advertising Ad Exchange platforms are a mature segment in the US, with multiple platforms in existence This observation leads us to shift our focus to understanding the business model of online advertising. Taking the search engine industry for example, it is commonly understood that search engines generate advertising revenues by responding with paid-advertisements to relevant search queries; therefore, the more online advertisers and search traffic there is, the greater the revenue potential for search engines. There is, however, another role a search engine platform can play that is less intuitive to an Internet user, commonly known as Ad Exchange or Real-Time Bidding (RTB). Ad Exchange platforms effectively connect online advertisers to a much broader set of Internet users through networks of Internet properties. Such Ad Exchange platforms are already in large scale commercial service in mature Internet markets such as the US, including Yahoo’s Right Media, Google’s DoubleClick, Facebook’s FBX, and independent platforms OpenX Market and AppNexus; these are all established exchanges for buying and selling website advertising space. EXHIBIT 1: Major US Ad Exchange platforms Source: Company websites Ad Exchange – efficiently connecting advertisers to Internet properties The first step in connecting users to advertising is to efficiently connect advertisers to relevant Internet properties, ie websites. An Ad Exchange platform fulfils such a purpose by making a large number of Internet properties available to advertisers; effectively an online market place where Internet properties can auction off their advertising spaces to the highest bidders. Within the Google framework, the bidding is done via its AdSense interface. The value of the service increases when the Ad Exchange platform enables advertisers to better target the audience by finely segmenting the various Internet properties based on their respective user attributes, enabling the serving of advertising to the most relevant audience. Exhibit 2 is an illustration by Google to describe the structure of the Ad Exchange. The buyers are advertisers that are seeking ad space. The sellers are Internet property owners that have ad space available for sale. The buyers become AdWords advertisers and interface with the Ad Exchange through Google’s AdWords interface; this is also knows as a Demand Side Platform (DSP). Advertisers have the ability to select a target audience through the AdWords interface. The sellers interface with the Ad Exchange through Google’s AdSense interface; this is also known as the Supply Side Platform (SSP). AdSense partners make their Internet properties available to the Ad Exchange platform that best fits the attributes of their user profile. Ad Exchange platforms efficiently connect advertisers to available ad space through real-time bidding In our view, while the various sub-segments within the Internet industry have unique business models, a common theme that encompasses much of the industry is the science of connecting users to advertising. More importantly, Ad Exchange can be utilised for various user platforms including PC and mobile Internet, as the Internet properties can include PC websites, mobile apps, video content, etc. This vastly expands the potential for online advertising, along with the number and variety of Internet properties available. 6 BNP PARIBAS 27 JUNE 2013 Alen Lin China Internet In China, Ad Exchange platforms have started to take shape in recent quarters with Baidu’s Baidu Exchange Service (BES), Tencent’s AdX, Sina’s Sina Ad Exchange (SAX), Alibaba’s Alimama, and Taobao’s Taobao Ad Network & Exchange (tanx). We believe Ad Exchange presents a new business opportunity for the Chinese Internet industry and has the potential to become a new growth driver. A key criterion for a viable Ad Exchange platform is having large bases of advertisers and Internet properties that target a wide range of demographics and interests. A critical hurdle is to educate both the advertisers and Internet property owners on the value of the Ad Exchange platform. We believe the Ad Exchange platform is an over-arching business model that encompasses the various sub-sectors of the Internet industry and will continue to displace traditional media as the preferred advertising mechanism. A number of Ad Exchange platforms emerging in China, backed by Baidu, Tencent, Sina, Alibaba, and others EXHIBIT 2: Ad Exchange platform Source: Google EXHIBIT 3: Major Ad Exchanges in China Source: Company websites 7 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin EXHIBIT 4: China RTB advertising technology landscape Source: RTB China, http://www.rtbchina.com/wp-content/uploads/2013/03/China-RTB-Ad-Tech-Landscape_V-March-20131.jpg Monetisation of online advertising Internet property owners participate in the Ad Exchange platform by becoming members in the particular exchange. In the case of Google, they become Google Network Members and use AdSense to put their advertising space on the exchange for sale. Baidu has also started to build up Baidu Union, where union members make their advertising space available. The Internet properties can be websites, mobile websites/apps, and other platforms where Internet content is consumed by users. According to Google, it has over 2m Google Network Members participating in its AdSense programme; in terms of revenue contribution, Google generated 28.5% of its advertising revenue from Google Network Members’ websites (Exhibit 5). Internet properties become members of Ad Exchange networks, such as Google Network and Baidu Union, and make their ad space available through the exchange platforms Baidu has also stated that Baidu Union members number in the hundreds of thousands; in our view, the low base provides significant upside potential for the Ad Exchange business model as there are 13.4m domain names registered in China (end of 2012); in terms of available websites, we expect it to be multiples of the domain names registered. We believe this could be a strategic growth area for the company and is effectively a new product offering for its advertising customers. According to China Internet Watch, total spending on online advertising in 2012 was RMB79.1b vs RMB29.3b for search advertising (Exhibit 6); Ad Exchange platforms can become participants in the RMB50b display advertisement market. Through the Ad Exchange platforms, advertisers bid for advertising spots on various Internet properties. Advertisers can choose placements of their advertisements based on various criteria, such as demographics, location and other specific attributes, enabling advertisers to target very specific Internet users and context, improving the effectiveness of ad placements. Ad Exchange platforms maximise efficiency of advertising pricing through numerous pricing strategies. The following are basic types of pricing structures: 8 BNP PARIBAS Advertisers bid for advertising spots in real time based on various attributes of the websites and Internet users 27 JUNE 2013 China Internet Alen Lin § Cost-per-thousand-impressions (CPM) – bidding price is based on the frequency the advertisement is displayed. § Cost-per-click (CPC) – bidding price is based on the frequency Internet search users click on an advertisement. § Cost-per-acquisition (CPA) – bidding price is based on the frequency an advertisement leads to an actual transaction, or number of conversions. Advertising prices are ultimately determined by how effective the advertisement is, as measured by the click-through-rate (CTR). The serving of advertisements to the most relevant Internet users would maximise the CTR, potentially leading to a higher conversion rate. EXHIBIT 5: Google’s advertising revenue breakdown (RMB b) Google websites EXHIBIT 6: China’s online advertising spending Google Network Members' websites 50,000 (RMB b) 90 45,000 80 40,000 70 35,000 60 30,000 Search Ads 50 25,000 40 20,000 30 15,000 20 10,000 10 5,000 0 0 2010 Source: Google Online Ads 2011 2008 2012 2009 2010 2011 2012 Source: China Internet Watch Focus on click-through rate – a measure of effectiveness and value of online advertising A key measure of online advertising’s effectiveness is the frequency with which users interact with a particular advertisement, the click-through-rate (CTR). While the range of CTR can be wide depending on where the advertisement is displayed, this is a practical measure. The CTR is a key determinant for the cost of advertising, as a high CTR would translate into higher advertising revenue for the Internet property owners, and higher probability of converting the impression (display of advertisement) into a transaction for the advertiser. CTR ultimately reflects the quality of the advertisement, and the relevance of the advertisement to the user; as such, advertising platforms’ knowledge of the users can greatly increase the relevance of advertisement to the users. Such characteristics favour platforms that have extensive data on their users. Through the Ad Exchange mechanism, the supply and demand of advertising space would determine the price of advertising placements in various contexts. According to discussions on Google’s AdWords community, the CTR for a typical Internet search generated advertisement could be in the range of 1-5% (Exhibit 8). The same advertisement placed through “display advertisement” on typical websites could be <1% (Exhibit 7 – display ad in top right corner of a web page); however, the number of impressions is generally much greater. The difference in CTR is because advertisements generated through Internet search are effective in targeting users that have interests in the specific topic as the search is user initiated. The same comparison can be made in other sub-sectors such as e-commerce; an advertisement generated through a user’s merchandise search is likely to have significantly higher CTR than a display advertisement that is served through a generic shopping portal. 9 BNP PARIBAS Effectiveness of advertising measured by frequency with which users click on specific ads (clickthrough-rate); a key measure for pricing online ads Internet search ads have higher CTR than display ads; however, display ads deliver a greater number of impressions 27 JUNE 2013 Alen Lin China Internet EXHIBIT 7: Sample of common display advertisement Source: hao123.com EXHIBIT 8: Sample of common search advertisement Source: Baidu Visualising the Ad Exchange auction process We use an auction example of Ad Exchange to illustrate the mechanics of online advertising placement. Referencing Exhibit 9, when an Internet user visits a website that belongs to a Supply Side Platform (SSP), ie, Baidu Union or Google Network Member, information of the user is sent to the Data-Management Platform (DMP) of the Ad Exchange. This may include the user profile and any relevant information about the user that may be extracted from the cookies stored on the user’s web browser. Based on the website visited by the user, user profile, cookies, and other contextual information, the DMP determines the advertisements that are most suitable for the user in the current context. Information taken into account by the advertiser during a bidding process includes attributes of the website, Internet user profile, tracking cookies, and other contextual information On the other side of the Ad Exchange platform, the Demand Side Platform (DSP), ie, the platform on which advertisers bid for advertising space, has bidding information for the specific advertising space. The Ad Exchange takes into account the bidding price and sometimes Quality Score for the particular advertisement based on historical CTR, to determine the winner of the particular auction; Quality Score is a quantitative measure by the Ad Exchange to gauge the effectiveness of an ad. In this auction example, Advertiser-1 wins the auction at USD8, assuming the three advertisers have similar Quality Scores; the advertisement is displayed on the Internet property for the user to see instantly. If the auction was done based on CPM pricing, the advertiser would pay the price per impression for the advertisement. For CPC pricing, the advertiser would only pay if the user clicks on the advertisement, and similarly for the CPA pricing model. Regardless of the pricing model, the underlying determinant of the price is the probability of converting an advertisement into a transaction for the advertiser. 10 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin EXHIBIT 9: Ad Exchange auction process Demand Side Platform (DSP) Supply Side Platform (SSP) Internet user Winner gets the placement – advertisement-1 is placed on the web page Heads to a page on a publisher’s website Advertiser-1 Bids $8 Winner of this auction Web Page Sends user profile and associated set of attributes (cookies) Ad Exchange Platform Advertiser-2 Bids $3 Targets advertisers based on user profile and attributes Sends out bid requests Advertiser-3 Bids $5 Source: BNP Paribas A self-optimising multi-platform advertising mechanism The structure of the Ad Exchange platform is conducive for adapting advertisements for wide-ranging types of device platforms and user segments; the exchange enables different types of advertisements and different types of Internet properties to come together. As the price of the advertising service will reflect the effectiveness of the ad placements on various platforms in an efficient exchange, we believe the ecosystem will self-optimise to maximise value creation. In our view, such self-optimising occurs when the supply and demand for advertising reaches critical mass, enabling full efficiency of the platform. Ad Exchange platforms enable various types of advertisements and Internet properties to come together efficiently; potentially cutting across multiple user access devices This could be an important mechanism for mobile advertising to become a more meaningful contributor to the overall online advertising Industry. Given a more fragmented mobile application environment, where apps dominate user access to mobile Internet as opposed to the typical browser access on a PC, we believe the Ad Exchange advertising structure will play a very important role in the serving of advertisements to users through various mobile apps. Such a mechanism enables app developers to tap into the vast amount of advertising demand available. Ad Exchange just getting off the ground in China Globally, the Ad Exchange segment has been growing rapidly; total Ad Exchange display ad spend for 2012 reached USD2.97b, +119% y-y (Exhibit 10). IDC forecasts a four-year CAGR through 2016 of 47%, reaching USD13.9b. As US companies have been the pioneers of platforms, the US represented 73.4% of global Ad Exchange revenue in 2012; this is expected by IDC to decline to 64% by 2016 as other markets begin to adopt the business model. Based on IDC estimates, by 2016, Western Europe will represent 18.1% of global revenue and China 4.5%. 11 BNP PARIBAS Global Ad Exchange CAGR through 2016 forecast at 47%; US currently the key market, representing 73.4% of global revenue in 2012 27 JUNE 2013 China Internet Alen Lin EXHIBIT 10: Worldwide Ad Exchange display ad sales Worldwide total RTB-based display ad sales (LHS) (USD b) Growth (RHS) 16,000 EXHIBIT 11: Contribution to Ad Exchange display ad sales United States (y-y) 100% 250% 90% 200% 12,000 10,000 150% 8,000 100% 6,000 China Others 2014 2015 2016 70% 60% 50% 40% 30% 4,000 50% 2,000 20% 10% 0 0% 2010 2011 2012 2013 2014 2015 2016 0% 2010 2011 2012 2013 Source: IDC Source: IDC The US is by far the largest market for Ad Exchange, with 2012 total industry ad spend of USD2.18b, +105% y-y (Exhibit 13). Major Ad Exchanges include Google’s DoubleClick, Yahoo’s Right Media, AppNexus (affiliated with Microsoft), and OpenX Market (unaffiliated with major Internet entities). In 2012, Facebook launched its Facebook Exchange (FBX) to monetise its extensive user base. Of all display advertising spend, the share of Ad Exchange spend is forecast by IDC to rise from 15.7% in 2012 to 32.6%% by 2016, indicating the increasing importance of Ad Exchange platforms. EXHIBIT 12: US total display ad sales (USD m) (y-y) 30% US total RTB-based display ad sales (LHS) (USD m) y-y growth (RHS) 10,000 25,000 25% 20,000 20% 15,000 15% 10,000 10% 5,000 5% 0 0% 2010 2011 2012 2013 2014 US Ad Exchange spend contributed 15.7% of total display ad spend in 2012, forecast by IDC to rise to 32.6% by 2016 EXHIBIT 13: US Ad Exchange display ad sales US total display ad sales (LHS) Growth (RHS) 30,000 2015 250% Share of total display ad sales (RHS) 8,000 200% 6,000 150% 4,000 100% 2,000 50% 0 0% 2010 2016 2011 2012 2013 2014 2015 2016 Source: IDC The Ad Exchange market in China is just starting to get off the ground, with 2012 revenue of USD21m, 0.7% of the total market display ad spend of USD2.85b (Exhibit 14). IDC forecasts a four-year Ad Exchange ad spend CAGR of 145% through 2016, reaching USD627m by 2016. Despite the strong growth forecast for the next few years, IDC expects Ad Exchange display ad revenue to still only represent 11.1% of total display ad revenue for the market; we believe there could be further upside from IDC’s forecast. We see great potential for the online advertising category as Ad Exchanges become more efficient and advertisers and Internet property owners become more educated about their effectiveness. 12 Japan 80% 14,000 Source: IDC Western Europe BNP PARIBAS China Ad Exchange spend contributed 0.7% of total display ad spend in 2012, forecast by IDC to increase to 11.1% by 2016, CAGR of 145% 27 JUNE 2013 China Internet Alen Lin EXHIBIT 14: China total display ad sales EXHIBIT 15: China Ad Exchange display ad sales (RMB m) China total display ad sales (LHS) (y-y) 6,000 Growth (RHS) 35% 4,000 700 25% 600 300% 500 250% 400 200% 300 150% 200 100% 5% 100 50% 0% 0 20% 15% 2,000 10% 1,000 0 Source: IDC 13 2012 2013 2014 2015 Share of total display ad sales (RHS) 30% 3,000 2011 y-y growth (RHS) (RMB m) 5,000 2010 China total RTB-based display ad sales (LHS) 2016 350% 0% 2012 2013 2014 2015 2016 Source: IDC BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin Internet search engine – a high-performance subset of Ad Exchange We view the Internet search engine industry as a subset of the Ad Exchange mechanism, as search engines are capable of serving highly targeted advertisements to Internet users with particular interest in a subject. As discussed in earlier sections, advertisements served through Internet search platforms have significantly higher CTR than display advertisements. This makes large Internet search platforms highly valuable Internet properties for advertising purposes. Internet search engines enable the targeting of Internet users with relevant ads, leading to high CTR with potential to extend into new products In markets outside of China, Google is the recognised leader in Internet search. In the US, it represented 66.5% of search market share in April 2013, far ahead of the second largest search platform, Microsoft’s Bing with 17.3% market share (according to comScore). Google’s success, however, has not been replicated in China, where Baidu remains the dominant search platform with 70% market share (by usage) in 1Q13; we attribute Baidu’s dominance to its strength in Chinese language search and government policies that potentially put Google in a weaker market position in China. We believe Baidu will continue to be a dominant search platform for the China market, given the strong positive feedback loop effect of the search industry (Exhibit 19). Beyond the core Internet search activities of the existing search platforms in China, we also expect them to begin to develop the Ad Exchange business model and increasingly become broader online advertising platforms. This should create new revenue opportunities for Chinese search platforms as they already have existing online advertising customers and have the potential to rapidly tap into new inventories of Internet properties through Ad Exchanges. In 2012, Google Network Members generated 28.5% of Google’s total advertising revenue; Baidu still depends on its core search business for over 90% of revenue. China Internet search engine – industry still on a strong upward trend China’s Internet penetration has risen from 16% at the end of 2007 to 42.1% in 2012; over the same period, Internet search industry revenue has increased from RMB2.9b to RMB29.3b. The five-year CAGR for China’s Internet user growth through 2012 was 21.8%, while the search engine industry revenue CAGR was 58.8% (Exhibit 16). This implies that while the number of Internet users in China rose rapidly, revenue generated per user also rose from RMB13.8 to RMB52. This indicates that the industry has already hit an inflection point; in the early phase of the search engine industry, the number of Internet users was the primary driver of search industry revenue; however, the main growth driver has now shifted to search revenue per Internet user, driven by the increasing number of online advertisers. While Internet penetration growth in China has been impressive in recent years, we see no signs of this tapering off, as penetration of 42.1% at the end of 2012 remains low compared with that in mature markets such as Japan and Korea; both are at c80% penetration, and Taiwan, Singapore, and Hong Kong are at c75% penetration (Exhibit 17). The US has the lowest penetration of the mature markets at 78%. We believe China’s aggressive upgrades of its fixed-line broadband network in recent years will continue to drive up Internet penetration; in addition, the rise in smartphone adoption presents incremental opportunities for the Internet search industry. In our view, these trends set the foundation for the continuing growth of China’s search engine industry. 14 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin EXHIBIT 16: Number of Internet users and search industry revenue EXHIBIT 17: Internet penetration comparison (2011) (Percentage of individuals using the Internet) Number of Internet users (LHS) 100 China 10 5 0 0 2007 2008 2009 2010 Sources: CEIC; China Internet Watch; BNP Paribas 2011 2012 Korea 15 200 Germany 20 300 France 25 400 Japan 30 U.S. 500 Hong Kong 35 Taiwan 600 90 80 70 60 50 40 30 20 10 0 Singapore Search industry revenue (RHS) (%) (RMB b) Spain (m) Source: ITU Industry growth should continue to outpace Internet user growth In our view, the increasing maturity and efficiency of the Internet search industry provide the fundamental support for search industry revenue growth to outpace Internet user growth in China. We illustrate the value creation process of the industry in Exhibit 18. The products being sold by search engines are their inventories of user-specific search queries. Based on the user profiles and activities, search engines create value by funnelling users with specific profiles to paying advertisers that seek a specific audience. The ultimate customers of the search engine industry are the advertisers that promote their products and services to targeted sets of Internet users. Search engines’ positive feedback loop, with increasing search users and increasing advertisers, should continue to drive industry value Internet search industry dynamics closely follow the academic definition of an information economy, where at the macro level the value of the industry increases exponentially as the number of industry participants increases. As illustrated in Exhibit 19, the rising number of Internet users drives growth of online advertisers. This leads to rising quality of search results, which in turn drives the growth of Internet users; as the numbers of products (Internet users and Internet properties) and customers (advertisers) rise, the probability of matching the right product to the right customer also rises, leading to higher revenue per product offered. Specific search engines also play an important role in raising the quality of search through enhanced profiling of Internet users. This can be achieved by collecting indepth data about the particular user and analysing the data to better understand and predict a user’s interests, habits, and behaviour. Such data collections and analytics can further enhance the matching of Internet users to paying advertisers, leading to higher conversion rates, or click-thru-rates (CTR). This provides more potential for raising the value of the industry as advertisers are willing to pay more to gain exposure to the relevant audience. Search engines create incremental value by applying user data and analytics in search algorithms to maximise CTR In our view, the combination of a rising number of Internet users and improving search quality through data analytics provides a solid foundation for continuing industry revenue growth, as evident in Exhibit 16. 15 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin EXHIBIT 18: Internet search engine value chain Inventory: search queries and user profiles Value Creation Search Engines Advertising Distributors Advertisers Source: BNP Paribas EXHIBIT 19: Positive feedback loop for Internet search value chain Rising number of Internet users drives higher number of advertisers Number Internet users Number advertisers Rising number of advertisers drives higher number of Internet users Source: BNP Paribas Industry outlook – no sign of a slowdown To analyse the relationship between search industry revenue and general Internet usage, and to quantify industry potential, we have developed a regression model. As the model is based on historical data, the potential contributions from Ad Exchange and mobile Internet are not fully factored in as they are still new business models in China; they could offer upside potential as they gain momentum. We develop a regression model to project search industry revenue growth – we forecast a 2012-15 CAGR of 31.9% Our regression model forecasts an industry revenue CAGR of 31.9% between 2012 and 2015, vs 58.8% for the five years between 2007 and 2012, as the comparison base has grown significantly in recent years (Exhibit 20). A major component of this is search revenue per Internet user, for which we forecast a CAGR of 25.8% over the period vs CAGR of 30.3% over the past five years. This is consistent with our earlier discussion that the rising numbers of Internet users and online advertisers will drive growth in advertising revenue per user. This is also consistent with Baidu’s disclosure that revenue per online advertising customer rose to RMB37,300 in 2012, from RMB5,800 in 2006, a CAGR of 45.1%. Our regression analysis indicates that industry revenue is most sensitive to Internet search user penetration and the number of Internet search users, while the hours of Internet use per week, seasonality, and number of Internet users are also strong predictors. The number of mobile Internet users yields the lowest t-ratio in our model, suggesting that the growth of mobile Internet users has so far not contributed significantly to industry revenue. This could be a new growth driver as the industry finds mechanisms for monetising mobile Internet usage; we believe the Ad Exchange platform could be a powerful tool with which to more effectively monetise mobile Internet. 16 BNP PARIBAS Regression analysis indicates high sensitivity to search user penetration and number of search users; Internet usage time, seasonality, and number of Internet users also important 27 JUNE 2013 China Internet Alen Lin EXHIBIT 20: Search industry revenue forecasts (RMB b) Search industry revenue (LHS) 40 Search revenue/Internet user (RHS) (m) 60 35 50 30 40 25 20 30 15 20 10 10 5 2H15E 1H15E 2H14E 1H14E 2H13E 1H13E 2H12 1H12 2H11 1H11 2H10 1H10 2H09 1H09 2H08 0 1H08 0 Sources: BNP Paribas estimates, CEIC and China Internet Watch Industry value drivers – combination of search users, advertisers, and analytics A search engine is effectively a factory that packages its inventory, the Internet search users, and sells this inventory to its customers, the advertisers. The search engine’s role ends when it connects the inventory to the customer. Any interactions or transactions that occur beyond this point are made outside the initial connection created by the search engine. There are a number of factors that affect the probability of an interaction taking place after the initial search connection, including the size of the inventory, the number of customers, and the analysis performed by the search algorithm that matched a particular search query/user to a particular advertiser (Exhibit 21). The value of the inventory depends on the quantity and quality of search users (subjective measure used by advertisers, based on user profile and attributes); higher quality search users increase the probability of a transaction between the user and advertiser. Search engines create value when inventory – search users – are sold to customers – the advertisers EXHIBIT 21: Internet search industry value drivers Sourcing/inventory of Internet search users Direct search queries www.baidu.com www.so.com www.sogou.com www.soso.com www.google.com Navigation sites Web/news portals www.sina.com.cn news.sogou.com news.soso.com hao.360.cn www.hao123.com Social Networking Internet utilities Email applications E-Commerce sites Video/entertainment sites Web browsers Etc. Search Engine Distribution of advertising service Advertising distributors Advertisers Advertisers Advertisers Buyers of advertising service Source: BNP Paribas 17 BNP PARIBAS 27 JUNE 2013 Alen Lin China Internet The battle to dominate Internet search traffic A starting point for a search engine service is to maintain a stream of search queries, ie, having a large pool of inventory; this maximises the traffic for which advertisers pay. This can be done through capturing direct search queries through search engine websites such as google.com, baidu.com, so.com, and others, or through other Internet services (Exhibit 22). While direct search queries at a search engine website are the most direct mechanism for search engines to acquire users, traffic tends to flow to the dominant search engine brands and be less effective for the secondary search engine brands. Secondary search engines must utilise other mechanisms to acquire search traffic. Aside from direct search queries, search engines tap into other sources of search traffic such as Internet navigation sites including Baidu’s hao123.com and Qihoo 360’s hao.360.cn (Exhibits 22-23). These navigation sites are collections of Internet resources that users frequently utilise to navigate the Internet and often serve as the starting point for users. According to statistics compiled by China Internet Watch, hao.360.cn contributed 11% of search engine traffic in November 2012, while hao123.com contributed 8.2% (Exhibit 26). Search users reach search engines through direct access or traffic referrals; Internet navigation sites contribute a significant amount of search traffic Popular navigation websites are powerful as users frequently utilise the default search engines within these websites. In 3Q12, Qihoo 360 changed the default search engine of its popular navigation website from Baidu to its own (Exhibit 24), and pushed Baidu to a lower position on the list of search engine selections; this had an immediate impact on Qihoo 360’s search engine market share (Exhibit 25). These navigation sites, while serving as a source for search engine traffic, may also be revenue generating as a standalone service. Qihoo 360’s hao.360.cn contributes the majority of its online advertising revenue currently, according to the company. EXHIBIT 22: Qihoo 360 navigation site EXHIBIT 23: Baidu navigation site Source: hao.360.cn Source: www.hao123.com 18 BNP PARIBAS 27 JUNE 2013 Alen Lin China Internet EXHIBIT 24: Qihoo 360 navigation site default search engine EXHIBIT 25: China Internet search market share by page views 100% Baidu Qihoo Soso Google Sogou 80% 60% 40% Chart or other exhibit 20% Source: hao.360.cn 2Q13 1Q13 4Q12 3Q12 2Q12 1Q12 4Q11 3Q11 2Q11 1Q11 4Q10 3Q10 0% Source: cnzz.com Other sources of search traffic come from online video/entertainment, social networking, news portals, and others. Exhibit 27 lists the popular Internet services in China that can serve as sources of search traffic. In our view, the popularity of online video services is a major reason for Baidu owning video service providers iQiyi and PPS; while these services may not be profitable as standalone entities currently, they could be important search traffic sources when integrated into Baidu’s product portfolio. In addition, premium content could generate additional revenues as the online entertainment industry matures. Internet services such as entertainment, social networking and news portals also contribute to search traffic In addition to having a strong position in navigation websites, Qihoo 360 also has a solid market share in China’s web browsers; in recent quarters, Qihoo 360’s Secure Browser maintained a market share of c25%, according to cnzz.com. These browsers, when installed, default the home page to Qihoo’s hao.360.cn unless the user changes the home page, but it also prompts users periodically to set Qihoo 360’s navigation as the home page. In conjunction with Qihoo’s popular antivirus software, Qihoo has a dominant position on Internet users’ desktops. Despite the initial success in gaining market share in 3Q-4Q12, the company’s ability to continue converting navigation site users to its own search engine users is yet to be seen. EXHIBIT 26: Navigation websites’ contribution to search engine traffic Contribution to search engines (Nov 2012) (%) hao.360.cn 11 hao123.com 8.20 web.sogou.com 6.50 123.duba.net 4.20 2345.com 1.40 hao.uc.cn 0.90 Source: China Internet Watch 19 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin EXHIBIT 27: Top Internet services in China Top Internet services in China (Jan 2013, million users) Search service 454.0 Online video 452.0 Community friends 439.0 News information 413.0 E-commerce 404.4 Entertainment 403.8 IT digital products 395.0 Media domain 362.1 Financial service 361.5 Email 356.0 Source: China Internet Watch Expanding online advertising footprint Another key value-driver for the search engine industry is its advertising footprint. While search engines are designed to operate without geographical boundaries, a search engine service provider’s ability to monetise its service depends on its ability to distribute its advertising service. The search engine service provider has the ability to reach advertising customers through direct channels or via advertising agencies that act as distributors of the service. Expansion of advertising footprint critical to monetising search services; search engines leverage direct sales force and ad agencies to expand footprint China’s online advertising market continued to expand in recent quarters, with 1Q13 growing 39% y-y as advertisers increasingly shifted marketing budgets toward online media, according to China Internet Watch data. Within the overall online advertising market, search engine remains the largest category of spending, representing 34% of total industry revenue in 1Q13; the category grew 36% y-y (Exhibit 28). The second largest category of online advertising is vertical search, which represented 23% of total industry revenue in 1Q13, +84.3% y-y. While growth of the online industry has been strong, we see no slowdown as the number of Internet users continues to rise and advertisers increasingly find online advertising to be more effective in targeting customers. Despite the benefits of online advertising, we believe the number of advertisers can only grow as fast as the major online platforms can expand and reach them. At the end of 1Q13, Baidu had 410,000 online advertising customers, +28% y-y; Qihoo 360 currently has less than 10,000 as it had just begun to monetise its Internet search business. While Qihoo 360 aims to rapidly increase its search advertising customers, its scale could be limited by its search users. On the other hand, given Qihoo 360’s smaller search market share, its lower advertising price could serve a niche market of smaller advertisers. Both Baidu and Qihoo 360 leverage direct internal sales forces in major markets and external advertising agencies in other markets to reach additional advertising customers. In the case of Baidu, we estimate the top three markets, Beijing, Shanghai, and Guandong, represent over 50% of advertising revenue and are mostly covered by its direct sales force. While Qihoo is still building up its customer base, it is also primarily targeting tier-1 markets with its internal sales force. 20 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin EXHIBIT 28: China online advertising market in 1Q13 Search Engine (RMB b) 25 Video Vertical Search Rich Media EXHIBIT 29: Baidu’s online advertising statistics No of online marketing customers (LHS) Brand Graphic (nos) Other 20 Average revenue per customer (RHS) 40,000 600,000 35,000 30,000 500,000 15 (RMB) 700,000 25,000 400,000 20,000 300,000 10 15,000 200,000 5 10,000 100,000 5,000 0 0 Source: China Internet Watch 0 2006 2007 2008 2009 2010 2011 2012 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 Source: Baidu Improving advertising performance through Internet user profiles In an effort to improve the effectiveness of search results and search advertising, search engines could leverage contextual data about the user in the search algorithm. Such data could include a user’s location, historical Internet activities or other personal data. When utilised in online advertising, this leads to more targeted ads. The popular Internet services Baidu operates could serve the purpose of collecting user data and profiling users; these services include Baidu Knows, Baidu Post Bar, Qunar travel service, and the numerous video services such as iQiyi and the recently acquired PPS. Fine-tuning search performance through extensive user data collection and analytics; breadth of popular Internet services provides an advantage Partnerships with Internet services that have access to detailed user information can also be powerful. Theoretically, search engines such as Baidu could partner with Internet social networking services to exchange user profiles and improve their advertising effectiveness. This could ultimately lead to higher prices for advertising on the particular search platform. 21 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin Competitive landscape in the China search engine industry EXHIBIT 30: Porter's five-forces for Internet search engine Threat of new entrants: Low The challenges for new entrants to reach critical mass in a networked industry inhibits expansion of new entrants as scale in terms of market share is necessary to gain new Internet users and advertising customers. Such dynamics is prohibitive for new entrants to compete against incumbents. Supplier bargaining power: Medium Internet services that feed search users to search engines are the key suppliers; as they are key sources of user traffic, they could be in the position to charge the search engines Through referral fee structure. Others are infrastructure suppliers such as computer servers and network operators which have relatively lower bargaining power. Internal rivalry: Medium Internet search industry conforms closely with the positive feedback characteristics of a networked industry; as the number of users who use its service increases, the value of its service increases. Such dynamic tends to limit smaller rivals to serving niche market segments. Buyer bargaining power: Medium The price of service provided by search engines is generally determined by bidding process where advertising customers bid for search words. The price depends on the search engines’ user market share and its ability to match users to the advertisers. Threat of substitutes: Medium Traditional media are substitutes for online advertising services; however, they have been losing ground to online advertising channels in recent years. Other online channels also compete for advertising budget; they include Internet portals, social networking services, etc. In addition, in mobile Internet environment, browser based search is being displaced by mobile applications. Source: BNP Paribas Top-3 search engines account for over 90% of search traffic The China search engine industry is dominated by the top three search engines, Baidu, Qihoo 360, and Sogou, with a combined search market share of over 90% in the past year, as measured by search usage by cnzz.com. Baidu remains the most dominant player, with over 70% market share in 1Q13; this is a decline from nearly 80% a year earlier. Since August 2012, Qihoo 360 has gained market share to reach nearly 15% as the company leveraged its strong presence on PCs to direct search traffic to its own search engine, instead of to Baidu as it did previously. This was effective for Qihoo 360 as its browser and navigation site have an established user base, and they serve as effective sources of search traffic. The number three search engine, as measured by search usage, Sohu’s Sogou, leverages the many Internet services provided by Sohu, such as sohu.com, as sources of search traffic to achieve a search market share of 7-8%. We expect traffic share among the top three to shift as the two smaller players look for ways to gain traffic share both organically and through partnerships with other Internet services. In May 2013, Qihoo 360 and Alibaba announced a partnership to launch 360.etao.com to position the search service for online shopping; this is intended to both increase vertical search traffic for Qihoo 360 and to raise brand awareness of Qihoo 360 search by leveraging Taobao’s dominance in e-tailing. Higher brand awareness could lead to higher direct search queries via Qihoo 360’s so.com. Qihoo 360 has publicly stated its target of 20% traffic share by the end of 2013. Highly concentrated industry with top three in China representing over 90% of search usage Smaller search engines likely to seek partnerships or acquisitions to increase market share; they could target niche verticals to differentiate themselves from the dominant player It has also been widely reported that Sohu could be seeking to sell its Sogou search engine; in our view, an acquisition could be more complementary to entities other than the existing large search platforms. While the top search engine platforms continue to compete for traffic share, we believe Baidu will maintain its revenue share as its scale has enabled it to gain substantially more online advertising customers than rivals have; Baidu’s 2012 revenue market share was 79% according to China Internet Watch (Exhibit 32). While we believe Qihoo 360 and Sogou could 22 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin gain traffic share in the near term, their ability to fully monetise the service could be more limited and would take time. Beyond the top three, search engines such as Tencent’s Soso, Google, Microsoft’s Bing, and Yahoo have combined market share in China of less than 10%. As we believe the Internet search market strongly exhibits the characteristics of a networked industry, market share gains for smaller players could therefore be limited in the near term. Sourcing of search traffic Key suppliers for the Internet search industry are other Internet service providers that have established user bases, such as Internet portals, navigation sites, social networking services, e-commerce sites, Internet utilities and entertainment sites. Baidu has built up a large number of services with a large user base that enables the company to direct its user base to its own search engine. It also pays search referral fees to partners for search traffic (Exhibit 31); we believe the recent rise in traffic acquisition costs could include the effect of the Ad Exchange product. Other search engine platforms such as Qihoo 360 mostly depend on traffic from their respective navigation sites. However, as discussed earlier, Qihoo 360 is partnering with other Internet platforms such as Alibaba to source search traffic along specific industry verticals. This could imply that the company has exhausted internal resources. While the partnership terms with Alibaba are undisclosed, we believe that as Qihoo 360 aims to become a more significant player, it will increasingly need to source traffic from other platforms in exchange for referral fees. In our view, such an arrangement could be based on Qihoo 360’s ability to successfully monetise its Internet search service. EXHIBIT 31: Baidu's traffic acquisition cost (RMB m) Traffic acquisition cost (LHS) 700 Traffic acquisition cost/Revenue (RHS) 14% 600 12% 500 10% 400 8% 300 6% 200 4% 100 2% 0 0% 1Q10 2Q10 3Q10 4Q10 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 Source: Company Monetisation of search Once the traffic for a search engine reaches critical mass, search platforms need to address the monetisation aspect of the business. The value of search increases in line with the number of Internet search users; this has been shown in empirical data, as illustrated in Exhibit 33, where search revenue per user rises as the number of search users increases. A similar relationship is found for Baidu, where revenue per online advertising customer increases in tandem with the rising number of online advertising customers. We expect such a pattern to continue for the foreseeable future as the perceived utility of a search engine increases from a user perspective; this should continue to drive online advertising revenue in a self-reinforcing cycle. Empirical data has shown search revenue per user increases with a rising number of search users and number of advertising customers In terms of monetising Internet search, market leader Baidu shows a consistent link, with market share by revenue of 78.5% in 2012 vs its page view market share of 77.6%. As the field of search users approaches saturation, we believe the dominant player, Baidu, will shift its focus to extending its advertising reach to grow its advertising customer base, leading to a rise in revenue per search user. In addition, 23 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin its vast amount of user data to which it has access could enable the search engine to deliver more targeted advertisements, leading to higher advertising fees. Qihoo 360 also needs to expand its online advertising customer base, which is currently less than 10,000. Qihoo 360 has partnered with Google, where the former generates search traffic and the latter provides broad exposure to online advertising customers; as such, Google’s revenue market share in 2012 was 15.6%, despite its limited direct exposure to the search users (Exhibit 32). As Qihoo 360 builds up its own online advertising customer base, it plans to reduce its dependence on Google and capture a larger proportion of the search revenue generated. EXHIBIT 32: Search market share by revenue (2012) SOSO 1.5% Youdao 0.3% Others 1.0% Sogou 3.1% Qihoo 360 starting to build up advertising customers to monetise search services; currently rely on partnership with Google EXHIBIT 33: Continuing growth of revenue/search user (m) Number of search users (LHS) 500 Revenue/search user (RHS) (RMB) 70 450 60 400 Google 15.6% 350 50 300 40 250 Baidu 78.5% 200 30 150 20 100 10 50 0 0 2007 Source: China Internet Watch 2008 2009 2010 2011 2012 Sources: China Internet Watch; CEIC; BNP Paribas Maximising search advertising revenue through high performance searches The Internet search industry primarily generates revenue through online advertising. Given the extensive and dynamic information that exists about Internet users and advertisers, search platforms can maximise advertising price/revenue through numerous pricing strategies to improve effectiveness. The pricing mechanism for such advertising is similar to that for Ad Exchange platforms discussed previously, where advertisers bid for advertising spots based on CPM, CPC, and CPA. Industry value can be increased by optimising search quality – ie, using user data analytics to improve relevance of advertising for search users The placement of a particular advertisement is generally determined by the relevance and quality of an advertisement in a particular search query (in Google terms – Quality Score), and bidding prices from advertisers for the particular search query; these two factors combined determine the position of the particular advertisement in the particular search query (Exhibit 7). Advertising prices are ultimately determined by how effective the advertisement is as measured by the click-through-rate (CTR); we believe improving the advertising relevance can improve the CTR, and this is where search platforms can continue to optimise and differentiate. Exhibit 33 provides a sample for CPC prices for different keywords related to “Used cars”. Such industry dynamics should enable search platforms with large amounts of user data to better monetise Internet search services. Data collected through the various Internet services can be used to optimise search results for the particular user, leading to higher a CTR; this would directly lead to higher market prices for advertising placements. As we discussed in an earlier section, CTRs are generally in the range of 1-5% for search advertisements. 24 BNP PARIBAS 27 JUNE 2013 Alen Lin China Internet EXHIBIT 34: Sample keyword bidding prices Source: adwords.google.com Display advertisements – a different type of advertisements Another type of common advertisement on the Internet is a display advertisement. These are advertisements placed on various websites. While a less sophisticated advertisement may be static (like the one shown in Exhibit 7, more intelligence can be built into the placement of the advertisements by taking into account contextual information such as the user profile, geographical location, time of day, etc.; this approach could have the effect of increasing the click-through-rate. As display advertisements are considered to be “push” advertisements – i.e. pushed to the users by advertisers – CTR for typical display advertisement is <1%. While the CTR is lower than for a search advertisement, the frequency of advertising display is significantly higher for display advertisements. Techniques for search ads can be applied to display ads to improve CTR; Ad Exchange platforms enable search engines to expand into display ads Advertising space inventory can be significantly increased when a search platform leverages the Ad Exchange platform to include partner Internet websites. According to Google, it has over 2m partners in its Google Network Members; Baidu also has partners in its Baidu Union in the range of hundreds of thousands. We believe the ramping up of Baidu Union is a major step for Baidu to build up an Ad Exchange platform as efficiency of the platform increases with the increasing number of participants. We view the Ad Exchange platform as a new product for search engine platforms that has implications on future growth potential. Mobile Internet – the next frontier As the number of mobile Internet users has grown rapidly in recent years with the rise of smartphones and mobile broadband services in China, Internet companies are still adapting their business models to tap into this new opportunity. Exhibit 35 shows the number of overall Internet users 5-year CAGR through 2012 was 21.8% vs. the number of mobile Internet users CAGR of 52.8%. While the rise of mobile Internet usage presents new opportunities to Internet companies, it also poses challenges as Internet companies have to adapt to mobile Internet monetisation. The growth of mobile Internet users has far outpaced Internet companies’ ability to monetise mobile Internet users’ utility for the services. Access to mobile Internet more fragmented than PC Internet; Ad Exchange platforms can also be adapted for service mobile ads Internet companies that have optimised their products for PC Internet advertising are still adjusting their products for the mobile Internet environment, where a major limitation is the real estate available on much smaller screens compared to PC monitors. Taking search engines for example, for each search query a number of paid-advertisements can be served to the PC user; however, on a mobile device limited space is available to serve advertisements. Therefore, for mobile Internet 25 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin advertising to be meaningful the frequency of advertisements must be increased; this would require Internet companies to expand advertising inventory across multiple applications on mobile devices. The user behaviour for mobile Internet is very different from PC Internet. On PCs, users mostly access the Internet through web browsers; this is how most Internet companies that rely on advertising have developed their products. On a mobile device, users are more likely to access information using applications for specific activities; in many cases Internet companies that have succeeded in PC Internet have limited presence in mobile applications. Based on comScore data for May 2013, browsers only represented 20% of smartphone and tablet usage based on time spent. A 2012 Google survey of China’s smartphone users indicated that only 34% of users noticed advertisements appearing in an app; users mostly noticed advertisements on mobile devices while using search engines (49%) and on a website (54%). In our view, this suggests a significant opportunity for in-app advertisements. Exhibit 36 is a summary of popular services being used by China’s mobile users where instant messaging, mobile search, news, Weibo, and music round out the top 5 applications. Browser access on mobile Internet only represents 20% of usage time; only 34% of users notice ads in mobile apps – presenting significant opportunities In our view, the Ad Exchange could be an effective platform for mobile advertising to proliferate as mobile applications can become members of the Ad Exchange network just like other Internet properties and make their mobile applications available through the Ad Exchange. While the business model for mobile advertising still needs to be optimised, we believe the potential exists as mobile devices offer significantly more contextual information about the user that can be used for more targeted advertising. In our view, this places even more importance on the Ad Exchange platform we discussed in earlier sections as it has the potential to serve targeted contextual advertisements to mobile users across multiple applications. 0 2006 Source: CEIC 26 2007 2008 2009 2010 2011 2012 Post & comment 100 Mobile e-mail 200 Network video 300 Social network 400 Network literature 500 Network music 90 80 70 60 50 40 30 20 10 0 Weibo (%) Total number of mobile Internet users News browse Total number of Internet users Instant messaging (m) 600 Mobile search EXHIBIT 36: China mobile users’ usage of mobile apps (April 2013) EXHIBIT 35: Number of China Internet users Source: China Internet Watch BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin Regulatory control; a significant force in the search engine industry The Internet is a very effective platform for removing geographical boundaries for various business models, and also a very efficient medium for the dissemination of information in real-time. As such, the Internet industry in China is subject to complex supervision by various regulatory entities of the Chinese government; Exhibit 37 illustrates the full scope of regulatory framework for the industry. Internet search engines regulated by 13 agencies covering numerous Internet activities Industry regulations cover a wide variety of activities including the following: § Cultural activities § News reporting § Payment services § Internet publishing, online games, and online game virtual currency § Value-added telecommunication services § Audio and video broadcasting § Internet map services § Advertisements § Intellectual Property Rights § Information security As these regulatory areas fall under various government agencies, the industry has to comply with all relevant rules to ensure full compliance. In our view, the area that most concerns the Chinese government is the dissemination of information, which is often subject to censorship. Internet censorship in China China’s Internet censorship comprises over sixty Internet regulations which have been executed by provincial branches of state-owned Internet service providers, companies, and other organisations. These authorities are involved in blocking website content and monitoring Internet access of individuals in order to eliminate leakage of illegal information on state security and public interests, and also curb Internet pornography and other socially disruptive content. In addition to citizens, foreign individuals and firms are also brought under the purview of these laws. The three main Internet censorship regulations of China include: § Temporary Regulation for the Management of Computer Information Network International Connection: The regulation mandates that; a) Internet service providers should be licensed, and b) that Internet traffic should be routed through ChinaNet, GBNet, CERNET or CSTNET. § Ordinance for Security Protection of Computer Information Systems: According to the ordinance, the Ministry of Public Security will have complete responsibility for Internet security protection. § Security Management Procedures in Internet Accessing: "Harmful information" and "harmful activities" regarding Internet usage are defined in this regulation. Internet censorship in China comprises over 60 regulations that are executed by state-owned provincial ISPs Content restrictions became effective from 2000 In September 2000, content restrictions for Internet content providers were first created by the State Council, including: § China-based websites require separate approvals to link to overseas news websites or distribution from overseas media. § Only “licensed print publishers” are authorised to deliver news online. 27 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin § Non-licensed websites can only publish/disseminate information already available in the public. § The approvals should be attained from state information offices and the State Council Information Agency. After the 2001 APEC summit in Shanghai, overseas news websites including CNN, NBC, and the Washington Post that were blocked previously were relaxed permanently except for special occasions. State Council Information Office established separate authorities to manage Internet content In 2011 the State Internet Information Office and the Internet News Coordination Bureau were established by the State Council Information Office to manage Internet content more effectively. The authorities enforce restrictions largely through Internet service providers and impose direct penalties in case of violations. Hence, the service providers have been entrusted with the responsibility to monitor consumer content. State Internet Information Office and the Internet News Coordination Bureau were created specifically to regulate Internet content Extent of Internet censorship in websites Search engines Generally in search engine censorship, search results of certain terms are filtered leading to little or no results. Some of the examples that can be cited include: § Searching for information about “Falun Gong”, “1989 protests in Tiananmen Square”, “Urumqi riots”, and “Tibetan independence”. § Search engine Bing censored search results in mainland China, but not in Hong Kong, Macau and Taiwan. Google started blocking websites which the Chinese government considered illegal in February 2006, in exchange for remaining in mainland China. However, in 2010 Google stopped censorship and closed Google China. Currently, search traffic for Google is routed to the Google Search Hong Kong site which is uncensored; however, sensitive keyword searches continue to be blocked for Chinese users. In support of the Chinese government’s censorship mandate, Baidu search engine has been the most compliant, given that it has a long list of censored topics and blocked websites. Discussion forums According to British Standpoint magazine, the Chinese government mandated a shutdown of Internet portals, forums, and discussion boards for maintenance ahead of the twentieth anniversary of the Tiananmen Square event. Due to the outage, Chinese Internet users were unable to access popular websites like Twitter, Hotmail and Flickr, among others. Social media websites § In spite of the government using Sina Weibo (micro blogging service) for disseminating information, the content is supervised and censored by various self-censorship policies and methods. § According to an article from the Harvard Law School – In Washington Post op-ed, Berkman Center directors discuss Twitter revolution in Iran on 22 June, 2009 – social media websites such as Facebook and Twitter have been blocked since the second half of 2009 due to violations of social or political commentary rules. § Generally, messages agitating collective action are blocked. For example, discussions around Chinese human rights activist Liu Xiaobo, the Nobel Peace Prize winner, were prohibited. However, discussions around sensitive topics and controversies such as the Li Gang incident, Wenzhou train collision, and 2010 Shanghai fire incidents on Weibo increased. 28 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin EXHIBIT 37: Regulatory framework for Internet industry The Ministry of Culture The State Council News Office The Ministry of Commerce The State Food and Drug Administration People’s Bank of China General Adminis-tration of Press and Publication (GAPP) MIIT The National People’s Congress PRC State Council The State Administration of Radio Film and Television National Administration of Surveying, Mapping and Geo-information The State Administration for Industry and Commerce • Regulations on Internet Culture Activities • Regulations on Online Game Virtual Currency • Regulations on News Display • Restrictions on Foreign Ownership in ValueAdded Telecommunications Services • Regulations on Internet Drug Information Services • Regulations on Payment Services by Nonfinancial Institutions • Regulations on Internet Publishing • Regulations on Online Games • Regulations on Value-Added Telecommunications Services and Internet Content Services • Regulation on Broadcasting Audio/Video Programs through the Internet • Regulations on Internet Map Services • Regulations on Advertisements • Regulations on Intellectual Property Rights The State Intellectual Property Office The State Secrecy Bureau • Regulations on Information Security The Ministry of Public Security Source: Baidu 29 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin Ad Exchange favours participants with large user bases and advertisers While Ad Exchange is a potential new product for many Internet players, we believe the business model favours players with either large user bases or large online advertiser bases. The large number of Internet properties available through partnerships such as the Baidu Union ultimately translates to expansion of the user base. Within the search engine industry, we believe Baidu has a solid lead in key measures compared to smaller competitors. In our view, this puts Baidu ahead of the pack for building a large scale Ad Exchange platform among search engine players with its Baidu Exchange Service. Large search engine platforms leverage existing advertising customer bases to expand into Ad Exchange market We also believe industry vertical specific Ad Exchange platforms can exist as they cater to specific market segments, both for users and Internet properties. One such industry vertical could be the e-Commerce segment where Alibaba/Taobao leverages their respective user base. However, we believe their core strengths in user base and Internet properties are primarily in the e-tailing space in the early stage as this is the core segment of their underlying business. In our view, Ad Exchange platforms that target different market segments can co-exist in the long-run. We believe Baidu Exchange Service holds significant potential for Baidu as it leverages its existing core strengths. More importantly, not only does BES expand Baidu’s inventory of PC Internet properties, it could also enable Baidu to tap into mobile Internet by including mobile apps in Baidu Union. While it would still take time to educate the market, and for the platform to adjust for advertising on mobile apps, we believe the efficiency of the Ad Exchange will self-optimise based on supply and demand. EXHIBIT 38: Valuation comparison BBG Share Mkt ------ PE ------ --EV/EBITDA-- ---- P/BV ---- code price cap 2013 2014 2013 2014 2013 (LC) (USD m) (x) (x) (x) (x) (x) 5.4 ----- Yield ----- ----- ROE ----- 2014 2013 2014 2013 2014 (x) (x) (x) (x) (x) 4.0 0.0 0.0 28.4 27.2 BAIDU INC - SPON ADR* BIDU US 113.0 31,435 19.2 14.6 14.7 10.7 QIHOO 360 TECHNOLOGY CO-ADR* QIHU US 44.2 5,336 125.7 43.2 67.6 26.3 9.2 7.1 0.0 0.0 7.3 16.3 SINA CORP SINA US 52.6 3,504 75.5 32.0 47.8 20.2 2.6 2.4 0.0 0.0 4.5 9.1 SOHU.COM INC SOHU US 58.9 2,252 21.4 15.7 5.4 4.3 1.8 1.7 0.0 0.0 8.4 10.4 700 HK 283.4 67,654 25.6 21.0 18.3 15.3 7.5 5.7 0.3 0.4 33.0 30.1 TENCENT HOLDINGS LTD NETEASE INC-ADR NTES US 56.2 7,343 10.4 9.2 6.5 5.5 2.3 1.9 0.9 0.9 24.1 21.8 GOOGLE INC-CL A GOOG US 869.8 288,567 18.8 16.2 11.1 9.4 3.4 2.9 0.0 0.0 17.8 17.7 YAHOO! INC YHOO US 24.1 26,054 18.2 16.4 13.1 12.5 1.9 1.7 0.0 0.0 10.5 10.1 YOUKU TUDOU INC YOKU US 18.0 2,961 NA 51.2 NA 23.9 2.0 1.9 0.0 0.0 (3.4) 5.0 RENREN INC-ADR RENN US 2.8 1,042 Internet service - Mcap Weighted Average (x) NA NA NA NA 1.0 1.0 0.0 0.0 (5.7) (2.5) 20.0 17.3 11.3 9.6 4.1 3.3 0.0 0.0 20.8 20.2 As at 25 June 2013 Sources: * BNP Paribas estimates; all others (Not rated) are Bloomberg consensus estimates 30 BNP PARIBAS 27 JUNE 2013 27 JUNE 2013 INITIATION / INTERNET CHINA BAIDU BIDU US BUY TARGET PRICE USD113.00 CLOSE HOW WE DIFFER FROM CONSENSUS MARKET RECS TARGET PRICE (%) 7.66 POSITIVE 15 EPS 2013 (%) (2.8) NEUTRAL 11 EPS 2014 (%) 2.6 NEGATIVE 4 USD91.09 UP/DOWNSIDE +24.1% KEY STOCK DATA Beyond search advertising n n n Profit growth to return in 2014: initiate with BUY We initiate on Baidu, the largest web services company in China, with a BUY and TP of USD113. Baidu’s Internet search market share will remain dominant. We expect revenue growth to remain strong at >30% y-y through 2015, and NP growth to return to +31.6% in 2014. Growth beyond PC search advertising We believe Baidu will increasingly tap into display advertising with its Baidu Exchange Service (BES). Leveraging its Baidu Union and advertising customers, we believe BES enables Baidu to tap into vast display ad opportunities and mobile internet monetization. DCF-based TP of USD113 Our DCF-based USD113 TP implies 2013-15 PE multiples of 23.8x/18.1x/13.5x. Key downside risks include: 1) intensifying competition in Baidu’s core search segment; 2) new growth drivers such as mobile Internet and BES fails to gain momentum. YE Dec (RMB m) 2012A 2013E 2014E 2015E Revenue 22,306 30,659 40,594 53,851 Rec. net profit 10,456 10,307 13,563 18,144 Recurring EPS (RMB) 299 295 388 519 EPS growth (%) 57.3 (1.5) 31.6 33.8 Recurring P/E (x) 18.7 19.0 14.4 10.8 Dividend yield (%) EV/EBITDA (x) 0.0 0.0 0.0 0.0 15.6 15.0 11.0 7.5 Price/book (x) 7.5 5.4 3.9 2.9 Net debt/Equity (%) (8.0) (37.1) (57.4) (70.8) ROE (%) 50.6 33.1 31.5 30.8 Jun-12 Sep-12 Dec-12 Mar-13 Jun-13 23 142 13 132 3 122 112 (7) 102 (17) 92 (27) 82 (37) 72 n Ad Exchange – an emerging new product segment We believe the Ad Exchange platform will complement Baidu’s existing search advertising business and provide incremental growth for the company. While a proven business model in mature markets, the category currently contributes <1% of China’s display ad spend. (USD) (47) BAIDU Share price performance Rel to MSCI China 1 Month Absolute (%) Relative to country (%) (%) 3 Month 12 Month (5.6) 6.8 (18.0) 9.2 21.8 (16.0) Next results July, 2013 Mkt cap (USD m) 3m avg daily turnover (USD m) Key chart: Continuing growth of revenue/search user (m) 500 31,853 (RMB) 80 Number of search users (LHS) Revenue/search user (RHS) 400 60 383.7 Free float (%) Major shareholder 12m high/low (USD) 70 Robin Yanhong Li (16%) 133.98/83.59 3m historic vol. (%) 300 40 ADR ticker 20 ADR closing price (USD; 25 Jun 2013) 200 100 Issued shares (m) 0 0 2007 2008 2009 Sources: Bloomberg consensus; BNP Paribas estimates Alen Lin [email protected] +852 2825 1801 31 2010 2011 2012 Sources: Bloomberg consensus; BNP Paribas estimates 35.7 BIDU US 91.09 35 BAIDU BIDU US Alen Lin Investment thesis Catalyst We initiate coverage of Baidu with a BUY recommendation and a DCF-based target price of USD113. We forecast Baidu’s 2013 net profit will decline 1.3% as it is investing in new growth areas and net margin will contract on productmix change. However, we forecast profit growth will resume in 2014. We estimate net profit will increase 31.6% in 2014 and 33.8% in 2015 as new products begin to contribute and margins reach a stable and sustainable level of about 33%. We expect revenue growth to remain at over 30% over our three-year forecast period. We believe Baidu is in good position to build up its Ad Exchange platform by leveraging its large base of online advertisers. This would significantly expand the company’s reach to non-search online advertising. In addition, we believe the Ad Exchange platform will enable Baidu to tap into mobile Internet by including mobile apps in the Baidu Union. As contributions from these new product offerings begin to materialize, we believe they will pose upside to the company’s growth outlook. Risks to our call Internet search advertising spending currently represents 37% of total online advertising spending in China. Thus, there is significant upside potential for Baidu as it builds its BES platform to tap into non-search online advertising. Baidu currently generates over 90% of its revenue from Internet search advertising. In our view, key metrics to track are the number of online advertising customers and the number of Baidu Union members; the latter being an indication of Baidu’s inventory of advertising space. Company background Key risks to our BUY recommendation include: 1) more intense competition in Baidu’s core Internet search segment, which could slow search advertising revenue and erode margins; and 2) inability to generate sufficient revenue growth beyond the core search segment including its BES platform, online video services and mobile Internet. Key assumptions Baidu is the largest web service company in China mainly focus on search engine for website. On Aug 5, 2005, Baidu was listed on The NASDAQ National Market with symbol "BIDU". In addition to its core web search product Baidu.com, the company powers a broad range of products and services, including Baidu Postbar (an interactive experience sharing community), Baidu Encyclopaedia and Baidu Map. 2012 Number of online marketing customers Average revenue per customer (RMB) 2013E 2014E 2015E 596,000 507,500 609,000 730,800 37,300 16,895 18,585 20,443 50.5 64.5 66.6 66.6 Total operating costs and expenses/ revenue ratio (%) Sources: Company data; BNP Paribas estimates Principal activities by revenue (2013E) Earnings sensitivity ------- Base ------- Other services, 0.6% 2013E 2014E 2013E 2014E Total revenue 30,659 40,594 33,725 44,653 27,594 36,534 EPS (RMB) 294.74 387.85 321.12 420.80 268.37 354.90 8.9 8.5 (8.9) (8.5) Operating expenses/revenue ratio (%) EPS (RMB) 64.5 66.6 61.3 63.3 67.7 69.9 294.74 387.85 331.90 437.19 257.59 338.52 12.6 12.7 (12.6) (12.7) Change (%) Key executives Change in total revenue (%) 0 0 10 10 (10) (10) Change in operating expenses 0 0 (5 ppt) (5 ppt) 5 ppt 5 ppt Source: BNP Paribas estimates Age Since Title Robin Yanhong Li 44 2000 Chairman and Chief Executive Officer Jennifer Xinzhe Li 45 2008 Chief Financial Officer http://www.baidu.com 32 ------- Worst ------- 2014E Change (%) Online marketing services, 99.4% ------- Best ------- 2013E The key earnings drivers include total revenue and operating expenses/revenue ratio A 10% increase in total revenue raises 2013E/2014E EPS by 8.9%/8.5%; a 10% decrease reduces 2013E/2014E EPS by 8.9%/8.5% A 5ppt fall in operating expenses/revenue ratio lifts 2013E/ 2014E EPS by 12.6%/12.7%; a 5ppt rise decreases 2013E/2014E EPS by 12.6%/12.7% BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin Initiate coverage with BUY on emerging new revenue drivers We initiate coverage of Baidu with a BUY recommendation and a DCF-based target price of USD113. Baidu offers traditional Internet search advertising services, but it is now changing its product mix to tap into new growth opportunities. Internet search advertising spending currently represents 37% of total online advertising spending. Thus, there is significant upside potential for Baidu as it builds its BES platform to tap into nonsearch online advertising. Google’s (GOOG US, Not rated) revenue structure could be a benchmark for Baidu. Google currently generates 28.5% of its advertising revenue from Google Network Members’ websites, while Baidu generates over 90% of its revenue from Internet search advertising. We believe Bloomberg consensus forecasts do not fully factor in the potential from the new product, BES, which could offer meaningful upside in 2014-15. In our view, key metrics to track are the number of online advertising customers and the number of Baidu Union members; the latter being an indication of Baidu’s inventory of advertising space. Changing product mix to tap into new growth opportunities We forecast Baidu’s 2013 net profit will decline 1.3% as it is investing in new growth areas and net margin will contract on product-mix change. But, profit growth should resume in 2014. We estimate net profit will increase 31.6% in 2014 and 33.8% in 2015 as new products begin to contribute and as margins reach a sustainable level of about 33%. We expect revenue growth to remain at over 30% over our three-year forecast period. Internet search is the revenue generator Baidu delivered revenue CAGR of 62.5% over 2008-12. Internet search advertising remains its core business, contributing over 90% to total revenue during this period. The company’s revenue base has grown 6x from 2008 to RMB22.3b in 2012, and we forecast revenue CAGR of 34.2% over 2012-15 (Exhibit 1). We believe online advertising growth will be driven by an increase in the number of advertising customers and higher revenue per advertising customer as Internet users are increasingly dependent on Internet search for information. We estimate the number of advertising customers will increase from 410,000 in 1Q13 to 507,000 by end-2013 and further to 730,000 by end-2015. In our view, the number of online advertisers will continue to grow as they become more aware about the effectiveness of this media channel. Internet search advertising remains Baidu’s core business In addition to Baidu’s traditional Internet search platform for advertising, it is focusing on the Ad Exchange platform as a new product offering for incremental growth. We believe this can significantly expand the company’s inventory of advertising space. Baidu’s increasing emphasis on Baidu Union is setting the stage for the Ad Exchange platform; Baidu Union members make their Internet properties available for ad placements. The increasing availability of advertising space should also help to expand the number of advertising customers. While we do not estimate revenue contribution from the Ad Exchange platform separately, we believe this would become an important segment in the future. 33 BNP PARIBAS 27 JUNE 2013 BIDU US BAIDU Alen Lin EXHIBIT 1: Baidu revenue trend (RMB m) Total revenue (LHS) Revenue growth rate (RHS) (%) 90 60,000 80 50,000 70 40,000 60 50 30,000 40 20,000 30 20 10,000 10 0 0 2009 2010 2011 2012 2013E 2014E 2015E Sources: Company data; BNP Paribas estimates Significant upside potential for Internet search As highlighted in our accompanying sector report, Connecting users to advertising, we believe the Internet search industry will grow strongly. This is because both the number of Internet search users and online advertising customers will continue to rise (the latter directly impacting search advertising revenue), thereby forming a positive feedback loop in driving value for the service. We expect industry revenue growth to outpace both the number of Internet search users and online advertisers. In our regression model for projecting search industry revenue, we estimate industry CAGR of 31.9% over 2012-15, vs 58.8% CAGR over 2007-2012. We assume revenue/Internet user CAGR of 25.8% for our three-year forecast period, implying slower Internet user growth. The regression model is based on historical data, thus it does not take into account fully the potential contribution from new growth drivers such as mobile Internet and Ad Exchange platform, which could provide incremental upside for Baidu. While Baidu has been facing competitive pressure from Qihoo 360 and has lost market share, we believe incremental gains for Qihoo 360 will moderate as organic market-share expansion becomes more challenging. Aside from Qihoo 360, we believe no other viable Internet search platform exists to compete against Baidu’s large scale given the current industry structure. In spite of the recent market share shift, we believe Baidu can remain the dominant Internet search service provider as it has built a strong reputation for its products. In our view, the many popular nonsearch products offered by Baidu can create stickiness for its services. In addition, we believe there is significant market share inertia as the value created by Baidu from the positive feedback loop (discussed in details in our sector report) makes major shifts in market share less likely. Only one viable competitor: Qihoo 360 Baidu Exchange Service – a new product offering While Baidu Union has been around for a few years, we believe it will increasingly become a more significant part of Baidu’s growth driver as it taps into its Ad Exchange platform – Baidu Exchange Service (BES). At a high level, online advertising through BES is similar to Baidu’s traditional Internet search advertising, but it provides a new product line. Internet properties – advertising space inventories – offered to advertisers through Baidu were its own properties previously. With BES, the company is able to offer new inventory from its Baidu Union members to its advertising customers. This vastly expands the amount of advertising space that can be sold. This is a new product and contribution is currently limited, thus we believe the Street has not fully factored in the upside potential for Baidu. Based on IDC forecast, Ad Exchange display ad spend could reach RMB2.5b in China by 2015. Baidu’s ability to capitalize on this opportunity would provide incremental upside. In addition, we 34 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin believe IDC’s forecast may be conservative as it implies that Ad Exchange ad spend would represent just 7.8% of total display ad spend. We view inventories sold through the BES platform as a new product for Baidu, as it acts as an exchange operator rather than a seller of its own Internet properties. While the company’s blended margins will be impacted as the product mix changes, we think this is a natural evolution when offerings are extended (Exhibit 2). A change in product mix would be reflected in Baidu’s traffic acquisition cost as the BES platform leverages Baidu Union members to offer more online advertising space. Because of the product-mix change in 2013, we forecast Baidu’s operating margin will decline from 37% in 1Q13 to 33.7% in 4Q13 (Exhibit 9). Despite the potential decline in blended profit margins, the new product is a new source of revenue. EXHIBIT 2: Baidu product offerings Internet users Internet search users Users of other Baidu Internet properties Users of Baidu Union member Internet properties Search Engine Display Ad Platform Ad Exchange Advertising space/product inventory offered Advertosers Advertisers Advertisers Advertisers Consumers of advertising service Source: BNP Paribas Mobile Internet monetization – potential expansion through Baidu Union As discussed in our sector report, access of mobile Internet tends to be more fragmented than PC Internet where users mostly access it via the browser. This is a challenge to monetization as no one dominates access of the mobile Internet. Baidu is making efforts to make its Internet services more relevant on mobile devices through its mobile apps such as Baidu Search and Baidu Map. The eventual expansion of Baidu Union to include non-Baidu mobile apps will vastly expand the company’s advertising presence on mobile platforms and enable it to monetize mobile Internet, in our view. Making its services more relevant on mobile devices Given the fragmentation of mobile applications, we believe Ad Exchange is an effective platform to aggregate mobile advertising opportunities to make them available to a large set of advertisers. More importantly, we believe a common Ad Exchange platform for PC Internet and mobile Internet will increase the efficiency of the value chain, including that for advertisers, Internet property owners and exchange operators. Baidu has a strong presence in PCs, thus a strong Ad Exchange platform should enable it to leverage its strengths across platforms. We believe the development of mobile advertising will be gradual as advertisers become more aware about the value of mobile advertising. 35 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin Normalizing return on equity We estimate Baidu’s ROAE to approach 30% through our forecast period, from a peak of 55.8% in 2011. We expect net margin to decline primarily due to a change in product mix. The expansion of inventories to include its Baidu Union partners effectively adds a new product offering from its traditional dependence on Internet search advertising. We forecast Baidu’s asset turnover will remain stable at around 50%. The expansion of Baidu’s equity base will lead to decline in its assets/equity ratio; this also contributes to the lower ROE in the future. In our view, this can be enhanced by restructuring its capital structure to lower its equity base. EXHIBIT 3: Baidu's ROAE decomposition (%) 2009 2010 2011 2012 2013E 2014E 2015E NI/Sales 33.4 44.5 45.8 46.9 33.6 33.4 33.7 Sales/Assets 72.2 71.6 62.1 48.8 50.7 51.0 52.2 Assets/Equity 157.0 167.9 196.2 219.7 193.1 184.1 174.8 37.9 53.6 55.8 50.3 32.9 31.4 30.7 ROAE Sources: Company data; BNP Paribas estimates Strong balance sheet and rising FCF yield We forecast Baidu will invest in new business platforms in 2013 and that profit growth will resume in 2014. Thus, the company’s cash position will expand rapidly from a debt/equity ratio of -0.1% in 2012 to -71.3% in 2015E. During the same period, we forecast FCF yield will rise from 5.7% in 2012 to 10.1% in 2015. We assume capex will remain close to 10% of revenue as Baidu continues to expand PP&E to accommodate more Internet traffic. EXHIBIT 4: Baidu's balance sheet (RMB m) 2009 2010 2011 2012 2013E 2014E 2015E Cash & cash equivalents + Bank deposits 4,180 662 7,782 4,127 11,881 23,584 38,995 58,657 1,000 11,721 22,793 23,490 24,295 25,315 4,843 8,782 15,848 34,674 47,074 63,290 83,972 998 1,622 2,694 3,888 5,578 7,801 10,358 6,157 11,048 23,341 45,669 60,261 79,115 102,701 - - 172 2,171 2,171 2,171 2,171 Current Liabilities 1,400 2,552 4,235 6,066 10,398 15,702 21,155 Total Current Liabilities 1,400 2,552 4,407 8,237 12,569 17,873 23,325 LT borrowings 4 86 2,278 9,693 9,693 9,693 9,693 Other LT Liabilities - 5 331 524 524 524 524 Total Liabilities 1,404 2,643 7,015 18,454 22,787 28,090 33,543 Share Capital 1,426 1,557 1,772 2,095 2,095 2,095 2,095 Reserves 3,441 6,966 13,604 24,038 34,309 47,872 66,016 - - 1,034 1,160 1,160 1,160 1,160 Current Assets Total Current Assets Non-Current Assets Total Assets ST borrowings / Current portion of LT borrowings Minority Interest Total Equity 4,753 8,406 15,292 26,055 36,314 49,865 67,998 Total Equity & Liabilities 6,157 11,048 23,341 45,669 60,261 79,115 102,701 Sources: Company data; BNP Paribas estimates 36 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin EXHIBIT 5: Baidu's cash flow RMB m 2009 2010 2011 2012 2013E 2014E 2015E Net Profits 1,485 3,525 6,639 10,456 10,307 13,563 18,144 Depreciation & Amortization 317 441 885 1,515 1,740 2,314 3,070 Change in working cap 392 746 592 781 3,636 4,499 4,432 11 58 464 1,279 0 0 0 2,205 4,770 8,580 14,031 15,684 20,375 25,646 33 67 336 759 1,235 2,400 3,360 Other operating cashflow Operating cash flow Interest Inc / (Exp) Taxes paid (198) (536) (1,189) (1,574) (1,847) (2,393) (3,202) Capex (439) (963) (1,866) (2,340) (3,060) (4,080) (5,100) Free cash flow 1,766 3,807 6,714 11,692 12,624 16,295 20,546 Free cash flow yield (%) 0.9 1.9 3.4 5.9 6.4 8.3 10.4 Dividend yield (%) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Capex/sales (%) 9.9 12.2 12.9 10.5 10.0 10.1 9.5 (87.9) (91.6) (10.6) (8.0) (37.1) (57.4) (70.8) Net debt to equity (%) Sources: Company data; BNP Paribas estimates EXHIBIT 6: Baidu's income statement RMB m Revenue Cost of goods sold Gross profit OpEx 2009 2010 2011 2012 2013E 2014E 2015E 4,448 7,915 14,501 22,306 30,659 40,594 53,851 (1,616) (2,149) (3,897) (6,449) (11,303) (15,672) (20,787) 2,832 5,766 10,604 15,857 19,356 24,922 33,065 (1,227) (1,807) (3,027) (4,806) (8,473) (11,366) (15,078) Operating income 1,605 3,959 7,577 11,051 10,883 13,556 17,986 Profit before tax & minority interest 1,683 4,061 7,809 11,965 12,118 15,956 21,346 Tax (198) (536) (1,189) (1,574) (1,847) (2,393) (3,202) Minority interest Net income Depreciation & Amortization EBITDA Operating margin (%) - - (18) (65) (36) - - 1,485 3,525 6,639 10,456 10,307 13,563 18,144 317 441 885 1,515 1,740 2,314 3,070 1,911 4,390 8,396 12,333 12,623 15,870 21,056 36.1 50.0 52.2 49.5 35.5 33.4 33.4 EBITDA margin (%) 43.2 55.6 58.4 56.3 41.2 39.1 39.1 Pre-tax margin (%) 37.8 51.3 53.9 53.6 39.5 39.3 39.6 Net margin (%) 33.4 44.5 45.8 46.9 33.6 33.4 33.7 Sources: Company data; BNP Paribas estimates Strong growth outlook – BUY with a target price of USD113.00 We estimate Baidu’s EPS at RMB294.74 for 2013, RMB387.85 for 2014 and RMB518.88 for 2015. We set our DCF-based target price at USD113.00 per ADS. At our target price, the stock would trade at P/E multiples of 23.8x in 2013, 18.1x in 2014 and 13.5x in 2015. The stock is trading at P/E multiples of 19.2x, 14.6x and 10.9x, respectively, over the same period. We believe Baidu is transitioning from its traditional Internet search business and expanding into a larger online advertising platform. Within the China market, we view Baidu as a pioneer for some of the business models that have been proven successful in more mature markets. Particularly, we believe the Ad Exchange platform – Baidu Exchange Service – will enable Baidu to expand the scale of its online advertising operations significantly to include Internet properties that cut across multiple device platforms, including PC and mobile Internet. 37 BNP PARIBAS Transitioning from traditional internet search business to a larger online ad platform 27 JUNE 2013 BAIDU BIDU US Alen Lin Our DCF model assumes mid-term revenue CAGR of 20.3% and terminal revenue growth rate of 5%. We assume average mid-term EBIT margin of 30.8% and terminal EBIT margin of 30%. We use WACC of 11.5%. Key downside risks to our investment view are: § An increase in competition in Baidu’s core Internet search segment, which would slow search advertising revenue growth and erode margins; § Failure to generate sufficient revenue growth beyond the core search segment; key areas include BES platform, online video services and mobile Internet; EXHIBIT 7: DCF assumptions and derivation of TP WACC Mid-term Terminal Revenue growth rate (%) 20.3 5.0 EBIT margin (%) 30.8 25.0 WACC (%) 11.5 Source: BNP Paribas estimates EXHIBIT 8: Baidu’s P/E band EXHIBIT 9: Baidu’s P/BV band (HKD) (HKD) 250 350 40x 200 30x 150 300 15.0x 250 12.5x 200 10.0x 150 7.5x 100 5.0x 50 2.5x 20x 100 10x 50 0 Aug-05 Feb-07 Aug-08 Feb-10 Aug-11 0 Aug-05 Feb-13 Sources: Bloomberg; BNP Paribas estimates Feb-07 Aug-08 Feb-10 Aug-11 Feb-13 Sources: Bloomberg; BNP Paribas estimates EXHIBIT 10: Baidu quarterly P&L 1Q13A Total revenues Cost of revenues 2Q13E 3Q13E 4Q13E 2013E 5,969 7,453 8,521 8,718 30,659 (2,099) (2,661) (3,170) (3,374) (11,303) Gross profit 3,869 4,792 5,351 5,344 19,356 Selling, general and administrative (848) (1,043) (1,193) (1,220) (4,305) Research and development (811) (1,014) (1,159) (1,186) (4,169) Operating profit 2,210 2,735 2,999 2,938 10,883 Income before income taxes 2,396 3,085 3,349 3,288 12,118 Net income 2,007 2,622 2,847 2,795 10,271 Gross margin (%) 64.8 64.3 62.8 61.3 63.1 Operating margin (%) 37.0 36.7 35.2 33.7 35.5 Net margin (%) 34.2 35.2 33.4 32.1 33.6 Margins (%) Sources: Company data; BNP Paribas estimates 38 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin Company Profile History and development of the company Incorporated in January 2000, Baidu Inc is a leading Chinese-language internet search provider and it conducts operations mainly through Baidu Online and Baidu Netcom. The company’s ADS have been trading on The NASDAQ Global Select Market under the symbol “BIDU” since August 2005. The conversion between ADS and ordinary share is 10 ADS for every 1 Class A ordinary share. In December 2007, Baidu was included in the NASDAQ-100 index. The company acquired the domain name hao123.com for its Internet navigation site in August 2004, online travel search service provider, Qunar (www.qunar.com), in July 2011 and the online video business of Internet video provider, PPS, in May 2013. In November 2012, Baidu obtained controlling interest in iQIYI.COM, an online television and movie portal founded by Baidu and Providence Equity Partners in January 2010. Business overview In addition to its core web search product Baidu.com, the company powers a broad range of products and services. Popular services include: § Baidu PostBar (tieba.baidu.com), an interactive experience-sharing community § Baidu Knows (zhidao.baidu.com), a query-based searchable platform § Baidu Encyclopaedia (baike.baidu.com) § Baidu Map § Baidu image and video Search and Video Search. While Baidu provides many popular products, it generated 99.7% of its revenue in 2012 from online marketing services, thanks to its 596,000 active online marketing customers in 2012. A complete list of Baidu’s product and service offerings is shown in Exhibit 11. Industry competition Baidu faces competition primarily from Google and other Chinese search engines such as Qihoo 360, Sogou and SOSO. Baidu competes with these entities based on user traffic, quality and quantity of search results, availability and ease of use of products and services, the number of online advertising customers, distribution channels, and the number of associated third-party websites. China search engine market data released by CNZZ show that Baidu had 71.1% search engine market share based on usage, Qihoo 360 had 14.1%, Sougou for 8.0%, SOSO for 3.4% and Google 2.3% as at May 2013. 39 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin EXHIBIT 11: China search engine market share by usage (May 2013) Soso 3.4% Google 2.3% Others 1.1% Sogou 7.9% Qihoo 14.1% Baidu 71.1% Source: CNZZ Subsidiaries Baidu has several wholly owned subsidiaries which offer specific products and provide support to Baidu’s operations (Exhibit 8). Out of the list of wholly owned subsidiaries, Qunar and qiyi.com offer Baidu’s travel services and online video services,. Other entities including Baidu Cloud Computing Technology are subsidiaries that support Baidu’s operations. EXHIBIT 12: Baidu's principal subsidiaries and consolidated affiliated entities Name of entity Baidu Online Network Technology (Beijing) Co Ltd Baidu Holdings Ltd Beijing Baidu Netcom Science Technology Co Ltd Baidu (China) Co Ltd Baidu.com Times Technology (Beijing) Co Ltd Beijing Perusal Technology Co Ltd Baidu Japan Inc Baidu (Hong Kong) Ltd Beijing BaiduPay Science and Technology Co Ltd Baidu HR Consulting (Shanghai) Co Ltd Qunar Cayman Islands Ltd Qiyi.com Inc B.D. Mobile Telecommunications Ltd Baidu Cloud Computing Technology (Shanxi) Co Ltd Baidu Cloud Computing Technology (Beijing) Co Ltd Source: Company data Directors and principal Shareholders Baidu has five directors on its board: one non-independent director and four independent directors. Baidu’s CEO, Mr Robin Li, serves on the board as chairman. Mr Robin Li owns 16% of the company’s outstanding shares. Out of the four independent Directors, Mr William Decker also serves as chairman of the audit committee. Ms Jennifer Li and the four independent directors have combined share ownership of 0.8%. From the principal shareholders, Handsome Reward Ltd is a company wholly owned and controlled by Mr Robin Li. 15.7% of Baidu’s ordinary shares owned by Handsome Reward are included in Mr. Robin Li’s individual ownership. 40 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin EXHIBIT 13: Directors and executive officers Directors and Executive Officers: Position/Title Shares Beneficially Owned (%) Robin Yanhong Li Chairman and Chief Executive Officer 16.0% Jennifer Xinzhe Li Chief Financial Officer 0.8% (combined with other independent directors) William Decker Independent director James Ding Independent director Nobuyuki Idei Independent director Greg Penner Independent director Source: Company EXHIBIT 14: Principal shareholders Principal Shareholders: Shares Owned (%) Handsome Reward Limited 15.7 Baillie Gifford & Co (Scottish partnership) 8.1 T. Rowe Price Associates, Inc. 5.4 Source: Company 41 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin EXHIBIT 15: Baidu’s product offerings Source: Company data 42 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin Internet search engine – a high-performance subset of Ad Exchange We view the Internet search engine industry as a subset of the Ad Exchange mechanism, as search engines are capable of serving highly targeted advertisements to Internet users with particular interest in a subject. As discussed in earlier sections, advertisements served through Internet search platforms have significantly higher CTR than display advertisements. This makes large Internet search platforms highly valuable Internet properties for advertising purposes. Internet search engines enable the targeting of Internet users with relevant ads, leading to high CTR with potential to extend into new products In markets outside of China, Google is the recognised leader in Internet search. In the US, it represented 66.5% of search market share in April 2013, far ahead of the second largest search platform, Microsoft’s Bing with 17.3% market share (according to comScore). Google’s success, however, has not been replicated in China, where Baidu remains the dominant search platform with 70% market share (by usage) in 1Q13; we attribute Baidu’s dominance to its strength in Chinese language search and government policies that potentially put Google in a weaker market position in China. We believe Baidu will continue to be a dominant search platform for the China market, given the strong positive feedback loop effect of the search industry (Exhibit 19). Beyond the core Internet search activities of the existing search platforms in China, we also expect them to begin to develop the Ad Exchange business model and increasingly become broader online advertising platforms. This should create new revenue opportunities for Chinese search platforms as they already have existing online advertising customers and have the potential to rapidly tap into new inventories of Internet properties through Ad Exchanges. In 2012, Google Network Members generated 28.5% of Google’s total advertising revenue; Baidu still depends on its core search business for over 90% of revenue. China Internet search engine – industry still on a strong upward trend China’s Internet penetration has risen from 16% at the end of 2007 to 42.1% in 2012; over the same period, Internet search industry revenue has increased from RMB2.9b to RMB29.3b. The five-year CAGR for China’s Internet user growth through 2012 was 21.8%, while the search engine industry revenue CAGR was 58.8% (Exhibit 16). This implies that while the number of Internet users in China rose rapidly, revenue generated per user also rose from RMB13.8 to RMB52. This indicates that the industry has already hit an inflection point; in the early phase of the search engine industry, the number of Internet users was the primary driver of search industry revenue; however, the main growth driver has now shifted to search revenue per Internet user, driven by the increasing number of online advertisers. While Internet penetration growth in China has been impressive in recent years, we see no signs of this tapering off, as penetration of 42.1% at the end of 2012 remains low compared with that in mature markets such as Japan and Korea; both are at c80% penetration, and Taiwan, Singapore, and Hong Kong are at c75% penetration (Exhibit 17). The US has the lowest penetration of the mature markets at 78%. We believe China’s aggressive upgrades of its fixed-line broadband network in recent years will continue to drive up Internet penetration; in addition, the rise in smartphone adoption presents incremental opportunities for the Internet search industry. In our view, these trends set the foundation for the continuing growth of China’s search engine industry. 43 BNP PARIBAS 27 JUNE 2013 BIDU US Alen Lin EXHIBIT 16: Number of Internet users and search industry revenue EXHIBIT 17: Internet penetration comparison (2011) (Percentage of individuals using the Internet) Number of Internet users (LHS) 15 200 100 China 10 5 0 0 2007 2008 2009 2010 Sources: CEIC; China Internet Watch; BNP Paribas 2011 2012 Korea 20 300 Germany 25 400 France 30 Japan 500 U.S. 35 Hong Kong 600 90 80 70 60 50 40 30 20 10 0 Taiwan Search industry revenue (RHS) (%) (RMB b) Singapore (m) Spain BAIDU Source: ITU Industry growth should continue to outpace Internet user growth In our view, the increasing maturity and efficiency of the Internet search industry provide the fundamental support for search industry revenue growth to outpace Internet user growth in China. We illustrate the value creation process of the industry in Exhibit 18. The products being sold by search engines are their inventories of user-specific search queries. Based on the user profiles and activities, search engines create value by funnelling users with specific profiles to paying advertisers that seek a specific audience. The ultimate customers of the search engine industry are the advertisers that promote their products and services to targeted sets of Internet users. Search engines’ positive feedback loop, with increasing search users and increasing advertisers, should continue to drive industry value Internet search industry dynamics closely follow the academic definition of an information economy, where at the macro level the value of the industry increases exponentially as the number of industry participants increases. As illustrated in Exhibit 19, the rising number of Internet users drives growth of online advertisers. This leads to rising quality of search results, which in turn drives the growth of Internet users; as the numbers of products (Internet users and Internet properties) and customers (advertisers) rise, the probability of matching the right product to the right customer also rises, leading to higher revenue per product offered. Specific search engines also play an important role in raising the quality of search through enhanced profiling of Internet users. This can be achieved by collecting indepth data about the particular user and analysing the data to better understand and predict a user’s interests, habits, and behaviour. Such data collections and analytics can further enhance the matching of Internet users to paying advertisers, leading to higher conversion rates, or click-thru-rates (CTR). This provides more potential for raising the value of the industry as advertisers are willing to pay more to gain exposure to the relevant audience. Search engines create incremental value by applying user data and analytics in search algorithms to maximise CTR In our view, the combination of a rising number of Internet users and improving search quality through data analytics provides a solid foundation for continuing industry revenue growth, as evident in Exhibit 16. 44 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin EXHIBIT 18: Internet search engine value chain Inventory: search queries and user profiles Value Creation Search Engines Advertising Distributors Advertisers Source: BNP Paribas EXHIBIT 19: Positive feedback loop for Internet search value chain Rising number of Internet users drives higher number of advertisers Number Internet users Number advertisers Rising number of advertisers drives higher number of Internet users Source: BNP Paribas Industry outlook – no sign of a slowdown To analyse the relationship between search industry revenue and general Internet usage, and to quantify industry potential, we have developed a regression model. As the model is based on historical data, the potential contributions from Ad Exchange and mobile Internet are not fully factored in as they are still new business models in China; they could offer upside potential as they gain momentum. We develop a regression model to project search industry revenue growth – we forecast a 2012-15 CAGR of 31.9% Our regression model forecasts an industry revenue CAGR of 31.9% between 2012 and 2015, vs 58.8% for the five years between 2007 and 2012, as the comparison base has grown significantly in recent years (Exhibit 20). A major component of this is search revenue per Internet user, for which we forecast a CAGR of 25.8% over the period vs CAGR of 30.3% over the past five years. This is consistent with our earlier discussion that the rising numbers of Internet users and online advertisers will drive growth in advertising revenue per user. This is also consistent with Baidu’s disclosure that revenue per online advertising customer rose to RMB37,300 in 2012, from RMB5,800 in 2006, a CAGR of 45.1%. Our regression analysis indicates that industry revenue is most sensitive to Internet search user penetration and the number of Internet search users, while the hours of Internet use per week, seasonality, and number of Internet users are also strong predictors. The number of mobile Internet users yields the lowest t-ratio in our model, suggesting that the growth of mobile Internet users has so far not contributed significantly to industry revenue. This could be a new growth driver as the industry finds mechanisms for monetising mobile Internet usage; we believe the Ad Exchange platform could be a powerful tool with which to more effectively monetise mobile Internet. 45 BNP PARIBAS Regression analysis indicates high sensitivity to search user penetration and number of search users; Internet usage time, seasonality, and number of Internet users also important 27 JUNE 2013 BIDU US BAIDU Alen Lin EXHIBIT 20: Search industry revenue forecasts (RMB b) Search industry revenue (LHS) 40 Search revenue/Internet user (RHS) (m) 60 35 50 30 40 25 20 30 15 20 10 10 5 2H15E 1H15E 2H14E 1H14E 2H13E 1H13E 2H12 1H12 2H11 1H11 2H10 1H10 2H09 1H09 2H08 0 1H08 0 Sources: BNP Paribas estimates, CEIC and China Internet Watch Industry value drivers – combination of search users, advertisers, and analytics A search engine is effectively a factory that packages its inventory, the Internet search users, and sells this inventory to its customers, the advertisers. The search engine’s role ends when it connects the inventory to the customer. Any interactions or transactions that occur beyond this point are made outside the initial connection created by the search engine. There are a number of factors that affect the probability of an interaction taking place after the initial search connection, including the size of the inventory, the number of customers, and the analysis performed by the search algorithm that matched a particular search query/user to a particular advertiser (Exhibit 21). The value of the inventory depends on the quantity and quality of search users (subjective measure used by advertisers, based on user profile and attributes); higher quality search users increase the probability of a transaction between the user and advertiser. Search engines create value when inventory – search users – are sold to customers – the advertisers EXHIBIT 21: Internet search industry value drivers Sourcing/inventory of Internet search users Direct search queries www.baidu.com www.so.com www.sogou.com www.soso.com www.google.com Navigation sites Web/news portals www.sina.com.cn news.sogou.com news.soso.com hao.360.cn www.hao123.com Social Networking Internet utilities Email applications E-Commerce sites Video/entertainment sites Web browsers Etc. Search Engine Distribution of advertising service Advertising distributors Advertisers Advertisers Advertisers Buyers of advertising service Source: BNP Paribas 46 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin The battle to dominate Internet search traffic A starting point for a search engine service is to maintain a stream of search queries, ie, having a large pool of inventory; this maximises the traffic for which advertisers pay. This can be done through capturing direct search queries through search engine websites such as google.com, baidu.com, so.com, and others, or through other Internet services (Exhibit 22). While direct search queries at a search engine website are the most direct mechanism for search engines to acquire users, traffic tends to flow to the dominant search engine brands and be less effective for the secondary search engine brands. Secondary search engines must utilise other mechanisms to acquire search traffic. Aside from direct search queries, search engines tap into other sources of search traffic such as Internet navigation sites including Baidu’s hao123.com and Qihoo 360’s hao.360.cn (Exhibits 22-23). These navigation sites are collections of Internet resources that users frequently utilise to navigate the Internet and often serve as the starting point for users. According to statistics compiled by China Internet Watch, hao.360.cn contributed 11% of search engine traffic in November 2012, while hao123.com contributed 8.2% (Exhibit 26). Search users reach search engines through direct access or traffic referrals; Internet navigation sites contribute a significant amount of search traffic Popular navigation websites are powerful as users frequently utilise the default search engines within these websites. In 3Q12, Qihoo 360 changed the default search engine of its popular navigation website from Baidu to its own (Exhibit 24), and pushed Baidu to a lower position on the list of search engine selections; this had an immediate impact on Qihoo 360’s search engine market share (Exhibit 25). These navigation sites, while serving as a source for search engine traffic, may also be revenue generating as a standalone service. Qihoo 360’s hao.360.cn contributes the majority of its online advertising revenue currently, according to the company. EXHIBIT 22: Qihoo 360 navigation site EXHIBIT 23: Baidu navigation site Source: hao.360.cn Source: www.hao123.com 47 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin EXHIBIT 24: Qihoo 360 navigation site default search engine EXHIBIT 25: China Internet search market share by page views 100% Baidu Qihoo Soso Google Sogou 80% 60% 40% Chart or other exhibit 20% Source: hao.360.cn 2Q13 1Q13 4Q12 3Q12 2Q12 1Q12 4Q11 3Q11 2Q11 1Q11 4Q10 3Q10 0% Source: cnzz.com Other sources of search traffic come from online video/entertainment, social networking, news portals, and others. Exhibit 27 lists the popular Internet services in China that can serve as sources of search traffic. In our view, the popularity of online video services is a major reason for Baidu owning video service providers iQiyi and PPS; while these services may not be profitable as standalone entities currently, they could be important search traffic sources when integrated into Baidu’s product portfolio. In addition, premium content could generate additional revenues as the online entertainment industry matures. Internet services such as entertainment, social networking and news portals also contribute to search traffic In addition to having a strong position in navigation websites, Qihoo 360 also has a solid market share in China’s web browsers; in recent quarters, Qihoo 360’s Secure Browser maintained a market share of c25%, according to cnzz.com. These browsers, when installed, default the home page to Qihoo’s hao.360.cn unless the user changes the home page, but it also prompts users periodically to set Qihoo 360’s navigation as the home page. In conjunction with Qihoo’s popular antivirus software, Qihoo has a dominant position on Internet users’ desktops. Despite the initial success in gaining market share in 3Q-4Q12, the company’s ability to continue converting navigation site users to its own search engine users is yet to be seen. EXHIBIT 26: Navigation websites’ contribution to search engine traffic Contribution to search engines (Nov 2012) (%) hao.360.cn 11 hao123.com 8.20 web.sogou.com 6.50 123.duba.net 4.20 2345.com 1.40 hao.uc.cn 0.90 Source: China Internet Watch 48 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin EXHIBIT 27: Top Internet services in China Top Internet services in China (Jan 2013, million users) Search service 454.0 Online video 452.0 Community friends 439.0 News information 413.0 E-commerce 404.4 Entertainment 403.8 IT digital products 395.0 Media domain 362.1 Financial service 361.5 Email 356.0 Source: China Internet Watch Expanding online advertising footprint Another key value-driver for the search engine industry is its advertising footprint. While search engines are designed to operate without geographical boundaries, a search engine service provider’s ability to monetise its service depends on its ability to distribute its advertising service. The search engine service provider has the ability to reach advertising customers through direct channels or via advertising agencies that act as distributors of the service. Expansion of advertising footprint critical to monetising search services; search engines leverage direct sales force and ad agencies to expand footprint China’s online advertising market continued to expand in recent quarters, with 1Q13 growing 39% y-y as advertisers increasingly shifted marketing budgets toward online media, according to China Internet Watch data. Within the overall online advertising market, search engine remains the largest category of spending, representing 34% of total industry revenue in 1Q13; the category grew 36% y-y (Exhibit 28). The second largest category of online advertising is vertical search, which represented 23% of total industry revenue in 1Q13, +84.3% y-y. While growth of the online industry has been strong, we see no slowdown as the number of Internet users continues to rise and advertisers increasingly find online advertising to be more effective in targeting customers. Despite the benefits of online advertising, we believe the number of advertisers can only grow as fast as the major online platforms can expand and reach them. At the end of 1Q13, Baidu had 410,000 online advertising customers, +28% y-y; Qihoo 360 currently has less than 10,000 as it had just begun to monetise its Internet search business. While Qihoo 360 aims to rapidly increase its search advertising customers, its scale could be limited by its search users. On the other hand, given Qihoo 360’s smaller search market share, its lower advertising price could serve a niche market of smaller advertisers. Both Baidu and Qihoo 360 leverage direct internal sales forces in major markets and external advertising agencies in other markets to reach additional advertising customers. In the case of Baidu, we estimate the top three markets, Beijing, Shanghai, and Guandong, represent over 50% of advertising revenue and are mostly covered by its direct sales force. While Qihoo is still building up its customer base, it is also primarily targeting tier-1 markets with its internal sales force. 49 BNP PARIBAS 27 JUNE 2013 BIDU US BAIDU Alen Lin EXHIBIT 28: China online advertising market in 1Q13 Search Engine (RMB b) 25 Video Vertical Search Rich Media EXHIBIT 29: Baidu’s online advertising statistics No of online marketing customers (LHS) Brand Graphic (nos) Other 20 Average revenue per customer (RHS) 40,000 600,000 35,000 30,000 500,000 15 (RMB) 700,000 25,000 400,000 20,000 300,000 10 15,000 200,000 5 10,000 100,000 5,000 0 0 Source: China Internet Watch 0 2006 2007 2008 2009 2010 2011 2012 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 Source: Baidu Improving advertising performance through Internet user profiles In an effort to improve the effectiveness of search results and search advertising, search engines could leverage contextual data about the user in the search algorithm. Such data could include a user’s location, historical Internet activities or other personal data. When utilised in online advertising, this leads to more targeted ads. The popular Internet services Baidu operates could serve the purpose of collecting user data and profiling users; these services include Baidu Knows, Baidu Post Bar, Qunar travel service, and the numerous video services such as iQiyi and the recently acquired PPS. Fine-tuning search performance through extensive user data collection and analytics; breadth of popular Internet services provides an advantage Partnerships with Internet services that have access to detailed user information can also be powerful. Theoretically, search engines such as Baidu could partner with Internet social networking services to exchange user profiles and improve their advertising effectiveness. This could ultimately lead to higher prices for advertising on the particular search platform. 50 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin Competitive landscape in the China search engine industry EXHIBIT 30: Porter's five-forces for Internet search engine Threat of new entrants: Low The challenges for new entrants to reach critical mass in a networked industry inhibits expansion of new entrants as scale in terms of market share is necessary to gain new Internet users and advertising customers. Such dynamics is prohibitive for new entrants to compete against incumbents. Supplier bargaining power: Medium Internet services that feed search users to search engines are the key suppliers; as they are key sources of user traffic, they could be in the position to charge the search engines Through referral fee structure. Others are infrastructure suppliers such as computer servers and network operators which have relatively lower bargaining power. Internal rivalry: Medium Internet search industry conforms closely with the positive feedback characteristics of a networked industry; as the number of users who use its service increases, the value of its service increases. Such dynamic tends to limit smaller rivals to serving niche market segments. Buyer bargaining power: Medium The price of service provided by search engines is generally determined by bidding process where advertising customers bid for search words. The price depends on the search engines’ user market share and its ability to match users to the advertisers. Threat of substitutes: Medium Traditional media are substitutes for online advertising services; however, they have been losing ground to online advertising channels in recent years. Other online channels also compete for advertising budget; they include Internet portals, social networking services, etc. In addition, in mobile Internet environment, browser based search is being displaced by mobile applications. Source: BNP Paribas Top-3 search engines account for over 90% of search traffic The China search engine industry is dominated by the top three search engines, Baidu, Qihoo 360, and Sogou, with a combined search market share of over 90% in the past year, as measured by search usage by cnzz.com. Baidu remains the most dominant player, with over 70% market share in 1Q13; this is a decline from nearly 80% a year earlier. Since August 2012, Qihoo 360 has gained market share to reach nearly 15% as the company leveraged its strong presence on PCs to direct search traffic to its own search engine, instead of to Baidu as it did previously. This was effective for Qihoo 360 as its browser and navigation site have an established user base, and they serve as effective sources of search traffic. The number three search engine, as measured by search usage, Sohu’s Sogou, leverages the many Internet services provided by Sohu, such as sohu.com, as sources of search traffic to achieve a search market share of 7-8%. We expect traffic share among the top three to shift as the two smaller players look for ways to gain traffic share both organically and through partnerships with other Internet services. In May 2013, Qihoo 360 and Alibaba announced a partnership to launch 360.etao.com to position the search service for online shopping; this is intended to both increase vertical search traffic for Qihoo 360 and to raise brand awareness of Qihoo 360 search by leveraging Taobao’s dominance in e-tailing. Higher brand awareness could lead to higher direct search queries via Qihoo 360’s so.com. Qihoo 360 has publicly stated its target of 20% traffic share by the end of 2013. Highly concentrated industry with top three in China representing over 90% of search usage Smaller search engines likely to seek partnerships or acquisitions to increase market share; they could target niche verticals to differentiate themselves from the dominant player It has also been widely reported that Sohu could be seeking to sell its Sogou search engine; in our view, an acquisition could be more complementary to entities other than the existing large search platforms. While the top search engine platforms continue to compete for traffic share, we believe Baidu will maintain its revenue share as its scale has enabled it to gain substantially more online advertising customers than rivals have; Baidu’s 2012 revenue market share was 79% according to China Internet Watch (Exhibit 32). While we believe Qihoo 360 and Sogou could 51 BNP PARIBAS 27 JUNE 2013 BIDU US BAIDU Alen Lin gain traffic share in the near term, their ability to fully monetise the service could be more limited and would take time. Beyond the top three, search engines such as Tencent’s Soso, Google, Microsoft’s Bing, and Yahoo have combined market share in China of less than 10%. As we believe the Internet search market strongly exhibits the characteristics of a networked industry, market share gains for smaller players could therefore be limited in the near term. Sourcing of search traffic Key suppliers for the Internet search industry are other Internet service providers that have established user bases, such as Internet portals, navigation sites, social networking services, e-commerce sites, Internet utilities and entertainment sites. Baidu has built up a large number of services with a large user base that enables the company to direct its user base to its own search engine. It also pays search referral fees to partners for search traffic (Exhibit 31); we believe the recent rise in traffic acquisition costs could include the effect of the Ad Exchange product. Other search engine platforms such as Qihoo 360 mostly depend on traffic from their respective navigation sites. However, as discussed earlier, Qihoo 360 is partnering with other Internet platforms such as Alibaba to source search traffic along specific industry verticals. This could imply that the company has exhausted internal resources. While the partnership terms with Alibaba are undisclosed, we believe that as Qihoo 360 aims to become a more significant player, it will increasingly need to source traffic from other platforms in exchange for referral fees. In our view, such an arrangement could be based on Qihoo 360’s ability to successfully monetise its Internet search service. EXHIBIT 31: Baidu's traffic acquisition cost (RMB m) Traffic acquisition cost (LHS) 700 Traffic acquisition cost/Revenue (RHS) 14% 600 12% 500 10% 400 8% 300 6% 200 4% 100 2% 0 0% 1Q10 2Q10 3Q10 4Q10 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 Source: Company Monetisation of search Once the traffic for a search engine reaches critical mass, search platforms need to address the monetisation aspect of the business. The value of search increases in line with the number of Internet search users; this has been shown in empirical data, as illustrated in Exhibit 33, where search revenue per user rises as the number of search users increases. A similar relationship is found for Baidu, where revenue per online advertising customer increases in tandem with the rising number of online advertising customers. We expect such a pattern to continue for the foreseeable future as the perceived utility of a search engine increases from a user perspective; this should continue to drive online advertising revenue in a self-reinforcing cycle. Empirical data has shown search revenue per user increases with a rising number of search users and number of advertising customers In terms of monetising Internet search, market leader Baidu shows a consistent link, with market share by revenue of 78.5% in 2012 vs its page view market share of 77.6%. As the field of search users approaches saturation, we believe the dominant player, Baidu, will shift its focus to extending its advertising reach to grow its advertising customer base, leading to a rise in revenue per search user. In addition, 52 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin its vast amount of user data to which it has access could enable the search engine to deliver more targeted advertisements, leading to higher advertising fees. Qihoo 360 also needs to expand its online advertising customer base, which is currently less than 10,000. Qihoo 360 has partnered with Google, where the former generates search traffic and the latter provides broad exposure to online advertising customers; as such, Google’s revenue market share in 2012 was 15.6%, despite its limited direct exposure to the search users (Exhibit 32). As Qihoo 360 builds up its own online advertising customer base, it plans to reduce its dependence on Google and capture a larger proportion of the search revenue generated. EXHIBIT 32: Search market share by revenue (2012) SOSO 1.5% Youdao 0.3% Others 1.0% Sogou 3.1% Qihoo 360 starting to build up advertising customers to monetise search services; currently rely on partnership with Google EXHIBIT 33: Continuing growth of revenue/search user (m) Number of search users (LHS) 500 Revenue/search user (RHS) (RMB) 70 450 60 400 Google 15.6% 350 50 300 40 250 Baidu 78.5% 200 30 150 20 100 10 50 0 0 2007 Source: China Internet Watch 2008 2009 2010 2011 2012 Sources: China Internet Watch; CEIC; BNP Paribas Maximising search advertising revenue through high performance searches The Internet search industry primarily generates revenue through online advertising. Given the extensive and dynamic information that exists about Internet users and advertisers, search platforms can maximise advertising price/revenue through numerous pricing strategies to improve effectiveness. The pricing mechanism for such advertising is similar to that for Ad Exchange platforms discussed previously, where advertisers bid for advertising spots based on CPM, CPC, and CPA. Industry value can be increased by optimising search quality – ie, using user data analytics to improve relevance of advertising for search users The placement of a particular advertisement is generally determined by the relevance and quality of an advertisement in a particular search query (in Google terms – Quality Score), and bidding prices from advertisers for the particular search query; these two factors combined determine the position of the particular advertisement in the particular search query (Exhibit 7). Advertising prices are ultimately determined by how effective the advertisement is as measured by the click-through-rate (CTR); we believe improving the advertising relevance can improve the CTR, and this is where search platforms can continue to optimise and differentiate. Exhibit 33 provides a sample for CPC prices for different keywords related to “Used cars”. Such industry dynamics should enable search platforms with large amounts of user data to better monetise Internet search services. Data collected through the various Internet services can be used to optimise search results for the particular user, leading to higher a CTR; this would directly lead to higher market prices for advertising placements. As we discussed in an earlier section, CTRs are generally in the range of 1-5% for search advertisements. 53 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin EXHIBIT 34: Sample keyword bidding prices Source: adwords.google.com Display advertisements – a different type of advertisements Another type of common advertisement on the Internet is a display advertisement. These are advertisements placed on various websites. While a less sophisticated advertisement may be static (like the one shown in Exhibit 7, more intelligence can be built into the placement of the advertisements by taking into account contextual information such as the user profile, geographical location, time of day, etc.; this approach could have the effect of increasing the click-through-rate. As display advertisements are considered to be “push” advertisements – i.e. pushed to the users by advertisers – CTR for typical display advertisement is <1%. While the CTR is lower than for a search advertisement, the frequency of advertising display is significantly higher for display advertisements. Techniques for search ads can be applied to display ads to improve CTR; Ad Exchange platforms enable search engines to expand into display ads Advertising space inventory can be significantly increased when a search platform leverages the Ad Exchange platform to include partner Internet websites. According to Google, it has over 2m partners in its Google Network Members; Baidu also has partners in its Baidu Union in the range of hundreds of thousands. We believe the ramping up of Baidu Union is a major step for Baidu to build up an Ad Exchange platform as efficiency of the platform increases with the increasing number of participants. We view the Ad Exchange platform as a new product for search engine platforms that has implications on future growth potential. Mobile Internet – the next frontier As the number of mobile Internet users has grown rapidly in recent years with the rise of smartphones and mobile broadband services in China, Internet companies are still adapting their business models to tap into this new opportunity. Exhibit 35 shows the number of overall Internet users 5-year CAGR through 2012 was 21.8% vs. the number of mobile Internet users CAGR of 52.8%. While the rise of mobile Internet usage presents new opportunities to Internet companies, it also poses challenges as Internet companies have to adapt to mobile Internet monetisation. The growth of mobile Internet users has far outpaced Internet companies’ ability to monetise mobile Internet users’ utility for the services. Access to mobile Internet more fragmented than PC Internet; Ad Exchange platforms can also be adapted for service mobile ads Internet companies that have optimised their products for PC Internet advertising are still adjusting their products for the mobile Internet environment, where a major limitation is the real estate available on much smaller screens compared to PC monitors. Taking search engines for example, for each search query a number of paid-advertisements can be served to the PC user; however, on a mobile device limited space is available to serve advertisements. Therefore, for mobile Internet 54 BNP PARIBAS 27 JUNE 2013 BIDU US BAIDU Alen Lin advertising to be meaningful the frequency of advertisements must be increased; this would require Internet companies to expand advertising inventory across multiple applications on mobile devices. The user behaviour for mobile Internet is very different from PC Internet. On PCs, users mostly access the Internet through web browsers; this is how most Internet companies that rely on advertising have developed their products. On a mobile device, users are more likely to access information using applications for specific activities; in many cases Internet companies that have succeeded in PC Internet have limited presence in mobile applications. Based on comScore data for May 2013, browsers only represented 20% of smartphone and tablet usage based on time spent. A 2012 Google survey of China’s smartphone users indicated that only 34% of users noticed advertisements appearing in an app; users mostly noticed advertisements on mobile devices while using search engines (49%) and on a website (54%). In our view, this suggests a significant opportunity for in-app advertisements. Exhibit 36 is a summary of popular services being used by China’s mobile users where instant messaging, mobile search, news, Weibo, and music round out the top 5 applications. Browser access on mobile Internet only represents 20% of usage time; only 34% of users notice ads in mobile apps – presenting significant opportunities In our view, the Ad Exchange could be an effective platform for mobile advertising to proliferate as mobile applications can become members of the Ad Exchange network just like other Internet properties and make their mobile applications available through the Ad Exchange. While the business model for mobile advertising still needs to be optimised, we believe the potential exists as mobile devices offer significantly more contextual information about the user that can be used for more targeted advertising. In our view, this places even more importance on the Ad Exchange platform we discussed in earlier sections as it has the potential to serve targeted contextual advertisements to mobile users across multiple applications. 0 2006 Source: CEIC 55 2007 2008 2009 2010 2011 2012 Post & comment 100 Mobile e-mail 200 Network video 300 Social network 400 Network literature 500 Network music 90 80 70 60 50 40 30 20 10 0 Weibo (%) Total number of mobile Internet users News browse Total number of Internet users Instant messaging (m) 600 Mobile search EXHIBIT 36: China mobile users’ usage of mobile apps (April 2013) EXHIBIT 35: Number of China Internet users Source: China Internet Watch BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin Financial statements BAIDU Profit and Loss (RMB m) Year Ending Dec 2011A 2012A 2013E 2014E 2015E Revenue 14,501 22,306 30,659 40,594 53,851 Cost of sales ex depreciation (3,078) (5,167) (9,563) (13,358) (17,717) Gross profit ex depreciation 11,423 17,139 21,097 27,236 36,134 0 0 0 0 0 (3,027) (4,806) (8,473) (11,366) (15,078) Other operating income Operating costs Operating EBITDA 8,396 12,333 12,623 15,870 21,056 Depreciation (819) (1,281) (1,740) (2,314) (3,070) Goodwill amortisation Operating EBIT Net financing costs Associates Recurring non operating income Non recurring items 0 0 0 0 0 7,577 11,051 10,883 13,556 17,986 3,360 334 754 1,234 2,400 (179) (294) (5) 0 0 78 454 6 0 0 0 0 0 0 0 7,809 11,965 12,118 15,956 21,346 (1,189) (1,574) (1,847) (2,393) (3,202) 6,620 10,391 10,271 13,563 18,144 18 65 36 0 0 Preferred dividends 0 0 0 0 0 Other items 0 0 0 0 0 6,639 10,456 10,307 13,563 18,144 Profit before tax Tax Profit after tax Minority interests Reported net profit Non recurring items & goodwill (net) 0 0 0 0 0 6,639 10,456 10,307 13,563 18,144 Recurring EPS * 190 299 295 388 519 Reported EPS 190 299 295 388 519 DPS 0.00 0.00 0.00 0.00 0.00 Revenue (%) 83.2 53.8 37.4 32.4 32.7 Operating EBITDA (%) 91.3 46.9 2.4 25.7 32.7 Operating EBIT (%) 91.4 45.9 (1.5) 24.6 32.7 Recurring EPS (%) 87.9 57.3 (1.5) 31.6 33.8 Reported EPS (%) 87.9 57.3 (1.5) 31.6 33.8 Gross margin inc depreciation (%) 73.1 71.1 63.1 61.4 61.4 Operating EBITDA margin (%) 57.9 55.3 41.2 39.1 39.1 Operating EBIT margin (%) 52.2 49.5 35.5 33.4 33.4 Net margin (%) 45.8 46.9 33.6 33.4 33.7 Effective tax rate (%) 15.2 13.2 15.2 15.0 15.0 0.0 0.0 0.0 0.0 0.0 - - - - - 0.0 0.0 0.0 0.0 0.0 Recurring net profit Per share (RMB) Growth Operating performance Dividend payout on recurring profit (%) Interest cover (x) Inventory days Debtor days 11.3 15.2 18.2 19.0 19.1 Creditor days 229.1 224.4 199.7 229.2 245.0 Operating ROIC (%) 148.5 77.3 62.3 88.2 135.4 ROIC (%) 95.6 55.4 43.8 58.5 83.7 ROE (%) 56.0 50.6 33.1 31.5 30.8 ROA (%) 36.6 27.9 17.1 16.0 16.3 *Pre exceptional, pre-goodwill and fully diluted Revenue By Division (RMB m) 2011A 2012A 2013E 2014E 2015E Online marketing services 14,490 22,246 30,463 40,201 53,065 11 60 197 393 786 - - - - - Other services Sources: Baidu Inc - ADR; BNP Paribas estimates 56 BNP PARIBAS 27 JUNE 2013 BAIDU BIDU US Alen Lin Financial statements BAIDU Cash Flow (RMB m) Year Ending Dec Recurring net profit 2011A 2012A 2013E 2014E 2015E 6,639 10,456 10,307 13,563 18,144 3,070 Depreciation 819 1,281 1,740 2,314 Associates & minorities (18) (65) (36) 0 0 Other non-cash items 147 (458) 342 427 498 Recurring cash flow 7,586 11,215 12,353 16,303 21,711 592 781 3,636 4,499 4,432 0 0 0 0 0 (1,866) (2,340) (3,060) (4,080) (5,100) Change in working capital Capex - maintenance Capex - new investment Free cash flow to equity 6,313 9,656 12,929 16,722 21,044 (1,946) (821) 0 0 0 0 0 0 0 0 (10,438) (10,590) (872) (872) (872) Net cash flow (6,072) (1,754) 12,057 15,849 20,171 Equity finance 67 157 0 0 0 2,359 9,361 0 0 0 (3,646) 7,765 12,057 15,849 20,171 Recurring cash flow per share 217 321 353 466 621 FCF to equity per share 181 276 370 478 602 2011A 2012A 2013E 2014E 2015E Net acquisitions & disposals Dividends paid Non recurring cash flows Debt finance Movement in cash Per share (RMB) Balance Sheet (RMB m) Year Ending Dec Working capital assets 11,721 22,793 23,490 24,295 25,315 Working capital liabilities (4,281) (8,237) (12,569) (17,873) (23,325) Net working capital 7,440 14,557 10,921 6,422 1,990 Tangible fixed assets 2,694 3,888 5,578 7,801 10,358 10,134 18,445 16,499 14,223 12,348 2,420 3,878 3,878 3,878 3,878 Other intangible assets 979 1,588 2,089 2,505 2,851 Investments 734 803 803 803 803 Other assets 666 838 838 838 838 Invested capital 14,932 25,552 24,107 22,247 20,718 Cash & equivalents (4,127) (11,881) (23,584) (38,995) (58,657) 126 0 0 0 0 2,278 9,693 9,693 9,693 9,693 Operating invested capital Goodwill Short term debt Long term debt * Net debt (1,724) (2,187) (13,891) (29,302) (48,964) Deferred tax 132 289 289 289 289 Other liabilities 179 44 44 44 44 15,292 26,055 36,314 49,865 67,998 Total equity Minority interests 1,034 1,160 1,160 1,160 1,160 Invested capital 14,932 25,552 24,107 22,247 20,718 * includes convertibles and preferred stock which is being treated as debt Per share (RMB) Book value per share 438 746 1,038 1,426 1,945 Tangible book value per share 341 589 868 1,243 1,752 (10.6) (8.0) (37.1) (57.4) (70.8) (7.4) (4.8) (23.1) (37.0) (47.7) 3.6 4.2 3.7 3.5 3.6 - - - - - 2011A 2012A 2013E 2014E 2015E 10.8 Financial strength Net debt/equity (%) Net debt/total assets (%) Current ratio (x) CF interest cover (x) Valuation Recurring P/E (x) * 29.4 18.7 19.0 14.4 Recurring P/E @ target price (x) * 36.5 23.2 23.6 17.9 13.4 Reported P/E (x) 29.4 18.7 19.0 14.4 10.8 Dividend yield (%) 0.0 0.0 0.0 0.0 0.0 P/CF (x) 25.7 17.4 15.8 12.0 9.0 P/FCF (x) 30.9 20.3 15.1 11.7 9.3 Price/book (x) 12.8 7.5 5.4 3.9 2.9 Price/tangible book (x) 16.4 9.5 6.4 4.5 3.2 EV/EBITDA (x) ** 23.0 15.6 15.0 11.0 7.5 EV/EBITDA @ target price (x) ** 28.7 19.3 18.7 14.0 9.7 EV/invested capital (x) 13.0 7.6 7.6 7.5 7.1 * Pre exceptional, pre-goodwill and fully diluted ** EBITDA includes associate income and recurring non-operating income Sources: Baidu Inc - ADR; BNP Paribas estimates 57 BNP PARIBAS 27 JUNE 2013 27 JUNE 2013 INITIATION / INTERNET CHINA QIHOO 360 TECHNOLOGY QIHU US MARKET RECS HOW WE DIFFER FROM CONSENSUS HOLD TARGET PRICE USD44.20 CLOSE USD44.14 UP/DOWNSIDE +0.1% TARGET PRICE (%) (11) POSITIVE 11 EPS 2013 (%) (47) NEUTRAL 3 EPS 2014 (%) (17) NEGATIVE 0 KEY STOCK DATA Demanding valuation n n n n YE Dec (USD m) We initiate at HOLD, valuations are stretched We initiate on Qihoo 360, a provider of PC anti-virus software, Internet browser, navigation website, and recently added Internet search, with a HOLD and TP of USD44.20. While we forecast strong growth in 2014, we believe valuations are stretched. High expectations for Qihoo 360’s operating leverage We believe Qihoo 360’s current high valuations reflect high market expectations for the company’s operating leverage. However, we think such expectations are vulnerable to marketing and R&D expenses remaining high, given Qihoo 360’s current growth phase. DCF-based Target Price of USD44.20 We have a DCF-based TP of USD44.20. While we forecast strong NP growth in 2014, we believe upside is fully in the price (as seen in its high valuation). The key upside risk to our call is the company’s ability to expand market share margins. The key downside risk is higher operating expenses to grow market share. Organic growth to become more challenging We believe, given the company’s current expansion phase, the effect of operating leverage will be limited. This is particularly true as we believe organic traffic growth will become more challenging, and more resources will be required for Qihoo 360 to gain incremental search market share. 2013E 2014E 2015E 329 592 844 1,046 38 42 121 216 Rec. net profit 0.22 0.23 0.68 1.21 EPS growth (%) 108.0 7.5 191.3 77.7 Recurring P/E (x) 135.2 125.7 43.2 24.3 0.0 0.0 0.0 0.0 EV/EBITDA (x) 69.1 66.8 26.6 15.0 Price/book (x) 10.9 9.2 7.1 5.2 (79.6) (54.8) (59.5) (66.2) 7.9 18.5 24.6 Recurring EPS (USD) Dividend yield (%) Net debt/Equity (%) 9.1 ROE (%) Jun-12 Sep-12 Dec-12 Mar-13 Jun-13 169 49 44 119 39 34 69 29 24 19 19 14 (USD) (31) QIHOO 360 TECHNOLOGY Share price performance Rel to MSCI China 3 Month 12 Month 2.5 44.7 158.0 17.3 59.8 159.9 Absolute (%) Relative to country (%) Next results August, 2013 5,238 Mkt cap (USD m) 71.9 3m avg daily turnover (USD m) Major shareholder Cost of revenues Selling and marketing G&A Product development 40% (%) 1 Month 40 Free float (%) Qihoo 360’s expense ratio trends 50% Revenue 2012A Hongyi Zhou (0%) 12m high/low (USD) 3m historic vol. (%) ADR ticker 48.38/14.63 45.7 QIHU US 30% ADR closing price (USD; 25 Jun 2013) 20% Issued shares (m) 10% Sources: Bloomberg consensus; BNP Paribas estimates 0% 2010 2011 2012 Sources: Bloomberg consensus; BNP Paribas estimates Alen Lin [email protected] +852 2825 1801 58 2013E 2014E 2015E 44.14 178 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Investment thesis Catalyst We initiate coverage of Qihoo 360 with HOLD recommendation and target price of USD44.20. Qihoo 360 shares are trading at 122x our 2013 earnings and 42x our 2014 earnings. We believe these valuations are stretched. We forecast strong profit growth in 2014-15, but believe market expectations for Qihoo 360’s operating leverage are too high and that this presents downside risk to valuations. We forecast 2013-15 y-y revenue growth of 80%, 42.5% and 24%. We forecast GAAP net profit growth of -10.3%, +191.2% and +77.7%, and non-GAAP net profit growth (excluding share-based compensation) of -4.9%, +86.4% and +54.9%. We expect its net margin to expand from 7% in 2013 to 20.6% in 2015. With high market expectations for Qihoo 360’s operating leverage, we believe the downside risks outweigh the upside risks. We expect rapid revenue growth to be accompanied by high sales, marketing and R&D expenses. We expect any disappointment in margins to lead to share price weakness given such high expectations. Risks to our call Key upside risks are: 1) the company managing to grow its user base (online gaming, navigation site and search) more than we expect without significant margin contraction; 2) expansion of online advertising customers for its search product; and 3) share price volatility with the shares tracking search market share changes. The key downside risk is: weaker-than-expected margins from the company’s aggressive push to gain user market share. Key assumptions Company background Qihoo is a software company in China mainly focus on antivirus software and web browser. On March 30, 2011, Qihoo was listed on New York Stock Exchange with symbol "QIHU". The company offers Internet and mobile security products for free, which has resulted in the company developing a large and loyal user base. Qihoo 360 monetizes this primarily by offering different forms of online marketing and Internet value-added services. 2012 Web game paying users 2013E 2014E 2015E 238,000 404,600 525,980 683,774 Revenue per paying user 485 556 516 516 Security software monthly active users (m) 456 474 493 513 Avg daily UV to 360 navigation page (m) 91 109 126 138 0 95.0 230.0 310.0 Cost of revenues/revenue ratio (%) 10.0 12.7 12.7 12.7 Operating expenses/revenue ratio (%) 76.0 79.8 70.7 63.0 Advertising revenue - search engine Sources: Company data; BNP Paribas estimates Earnings sensitivity Principal activities by revenue (2013E) ------ Base ------ Other services 0.6% Internet valueadded services 37.7% 2013E 2014E Total revenue 30,659 40,594 33,725 44,653 27,594 36,534 EPS (RMB) 294.74 387.85 321.12 420.80 268.37 354.90 8.9 8.5 (8.9) (8.5) 64.5 66.6 61.3 63.3 67.7 69.9 294.74 387.85 331.90 437.19 257.59 338.52 12.6 12.7 (12.6) (12.7) Change in total revenue (%) 0 0 10 10 (10) (10) Change in operating expenses 0 0 (5 ppt) (5 ppt) 5 ppt 5 ppt Change (%) Sources: BNP Paribas estimates Age Since Title 42 2006 Chairman of the board and chief executive officer Xiangdong Qi 48 2006 Director and president Shu Cao 37 2006 Director and chief engineer Alex Zuoli Xu 44 2011 Co-chief financial officer Jue Yao 39 2012 Co-chief financial officer 59 2014E EPS (RMB) Hongyi Zhou http://www.360.cn 2013E Operating expenses/revenue ratio Key executives ------ Worst ------ 2014E Change (%) Online advertising 61.7% ------ Best ------ 2013E Key earnings drivers include total revenue and operating expenses/revenue ratio A 10% increase in total revenue would raise 2013E/2014E EPS by 6.8%/10.0%, while a 10% decrease in total revenue would reduce 2013E and 2014E EPS by 10.3% and 9.1% A 5ppt decline in operating expenses/revenue ratio would lift 2013E and 2014E EPS by 58.1% and 29.1%, while a 5ppt rise in operating expenses/revenue ratio would decrease 2013E/2014E EPS by 57.3%/28.1% BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Revenue growth driven by online advertising and gaming Since Qihoo 360 was incorporated in 2005, the company has successfully built up its user base for its various product offerings and continues to enhance its ability to monetize services. Key revenue contributors are its online advertising, which is primarily driven by its Internet navigation site, and Internet value-added services (VAS), which is primarily driven by PC online games. During 1Q13, online advertising contributed 57.7% of Qihoo 360’s revenue, and Internet VAS contributed 41.7% (up from 34.2% in 4Q12 with a number of new online games stimulating user activity). EXHIBIT 1: Qihoo 360 revenue trends (%) Online advertising Internet value-added services EXHIBIT 2: Qihoo 360 revenue contribution (1Q13) Others Others 0.6% 120 100 80 Internet value-added services 41.7% 60 40 Online advertising 57.7% 20 0 4Q10 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 Source: Company Source: Company Organic user growth reaching a plateau Qihoo 360 has built a strong presence in the China PC market, initially based on its free-to-download PC antivirus software since 2008 (this software, according to the company, currently runs on 96% of China’s PCs with 457m active users as of 1Q13). Through this solid penetration, the company has managed to attract 332m users to install its 360 Safe Browser. Through these browsers, Qihoo 360 directs traffic to its own Internet navigation site; the browsers effectively provide users a window into Qihoo 360 Internet content. Prior to August 2012, the 360 navigation site had Baidu as the default search engine for Internet users that initiate search queries from the site. However, in August 2012, Qihoo launched its own search engine and changed the default search engine to its own, and its search market share (by usage) has been 15% in the past few quarters (Exhibit 5). Qihoo 360 has a strong presence in China through its anti-virus software, browser, and navigation website; organic market share gain in Internet search could be challenging We believe further organic growth of its search engine market share will become more difficult. As we follow the user conversion process (Exhibit 3), it starts with security software users and, as it has already penetrated 96% of PCs in China, we believe upside potential through this avenue is limited. At the browser level, Qihoo’s market share by browser usage has remained at around 25% since 4Q11 (Exhibit 4). We do not see any catalysts to drive usage share significantly higher and believe this will limit the incremental growth of navigation page users. As Qihoo’s navigation page is a major source for its search engine traffic, we believe organic expansion of search market share depends on users being more aware of the Qihoo 360 search engine and switching from competing search engines. Such a change in user behaviour could take time. As such, we believe near-term incremental gain in search market share can only come through partnerships with other Internet businesses, or through acquisitions of popular Internet businesses. In either case, the associated costs would disrupt the company’s cost structure. 60 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin EXHIBIT 3: Qihoo 360 user base EXHIBIT 4: China web browser market share by usage Number of security software active users 457m (96% penetration) 90% IE Qihoo Safari Chrome Sogou 80% 70% Number of 360 safe browser users 332m (70% penetration; ~25% usage share) 60% 50% 40% 30% Number of visitors to 360 navigation page – 94m 20% 10% 2Q13 1Q13 4Q12 3Q12 2Q12 1Q12 4Q11 3Q11 2Q11 1Q11 4Q10 3Q10 1Q10 Source: BNP Paribas estimates 2Q10 0% Search usage market share 15% Source: CNZZ EXHIBIT 5: China Internet search engine market share by usage Baidu Qihoo Sogou Soso Google 100% 80% 60% 40% 20% 0% 3Q10 4Q10 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 2Q13 Source: CNZZ Monetization of Internet navigation site not easy to scale up The company’s current ability to monetize its Internet services is primarily through advertising revenues generated from its 360 navigation website and search engine. However, as the company’s search engine is a new enterprise, its contribution is still just <5% of Qihoo 360’s total revenue (as of 1Q13) as it is still building up its advertising customer base. Given the current deficiency in company’s ability to monetize its search traffic, it has partnered with Google to leverage Google’s online advertising customer base. The company has indicated that revenue sharing between the two partners exists, however details of the arrangement have not been publicly disclosed. We expect the revenue profile of Qihoo 360’s Internet search business to change, but for this to be a gradual process. We assume search engine revenue represents 26% of Qihoo 360’s total online advertising revenue in 2013, and rises to 43% in 2014. The company’s Internet services are monetized primarily through Internet navigation website ad revenues Advertising revenue from Qihoo 360’s navigation site has been the company’s key contributor. However, we believe growth could slow from +80% y-y in 2012 to 22% yy in 2013. Our analysis suggests that, while both the growth in the number of unique visitors and of daily clicks remained robust at +22.1%/+65.8% y-y in 1Q13 (Exhibit 6), the quality of growth has slowed significantly. Advertising revenue/unique visitor has slowed y-y since 2Q12, to 5.4% growth in 1Q13. Advertising revenue/click has also slowed over the same period, to decline 22.4% in 1Q13 (Exhibit 7). In our view, this suggests that strong advertising revenue growth can only be achieved by the continued expansion of its navigation site user base, but we expect this to actually decelerate. 61 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin EXHIBIT 6: Qihoo 360 navigation site traffic (m) Avg daily unique visitors EXHIBIT 7: Qihoo 360 navigation site revenue metrics Avg daily clicks (RMB) 600 Ad revenue/click Ad revenue/UV 1.0 500 0.8 400 0.6 300 0.4 200 0.2 100 0 0.0 4Q10 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 Source: Company 4Q10 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 Sources: Company; BNP Paribas Non-organic methods for driving more traffic We believe Qihoo 360 will explore other methods for driving user traffic, either through partnerships or acquisitions. In May 2013, the company partnered with Alibaba (not listed) to create a Qihoo 360 variant of the www.etao.com e-commerce website, called 360.etao.com. While the merchandise offered for the two locations is practically the same, the 360.etao.com website includes links to Qihoo 360’s other Internet properties, including the search engine. In our view, the intent is to both build awareness of the Qihoo 360 brand name and to direct search traffic from the etao.com website to Qihoo 360’s search engine. Exploring partnerships or acquisitions to drive incremental search traffic In our view, any partnership or acquisition for Qihoo 360’s search business would be for the purpose of expanding user traffic, or expanding advertising customer base. Through either approach, however, the incremental traffic would come at a cost either in the form of revenue sharing or upfront acquisition cost. Sohu (SOHU US; NR) has publically disclosed its intention to sell its search engine, www.sogou.com. We see limited synergies for Qihoo 360 as combining sites does not necessarily result in more search traffic without securing the referring websites. However, it may derive some synergies by expanding its advertising customer base – Qihoo 360 is still building its base up, so Sogou’s base could prove valuable. That said, separating the customer base between www.sohu.com and www.sogou.com might prove complicated as both depend on online advertising for revenue generation. Gaming – the key revenue growth driver for the near-term During 1Q13, the gaming segment contributed 33.5% of Qihoo 360’s total revenue, up from 25.6% in 2012. The gaming segment has registered triple digit revenue growth since 2011, and we estimate it will grow 112% in 2013. While this growth has been remarkable, we expect it to gradually decelerate in the coming years. We assume the company’s gaming revenue grows 34% in 2014 and 30% in 2015, with its revenue contribution falling from 30.2% in 2013E, to 28.5% in 2014E and 29.8% in 2015E. Gaming contributed 33.5% of 1Q13 total revenue Our analysis suggests that Qihoo 360’s gaming revenue mostly correlates with the number of games it offers (Exhibit 8); this has a direct impact on the number of paying users. We see less correlation between the number of games offered and gaming revenue per user as quarterly gaming revenue per user has fluctuated between RMB125 and RMB145 from 4Q11 to 1Q13 (Exhibit 9). This implies that in order for Qihoo 360’s gaming revenue to continue its growth, it needs to expand its portfolio of games from the 1Q13 level of 150. In our forecasts, we assume the number of paying users grows by 70% y-y in 4Q13 to 405k, up from 281k in 1Q13. We estimate growth normalizes at 30% y-y for 2014 and 2015. We assume quarterly revenue per paying user remains RMB130. 62 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin EXHIBIT 8: Number of games and paying users (nos) Web game paying users (LHS) 300,000 Number of games (RHS) EXHIBIT 9: Game quarterly revenue per paying user (nos) 160 140 250,000 120 (RMB m) 145 140 135 130 200,000 100 150,000 80 100,000 50,000 0 4Q11 1Q12 2Q12 3Q12 4Q12 125 120 60 115 40 110 20 105 0 100 4Q11 1Q13 1Q12 2Q12 3Q12 4Q12 1Q13 Sources: Company; BNP Paribas Source: Company The market has high expectations of Qihoo 360’s operating leverage As Qihoo 360 is expanding its monetization capabilities, we forecast its revenue to grow 80% y-y in 2013, 42.5% y-y in 2014 and 24% y-y in 2015. We expect such rapid revenue expansion to deliver operating leverage and lower its expense ratios for sales and marketing, product development, and G&A in 2014. We estimate its operating profit will decline a modest 2.2% in 2013 as its sales and marketing expense ratio increases to 24.1% from 17.7% in 2012. We estimate its operating profit will grow 202.5% y-y in 2014 and 79.3% y-y in 2015. While the growth outlook is strong, we believe expectations are high as reflected in its shares high valuation. Rapid revenue expansion to deliver operating leverage and lower expense ratios In our view, the key risks to such high expectations are its sales and marketing expense ratio and its product development expense ratio both remaining high. For sales and marketing expenses, this risk is that the company needs to increasingly depend on external sources for traffic growth. For product development expenses, the company could need to continue investing in product development to close the gap with larger competitors; and, given Qihoo 360’s relatively smaller scale, the expense ratio could remain high in the near term. While we forecast Qihoo 360’s operating margin will expand from 8% in 2013 to 24.7% in 2015, we believe the downside risks outweigh the upside risks given the current state of the company. In our view, any disappointment in margins will lead to share price weakness given the market’s high expectations. EXHIBIT 10: Qihoo 360 y-y growth rates 450% EXHIBIT 11: Qihoo 360’s expense ratio trends Revenue Operating profit Cost of revenues Selling and marketing Net profit Non-GAAP NP G&A Product development 50% 400% 45% 350% 40% 300% 35% 250% 30% 200% 25% 150% 20% 100% 15% 50% 10% 0% 5% (50%) 0% 2010 2011 2012 Sources: Company; BNP Paribas estimates 63 2013E 2014E 2015E 2010 2011 2012 2013E 2014E 2015E Sources: Company; BNP Paribas estimates BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Return on equity rising on operating leverage We forecast Qihoo 360’s ROE to be on an upward trend primarily driven by expanding net margin; we estimate its net margin will bottom out in 2013E at 7.0%, and expand to 20.6% in 2015E as operating leverage increases with revenue rising as the company begins to monetize its services from earlier investments. These efforts also contribute to our forecast rise in asset turnover. We expect it to rise from 47.7% in 2012 to 81% in 2013E, and to remain at this high level. This would lead to ROAE rising from 7.9% in 2013E to 24.6% in 2015E. EXHIBIT 12: Qihoo 360 ROAE decomposition 2011 NI/Sales 2012 2013E 2014E 2015E (%) (%) (%) (%) (%) 9.3 14.2 7.0 14.4 20.6 Sales/Assets 39.6 47.7 81.3 85.7 79.5 Assets/Equity 190.8 162.3 139.0 150.0 150.4 7.0 11.0 7.9 18.5 24.6 ROAE Sources: Company; BNP Paribas estimates 64 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Increasing FCF and expanding net cash position With Qihoo 360’s profitability rising, we forecast its FCF yield will reach 2.5% in 2015, up from -0.7% in 2013E. This would lead to the strengthening of its balance sheet with its net debt/equity ratio rising from -55% in 2013E to -66.2% in 2015E. Given the relatively small scale of Qihoo 360’s operations, this forecast is highly dependent on how its business model evolves; significantly weaker profitability could have a major impact on the company’s FCF outlook. We believe its strengthening net cash position could enable Qihoo 360 to explore expansion strategies through acquisitions. Strengthening net cash position could enable expansion strategies through acquisitions EXHIBIT 13: Qihoo 360's balance sheet 2010 2011 2012 2013E 2014E 2015E (USD m) (USD m) (USD m) (USD m) (USD m) (USD m) Cash & cash equivalents + Bank deposits 60 Current Assets 12 31 Total Current Assets 73 374 Non-Current Assets 15 50 Total Assets 88 424 2 Current Liabilities Total Current Liabilities ST borrowings / Current portion of LT borrowings LT borrowings Other LT Liabilites 344 381 312 442 667 55 96 148 193 435 409 589 861 254 320 396 456 690 729 985 1,317 0 - - - - 12 46 204 151 235 300 14 47 204 151 235 300 0 - - - - - 1 6 8 8 8 8 Total Liabilities 15 52 211 158 243 308 Share Capital 0 0 0 0 0 0 Reserves 2 371 478 570 742 1,008 Minority Interest 1 1 0 0 0 0 Total Equity 73 371 478 571 742 1,009 Total Equity & Liabilities 88 424 690 729 985 1,317 Source: Company EXHIBIT 14: Qihoo 360's cash flow 2010 2011 2012 2013E 2014E 2015E (USD m) (USD m) (USD m) (USD m) (USD m) (USD m) Net Profits 9 16 47 42 121 216 Depreciation 1 4 16 25 38 53 Change in working cap (4) (6) (15) (16) (15) Other operating (0) (3) 17 - - - Operating cash flow 10 13 74 52 144 254 0 3 7 7 7 7 Taxes paid (0) (11) (11) (9) (26) (46) Capex (3) (16) (69) (91) (114) (114) 7 (3) 6 (39) 30 140 Free cash flow yield (%) 0.3 (0.1) 0.1 (0.7) 0.5 2.5 Dividend yield (%) 0.0 0.0 0.0 0.0 0.0 0.0 Capex/sales (%) 4.9 9.7 20.9 15.4 13.5 10.9 (80.2) (92.4) (79.6) (54.7) (59.5) (66.2) Interest Inc / (Exp) Free cash flow Net debt to equity (%) Source: Company 65 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Operating leverage driving margin expansion We estimate Qihoo 360’s revenue will reach USD1,046m by 2015 (3.2x its 2012 revenue) and its operating margin will rise to 24.7% in 2015, from 14.8% in 2012 and 8% in 2013. We attribute the rapid margin expansion to strong revenue growth in online gaming (112% y-y in 2013 and 34% in 2014) and online advertising (65% and 48%). While we forecast strong margin expansion, we note this could vary significantly depending on the company’s implementation of strategies to gain market share in online gaming and Internet search. Rapid margin expansion due to strong revenue growth in online gaming and advertising EXHIBIT 15: Qihoo 360's income statement 2010 2011 2012 2013E 2014E 2015E (USD m) (USD m) (USD m) (USD m) (USD m) (USD m) Revenue 58 168 329 592 844 1,046 Cost Of Goods Sold (7) (19) (33) (75) (107) (133) Gross Profit 51 149 296 517 737 913 (42) (131) (250) (472) (597) (659) 9 18 49 48 144 258 OpEx Operating Income Profit before tax & Minority Interest 9 26 63 55 151 265 (0) (11) (11) (9) (26) (46) Minority Interest 0 1 0 (0) - - Net Income 9 16 47 42 121 216 Tax Share based compensation 4 48 51 51 51 51 12 63 97 92 172 266 1 4 16 25 38 53 10 22 65 73 182 312 15.6 10.8 14.8 8.0 17.1 24.7 EBITDA Margin (%) 18.0 13.3 19.8 12.3 21.6 29.8 Pre-tax Margin (%) 15.6 15.4 19.1 9.3 17.9 25.4 Net Margin (%) 14.8 9.3 14.2 7.0 14.4 20.6 Non-GAAP Net Income Depreciation & Amortization EBITDA Operating Margin (%) Source: Company We initiate at HOLD with a target price of USD44.20 We estimate Qihoo 360’s GAAP EPS at USD0.23 per ordinary share for 2013, USD0.68 for 2014 and USD1.21 for 2015; we estimate non-GAAP EPS at USD0.52, USD0.97, USD1.50 for 2013-15. We initiate coverage on the stock with a HOLD recommendation and DCF-based target price of USD44.2. Our target price implies 2013/14E PE multiples of 125.9x/43.2x. The stock is currently trading at 2013/14E PE multiples of 125.7.2x/43.2x. While we forecast Qihoo 360 will see significant earnings growth in 2014 (+191.2%) and 2015 (+77.7%), we believe this is fully in the price given the high PE multiple. Also, as the forecast implies significant margin expansion (8% in 2013 to 24.7% in 2015), we see downside risks as the company will increasingly need to expand its user base through non-organic methods such as partnerships or M&A. Such approaches are likely to have a material impact on its profitability or balance sheet. Key upside risks to our call are: the company managing to grow its user base (online gaming, navigation site and search) more than we expect without significant margin contraction; and expansion of online advertising customers for its search product. Key downside risks are: weaker-than-expected margins from its aggressive push to gain user market share. We also expect share price to be volatile and may respond to Internet search market share in the short-term. 66 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin In our DCF valuation, we assume a medium-term revenue CAGR of 14.9%, and terminal revenue growth rate of 5%. We assume an average medium-term EBIT margin of 27.5%, and terminal EBIT margin of 30%. We use a WACC of 10.5%. EXHIBIT 16: DCF assumptions and derivation of TP Mid-term WACC (%) Terminal 10.5 Revenue growth rate (%) 14.9 5.0 EBIT margin (%) 27.5 25.0 Source: Company EXHIBIT 17: Qihoo 360 P/E band EXHIBIT 18: Qihoo 360 P/B band (HKD) 80 (HKD) 60 100x 70 60 75x 50 40 50x 50 9.0x 40 7.5x 6.0x 30 4.5x 30 25x 20 3.0x 10 10 0 Mar-11 20 Mar-12 1.5x 0 Mar-11 Mar-13 Sources: Bloomberg; BNP Paribas estimates Mar-12 Mar-13 Sources: Bloomberg; BNP Paribas estimates EXHIBIT 19: Qihoo quarterly P&L 1Q13A 2Q13E 3Q13E 4Q13E 2013E (USD m) (USD m) (USD m) (USD m) (USD m) Total revenues 109.9 142.6 160.3 179.5 592.3 Cost of revenues (13.9) (18.1) (20.4) (22.8) (75.2) 96.0 124.5 140.0 156.7 517.1 Selling, general and administrative (39.0) (48.5) (53.7) (59.2) (200.4) Research and development (50.2) (64.2) (72.1) (80.8) (267.3) Operating profit 6.8 11.4 13.5 15.9 47.6 Income before income taxes 8.6 13.2 15.3 17.7 54.8 Gross profit Net income 5.4 10.2 12.0 14.0 41.7 17.5 10.2 12.0 14.0 92.3 87.3 87.3 87.3 87.3 87.3 Operating margin 6.2 8.0 8.4 8.9 8.0 Net margin 4.9 7.2 7.5 7.8 7.0 Non-GAAP net income (excludes sharebased compensation) Margins (%) Gross margin Sources: Company; BNP Paribas estimates 67 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Company Profile History and development of the company Qihoo Technology Company Limited, incorporated in June 2005, had its name changed to Qihoo 360 Technology Co Ltd on 31 December 2010. The company operates in China through its wholly-owned subsidiaries and affiliated entities (Exhibit 13). Qihoo 360 listed its ADS on the New York Stock Exchange under the ticker “QIHU” on 30 March 2011 and soon after, in April 2011, the company completed its initial public offering of its ADS. Business overview The company offers Internet and mobile security products for free, which has resulted in the company developing a large and loyal user base. Qihoo 360 monetizes this primarily by offering different forms of online marketing and Internet value-added services. Qihoo 360’s product offerings can be segmented into three layers of user applications (Exhibit 16): § Core security layer: Qihoo’s many PC and mobile security software products offered to users for free through which Qihoo 360 has acquired a solid user base and 96% penetration of China PCs. § Access layer: Internet products and services users use to access the Internet. Its key products are its hao.360.cn Internet navigation site and its 360 Secure browser. Qihoo 360’s current online advertising revenue is mostly generated through its navigation page. § Service layer: Various Internet services where there is potential for monetization. Key services include its 360 Search and online gaming. Qihoo 360 is expanding its search service to enhance its ability to monetize online advertising revenue. Given Qihoo 360’s smaller scale than large players such as Baidu, the company leverages partnership relationships to expand its product offerings. Currently, Qihoo 360 is expanding its business through co-operations with well-recognized vertical websites, such as Alibaba’s etao.com, xiami.com and amap.com, and forming partnerships with Nokia and Google for the mobile phone and search markets, respectively. EXHIBIT 20: Qihoo 360's core Internet and mobile products Source: Qihoo 360 Competition Besides Internet security product and service providers, Qihoo competes with PRCbased Internet companies. According to iReseach, Qihoo is the No 1 provider of Internet and mobile security products in China based on user base. At the end of 1Q13, its PC Internet security product had 457m monthly active users, representing a penetration rate of 96%, according to the company. Within the China Internet market, Qihoo’s primary competitors include Tencent (700 HK, NR), the largest Internet company in China in terms of revenue, which generates most of its revenues from online gaming. In the China Internet search industry, Baidu (BIDU US, BUY) and Sohu’s (SOHU US, NR) Sogou.com are the key competitors. 68 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Organizational structure and subsidiaries EXHIBIT 21: Qihoo 360 organizational structure and subsidiaries Qihoo 360 Technology Co. Ltd (Cayman Islands) 100% 100% 360 International Development Co. Limited (Hong Kong) 100% Qihu 360 Software Co., Limited (Hong Kong) Qiji International Development Limited (Hong Kong) 100% 100% Qizhi Software (Beijing) Co., Ltd) Tianjin Qisi Technology Co., Ltd 30% 100% Qifei International Development Co. Limited (Hong Kong) 100% Qifei Xiangyi (Beijing) Software Co., Ltd 70% Beijing Qichuang Yousheng Keji Co., Ltd Beijing Qihu Technology Co., Ltd (PRC) Beijing Star World Technology Co., Ltd. (PRC) Eleven Other VIEs (PRC) Source: Company 69 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Directors and principal shareholders The company’s board consists of three non-independent directors and six independent (a total of nine directors). Mr Hongyi Zhou is CEO of Qihoo and chairman of the board (he owns 17.7% of outstanding shares). Mr Xiangdong Qi and Mr Shu Cao are the other two non-independent directors (9.5% and 4.9% of shares respectively). Mr Neil Nanpeng Shen is an independent director and chairman of the Compensation Committee (5.4% of shares). Mr Xiaohong Shi is vice president of technology (1.2% of shares). Mr Alex Zuoli Xu, Mr Jue Yao and the other five independent directors have combined share ownership of 0.6%. Of the principal shareholders, Global Village Associates Limited is a company controlled by Mr Hongyi Zhou. The 17.7% of Qihoo’s ordinary shares owned by Global Village Associates Limited are included in Mr Hongyi Zhou’s individual ownership. Young Vision Group Limited and Sequoia Capital China I, LP are whollyowned by Mr Xiangdong Qi and Mr Neil Nanpeng Shen, respectively. Mr Shujun Li is the managing partner of Trustbridge Partners III, LP currently. EXHIBIT 22: Qihoo 360 Directors and Executive Officers Directors and Executive Officers: Position/Title Shares Beneficially Owned Hongyi Zhou Chairman of the board and chief executive officer Xiangdong Qi Director and president 9.5 Shu Cao Director and chief engineer 4.9 Neil Nanpeng Shen Independent Director 5.4 Wenjiang Chen Independent Director 0.6 (combined with other independent directors and CFOs) Hong Chuan Thor Independent Director Ming Huang Independent Director Shujun Li Independent Director (%) William Mark Evans Independent Director Alex Zuoli Xu Co-chief financial officer Jue Yao Co-chief financial officer Xiaohong Shi Vice president of technology 17.7 1.2 Source: Company EXHIBIT 23: Qihoo 360 principal shareholders Principal Shareholders: Shares Owned (%) Global Village Associates Limited 17.7 Young Vision Group Limited 8.6 Sequoia Capital China I, L.P. and its affiliates 5.4 Trustbridge Partners III, L.P. and its affiliates 5.4 Source: Company 70 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Internet search engine – a high-performance subset of Ad Exchange We view the Internet search engine industry as a subset of the Ad Exchange mechanism, as search engines are capable of serving highly targeted advertisements to Internet users with particular interest in a subject. As discussed in earlier sections, advertisements served through Internet search platforms have significantly higher CTR than display advertisements. This makes large Internet search platforms highly valuable Internet properties for advertising purposes. Internet search engines enable the targeting of Internet users with relevant ads, leading to high CTR with potential to extend into new products In markets outside of China, Google is the recognised leader in Internet search. In the US, it represented 66.5% of search market share in April 2013, far ahead of the second largest search platform, Microsoft’s Bing with 17.3% market share (according to comScore). Google’s success, however, has not been replicated in China, where Baidu remains the dominant search platform with 70% market share (by usage) in 1Q13; we attribute Baidu’s dominance to its strength in Chinese language search and government policies that potentially put Google in a weaker market position in China. We believe Baidu will continue to be a dominant search platform for the China market, given the strong positive feedback loop effect of the search industry (Exhibit 27). Beyond the core Internet search activities of the existing search platforms in China, we also expect them to begin to develop the Ad Exchange business model and increasingly become broader online advertising platforms. This should create new revenue opportunities for Chinese search platforms as they already have existing online advertising customers and have the potential to rapidly tap into new inventories of Internet properties through Ad Exchanges. In 2012, Google Network Members generated 28.5% of Google’s total advertising revenue; Baidu still depends on its core search business for over 90% of revenue. China Internet search engine – industry still on a strong upward trend China’s Internet penetration has risen from 16% at the end of 2007 to 42.1% in 2012; over the same period, Internet search industry revenue has increased from RMB2.9b to RMB29.3b. The five-year CAGR for China’s Internet user growth through 2012 was 21.8%, while the search engine industry revenue CAGR was 58.8% (Exhibit 24). This implies that while the number of Internet users in China rose rapidly, revenue generated per user also rose from RMB13.8 to RMB52. This indicates that the industry has already hit an inflection point; in the early phase of the search engine industry, the number of Internet users was the primary driver of search industry revenue; however, the main growth driver has now shifted to search revenue per Internet user, driven by the increasing number of online advertisers. While Internet penetration growth in China has been impressive in recent years, we see no signs of this tapering off, as penetration of 42.1% at the end of 2012 remains low compared with that in mature markets such as Japan and Korea; both are at c80% penetration, and Taiwan, Singapore, and Hong Kong are at c75% penetration (Exhibit 25). The US has the lowest penetration of the mature markets at 78%. We believe China’s aggressive upgrades of its fixed-line broadband network in recent years will continue to drive up Internet penetration; in addition, the rise in smartphone adoption presents incremental opportunities for the Internet search industry. In our view, these trends set the foundation for the continuing growth of China’s search engine industry. 71 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin EXHIBIT 24: Number of Internet users and search industry revenue EXHIBIT 25: Internet penetration comparison (2011) (Percent of individuals using the Internet) Number of Internet users (LHS) 100 China 10 5 0 0 2007 2008 2009 2010 Sources: CEIC; China Internet Watch; BNP Paribas 2011 2012 Korea 15 200 Germany 20 300 France 25 400 Japan 30 U.S. 500 Hong Kong 35 Taiwan 600 90 80 70 60 50 40 30 20 10 0 Singapore Search industry revenue (RHS) (%) (RMB b) Spain (m) Source: ITU Industry growth should continue to outpace Internet user growth In our view, the increasing maturity and efficiency of the Internet search industry provide the fundamental support for search industry revenue growth to outpace Internet user growth in China. We illustrate the value creation process of the industry in Exhibit 26. The products being sold by search engines are their inventories of user-specific search queries. Based on the user profiles and activities, search engines create value by funnelling users with specific profiles to paying advertisers that seek a specific audience. The ultimate customers of the search engine industry are the advertisers that promote their products and services to targeted sets of Internet users. Search engines’ positive feedback loop, with increasing search users and increasing advertisers, should continue to drive industry value Internet search industry dynamics closely follow the academic definition of an information economy, where at the macro level the value of the industry increases exponentially as the number of industry participants increases. As illustrated in Exhibit 27, the rising number of Internet users drives growth of online advertisers. This leads to rising quality of search results, which in turn drives the growth of Internet users; as the numbers of products (Internet users and Internet properties) and customers (advertisers) rise, the probability of matching the right product to the right customer also rises, leading to higher revenue per product offered. Specific search engines also play an important role in raising the quality of search through enhanced profiling of Internet users. This can be achieved by collecting indepth data about the particular user and analysing the data to better understand and predict a user’s interests, habits, and behaviour. Such data collections and analytics can further enhance the matching of Internet users to paying advertisers, leading to higher conversion rates, or click-thru-rates (CTR). This provides more potential for raising the value of the industry as advertisers are willing to pay more to gain exposure to the relevant audience. Search engines create incremental value by applying user data and analytics in search algorithms to maximise CTR In our view, the combination of a rising number of Internet users and improving search quality through data analytics provides a solid foundation for continuing industry revenue growth, as evident in Exhibit 24. 72 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin EXHIBIT 26: Internet search engine value chain Inventory: search queries and user profiles Value Creation Search Engines Advertising Distributors Advertisers Source: BNP Paribas EXHIBIT 27: Positive feedback loop for Internet search value chain Rising number of Internet users drives higher number of advertisers Number Internet users Number advertisers Rising number of advertisers drives higher number of Internet users Source: BNP Paribas Industry outlook – no sign of a slowdown To analyse the relationship between search industry revenue and general Internet usage, and to quantify industry potential, we have developed a regression model. As the model is based on historical data, the potential contributions from Ad Exchange and mobile Internet are not fully factored in as they are still new business models in China; they could offer upside potential as they gain momentum. We develop a regression model to project search industry revenue growth – we forecast a 2012-15 CAGR of 31.9% Our regression model forecasts an industry revenue CAGR of 31.9% between 2012 and 2015, vs 58.8% for the five years between 2007 and 2012, as the comparison base has grown significantly in recent years (Exhibit 28). A major component of this is search revenue per Internet user, for which we forecast a CAGR of 25.8% over the period vs CAGR of 30.3% over the past five years. This is consistent with our earlier discussion that the rising numbers of Internet users and online advertisers will drive growth in advertising revenue per user. This is also consistent with Baidu’s disclosure that revenue per online advertising customer rose to RMB37,300 in 2012, from RMB5,800 in 2006, a CAGR of 45.1%. Our regression analysis indicates that industry revenue is most sensitive to Internet search user penetration and the number of Internet search users, while the hours of Internet use per week, seasonality, and number of Internet users are also strong predictors. The number of mobile Internet users yields the lowest t-ratio in our model, suggesting that the growth of mobile Internet users has so far not contributed significantly to industry revenue. This could be a new growth driver as the industry finds mechanisms for monetising mobile Internet usage; we believe the Ad Exchange platform could be a powerful tool with which to more effectively monetise mobile Internet. 73 BNP PARIBAS Regression analysis indicates high sensitivity to search user penetration and number of search users; Internet usage time, seasonality, and number of Internet users also important 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin EXHIBIT 28: Search industry revenue forecasts (RMB b) Search industry revenue (LHS) 40 Search revenue/Internet user (RHS) (m) 60 35 50 30 40 25 20 30 15 20 10 10 5 2H15E 1H15E 2H14E 1H14E 2H13E 1H13E 2H12 1H12 2H11 1H11 2H10 1H10 2H09 1H09 2H08 0 1H08 0 Sources: BNP Paribas estimates, CEIC and China Internet Watch Industry value drivers – combination of search users, advertisers, and analytics A search engine is effectively a factory that packages its inventory, the Internet search users, and sells this inventory to its customers, the advertisers. The search engine’s role ends when it connects the inventory to the customer. Any interactions or transactions that occur beyond this point are made outside the initial connection created by the search engine. There are a number of factors that affect the probability of an interaction taking place after the initial search connection, including the size of the inventory, the number of customers, and the analysis performed by the search algorithm that matched a particular search query/user to a particular advertiser (Exhibit 29). The value of the inventory depends on the quantity and quality of search users (subjective measure used by advertisers, based on user profile and attributes); higher quality search users increase the probability of a transaction between the user and advertiser. Search engines create value when inventory – search users – are sold to customers – the advertisers EXHIBIT 29: Internet search industry value drivers Sourcing/inventory of Internet search users Direct search queries www.baidu.com www.so.com www.sogou.com www.soso.com www.google.com Navigation sites Web/news portals www.sina.com.cn news.sogou.com news.soso.com hao.360.cn www.hao123.com Social Networking Internet utilities Email applications E-Commerce sites Video/entertainment sites Web browsers Etc. Search Engine Distribution of advertising service Advertising distributors Advertisers Advertisers Advertisers Buyers of advertising service Source: BNP Paribas 74 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin The battle to dominate Internet search traffic A starting point for a search engine service is to maintain a stream of search queries, ie, having a large pool of inventory; this maximises the traffic for which advertisers pay. This can be done through capturing direct search queries through search engine websites such as google.com, baidu.com, so.com, and others, or through other Internet services (Exhibit 30). While direct search queries at a search engine website are the most direct mechanism for search engines to acquire users, traffic tends to flow to the dominant search engine brands and be less effective for the secondary search engine brands. Secondary search engines must utilise other mechanisms to acquire search traffic. Aside from direct search queries, search engines tap into other sources of search traffic such as Internet navigation sites including Baidu’s hao123.com and Qihoo 360’s hao.360.cn (Exhibits 30-31). These navigation sites are collections of Internet resources that users frequently utilise to navigate the Internet and often serve as the starting point for users. According to statistics compiled by China Internet Watch, hao.360.cn contributed 11% of search engine traffic in November 2012, while hao123.com contributed 8.2% (Exhibit 34). Search users reach search engines en through direct access or traffic referrals; Internet navigation sites contribute a significant amount of search traffic Popular navigation websites are powerful as users frequently utilise the default search engines within these websites. In 3Q12, Qihoo 360 changed the default search engine of its popular navigation website from Baidu to its own (Exhibit 32), and pushed Baidu to a lower position on the list of search engine selections; this had an immediate impact on Qihoo 360’s search engine market share (Exhibit 33). These navigation sites, while serving as a source for search engine traffic, may also be revenue generating as a standalone service. Qihoo 360’s hao.360.cn contributes the majority of its online advertising revenue currently, according to the company. EXHIBIT 30: Qihoo 360 navigation site EXHIBIT 31: Baidu navigation site Source: hao.360.cn Source: www.hao123.com 75 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin EXHIBIT 32: Qihoo 360 navigation site default search engine EXHIBIT 33: China Internet search market share by page views 100% Baidu Qihoo Soso Google Sogou 80% 60% 40% Chart or other exhibit 20% Source: hao.360.cn 2Q13 1Q13 4Q12 3Q12 2Q12 1Q12 4Q11 3Q11 2Q11 1Q11 4Q10 3Q10 0% Source: cnzz.com Other sources of search traffic come from online video/entertainment, social networking, news portals, and others. Exhibit 35 lists the popular Internet services in China that can serve as sources of search traffic. In our view, the popularity of online video services is a major reason for Baidu owning video service providers iQiyi and PPS; while these services may not be profitable as standalone entities currently, they could be important search traffic sources when integrated into Baidu’s product portfolio. In addition, premium content could generate additional revenues as the online entertainment industry matures. Internet services such as entertainment, social networking and news portals also contribute to search traffic In addition to having a strong position in navigation websites, Qihoo 360 also has a solid market share in China’s web browsers; in recent quarters, Qihoo 360’s Secure Browser maintained a market share of c25%, according to cnzz.com. These browsers, when installed, default the home page to Qihoo’s hao.360.cn unless the user changes the home page, but it also prompts users periodically to set Qihoo 360’s navigation as the home page. In conjunction with Qihoo’s popular antivirus software, Qihoo has a dominant position on Internet users’ desktops. Despite the initial success in gaining market share in 3Q-4Q12, the company’s ability to continue converting navigation site users to its own search engine users is yet to be seen. EXHIBIT 34: Navigation websites’ contribution to search engine traffic Contribution to search engines (Nov 2012) (%) hao.360.cn 11 hao123.com 8.20 web.sogou.com 6.50 123.duba.net 4.20 2345.com 1.40 hao.uc.cn 0.90 Source: China Internet Watch 76 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin EXHIBIT 35: Top Internet services in China Top Internet services in China (Jan 2013, million users) Search service 454.0 Online video 452.0 Community friends 439.0 News information 413.0 E-commerce 404.4 Entertainment 403.8 IT digital products 395.0 Media domain 362.1 Financial service 361.5 Email 356.0 Source: China Internet Watch Expanding online advertising footprint Another key value-driver for the search engine industry is its advertising footprint. While search engines are designed to operate without geographical boundaries, a search engine service provider’s ability to monetise its service depends on its ability to distribute its advertising service. The search engine service provider has the ability to reach advertising customers through direct channels or via advertising agencies that act as distributors of the service. Expansion of advertising footprint critical to monetising search services; search engines leverage direct sales force and ad agencies to expand footprint China’s online advertising market continued to expand in recent quarters, with 1Q13 growing 39% y-y as advertisers increasingly shifted marketing budgets toward online media, according to China Internet Watch data. Within the overall online advertising market, search engine remains the largest category of spending, representing 34% of total industry revenue in 1Q13; the category grew 36% y-y (Exhibit 36). The second largest category of online advertising is vertical search, which represented 23% of total industry revenue in 1Q13, +84.3% y-y. While growth of the online industry has been strong, we see no slowdown as the number of Internet users continues to rise and advertisers increasingly find online advertising to be more effective in targeting customers. Despite the benefits of online advertising, we believe the number of advertisers can only grow as fast as the major online platforms can expand and reach them. At the end of 1Q13, Baidu had 410,000 online advertising customers, +28% y-y; Qihoo 360 currently has less than 10,000 as it had just begun to monetise its Internet search business. While Qihoo 360 aims to rapidly increase its search advertising customers, its scale could be limited by its search users. On the other hand, given Qihoo 360’s smaller search market share, its lower advertising price could serve a niche market of smaller advertisers. Both Baidu and Qihoo 360 leverage direct internal sales forces in major markets and external advertising agencies in other markets to reach additional advertising customers. In the case of Baidu, we estimate the top three markets, Beijing, Shanghai, and Guandong, represent over 50% of advertising revenue and are mostly covered by its direct sales force. While Qihoo is still building up its customer base, it is also primarily targeting tier-1 markets with its internal sales force. 77 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin EXHIBIT 36: China online advertising market in 1Q13 Search Engine (RMB b) 25 Video Vertical Search Rich Media EXHIBIT 37: Baidu’s online advertising statistics No of online marketing customers (LHS) Brand Graphic (nos) Other 20 Average revenue per customer (RHS) 40,000 600,000 35,000 30,000 500,000 15 (RMB) 700,000 25,000 400,000 20,000 300,000 10 15,000 200,000 5 10,000 100,000 5,000 0 0 Source: China Internet Watch 0 2006 2007 2008 2009 2010 2011 2012 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 Source: Baidu Improving advertising performance through Internet user profiles In an effort to improve the effectiveness of search results and search advertising, search engines could leverage contextual data about the user in the search algorithm. Such data could include a user’s location, historical Internet activities or other personal data. When utilised in online advertising, this leads to more targeted ads. The popular Internet services Baidu operates could serve the purpose of collecting user data and profiling users; these services include Baidu Knows, Baidu Post Bar, Qunar travel service, and the numerous video services such as iQiyi and the recently acquired PPS. Fine-tuning search performance through extensive user data collection and analytics; breadth of popular Internet services provides an advantage Partnerships with Internet services that have access to detailed user information can also be powerful. Theoretically, search engines such as Baidu could partner with Internet social networking services to exchange user profiles and improve their advertising effectiveness. This could ultimately lead to higher prices for advertising on the particular search platform. 78 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Competitive landscape in the China search engine industry EXHIBIT 38: Porter's five-forces for Internet search engine Threat of new entrants: Low The challenges for new entrants to reach critical mass in a networked industry inhibits expansion of new entrants as scale in terms of market share is necessary to gain new Internet users and advertising customers. Such dynamics is prohibitive for new entrants to compete against incumbents. Supplier bargaining power: Medium Internet services that feed search users to search engines are the key suppliers; as they are key sources of user traffic, they could be in the position to charge the search engines Through referral fee structure. Others are infrastructure suppliers such as computer servers and network operators which have relatively lower bargaining power. Internal rivalry: Medium Internet search industry conforms closely with the positive feedback characteristics of a networked industry; as the number of users who use its service increases, the value of its service increases. Such dynamic tends to limit smaller rivals to serving niche market segments. Buyer bargaining power: Medium The price of service provided by search engines is generally determined by bidding process where advertising customers bid for search words. The price depends on the search engines’ user market share and its ability to match users to the advertisers. Threat of substitutes: Medium Traditional media are substitutes for online advertising services; however, they have been losing ground to online advertising channels in recent years. Other online channels also compete for advertising budget; they include Internet portals, social networking services, etc. In addition, in mobile Internet environment, browser based search is being displaced by mobile applications. Source: BNP Paribas Top-3 search engines account for over 90% of search traffic The China search engine industry is dominated by the top three search engines, Baidu, Qihoo 360, and Sogou, with a combined search market share of over 90% in the past year, as measured by search usage by cnzz.com. Baidu remains the most dominant player, with over 70% market share in 1Q13; this is a decline from nearly 80% a year earlier. Since August 2012, Qihoo 360 has gained market share to reach nearly 15% as the company leveraged its strong presence on PCs to direct search traffic to its own search engine, instead of to Baidu as it did previously. This was effective for Qihoo 360 as its browser and navigation site have an established user base, and they serve as effective sources of search traffic. The number three search engine, as measured by search usage, Sohu’s Sogou, leverages the many Internet services provided by Sohu, such as sohu.com, as sources of search traffic to achieve a search market share of 7-8%. We expect traffic share among the top three to shift as the two smaller players look for ways to gain traffic share both organically and through partnerships with other Internet services. In May 2013, Qihoo 360 and Alibaba announced a partnership to launch 360.etao.com to position the search service for online shopping; this is intended to both increase vertical search traffic for Qihoo 360 and to raise brand awareness of Qihoo 360 search by leveraging Taobao’s dominance in e-tailing. Higher brand awareness could lead to higher direct search queries via Qihoo 360’s so.com. Qihoo 360 has publicly stated its target of 20% traffic share by the end of 2013. Highly concentrated industry with top three in China representing over 90% of search usage Smaller search engines likely to seek partnerships or acquisitions to increase market share; they could target niche verticals to differentiate themselves from the dominant player It has also been widely reported that Sohu could be seeking to sell its Sogou search engine; in our view, an acquisition could be more complementary to entities other than the existing large search platforms. While the top search engine platforms continue to compete for traffic share, we believe Baidu will maintain its revenue share as its scale has enabled it to gain substantially more online advertising customers than rivals have; Baidu’s 2012 revenue market share was 79% according to China Internet Watch (Exhibit 40). While we believe Qihoo 360 and Sogou could 79 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin gain traffic share in the near term, their ability to fully monetise the service could be more limited and would take time. Beyond the top three, search engines such as Tencent’s Soso, Google, Microsoft’s Bing, and Yahoo have combined market share in China of less than 10%. As we believe the Internet search market strongly exhibits the characteristics of a networked industry, market share gains for smaller players could therefore be limited in the near term. Sourcing of search traffic Key suppliers for the Internet search industry are other Internet service providers that have established user bases, such as Internet portals, navigation sites, social networking services, e-commerce sites, Internet utilities and entertainment sites. Baidu has built up a large number of services with a large user base that enables the company to direct its user base to its own search engine. It also pays search referral fees to partners for search traffic (Exhibit 39); we believe the recent rise in traffic acquisition costs could include the effect of the Ad Exchange product. Other search engine platforms such as Qihoo 360 mostly depend on traffic from their respective navigation sites. However, as discussed earlier, Qihoo 360 is partnering with other Internet platforms such as Alibaba to source search traffic along specific industry verticals. This could imply that the company has exhausted internal resources. While the partnership terms with Alibaba are undisclosed, we believe that as Qihoo 360 aims to become a more significant player, it will increasingly need to source traffic from other platforms in exchange for referral fees. In our view, such an arrangement could be based on Qihoo 360’s ability to successfully monetise its Internet search service. EXHIBIT 39: Baidu's traffic acquisition cost (RMB m) Traffic acquisition cost (LHS) 700 Traffic acquisition cost/Revenue (RHS) 14% 600 12% 500 10% 400 8% 300 6% 200 4% 100 2% 0 0% 1Q10 2Q10 3Q10 4Q10 1Q11 2Q11 3Q11 4Q11 1Q12 2Q12 3Q12 4Q12 1Q13 Source: Company Monetisation of search Once the traffic for a search engine reaches critical mass, search platforms need to address the monetisation aspect of the business. The value of search increases in line with the number of Internet search users; this has been shown in empirical data, as illustrated in Exhibit 41, where search revenue per user rises as the number of search users increases. A similar relationship is found for Baidu, where revenue per online advertising customer increases in tandem with the rising number of online advertising customers. We expect such a pattern to continue for the foreseeable future as the perceived utility of a search engine increases from a user perspective; this should continue to drive online advertising revenue in a self-reinforcing cycle. Empirical data has shown search revenue per user increases with a rising number of search users and number of advertising customers In terms of monetising Internet search, market leader Baidu shows a consistent link, with market share by revenue of 78.5% in 2012 vs its page view market share of 77.6%. As the field of search users approaches saturation, we believe the dominant player, Baidu, will shift its focus to extending its advertising reach to grow its advertising customer base, leading to a rise in revenue per search user. In addition, 80 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin its vast amount of user data to which it has access could enable the search engine to deliver more targeted advertisements, leading to higher advertising fees. Qihoo 360 also needs to expand its online advertising customer base, which is currently less than 10,000. Qihoo 360 has partnered with Google, where the former generates search traffic and the latter provides broad exposure to online advertising customers; as such, Google’s revenue market share in 2012 was 15.6%, despite its limited direct exposure to the search users (Exhibit 40). As Qihoo 360 builds up its own online advertising customer base, it plans to reduce its dependence on Google and capture a larger proportion of the search revenue generated. EXHIBIT 40: Search market share by revenue (2012) SOSO 1.5% Youdao 0.3% Others 1.0% Sogou 3.1% Qihoo 360 starting to build up advertising customers to monetise search services; currently rely on partnership with Google EXHIBIT 41: Continuing growth of revenue/search user (m) Number of search users (LHS) 500 Revenue/search user (RHS) (RMB) 70 450 60 400 Google 15.6% 350 50 300 40 250 Baidu 78.5% 200 30 150 20 100 10 50 0 0 2007 Source: China Internet Watch 2008 2009 2010 2011 2012 Sources: China Internet Watch; CEIC; BNP Paribas Maximising search advertising revenue through high performance searches The Internet search industry primarily generates revenue through online advertising. Given the extensive and dynamic information that exists about Internet users and advertisers, search platforms can maximise advertising price/revenue through numerous pricing strategies to improve effectiveness. The pricing mechanism for such advertising is similar to that for Ad Exchange platforms discussed previously, where advertisers bid for advertising spots based on CPM, CPC, and CPA. Industry value can be increased by optimising search quality – ie, using user data analytics to improve relevance of advertising for search users The placement of a particular advertisement is generally determined by the relevance and quality of an advertisement in a particular search query (in Google terms – Quality Score), and bidding prices from advertisers for the particular search query; these two factors combined determine the position of the particular advertisement in the particular search query (Exhibit 15). Advertising prices are ultimately determined by how effective the advertisement is as measured by the click-through-rate (CTR); we believe improving the advertising relevance can improve the CTR, and this is where search platforms can continue to optimise and differentiate. Exhibit 41 provides a sample for CPC prices for different keywords related to “Used cars”. Such industry dynamics should enable search platforms with large amounts of user data to better monetise Internet search services. Data collected through the various Internet services can be used to optimise search results for the particular user, leading to higher a CTR; this would directly lead to higher market prices for advertising placements. As we discussed in an earlier section, CTRs are generally in the range of 1-5% for search advertisements. 81 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin EXHIBIT 42: Sample keyword bidding prices Source: adwords.google.com Display advertisements – a different type of advertisements Another type of common advertisement on the Internet is a display advertisement. These are advertisements placed on various websites. While a less sophisticated advertisement may be static (like the one shown in Exhibit 15, more intelligence can be built into the placement of the advertisements by taking into account contextual information such as the user profile, geographical location, time of day, etc.; this approach could have the effect of increasing the click-through-rate. As display advertisements are considered to be “push” advertisements – i.e. pushed to the users by advertisers – CTR for typical display advertisement is <1%. While the CTR is lower than for a search advertisement, the frequency of advertising display is significantly higher for display advertisements. Techniques for search ads can be applied to display ads to improve CTR; Ad Exchange platforms enable search engines to expand into display ads Advertising space inventory can be significantly increased when a search platform leverages the Ad Exchange platform to include partner Internet websites. According to Google, it has over 2m partners in its Google Network Members; Baidu also has partners in its Baidu Union in the range of hundreds of thousands. We believe the ramping up of Baidu Union is a major step for Baidu to build up an Ad Exchange platform as efficiency of the platform increases with the increasing number of participants. We view the Ad Exchange platform as a new product for search engine platforms that has implications on future growth potential. Mobile Internet – the next frontier As the number of mobile Internet users has grown rapidly in recent years with the rise of smartphones and mobile broadband services in China, Internet companies are still adapting their business models to tap into this new opportunity. Exhibit 43 shows the number of overall Internet users 5-year CAGR through 2012 was 21.8% vs. the number of mobile Internet users CAGR of 52.8%. While the rise of mobile Internet usage presents new opportunities to Internet companies, it also poses challenges as Internet companies have to adapt to mobile Internet monetisation. The growth of mobile Internet users has far outpaced Internet companies’ ability to monetise mobile Internet users’ utility for the services. Access to mobile Internet more fragmented than PC Internet; Ad Exchange platforms can also be adapted for service mobile ads Internet companies that have optimised their products for PC Internet advertising are still adjusting their products for the mobile Internet environment, where a major limitation is the real estate available on much smaller screens compared to PC monitors. Taking search engines for example, for each search query a number of paid-advertisements can be served to the PC user; however, on a mobile device limited space is available to serve advertisements. Therefore, for mobile Internet 82 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin advertising to be meaningful the frequency of advertisements must be increased; this would require Internet companies to expand advertising inventory across multiple applications on mobile devices. The user behaviour for mobile Internet is very different from PC Internet. On PCs, users mostly access the Internet through web browsers; this is how most Internet companies that rely on advertising have developed their products. On a mobile device, users are more likely to access information using applications for specific activities; in many cases Internet companies that have succeeded in PC Internet have limited presence in mobile applications. Based on comScore data for May 2013, browsers only represented 20% of smartphone and tablet usage based on time spent. A 2012 Google survey of China’s smartphone users indicated that only 34% of users noticed advertisements appearing in an app; users mostly noticed advertisements on mobile devices while using search engines (49%) and on a website (54%). In our view, this suggests a significant opportunity for in-app advertisements. Exhibit 44 is a summary of popular services being used by China’s mobile users where instant messaging, mobile search, news, Weibo, and music round out the top 5 applications. Browser access on mobile Internet only represents 20% of usage time; only 34% of users notice ads in mobile apps – presenting significant opportunities In our view, the Ad Exchange could be an effective platform for mobile advertising to proliferate as mobile applications can become members of the Ad Exchange network just like other Internet properties and make their mobile applications available through the Ad Exchange. While the business model for mobile advertising still needs to be optimised, we believe the potential exists as mobile devices offer significantly more contextual information about the user that can be used for more targeted advertising. In our view, this places even more importance on the Ad Exchange platform we discussed in earlier sections as it has the potential to serve targeted contextual advertisements to mobile users across multiple applications. 0 2006 Source: CEIC 83 2007 2008 2009 2010 2011 2012 Post & comment 100 Mobile e-mail 200 Network video 300 Social network 400 Network literature 500 Network music 90 80 70 60 50 40 30 20 10 0 Weibo (%) Total number of mobile Internet users News browse Total number of Internet users Instant messaging (m) 600 Mobile search EXHIBIT 44: China mobile users’ usage of mobile apps (April 2013) EXHIBIT 43: Number of China Internet users Source: China Internet Watch BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Financial statements QIHOO 360 TECHNOLOGY Profit and Loss (USD m) Year Ending Dec 2011A 2012A 2013E 2014E 2015E Revenue 168 329 592 844 1,046 Cost of sales ex depreciation (19) (33) (75) (107) (133) Gross profit ex depreciation 149 296 517 737 913 0 3 3 4 4 (127) (234) (447) (558) (606) Other operating income Operating costs Operating EBITDA 22 65 73 182 312 Depreciation (4) (16) (25) (38) (53) Goodwill amortisation 0 0 0 0 0 18 49 48 144 258 Net financing costs 9 7 7 7 7 Associates 0 0 0 0 0 (1) (1) 0 0 0 Operating EBIT Recurring non operating income Non recurring items 0 8 0 0 0 26 63 55 151 265 (11) (11) (9) (26) (46) 15 51 46 125 220 Minority interests 1 0 0 0 0 Preferred dividends 0 0 0 0 0 Other items 0 (5) (4) (4) (4) 16 47 42 121 216 0 (8) 0 0 0 16 38 42 121 216 Recurring EPS * 0.10 0.22 0.23 0.68 1.21 Reported EPS 0.10 0.26 0.23 0.68 1.21 DPS 0.00 0.00 0.00 0.00 0.00 Profit before tax Tax Profit after tax Reported net profit Non recurring items & goodwill (net) Recurring net profit Per share (USD) Growth Revenue (%) 191.1 96.0 80.0 42.5 24.0 Operating EBITDA (%) 115.8 191.3 11.8 150.5 70.9 79.3 Operating EBIT (%) 101.1 168.4 (2.2) 202.5 Recurring EPS (%) (12.2) 108.0 7.5 191.3 77.7 Reported EPS (%) (12.2) 153.1 (11.7) 191.3 77.7 Gross margin inc depreciation (%) 86.2 85.0 83.1 82.8 82.2 Operating EBITDA margin (%) 13.3 19.8 12.3 21.6 29.8 Operating EBIT margin (%) 10.8 14.8 8.0 17.1 24.7 9.3 11.7 7.0 14.4 20.6 42.1 18.1 16.7 17.1 17.2 0.0 0.0 0.0 0.0 0.0 - - - - - 0.0 0.0 0.0 0.0 0.0 Debtor days 27.1 22.4 21.5 25.5 29.0 Creditor days 70.0 72.2 52.6 66.7 78.3 1,633.9 (436.3) 83.2 90.6 128.8 ROIC (%) 72.4 68.6 25.7 50.2 78.6 ROE (%) 7.0 9.1 7.9 18.5 24.6 ROA (%) 2.2 5.6 4.9 13.3 18.1 2011A 2012A 2013E 2014E 2015E 167 329 592 844 1,046 1 0 0 0 0 Operating performance Net margin (%) Effective tax rate (%) Dividend payout on recurring profit (%) Interest cover (x) Inventory days Operating ROIC (%) *Pre exceptional, pre-goodwill and fully diluted Revenue By Division (USD m) Internet services Sale of third party anti-virus software Sources: QIHOO 360 TECHNOLOGY; BNP Paribas estimates 84 BNP PARIBAS 27 JUNE 2013 QIHOO 360 TECHNOLOGY QIHU US Alen Lin Financial statements QIHOO 360 TECHNOLOGY Cash Flow (USD m) Year Ending Dec 2011A 2012A 2013E 2014E 2015E 16 38 42 121 216 4 16 25 38 53 Associates & minorities (1) 0 0 0 0 Other non-cash items 50 51 51 51 51 Recurring cash flow 68 105 117 210 320 Change in working capital 14 4 (95) 33 20 0 0 0 0 0 (16) (69) (91) (114) (114) Recurring net profit Depreciation Capex - maintenance Capex - new investment Free cash flow to equity 66 41 (68) 129 226 Net acquisitions & disposals 0 0 0 0 0 Dividends paid 0 0 0 0 0 (18) (7) 0 0 0 Net cash flow 47 34 (68) 129 226 Equity finance 236 2 0 0 0 (3) 0 0 0 0 281 36 (68) 129 226 Recurring cash flow per share 0.46 0.60 0.66 1.18 1.80 FCF to equity per share 0.44 0.23 (0.38) 0.73 1.27 2011A 2012A 2013E 2014E 2015E Non recurring cash flows Debt finance Movement in cash Per share (USD) Balance Sheet (USD m) Year Ending Dec Working capital assets 31 55 96 148 193 Working capital liabilities (47) (204) (151) (235) (300) Net working capital (16) (149) (54) (87) (107) Tangible fixed assets 17 126 192 267 328 Operating invested capital 1 (23) 138 180 221 Goodwill 5 5 5 5 5 Other intangible assets 8 12 12 12 12 Investments 16 28 28 28 28 Other assets 5 84 84 84 84 34 105 266 308 349 Invested capital Cash & equivalents (344) (381) (312) (442) (667) Short term debt 0 0 0 0 0 Long term debt * 0 0 0 0 0 (344) (381) (312) (442) (667) Deferred tax 1 1 1 1 1 Other liabilities 0 0 0 0 0 371 478 570 742 1,008 Net debt Total equity Minority interests 1 0 0 0 0 Invested capital 34 105 266 308 349 * includes convertibles and preferred stock which is being treated as debt Per share (USD) Book value per share 2.49 2.71 3.20 4.17 5.67 Tangible book value per share 2.40 2.61 3.11 4.07 5.57 Financial strength Net debt/equity (%) (92.6) (79.6) (54.8) (59.5) (66.2) Net debt/total assets (%) (81.1) (55.2) (42.9) (44.8) (50.7) 8.0 2.1 2.7 2.5 2.9 - - - - - 2011A 2012A 2013E 2014E 2015E 24.3 Current ratio (x) CF interest cover (x) Valuation Recurring P/E (x) * 281.1 135.2 125.7 43.2 Recurring P/E @ target price (x) * 281.5 135.3 125.9 43.2 24.3 Reported P/E (x) 281.1 111.1 125.7 43.2 24.3 Dividend yield (%) 0.0 0.0 0.0 0.0 0.0 P/CF (x) 64.1 49.4 44.6 24.9 16.4 P/FCF (x) 66.5 127.7 (76.6) 40.5 23.2 Price/book (x) 11.8 10.9 9.2 7.1 5.2 Price/tangible book (x) 12.2 11.3 9.5 7.2 5.3 EV/EBITDA (x) ** 143.2 69.1 66.8 26.6 15.0 EV/EBITDA @ target price (x) ** 143.4 69.2 66.9 26.7 15.0 EV/invested capital (x) 120.6 45.8 18.5 15.6 13.1 * Pre exceptional, pre-goodwill and fully diluted ** EBITDA includes associate income and recurring non-operating income Sources: QIHOO 360 TECHNOLOGY; BNP Paribas estimates 85 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin NOTES NOTES \ \ \ \ \ \ \ \ \ \ \ 86 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin NOTES NOTES \ \ \ \ \ \ \ \ \ \ \ 87 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin NOTES NOTES \ \ \ \ \ \ \ \ \ \ \ 88 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin Disclaimers and Disclosures ANALYST(S) CERTIFICATION Alen Lin, BNP Paribas Securities (Asia) Ltd, +852 2825 1801, [email protected]. The analyst(s) or strategist(s) herein each referred to as analyst(s) named in this report certifies that (i) all views expressed in this report accurately reflect the personal view of the analyst(s) with regard to any and all of the subject securities, companies or issuers mentioned in this report; (ii) no part of the compensation of the analyst(s) was, is, or will be, directly or indirectly, relate to the specific recommendation or views expressed herein; and (iii) is not aware of any other actual or material conflicts of interest concerning any of the subject securities, companies or issuers referenced herein as of the time of this certification. Analysts mentioned in this disclaimer are employed by a non-US affiliate of BNP Paribas Securities Corp., and are not registered/ qualified pursuant to NYSE and/ or FINRA regulations. GENERAL DISCLAIMER This report was produced by BNP Paribas Securities (Asia) Ltd, a member company of the BNP Paribas Group. "BNP Paribas” is the marketing name for the global banking and markets business of BNP Paribas Group1. This report is for the use of intended recipients only and may not be reproduced (in whole or in part) or delivered or transmitted to any other person without our prior written consent. By accepting this report, the recipient agrees to be bound by the terms and limitations set forth herein. BNP Paribas analysts prudently perform analysis and create quantitative models and estimates derived from their own review of publicly available data without any assistance from any represented company. BNP Paribas analyst estimates and models reflect the analysts’ current judgment only; they are neither all-inclusive nor can they be guaranteed. The analysts’ analysis and models are subject to change based on various other factors. Valuations are based on internal quantitative models and qualitative interpretation. No representation or warranty, express or implied, is made that such information or analysis is accurate, complete or verified and it should not be relied upon as such. Analysts' compensation is not linked to investment banking or capital markets transactions performed by BNP Paribas or the profitability or revenues of particular trading desks. BNP Paribas analysts may participate in company events such as site visits and are prohibited from accepting payment by the company of associated expenses unless pre-approved by authorized members of Research management. This report does not constitute a personal recommendation or take into account the particular investment objectives, financial situations, or needs of individual clients. Customers are advised to use the information contained herein as just one of many inputs and considerations prior to engaging in any trading activity. This report does not constitute a prospectus or other offering document or an offer or solicitation to buy or sell any securities or other investments. This report is not intended to provide the sole basis of any evaluation of the subject securities and companies mentioned in this report. Information and opinions contained in this report are published for reference of the recipients and are not to be relied upon as authoritative or without the recipient’s own independent verification, or taken in substitution for the exercise of judgment by the recipient. Additionally, the products mentioned in this report may not be available for sale in certain jurisdictions. BNP Paribas is not aware of any other actual or material conflicts of interest concerning any of the subject securities, companies or issuers referenced herein as of the time of publication of the research report. As an investment bank with a wide range of activities, BNPP may face conflicts of interest, which are resolved under applicable legal provisions and internal guidelines. You should be aware, however, that BNPP may engage in transactions in a manner inconsistent with the views expressed in this document, either for its own account or for the account of its clients. Australia: This report is being distributed in Australia by BNP Paribas Sydney Branch, registered in Australia as ABN 23 000 000 117 at 60 Castlereagh Street Sydney NSW 2000. BNP Paribas Sydney Branch is licensed under the Banking Act 1959 and the holder of Australian Financial Services Licence no. 238043 and therefore subject to regulation by the Australian Securities & Investments Commission in relation to delivery of financial services. By accepting this document you agree to be bound by the foregoing limitations, and acknowledge that information and opinions in this document relate to financial products or financial services which are delivered solely to wholesale clients (in terms of the Corporations Act 2001, sections 761G and 761GA; Corporations Regulations 2001, division 2, reg. 7.1.18 & 7.1.19) and/or professional investors (as defined in section 9 of the Corporations Act 2001). Canada: The information contained herein is not, and under no circumstances is to be construed as, a prospectus, an advertisement, a public offering, an offer to sell securities described herein, or solicitation of an offer to buy securities described herein, in Canada or any province or territory thereof. Any offer or sale of the securities described herein in Canada will be made only under an exemption from the requirements to file a prospectus with the relevant Canadian securities regulators and only by a dealer properly registered under applicable securities laws or, alternatively, pursuant to an exemption from the dealer registration requirement in the relevant province or territory of Canada in which such offer or sale is made. The information contained herein is under no circumstances to be construed as investment advice in any province or territory of Canada and is not tailored to the needs of the recipient. To the extent that the information contained herein references securities of an issuer incorporated, formed or created under the laws of Canada or a province or territory of Canada, any trades in such securities must be conducted through a dealer registered in Canada. No securities commission or similar regulatory authority in Canada has reviewed or in any way passed judgment upon these materials, the information contained herein or the merits of the securities described herein, and any representation to the contrary is an offence. Hong Kong: This report is prepared for professional investors and is being distributed in Hong Kong by BNP Paribas Securities (Asia) Limited to persons whose business involves the acquisition, disposal or holding of securities, whether as principal or agent. BNP Paribas Securities (Asia) Limited, a subsidiary of BNP Paribas, is regulated by the Securities and Futures Commission for the conduct of dealing in securities, advising on securities and providing automated trading services. For professional investors in Hong Kong, please contact BNP Paribas Securities (Asia) Limited for all matters and queries relating to this report. INDIA: In India, this document is being distributed by BNP Paribas Securities India Pvt. Ltd. ("BNPPSIPL"), having its registered office at 5th floor, BNP Paribas House, 1 North Avenue, Maker Maxity, Bandra Kurla Complex, Bandra (East), Mumbai 400 051 (Tel. no. +91 22 3370 4000 / 6196 4000). BNPPSIPL is registered with the Securities and Exchange Board of India (“SEBI”) as a stockbroker in the Equities and the Futures & Options segments of National Stock Exchange of India Ltd. and Bombay Stock Exchange Ltd. (SEBI regn. nos. INB/INF231474835, INB/INF011474831). Indonesia: This report is distributed by PT BNP Paribas Securities Indonesia, having its registered office at Menara BCA, 35th Floor, Grand Indonesia, Jl. M.H.Thamrin No.1, Jakarta, 10310, Indonesia. PT BNP Paribas Securities Indonesia is a fully subsidiaries company of BNP Paribas SA and is licensed under Capital Market Law No. 8 of 1995 and the holder of broker-dealer and underwriter licenses issued by the Capital Market Supervisory Agency. PT BNP Paribas Securities Indonesia is also a member of Indonesia Stock Exchange. Japan: This report is being distributed to Japanese based firms by BNP Paribas Securities (Japan) Limited or by a subsidiary or affiliate of BNP Paribas not registered as a financial instruments firm in Japan, to certain financial institutions defined by article 17-3, item 1 of the Financial Instruments and Exchange Law Enforcement Order. BNP Paribas Securities (Japan) Limited is a financial instruments firm registered according to the Financial Instruments and Exchange Law of Japan and a member of the Japan Securities Dealers Association and the Financial Futures Association of Japan. BNP Paribas Securities (Japan) Limited accepts responsibility for the content of a report prepared by another non-Japan affiliate only when distributed to Japanese based firms by BNP Paribas Securities (Japan) Limited. Some of the foreign securities stated on this report are not disclosed according to the Financial Instruments and Exchange Law of Japan. Malaysia: This report is issued and distributed by BNP Paribas Capital (Malaysia) Sdn Bhd. The views and opinions in this research report are our own as of the date hereof and are subject to change. BNP Paribas Capital (Malaysia) Sdn Bhd has no obligation to update its opinion or the information in this research report. This publication is strictly confidential and is for private circulation only to clients of BNP Paribas Capital (Malaysia) Sdn Bhd. This publication is being provided to you strictly on the basis that it will remain confidential. No part of this material may be (i) copied, photocopied, duplicated, stored or reproduced in any form by any means or (ii) redistributed or passed on, directly or indirectly, to any other person in whole or in part, for any purpose 89 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin without the prior written consent of BNP Paribas Capital (Malaysia) Sdn Bhd. Singapore: This report is distributed in Singapore by BNP Paribas Securities (Singapore) Ptd Ltd ("BNPPSSL") and may be distributed in Singapore only to an Accredited or Institutional Investor, each as defined under the Financial Advisers Regulations ("FAR") and the Securities and Futures Act (Chapter 289) of Singapore, as amended from time to time. In relation to the distribution to such categories of investors, BNPPSSL and its representatives are exempted under Regulation 35 of the FAR from the requirements in Section 36 of the Financial Advisers Act of Singapore, regarding the disclosure of certain interests in, or certain interests in the acquisition or disposal of, securities referred to in this report. For Institutional and Accredited Investors in Singapore, please contact BNP Paribas Securities (Singapore) Ptd Ltd for all matters and queries relating to this report. South Africa: In South Africa, BNP Paribas Cadiz Securities (Pty) Ltd and BNP Paribas Cadiz Stock Broking (Pty) Ltd are licensed members of Johannesburg Stock Exchange and are authorised Financial Services Providers and subject to regulation by the Financial Services Board. Taiwan: Information on securities that trade in Taiwan is distributed by BNP Paribas Securities (Taiwan) Co., Ltd. Such information is for your reference only. The reader should independently evaluate the investment risks and is solely responsible for their investment decision. Information on securities that do not trade in Taiwan is for informational purposes only and is not to be construed as a recommendation or a solicitation to trade in such securities. BNP Paribas Securities (Taiwan) Co., Ltd. may not execute transactions for clients in these securities. This publication may not be distributed to the public media or quoted or used by the public media without the express written consent of BNP Paribas. Thailand: Research relating to Thailand and Thailand based issuers are produced pursuant to an arrangement between BNP Paribas Securities (Asia) Ltd and ACL Securities Co. Ltd. ACL Securities Co. Ltd. is otherwise unaffiliated with BNP Paribas. This report is being distributed outside Thailand by members of BNP Paribas. Turkey: This report is being distributed in Turkey by TEB Investment a member company of the BNP Paribas Group. Notice Published in accordance with “Communiqué Regarding the Principles on Investment Consultancy Activities and the Investment Consultancy Institutions” Series: V, No: 55 issued by the Capital Markets Board. The investment related information, commentary and recommendations contained herein do not constitute investment consultancy services. Investment consultancy services are provided in accordance with investment consultancy agreements executed between investors and brokerage companies or portfolio management companies or non-deposit accepting banks. The commentary and recommendations contained herein are based on the personal views of the persons who have made such commentary and recommendations. These views may not conform to your financial standing or to your risk and return preferences. Therefore, investment decisions based solely on the information provided herein may fail to produce results in accordance with your expectations. United States: This report may be distributed in the United States only to U.S. Persons who are “major U.S. institutional investors” (as such term is defined in Rule 15a-6 under the Securities Exchange Act of 1934, as amended) and is not intended for the use of any person or entity that is not a “major U.S. institutional investor”. U.S persons who wish to effect transactions in securities discussed herein must do so through BNP Paribas Securities Corp., a USregistered broker dealer and member of FINRA, SIPC, NFA, NYSE and other principal exchanges. Certain countries within the European Economic Area: This document may only be distributed in the United Kingdom to eligible counterparties and professional clients and is not intended for, and should not be circulated to, retail clients (as such terms are defined in the Markets in Financial Instruments Directive 2004/39/EC (“MiFID”)). This document will have been approved for publication and distribution in the United Kingdom by BNP Paribas London Branch, a branch of BNP Paribas SA whose head office is in Paris, France. BNP Paribas London Branch is registered in England and Wales under No. FC13447. Registered Office: 10 Harewood Avenue, London NW1 6AA. BNP Paribas SA is incorporated in France with limited liability with its registered office at 16 boulevard des Italiens, 75009 Paris. BNP Paribas London Branch is authorised and supervised by the Autorité de Contrôle Prudentiel and is authorised and subject to limited regulation by the Financial Services Authority ("FSA") for the conduct of its designated investment business in the United Kingdom. Details of the extent of its authorisation and regulation by the Financial Services Authority are available upon request. This report has been approved for publication in France by BNP Paribas, a credit institution licensed as an investment services provider by the Autorité de Contrôle Prudentiel whose head office is 16, Boulevard des Italiens 75009 Paris, France. This report is being distributed in Germany either by BNP Paribas London Branch or by BNP Paribas Niederlassung Frankfurt am Main, regulated by the Bundesanstalt für Finanzdienstleistungsaufsicht (BaFin). Other Jurisdictions: The distribution of this report in other jurisdictions or to residents of other jurisdictions may also be restricted by law, and persons into whose possession this report comes should inform themselves about, and observe, any such restrictions. By accepting this report you agree to be bound by the foregoing instructions. This report is not directed to, or intended for distribution to or use by, any person or entity that is a citizen or resident of or located in any locality, state, country, or other jurisdiction where such distribution, publication, availability or use would be contrary to law or regulation. All research reports are disseminated and available to all clients simultaneously through our internal client websites. For all research available on a particular stock, please contact the relevant BNP Paribas research team or the author(s) of this report. 1 No portion of this report was prepared by BNP Paribas Securities Corp (US) personnel and it is considered Third-Party Affiliate research under NASD Rule 2711. IMPORTANT DISCLOSURES The disclosure column in the following table lists the important disclosures applicable to each company that has been rated and/or recommended in this report: Company - Disclosure (as applicable) - BNP Paribas represents that: 1. Within the past year, it has managed or co-managed a public offering for this company, for which it received fees. 2. It had an investment banking relationship with this company in the last 12 months. 3. It received compensation for investment banking services from this company in the last 12 months. 4. It beneficially owns 1% or more or the market capitalization of this company. 5. It makes a market in securities in respect of this company. 6. The analyst(s) or an individual who assisted in the preparation of this report (or a member of his/her household) has a financial interest position in securities issued by this company or derivatives thereof. 7. The analyst (or a member of his/her household) is an officer, director, or advisory board member of this company. Additional Disclosures Within the next three months, BNP Paribas may receive or seek compensation in connection with an investment banking relationship with one or more of the companies referenced herein. Target price history, stock price charts, valuation and risk details, and equity rating histories applicable to each company rated in this report is available in our most recently published reports available on our website: http://eqresearch.bnpparibas.com, or you can contact the analyst named on the front of this note or your BNP Paribas representative. All share prices are as at market close on 25 June 2013 unless otherwise stated. 90 BNP PARIBAS 27 JUNE 2013 China Internet Alen Lin RECOMMENDATION STRUCTURE Stock Ratings Stock ratings are based on absolute upside or downside, which we define as (target price* - current price) / current price. BUY (B). The upside is 10% or more. HOLD (H). The upside or downside is less than 10%. REDUCE (R). The downside is 10% or more. Unless otherwise specified, these recommendations are set with a 12-month horizon. Thus, it is possible that future price volatility may cause a temporary mismatch between upside/downside for a stock based on market price and the formal recommendation. * In most cases, the target price will equal the analyst's assessment of the current fair value of the stock. However, if the analyst doesn't think the market will reassess the stock over the specified time horizon due to a lack of events or catalysts, then the target price may differ from fair value. In most cases, therefore, our recommendation is an assessment of the mismatch between current market price and our assessment of current fair value. Industry Recommendations Improving (é): The analyst expects the fundamental conditions of the sector to be positive over the next 12 months. Neutral (çè): The analyst expects the fundamental conditions of the sector to be maintained over the next 12 months. Deteriorating (ê): The analyst expects the fundamental conditions of the sector to be negative over the next 12 months. Country (Strategy) Recommendations Overweight (O). Over the next 12 months, the analyst expects the market to score positively on two or more of the criteria used to determine market recommendations: index returns relative to the regional benchmark, index sharpe ratio relative to the regional benchmark and index returns relative to the market cost of equity. Neutral (N). Over the next 12 months, the analyst expects the market to score positively on one of the criteria used to determine market recommendations: index returns relative to the regional benchmark, index sharpe ratio relative to the regional benchmark and index returns relative to the market cost of equity. Underweight (U). Over the next 12 months, the analyst does not expect the market to score positively on any of the criteria used to determine market recommendations: index returns relative to the regional benchmark, index sharpe ratio relative to the regional benchmark and index returns relative to the market cost of equity. RATING DISTRIBUTION (as at 26 June 2013) Total BNP Paribas coverage universe 657 Investment Banking Relationship (%) Buy 354 Buy 8.5 Hold 207 Hold 2.9 96 Reduce 3.1 Reduce Should you require additional information concerning this report please contact the relevant BNP Paribas research team or the author(s) of this report. © 2013 BNP Paribas Group 91 BNP PARIBAS 27 JUNE 2013 BANGKOK JAKARTA (In cooperation with BNP Paribas) ACL Securities Co Ltd 990 Abdulrahim Place, 12/F Rama IV Road, Bangrak Bangkok 10500 Thailand Tel (66 2) 611 3500 Fax (66 2) 611 3551 PT BNP Paribas Securities Indonesia Grand Indonesia, Menara BCA, 35/F JI. M.H. Thamrin No. 1 Jakarta 10310 Indonesia Tel (62 21) 2358 6586 Fax (62 21) 2358 7587 SEOUL SINGAPORE TAIPEI BNP Paribas Securities Korea Co Ltd 22/F, Taepyeongno Building 310 Taepyeongno 2-ga Jung-gu, Seoul 100-767 Korea Tel (82 2) 2125 0500 Fax (82 2) 2125 0593 BNP Paribas Securities (Singapore) Pte Ltd (Co. Reg. No. 199801966C) 10 Collyer Quay 34/F Ocean Financial Centre Singapore 049315 Tel (65) 6210 1288 Fax (65) 6210 1980 BNP Paribas Securities (Taiwan) Co Ltd 72/ F, Taipei 101 No. 7 Xin Yi Road, Sec. 5 Taipei, Taiwan Tel (886 2) 8729 7000 Fax (886 2) 8101 2168 ISTANBUL NEW YORK BASEL TEB Investment (A JV between TEB Bank and BNP Paribas) TEB Kampus D7 Saray Mahallesi Sokullu Sok No 7 Umraniye 34768 Istanbul Turkey Tel: (90 216) 636 44 44 Fax: (90 216) 631 44 00 BNP Paribas The Equitable Tower 787 Seventh Avenue New York NY 10019, USA Tel (1 212) 841 3800 Fax (1 212) 841 3810 BNP Paribas Aeschengraben 26 CH 4002 Basel Switzerland Tel (41 61) 276 5555 Fax (41 61) 276 5514 HONG KONG BEIJING BNP Paribas Securities (Asia) Ltd 63/F, Two International Finance Centre 8 Finance Street, Central Hong Kong SAR China Tel (852) 2825 1888 Fax (852) 2845 9411 BNP Paribas (China) Ltd Beijing Branch Room 2001, 20/F China World Tower 1 Jianguomenwai Avenue Beijing, China Tel: +86-10-6535 0888 Fax: +86-10-6535 0883 BNP Paribas Equities (Asia) Ltd Shanghai Representative Office Room 2630, 26/F Shanghai World Financial Center 100 Century Avenue Shanghai 200120, China Tel (86 21) 6096 9000 Fax (86 21) 6096 9018 KUALA LUMPUR MUMBAI BNP Paribas Capital (Malaysia) Sdn Bhd Vista Tower, Level 48C The Intermark, 182 Jalan Tun Razak 50400 Kuala Lumpur Malaysia Tel (60 3) 2179 6222 Fax (60 3) 2179 6226 BNP Paribas Securities India Pvt Ltd BNP Paribas House 1 North Avenue, Maker Maxity Bandra Kurla Complex Bandra East Mumbai 400 051 Tel (91 22) 3370 4000 Fax (91 22) 3370 4386 TOKYO CAPE TOWN SHANGHAI BNP Paribas Securities (Japan) Ltd GranTokyo North Tower 1-9-1 Marunouchi, Chiyoda-Ku Tokyo 100-6740 Japan Tel (81 3) 6377 2000 Fax (81 3) 5218 5970 BNP Paribas Cadiz Securities (Pty) Ltd Ground floor, Fernwood House The Oval, 1 Oakdale Road, Newlands Cape Town South Africa 7700 Tel (27 21) 657 8300 Fax (27 21) 657 8301 FRANKFURT GENEVA LONDON MADRID BNP Paribas Mainzer Landstrasse 16 60325 Frankfurt Germany Tel (49 69) 7193 6637 Fax (49 69) 7193 2520 BNP Paribas 2 Place de Hollande 1211 Geneva 11 Switzerland Tel (41 22) 787 7377 Fax (41 22) 787 8020 BNP Paribas 10 Harewood Avenue London NW1 6AA UK Tel (44 20) 7595 2000 Fax (44 20) 7595 2555 MILAN BNP Paribas SA, sucursal en Espana Hermanos Becquer 3 PO Box 50784 28006 Madrid Spain Tel (34 91) 745 9000 Fax (34 91) 745 8888 BNP Paribas Equities Italia SIM SpA Piazza San Fedele, 2 20121 Milan Italy Tel (39 02) 72 47 1 Fax (39 02) 72 47 6562 PARIS ZURICH MANAMA BNP Paribas Equities France Société de Bourse 20 boulevard des Italiens 75009 Paris France Tel (33 1) 4014 9673 Fax (33 1) 4014 0066 BNP Paribas Talstrasse 41 8022 Zurich Switzerland Tel (41 1) 229 6891 Fax (41 1) 267 6813 BNP Paribas Bahrain PO Box 5253 Manama Bahrain Tel (973) 53 3978 Fax (973) 53 1237 https://eqresearch.bnpparibas.com