China Internet 091205

Comment
Page 1
Equity Research Global
Industry
China Internet
Industry Overview
Richard Ji (Internet / Media - China)

+852 2848 6926 Hong Kong Richard.Ji@morganstanley.com
September 12, 2005
Mary Meeker (Internet - US / Global)

+1-212 761 8042 New York / Menlo Park Mary.Meeker@morganstanley.com
Creating Consumer Value in Digital China

We are initiating coverage of the China Internet industry with an attractive
view. We believe Chinese Internet companies that focus on creating consumer value have the highest potential to create shareholder value. China is No. 1 in the world in mobile subscribers and No. 1 in Internet users under the age of 30 - this evolving presence on the world stage should not be underestimated. We endeavor to find businesses with robust growth potential, high barriers to entry, scalability, network effects and low regulatory interference. We prefer management teams who are consumer-centric and innovative, and stocks whose share prices have overlooked upside plus margins of safety. We believe the click plus brick model may dominate ecommerce in China; content may prove to be king; foreign companies lagging in localization may lag in market penetration; and local companies may need to become savvy acquirers to outgrow rivals. Online gaming is our favorite segment pick. Local players with selfdevelopment and distribution capacity may emerge as long-term winners. We believe competition is breaking out in online brand advertising, which faces inventory oversupply and a secular shift toward performance-centric models. Paid search, on the other hand, is poised for robust expansion. We view mobile value-added services as an overlooked opportunity. Despite regulatory concerns, entry barriers are rising and a segment recovery is in sight. We see rising opportunities in online commerce including travel, instant messaging-related services and auctions. Emerging payment mechanisms will be key to more substantive growth. Our investment concerns include economic slowdown, regulatory risks, dearth of innovation, management quality and excessive capital inflow. Our top stock picks include Ctrip (Overweight-V), NetEase (Overweight-V), and Tencent (Overweight-V). See our related reports for these companies, plus Sina (Equal-weight-V), Sohu (Equal-weight-V), TOM Online (Equalweight-V) and 51job (Equal-weight-V). See page 73 for a slide presentation summarizing this report.
Contributors:
Andy Xie (Economics - China) andy.xie@morganstanley.com Mark Shuper (Telecom - China) mark.shuper@morganstanley.com Hani Abuali (Telecom - China) hani.abuali@morganstanley.com Lina Choi (Telecom - China) lina.choi@morganstanley.com Viktor Ma (Technology - China) viktor.ma@morganstanley.com Brian Fitzgerald (US - Internet) brian.fitzgerald@morganstanley.com Ramji Srinivasan (US - Internet) ramji.srinivasan@morganstanley.com Jenny Wu (China - Internet / Media) jenny.wu@morganstanley.com
All of Morgan Stanleys technology equity research reports are available on the Internet through Client Link at https://secure.ms.com. If you wish to receive this service please contact your institutional sales representative. This report and its complementary summary slide presentation (along with other technology overview reports and presentations) can also be downloaded from http://www.morganstanley.com/techresearch/
Morgan Stanley does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision.
Please see analyst certification and other important disclosures starting on page 113.
Page 2
Table of Contents Report Roadmap

In this report, a follow up to The China Internet Report http://www.morganstanley.com/institutional/techresearch/pdfs/China_Internet_Report0404.pdf (4/04), we review key trends for the Internet market in China, relative anomalies (we call them mysteries), plus key challenges and risks. We also lay out our investment framework and drill down on four key segments of the market: 1) mobile value-added services (MVAS); 2) online advertising; 3) online gaming and 4) online commerce. In addition, we highlight investment ideas. Under separate cover, we have initiated research coverage of Chinese Internet companies Ctrip, NetEase, Tencent, Sina, Sohu, TOM Online, and 51job. For more detail on the Internet and China, see The China Internet Report. Investment Thesis - Focus on Consumer Value will Drive Shareholder Value ................................................................... 3 China Internet Trends ............................................................................................................................................................. 4 1 - Growing Faster Than Other Marketswith More Potential.............................................................................................. 4 2 - Internet Brings About Cultural Evolution ......................................................................................................................... 5 3 - Unlike the Rest of the World, China Internet is Mobile-centric Rather Than PC-centric ................................................. 5 4 - Foreign Interests Accelerating .......................................................................................................................................... 6 5 - Hotspot for Mergers and Acquisitions ............................................................................................................................... 7 6 - A League of Big Players ................................................................................................................................................ 8 7 - Leadership Position May Not Necessarily be Secure......................................................................................................... 9 8 - Purer Plays Tend to Have Higher Market Share ................................................................................................................ 9 9 - Regional Focus Can Vary ................................................................................................................................................ 10 10 - Content Becoming King................................................................................................................................................. 11 China Internet Mysteries........................................................................................................................................................ 13 1 - Chinese Internet Players at Valuation Discounts to Global Internet Peers ...................................................................... 14 2 - Real Revenue Perhaps Significantly Higher than the Headline Revenue ....................................................................... 14 3 - Chinese Internet Players Have Much Higher Margins than Global Peers........................................................................ 14 4 - Ecommerce Ready to Take Off? ...................................................................................................................................... 15 5 - Why Can Click Plus Brick Outcompete? ...................................................................................................................... 16 6 - State-Owned Enterprises (SOEs) - Referees and Players?............................................................................................... 17 7 - Why Are Foreign Players Especially Challenged in China? ........................................................................................... 18 China Internet Challenges / Risks.......................................................................................................................................... 24 1 - Economic Hard landing or Soft Landing?........................................................................................................................ 24 2 - Expect the Unexpected! ................................................................................................................................................... 25 3 - SOE-Friendly versus Entrepreneur-Friendly.................................................................................................................... 25 4 - Too Much Capital, Too Few Good Ideas......................................................................................................................... 26 5 - To Buy or to Build? ......................................................................................................................................................... 27 6 - Long-Term Value versus Short-Term Profit .................................................................................................................... 27 7 - Dearth of Innovation ........................................................................................................................................................ 28 8 - Founder Mentality............................................................................................................................................................ 28 9 - Hollowing of Management Teams................................................................................................................................... 29 China Internet Investment Framework Business / Competence / Price .......................................................................... 33 China Internet Industry Analysis - Close to Consumers, Focus on Growth ...................................................................... 34 Mobile Value Added Services (MVAS) - Light at the End of the Tunnel ............................................................................. 37 Online Advertising - A Race for Eyeballs and Performance.................................................................................................. 46 Online Gaming - Pipeline and Distribution Fuel Hypergrowth.............................................................................................. 53 Online Commerce Solid Outlook .................................................................................................................................... 62 Global Internet Companies Price Performance .................................................................................................................. 72 Summary Slide Presentation .................................................................................................................................................. 73
China Internet September 12, 2005
Page 3
Investment Thesis - Focus on Consumer Value will Drive Shareholder Value

Summary and Investment Conclusion
We are initiating coverage of the China Internet industry with an Attractive view. China is the worlds fastest growing economy and the Internet is its fastest growing industry. China has more consumers than any other country, and the Internet may reach more Chinese consumers than any other business. We view the Internet as one of the best ways for investors to benefit from Chinas consumer market growth. That said, it is key to remember that investing in emerging markets (the Internet) in emerging markets (like China) is always fraught with risks and challenges and a portfolio approach to investing is crucial.
Key Factors for Long-Term Success
2) Competition is breaking out in online brand advertising (industry revenue was $220MM in 2004, and should expand at a sales CAGR of 24% to $420MM in 2007), which faces inventory oversupply and a secular shift toward performance-centric models. Paid search, on the other hand, is poised for robust expansion. 3) We view mobile value-added services (MVAS; industry sales for MVAS excluding those for mobile carriers was $770MM in 2004, and should grow at a CAGR of 28% to $1.6B in 2007) as an overlooked opportunity. Despite regulatory concerns, entry barriers are rising and a segment recovery in is in sight. 4) We see rising opportunities in online commerce, including travel, instant messaging-related services, and auctions. Emerging payment mechanisms will be key to more substantive growth.
Our Top Picks in China Internet
We endeavor to identify and support Chinese Internet companies that focus on creating consumer value: First, these companies should regard consumers as their top priority. To paraphrase a well-known American politician, they should ask not what consumers can do for them, but what they can do for consumers. Second, they should target markets with enormous consumer needs and significant growth potential. Third, they should be innovative in delivering key content, key services and key products to consumers. And finally, in our view ill-gotten money will not last. For companies to be truly successful in the long term, they must monetize their consumer services in ethical ways. We firmly believe only those companies focused on consumer value have the potential to create shareholder value over the long term. As Meg Whitman, eBays CEO, has said: Whoever wins China will win the world.
Our View of the China Internet Industries
While these companies are still early stage, risky and volatile, our top China Internet investment ideas are: 1) Ctrip (CTRP, Overweight-V) A primary beneficiary of rising consumption power and travel growth in China. Ctrips sharp consumer focus, scalable platform and superior online and offline operational ability help make it a dominant player in the service-intensive travel industry. 2) NetEase (NTES, Overweight-V) The leading innovator in the rapidly growing online gaming market in China. With its market savvy, focus on innovation, and strong management team, led by William Ding, we believe NetEase should be able to sustain market leadership. 3) Tencent (700.HK, Overweight-V) The powerhouse in Chinas rapidly expanding instant messaging market with 60%+ user share. Tencent has created a strong community with networking effects and high barriers to entry. The company has been successful at growing profits in Internet value-added services and MVAS and looks to continue to ramp monetization through Online Gaming and online advertising.
Four points summarize our view of the China Internet Industries: 1) Online gaming is our favorite segment pick. We estimate the industry revenue for online gaming at $390MM in 2004, growing at a three-year sales CAGR of 37% to $1B in 2007. Local players with selfdevelopment and distribution capacity may emerge as long-term winners.
Page 4
China Internet Trends

1 - Growing Faster Than Other Marketswith More Potential 2 - Internet Brings About Cultural Evolution 3 - Unlike the Rest of the World, China Internet is Mobile-centric Rather Than PC-centric 4 - Foreign Interests Accelerating 5 - Hotspot for Mergers and Acquisitions 6 - A League of Big Players 7 - Leadership Position May Not Necessarily be Secure 8 - Purer Plays Tend to Have Higher Market Share 9 - Regional Focus Can Vary 10 - Content Becoming King
1 - Growing Faster Than Other Marketswith More Potential
ICP? on page 39 rather than the ISPs that western investors are likely familiar with, such as EarthLink.). Third, CNNIC reported that the number of Chinas broadband users hit 53MM in June 2005, up 71% year-overyear (Y/Y). We forecast that Chinas broadband users may grow 53% Y/Y to 66MM by the end of 2005. Fourth, the major drivers for Chinas Internet services are robust. We project the industry revenues for MVAS, Online Gaming, and online advertising may expand at mid-20-40% compound annual growth rates (CAGR) for the next three years.
Exhibit 2
The China Internet Network Information Center (CNNIC) reports that, as of June 2005, Chinas Internet users surpassed 100MM and rank No. 2 in the world, behind 211MM users in the US. Notably, we estimate that China has more Internet users under the age of 30 (70MM+ or 71% of total Internet users) than any other nation in the world. This is especially relevant as the evolution of computing technology has shown that innovation is often most active in markets with the largest number of younger users.
Exhibit 1
China Leads the World in Mobile Subscribers and Ranks No. 2 in Internet Users
Mobile Phones Internet Users Mobile Phone to Country (MM) (MM) Internet User Ratio Installed PCs (MM)
Major Drivers for Chinese Internet Are Trending Up

By the End of 2004 Volume Internet users Broadband users Mobile users CAGR for next 3 years 13% 32% 12%
China US Japan Germany UK Italy S. Korea
363 177 88 69 54 54 37
100 211 78 51 37 32 32
3.6:1 0.8:1 1.1:1 1.4:1 1.5:1 1.7:1 1.2:1
53 207 55 39 26 16 27
94MM (18% Y/Y; <7% of population) 43MM (146% Y/Y; <3% of population) 335MM (24% Y/Y; 26% of population)
ARPU Internet users $5-$6 per month Broadband users $10-$15 subscription fee per month Industry revenue MVAS $770MM (89% Y/Y) Online Gaming $390MM (90% Y/Y) Online advertising $220MM (78% Y/Y) Note: ARPU- average revenue per user. Source: CNNIC, Morgan Stanley Research
Source: Euromonitor, CNNIC, Morgan Stanley Research (July 2005).
Second, China leads the world in mobile phone users (363MM, in June 2005). We estimate that the mobile valueadded services (MVAS) industry contributed around $5-6B in revenue, including sales generated by both Internet Service Providers (ISPs) and telecom carriers, in 2004, with revenue from short messaging services (SMS) contributing the bulk of the total (Note that, when we refer to ISPs throughout this report, we refer to companies as defined in the section What Models Shall We Focus On ISP or
28% 37% 24%
Last, but not least, the China Internet space has increasingly captured the attention of global investors and Internet companies. Note that in August, the share price of Baidu, the sponsored search leader in China, jumped 350% on its first day of trading. Also in August, Yahoo! purchased a
Page 5
40% stake (with 35% voting rights) in Alibaba for $1B in cash. The deal created a combined entity, with Jack Ma a the helm, which comprises Alibaba properties (Alibaba International, Alibaba China, AliPay, and Taobao) and Yahoo! properties (Yahoo! China portal with its communications / advertising services, Yahoo! Search Technology and 3721). The deal also created a new template for doing business in China: Yahoo! contributed its Yahoo! China business to Alibaba.com, effectively outsourcing Yahoo!s China business.
2 - Internet Brings About Cultural Evolution
gaps and draw large numbers of followers. For instance, Shandas leading online game, Legend of Mir 2, attracted average concurrent users of 300,000-400,000 in 2004. In another words, if you access Shandas top game at any given moment, you may have a virtual audience equivalent in size to all the students from 20 to 30 US colleges.
Exhibit 3
China Internet User Age Demographics

Ages Ages 36-40 Ages 31-35 10% Ages 18-24 38% 7% 41-50 7% Ages Ages 51-60 60+ 3% 1% Under 18 16%
We believe the Internet delivers many things that Chinese consumers have long been craving, such as: Free-floating information. Compared to content on traditional media, Internet content is less regulated in China. To date, the top ten newspapers and the top ten TV operators (except Phoenix TV) in China are state-owned and state-regulated. The Internet offers Chinese consumers an alternative source of information. For instance, Sina and Phoenix TV were the only two Chinese media companies that reported the terrorist attacks in the US on September 11, 2001 within five minutes of the tragedyother sources were delayed by hours. For 2004, Sina reported a daily average of 30MM unique readers for its portal content, more than the combined total for Chinas top ten newspapers. Interactivity. The Internet allows consumers to communicate more openly with one another. Open communication has been lacking in China due partly to the legacy of the Cultural Revolution and partly due to the countrys one-child per family policy. In 4Q2004, Tencent reported peak concurrent users for its instant messaging (IM) services of 13MM, a figure comparable to New Yorks population. Entertainment. Over the 10-year span of the Cultural Revolution (from 1966 to 1976), the primary source of entertainment for Chinese people was eight model movies, including Hong-se-niang-zi-jun (The Red Army Girls), Hong-deng-ji (Tale of The Red Lantern), and Bai-maoneeu (The White-haired Girl). Over the past decade, we have witnessed the significant development of Chinas entertainment business, including domestic production of blockbuster movies, music and TV dramas. However, even to date, quality entertainment is still scarce, especially in the second-tier cities in China. We believe there is a significant opportunity, particularly among those younger people with access to computers, for online entertainment to fill in the
Ages 25-30 17%
71% under 30 54% under 24
Source: CNNIC 16th Statistical Survey Report on the Internet Development in China, July 2005.
3 - Unlike the Rest of the World, China Internet is Mobile-centric Rather Than PC-centric
Unlike most other regions of the world, Chinas Internet market centers on mobile handsets rather than personal computers (PCs): China has over 360MM mobile phone users, 3.6 times more than its Internet users. The total number of Chinas mobile users is equal to the total of the next three nations combined. In the US, it is the reverse: the total number of Internet users in the US is equal to the total of the next three nations combined. We estimate ISPs (or Internet Service providers, including Sina, TOM Online, and Tencent) generated around $700-800MM from MVAS for 2004, two times more than Online Gaming and three times more than online advertising. There are six publicly traded companies (Sina, TOM Online, Tencent, KongZhong, Linktone and Hurray!) that generated the majority of their 2004 revenues from MVAS, versus three (Shanda, NetEase and The9) from Online Gaming and two (Sohu and Baidu) from online advertising / search.
Page 6
Chinas people are heavier users of mobile data services, relative to other countries. This trend is exemplified by SMS (short message service; a service that allows brief text messages to be sent and received on mobile phones) usage: there were 652 SMS per mobile phone sent in China vs. 139 SMS per mobile phone sent in the US, in 2004 (where China SMS data is from the Ministry of Information Industry, US SMS data is from SMS.ac, and mobile data is from Morgan Stanley Research). We believe this China trend is due to the immediacy and cost-effectiveness of phones. When we say cost-effectiveness, we dont mean on a per MB basis, but on an absolute dollar basis, as noted in section #4, since the cost of a computer and Internet connection outweighs that of a mobile phone and a text message in the short term. To make the convenience of mobiles relative to PCs concrete, consider the experience of Wang Yi, a 19-year-old bartender in Beijing, quoted in the 7/21/04 Wall Street Journal. She noted that, she infrequently uses a PC to communicate with her parents in rural China; instead she asks, My parents have e-mail, but why go online when I can just send them an SMS?
Yahoo! acquired a 40% economic (35% voting) stake in Alibaba, while folding its Yahoo! China unit into Alibaba Alliances between foreign companies and domestic players are abundant, such as SkypeTOM Online (In September 2005, Skype formed its first joint venture in the world with TOM Online to promote Skype services in Chinese domestic market.), NCsoft-Sina (joint venture for online game operations), and Disney-Sohu (for online content distribution). We believe that, if it were not for governmental restrictions or the potential withdrawal of news supply by the Chinese government, foreign companies might have purchased stakes in Sina by now, given its high level of free-floating shares.
Exhibit 4
Examples of Foreign Entries into Chinas Internet Market

Foreign Entrant eBay Yahoo! Amazon.com InterActive Monster.com Google Local Target Eachnet 3721.com Alibaba Joyo elong ChinaHR Baidu Date Jul-03 Nov-03 Aug-05 Aug-04 Jul-04 Feb-05 Jun-04 Type of Transaction Acquisition Acquisition 40% stake Acquisition 30% stake 40% stake 3% stake Market Position of Target* No. 1 in online auction No. 2 in paid search No. 1 in B2B marketplace A leading e-commerce provider No. 2 in online traveling No. 2 in online recruiting No. 1 in paid search
China, in many ways, is leading the development and monetization of mobile content (i.e. KongZhong, TOM Online) in ways that are still nascent in the US; we believe investors underappreciate the secular trend at work within mobile computing, and the value, both in terms of wealth and consumer satisfaction, that mobile value-added services can contribute.
* Refers to the market position at the time of transaction. Source: Company data, Morgan Stanley Research
4 - Foreign Interests Accelerating
In response to Chinas potential market opportunity, over the past two years Leading global Internet players, including eBay, and Amazon.com, have acquired leading China Internet companies. InterActive and Monster.com purchased sizeable stakes in their Chinese counterparts. Google opened its own operations in China, and launched Google Local in China (bendi.google.com)
Page 7
Exhibit 5
Exhibit 6
Alexa Global Web Site Traffic Rankings

Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Site Name Yahoo! MSN Google Yahoo! Japan baidu.com Sina Passport.net eBay 163.com (NetEase.com) Sohu Microsoft QQ / Tencent Amazon.com 3271 (Yahoo!) / Alibaba Naver.com Myspace.com Google Japan Google UK Nate Allyes.com AOL BBC Online Daum.net Tom.com CNN Country USA USA USA Japan China China USA USA China China USA China USA China Korea USA Japan UK Korea China USA UK Korea China USA
Alexa Chinese Web Site Traffic Rankings

Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Site Name Baidu Sina NetEase Sohu QQ / Tencent 3271 (Yahoo!) / Alibaba Allyes.com Tom.com TaoBao (Alibaba) Google Yisou (Yahoo!) / Alibaba ChinaRen / Sohu SoGou / Sohu China.com 21CN eBay EachNet Xinghuanet Hao123 126.com mop.com hc360.com pconline 265.com cmfu.com poptang.com Major Business Paid search Portal (MVAS and online advertising) Portal (online games) Portal (online advertising) Instant Messaging Paid search Online advertising Portal (MVAS) Online auction Paid search Paid search Online community Paid search Portal Portal Online auction Portal Portal email community Entertainment portal Industry Portal Online commerce Portal Literature portal Online game portal
Source: Alexa.com, September 6, 2005. Alexa data may be somewhat skewed given that its statistics are based on the usage patterns of users of its Alexa toolbar. Users are not randomly selected to receive the toolbar: they opt-in. Nevertheless, we believe Alexas data does have some directional significance for site traffic and interest.
Source: Alexa.com, September 6, 2005. Alexa data may be somewhat skewed given that its statistics are based on the usage patterns of users of its Alexa toolbar. Users are not randomly selected to receive the toolbar: they opt-in. Nevertheless, we believe Alexas data does have some directional significance for site traffic and interest.
5 - Hotspot for Mergers and Acquisitions
Internet-related merger and acquisition (M&A) activity in China has been robust: Horizontal integration. Sohu ramped its total web traffic by acquiring ChinaRen (the No. 1 online alumni club), 17173.com (the No. 1 online game portal), and Focus.net (the No. 1 real estate portal). Shanda acquired Bianfeng (casual games), Haofang (LAN-based PC games), and Digital Red (mobile games) to solidify its leadership in the gaming business. Vertical integration. Sina jumpstarted its MVAS business via the acquisitions of Memestar and Crillion, which contributed around 30% of Sinas
revenue in 2003 (Memestar alone) and in 2004 (Memestar plus Crillion). TOM Online secured its leading position in the interactive voice response (IVR) business by purchasing Puccini. Hostile effort. In February 2005, we witnessed the first hostile share purchase in China as Shanda acquired a 19.5% stake in Sina in the open market. Following Shandas share purchase, we saw probably the most publicized poison pill in China as Sina adopted a defense strategy against Shanda. Note that Sohu adopted Chinas first poison pill in 2001 to fend off an intended hostile takeover by Jade Bird, although at a much smaller scale relative to the Shanda-Sina encounter.
Page 8
Bull Case for Combination of Shanda and Sina

Sina is the best online content provider, based on its topranked portal traffic, in our view. This could broaden Shandas product offering beyond Online Gaming and fuel content growth for Shandas home entertainment strategy. Sinas management depth may help Shanda venture into new business territories. The consumer demographics of the two parties may complement each other. ACNielsen estimates around 52% of Sinas users have an annual income of Rmb25,000 (or US$3,000) or above, versus around 19% for general Internet users. IDC assesses that 73% of Chinese online gamers are below the age of 25 and likely have low spending power. Shanda could help NC-Sina expand the distribution of Lineage 2, while NC-Sina could add some potential hits to Shandas pipeline.
Bear Case for Combination of Shanda and Sina

Shandas core competence is in Online Gaming, not MVAS or online advertising. Shanda thus needs the buy-in of Sinas management to operate businesses outside of online gaming. Unfortunately, the poison pill put in place by Sina suggests the combination may start off on the wrong foot. This could make it hard for Shanda to increase its holdings without incurring significant share dilution. There is little business overlap between the two companies, thus resulting in low cost-saving synergies. Shandas balance sheet is stretched. Poison pill aside, Shanda could only add 8-10% of Sinas stock at a price range of $28-30 (with its current cash and cash equivalents). Moreover, Shanda may need to pay a premium above Sinas market price to further increase its stake. Shandas online game operation has relatively high barriers to entry and relatively low industry risks. If the combination goes through, Shanda may inevitably be exposed to the regulatory risks and the earnings volatility in MVAS. Shandas business model is undergoing a major transformation. In its core gaming business, rivals such as NetEase and The9 are closing the gaps. Shandas interactive TV initiatives may prove more thorny than rosy as Shanda needs to collaborate with state-owned local media groups for content and China Telecom/Netcom for broadband access. Will such alliances turn into partnerships similar to that between the MVAS providers and China Mobile? We could not rule out the possibility as telecom carriers and local media companies control the billing relationship with customers. In that case, Shandas prepaid cards may no longer work for the payment of the interactive TV content.
The transaction could transform Shandas public image from

an online game operator into a mainstream digital media player.
Sina and Shanda have different business cultures. Sina is the

oldest Internet company in China, with a team of professional managers running a diverse spectrum of business operations. Shanda, on the other hand, is a relatively new entity on the block and focuses on a single business line. Like a merger between an incumbent supermarket and an upcoming specialty store, a cultural clash could jeopardize their integration.
6 - A League of Big Players
In Chinas online markets, industry leaders, such as Sina and Sohu in online advertising and Shanda and NetEase in Online Gaming, accounted for 50%+ of total industry revenues in 2004. Tencent has 60%+ share of users in instant messaging (IM). We believe larger players have achieved their dominance through scale, brand awareness (Sina and Sohu), networking effects (Tencent, Shanda, and
NetEase), and abundant financial resources to acquire key targets and develop key technologies. MVAS represents an interesting case as the top three players, including Sina, TOM Online, and Tencent, assumed relatively small market share of 41%, which we attribute to quickly expanding new services (2.5G) and new players such as KongZhong, Linktone, and Hurray!, seizing market share.
Page 9
Exhibit 7
Exhibit 8
Online Leaders Have Lions Share of Chinese Market

Industry MVAS Companies Sina TOM Online Tencent Industry Rank Market Share No. 1 16% No. 2 15% No. 3 10% Total share =41% of industry rev No. 1 30% No. 2 25% Total share =55% of industry rev No. 1 39% No. 2 18% Total share =58% of industry rev No. 1 65% No. 2 29% Total share =94% of GMV No. 1 63% of IM users
Gaps Are Closing Between the Market Leaders and the Runner-Ups
(Revenue, $MM) MVAS Sina TOM Online as % of Sina Online Advertising Sina Sohu as % of Sina Online Gaming Shanda NetEase as % of Shanda 2Q2003 14 2Q2004 31 29 93% 2Q2005 23 41 180%
Online Advertising
Sina Sohu
10 7 72%
Online Gaming
Shanda NetEase
16 13 86%
20 17 83%
e-commerce
eBay EachNet Alibaba/ Taobao
17 4 24%
35 16 46%
56 40 70%
Source: Company data, Morgan Stanley Research. Instant Messaging Tencent Note: Estimates are based on 2004 results. GMV- gross merchandise value. Source: Company data, iResearch, Morgan Stanley Research.
8 - Purer Plays Tend to Have Higher Market Share
7 - Leadership Position May Not Necessarily be Secure
Unlike the Internet business in the US, where there are established leaders, such as Google, eBay, and Yahoo!, we are witnessing leadership changes and the narrowing gaps between the incumbent leaders and the runner-ups. We believe there are several contributing factors: Chinas Internet market is relatively new but expanding rapidly. Todays leaders may face new and different competition in the coming years and may suffer from shrinking market share as a result. For instance, eBay EachNet saw its market share in gross merchandise volume (GMV) erode within one year primarily due to aggressive competition from Taobao. Chinese Internet companies are still early stage entities with fluid cultures. Like any growing entity, they may make mistakes and thus fall behind. Regulatory uncertainties, such as those in MVAS, add complications to competition. By the end of 1Q2005, TOM Online overtook Sina, which suffered from a series of regulatory crackdowns, to become the sales leader in MVAS. Innovation (which is far from guaranteed) is key to outperformance. To date, NetEases Fantasy Westward Journey, a self-developed game, has replaced Shandas Mir 2, a licensed game from South Korea, as the most popular online game in China.
Online leaders typically derive the bulk of their revenue from one single business line. By being relatively purer players, these companies can have more dedicated resources for technology development, customer service, and marketing, thus forming a virtuous cycle.
Exhibit 9
Pure Plays Tend to Dominate the Internet Market

Industry Companies Rank in the industry As % of the company's revenue
MVAS TOM Online Online Advertising Sohu Online Gaming Shanda NetEase No. 1 95%
No. 2
66%
No. 1 No. 2
87% 83%
Note: Data are based on the company results in 2Q2005. Source: Company data, Morgan Stanley Research.
One exception is Sina, which has a diversified business mix but still led in MVAS and online advertising (based on 2004 results). We attribute such abnormality to: Sinas competitive advantages in news content, making it the preferred portal destination for white-collar readers and advertisers. Sinas acquisitions of MVAS companies Memestar and Crillion, which in aggregate contributed 30% of its 2004 revenue. However, in 1Q2005, Sina lost its
Page 10
leadership in MVAS to TOM Online, a much purer MVAS player.

9 - Regional Focus Can Vary
It is notable that advertising customers are highly concentrated in top-tier cities (or the top four cities with the highest GDP contribution in China, including Shanghai, Beijing, Guangzhou, and Shenzhen), which contributed roughly 50% of total advertising spending. As noted in Exhibit 10, Chinas Internet users are also concentrated in these areas. Ctrip estimates that the top five cities accounted for 60% of its hotel sales volume and 80% of its air ticket sales volume for 2004. In contrast, IDC assesses that 85% of the online gamers came from non-top tier cities (or the cities other than the top four cities in GDP contribution) in 2003. We believe top-tier cities may continue to dominate online brand advertising because: Top tier cities have more Internet users. Residents of top tier cities have more spending power. Most corporate headquarters and top advertising agencies are located in top tier cities. However, as paid search gradually grabs market share, we may see an increasing advertising contribution from secondtier cities as paid search services are tailored toward small and medium enterprises (SMEs), which are typically located in regions outside top-tier cities. It is thus important for investors to focus on top-tier cities for banner advertising sales and non-top-tier regions for paid search growth. Moreover, we believe a nationwide sales presence is instrumental for online gaming operations and MVAS because: A nationwide distribution platform is key for the sales of pre-paid point cards to online game players. ISPs need to have local sales forces work closely with regional mobile carriers to promote new MVAS products.
Local sales forces may help identify local consumer needs, which vary from region to region given Chinas cultural diversity. For instance, Shanghai and Guangdong are two of the most prosperous regions in China. But the people there speak different dialects (Shanghaiese versus Cantonese) and the two regions have distinctively different local cultures.
Page 11
Exhibit 10
China Internet Users Concentrated in Costal Regions and Major Metropolitan Areas
2.9% Beijing 4.3% 1.3% 3.6 % 0.3% 0.2% 0.1 % 5.6% 1.0% 2.2% 1.3% 2.8% 2.2 % 3.2 % 4.6% 3.3% Tianjin 2.1% 9.0% 7.0 % 2.6 % 5.7% 1.7 % 12.6% 3.5% Shanghai 4.7% 1.9% 3.4 %
1.0 %
West China 19%
3.0 %
Coastal China 56%

0.5%
Source: CNNIC 15th Statistical Survey Report on the Internet Development in China, Jan 2005.
10 - Content Becoming King
As telecom carriers start to roll out broadband and 3G services, we may see escalating demand for quality content by ISPs and consumers. The value of differentiated / quality content may rise. Note the following TOM Online pays out 40-50% of revenue to music providers, such as Sony Music, for the wireless revenue generated on such content. KongZhong paid five times more for the wireless distribution rights for the leading blockbuster movie in 2005 than it did two years earlier. Foreign online game developers command ten games higher payments for game licensing fee now than just a couple years ago. For instance, to license The World of Warcraft, The9 needs to pay Blizzard a $3MM upfront licensing fee plus 22% of revenue sharing on the face value of prepaid card sales with $51MM minimal guarantee over the four-year licensing term.
In 2004, the most popular song, Mouse Loves Rice, had 5MM downloads, with potential profits comparable to the best-selling CDs in China. We estimate that content providers, such as MTV, may generate more revenue from mobile phones than from pay-TV in China in a few years.
We believe several approaches may help ISPs counter these trends: User-generated content. Tencents IM services center on communication and interaction. As a result, we estimate nearly 70-80% of Tencents content is generated by its IM users. In addition, blog operators such as BlogChina, allow a user to post their thoughts on personal, publicly accessible website (weblog), free of charge. In-house development. NetEase has been focusing on internal development, resulting in minimal licensing fee to third-party game developers and a high operating margin.
Page 12
Scale / distribution. TOM Online aims to be a content aggregator. Its scale in distribution may help it to gain bargaining power over content providers.
Exhibit 11
China Most Frequently Used Online Services

Online Service Email News Search engine Browsing, other than news Online Music Instant Messanging BBS (community and forum) Online Film & TV School/classmate BBS File uploads/downloads Internet games Online chatroom Online shopping Personal websites Online banking Online recruiting Online education Blog E-magazine Online sales (including promotion and auction) IP telephone SMS and MSM Stock trading online Ticket / hotel reservation Others % of Survey Respondents 91% 79% 65% 57% 46% 45% 41% 38% 29% 26% 23% 21% 20% 17% 14% 12% 11% 11% 10% 7% 5% 5% 5% 3% 1%
Source: CNNIC 16th Statistical Survey Report on the Internet Development in China, July 2005. Includes data gathered on 15 previous surveys. The survey uses multiple survey methods including automatic searching with computer, an online component, an offline sampling component, and data reported from relevant organizations. Respondents were allowed to reply with multiple answers.
Page 13
China Internet Mysteries

1 - Chinese Internet Players at Valuation Discounts to Global Internet Peers 2 - Real Revenue Perhaps Significantly Higher than the Headline Revenue 3 - Chinese Internet Players Have Much Higher Margins than Global Peers 4 - Ecommerce Ready to Take Off? 5 - Why Can Click Plus Brick Outcompete? 6 - State-Owned Enterprises (SOEs) - Referees and Players? 7 - Why Are Foreign Players Especially Challenged in China?
1 - Chinese Internet Players at Valuation Discounts to Global Internet Peers
China Internet companies are currently trading at 40-60% discounts to their global counterparts. Interestingly, their EPS growth rates, based on our forecasts and consensus estimates, are higher than their US comparables.
Exhibit 12
the CEOs of Verizon Wireless, Sprint-Nextel, and AT&T exchange jobs. There has also been speculation that mobile carriers may potentially change their revenue sharing scheme with ISPs. Third, China Internet companies have been troubled by operational risks. NetEase was at one point nearly delisted due to accounting missteps. Sohus multimedia messaging service (MMS) business was suspended for a year due to billing malpractice. Sinas interactive voice response (IVR) business was sanctioned due to improper content. Fourth, the business models for China Internet companies are still unfamiliar to global investors. China leads the world in MVAS and Online Gaming in terms of volume and growth. Yet these businesses are not as prevalent in the US and have hardly any listed comparables. Anecdotally, few people stateside have heard of a color ringback tone, while in China, there are 44MM ring back tone subscribers alone on China Mobile. (A color ringback tone is the sound that callers hear instead of a normal ring when they dial your number.) These uncertainties increase the risk premium and hence the discounts of China Internet stocks. In the long run, we believe these risk premiums may trend downward because: Government and SOEs may learn from their past mistakes to make their policies more favorable to Internet companies. China Internet companies may optimize their operations and solidify their leadership. Overseas investors may gain more knowledge of Chinese Internet economy.
Price Discount of China Internet Firms versus US Peers

Top 5 Internet Companies China Average P/E multiple Average earnings growth Average operating margin 15-25x 25-30% 30-40% US 40-50x 25-30% 15-20%
Source: Morgan Stanley Research (Excluding stock-based compensation costs, the average operating margin for the top 5 US companies is 20-30%).
We believe there are several plausible reasons for such a valuation gap. First, the competitive dynamics may change faster in China than in other Internet economies. Global comparables, such as eBay, Google, Yahoo! and Yahoo! Japan, achieved their dominance in their respective areas with formidable barriers to entry and powerful core competencies. These companies are rewarded with leadership premiums by the stock market as they are measurably ahead of the competition. By comparison, it is still too early to call the final winners in Chinas Internet space as the competition is still fierce and the rankings of the leaders are still in flux (See Leadership Position May Not Necessarily Be Secure). Second, policy risk is higher for Chinas Internet players than for their global counterparts. For instance, in 2H2004, based on a sudden government mandate, the heads of the top three telecom carriers shuffled among one another. This decree is comparable to the US government mandating that
Page 14
2 - Real Revenue Perhaps Significantly Higher than the Headline Revenue
For 2004, we estimate the industry revenue for ISPs was around $770MM for MVAS (up 89% Y/Y), $390MM for online gaming (up 90% Y/Y) and $220MM for online advertising (up 78% Y/Y) (Please see Exhibit 2 for details). We assume the total 2004 headline revenue for Chinese Internet companies approximated $1.5-2.0B (also including ecommerce, search, and other Internet services), relative to $3.3B 2004 revenue for eBay, $3.2B for Google, and $3.6B for Yahoo!. However, we believe these headline figures may be misleading: Morgan Stanley economists estimate the difference between the US and China in purchasing power parity (PPP) is around 4.4x. Chinas total 2004 Internet industry revenue should thus be equivalent to $6.5-9B in terms of PPP. IDC estimates that for every $1 of online game revenue in 2003, the revenue multiplier, or the revenue that the entire industry value chain may generate, should be 11x (telecommunication industry 6x, IT industry- 3x, and media and publishing 2x). Based on such a ratio, the online game-related revenue could be around $3.5-4.5B for 2004. We estimate person-to-person (P-to-P) MVAS accounted for around 80% of total MVAS volume in 2004 and the revenue multiplier for MVAS should be around 5-6x (including hardware, software, bandwidth, transmissions, and content spending, etc.). The 2004 MVAS-related revenue should thus exceed $4-5B. In comparison, China Mobile generated around $3.8B revenue from wireless data and other new services in 2004. The bottom line figures for China Internet companies are more impressive than their top line numbers. Their operating margins typically range between 30-40%, nearly twice higher than those for their US counterparts (See Exhibit 13). We believe investors should focus on the growth potential of China Internet as MVAS, Online Gaming and online advertising may expand at 2540% revenue CAGR for next three years.
It is of note that the combined market capitalizations for Amazon.com, eBay, Google, Yahoo! and Yahoo! Japan were around $2B pre2000 IPO. We make this calculation as follows we take the initial offering price times the shares outstanding after the offering (from the IPO prospectus). For Google, we include $100MM in post-money valuation. Currently the combined market capitalizations of these companies have climbed to over $230B as of August 2005.
3 - Chinese Internet Players Have Much Higher Margins than Global Peers
We estimate that listed China Internet companies typically have average operating margins of around 30-40% for 2004, nearly two times higher than their US counterparts.
Exhibit 13
Comparison of Operating Margins Between Selected Chinese and the US Internet Companies
FY2004 Reported Operating Margin
China China Finance NetEase KongZhong Tencent Shanda Ctrip Sina Sohu Linktone TOM Online 51job Average 65% 50% 41% 40% 39% 38% 35% 33% 25% 25% 20% 37% US eBay Google Yahoo! InterActive* VeriSign Amazon.com
FY2004 Reported Operating Margin

32% 20% 19% 17% 11% 6%
Average
18%
Note: All data are based on the reported results for fiscal year 2004. * InterActives operating income excludes amortization expenses. Note that excluding share-based compensation costs, the average operating margins for the above Chinese Internet companies and the US companies would be 39% and 27%, respectively) Source: Company data.
Lower labor cost is the key to the high China margins. We estimate that 2004 operating expenses (opex) were $8,000 per employee for Ctrip and $27,000 for NetEase, versus $7,000 for Chinese companies listed in the Hang Seng index, $73,000 for S&P 500 companies and around $300,000 for Microsoft (based on FactSet data). We believe labor costs may remain low because: There is abundant supply of labor in China. For instance, the Ministry of Personnel reported that
Page 15
there were 3MM job openings for 7MM job seekers in China for 1Q2005 in selected regional markets that it monitored. We have observed that the average salary for college graduates remained flattish or even declined over the past five years, likely due to the oversupply of recent college graduates. We estimate there will be around 3.4MM college graduates for 2005, up 35% Y/Y.
Trust. Despite a centuries-old merchant culture, the Chinese have not arrived at a level of comfort with buying goods sight unseen. Unlike a physical store, the web allows for only limited visual interaction with pictures or other representations of actual goods. While this observation is not new, it is especially valid in China, where consumers are conditioned to see and touch goods before buying them. GDP per capita. Chinas nominal GDP was $1,649B in 2004, vs. $11,733B for the US, according to Morgan Stanley economists, while Chinas nominal GDP per capita was $1,269 vs. $39,934 for the US. People with lower incomes will purchase less, all else equal. We estimate the average salary for residents in top-tier cities, such as Shanghai and Beijing, is around $180 per month. Given that the entry ticket to ecommerce in China is the price of a computer ($400-$600 for average desktop PC) and an Internet connection ($15 to $20 per month) or, in the case of Internet caf users, hourly caf charges of $0.3 to $0.7 it is reasonable to assume that the average consumer in China may not have much disposable income remaining to spend on ecommerce. That said, we believe this factor is potentially the easiest to reverse as Chinas economy continues to grow, and as population growth remains low, per capita GDP levels could continue to rise. Payment. CNNIC reported that 42% of online consumers used online payment in 2004, up from 16% in 2001. Shanda and NetEase developed direct billing relationships with online gamers through prepaid card system and e-sales. In China, mobile carriers currently control the billing relationship with 360MM+ Chinese mobile users. Most MVAS providers use such mobile payment channels for bill collection and are thus subject to the regulatory measures by mobile carriers. Companies specializing in online payment have mushroomed. For instance, YeePay plans to roll out three new payment methods, including SMS payment, fixed line phone payment, and mobile prepayment. SmartPay, a payment startup based in Shanghai, works by linking users bank accounts to their mobile phone number, thus allowing them to pay for products and services electronically.
However, we are concerned about potential cost inflation in certain areas: Competition may boost compensation costs. For instance, Electronic Arts indicated in early 2005 that it planned to hire around 500 game developers/programmers in China. NetEase pays out around 500,000 Rmb per year (or around $60,000) per head to retain its key game development talents. Property prices in selected regions, especially in Shanghai, have skyrocketed over the past three years. Our survey indicated that it may become increasingly costly for Internet companies to recruit and retain talents in these areas as Chinese people often prefer to own property in their working cities. Marketing expenses are on the rise. The9 has committed no less than $13MM in marketing expenses over the next four years to promote World of Warcraft, nearly ten times more than the typical marketing costs for one online game. Content costs may continue its climb (See Content Becoming the King). R&D expenses are trending upward as Internet companies beef up their in-house development for Online Gaming (such as is the case with Shanda, NetEase, and Tencent) and the new 2.5G / 3G MVAS products.
4 - Ecommerce Ready to Take Off?
We consider trust, GDP per capita, payment, delivery / logistics, and credit / settlement as bottlenecks for Chinas ecommerce development. We notice some significant developments since our April 2004 China Internet Report:
Page 16
Tencent reports that close to 45% of its sales come from non-mobile channels such as credit card and prepaid point card sales. Alibaba / Taobao developed Alipay, an escrowbased payment system, in collaboration with leading domestic banks, including Industry and Commerce Bank of China (leading commercial bank) and China Merchant Bank (pioneer in online banking). To date, over 60% of the number of transactions on Taobao clear payment through Alipay. eBay launched PayPal in China in July 2005. PayPal has over 70MM users worldwide and may offer a massive online trading opportunity between Chinese Internet users and the global Internet user base. Presently, PayPal China is only used for payments in China by Chinese users due to the government restrictions on Rmb from leaving the country. However, as restrictions ease over the long run, PayPal may enable foreign customers to take advantages of the low-cost merchandisers in China and help eBay EachNets sellers, mostly SMEs, to sell goods to overseas customers.
to customers at airports and collect cash on delivery (COD). Credit/settlement. Historically, due to potential cost and liability exposure, online commerce providers offload credit risks to online buyers and sellers. This has resulted in fraud including leakage of personal information, misinformation of listed goods, buyers canceling orders without proper causes, and sellers selling substandard products. The Online Signature Law, which was released in April 2005, grants online signatures similar legal authority and liability as printed signatures. We believe this may help reduce online fraud and facilitate the development of credit systems in China. Alibaba / Taobaos Alipay and eBay EachNets An Fu Tong are two escrow-based payment systems that help to reduce settlement risks (they typically receive payment from buyers first and then inform sellers to send out goods to buyers. Once the buyers confirm the reception of goods, Alipay or Au Fu Tong will release the payment to sellers). iResearch estimates that online shopping market expanded 2.6x Y/Y to $544MM in 2004 and may grow 93% Y/Y to $1B in 2005. We believe 2005 will be an important year for ecommerce providers to develop payment, credit and logistics systems before ecommerce truly breaks out in 2006.
5 - Why Can Click Plus Brick Outcompete?
Logistics. There are two emerging camps of logistics systems for online commerce in China: Merchant models which integrate inventory and distribution systems. One example is Joyo.com, now a wholly owned online commerce subsidiary of Amazon.com. The company developed a threetier logistics/distribution system: 1) In the key cities, Joyo has its own distribution team to ensure timely and orderly delivery of goods; 2) it also hires express mail services to complement its delivery; 3) in the second-tier regions, it mails goods through post offices. Joyo plans to open a 30,000 square meters-sized warehouse for inventory storage. Agency models which control delivery without owing any inventory. For example, Ctrip sells hotel room nights and air tickets without owning any inventory, which substantially reduces its inventory risks given the volatility and seasonality of traveling business. For air ticket booking, Ctrip uses its ground delivery people to hand tickets over
Taking a microscopic view of Chinas Internet companies, one finds that they tend to be hybrids between online and offline operations rather than pure-breed online entities. We believe such click plus brick model is and will be, at least for the medium term, the best way to thrive in Chinas unique and fast-moving business environment, for several reasons. 1) An offline presence facilitates customer acquisition: Ctrip has around 2,000 call center staff to sell tickets and hotel room nights. We estimate the company generated 70% plus of its sales from its offline operations. We assess that three quarters of 51jobs clients are SMEs with less than 50 employees. They typically lack dedicated human resource functions and thus require close sales coverage. As a result, 51job doubled its sales force Y/Y in 2005 to acquire and service these SMEs.
Page 17
2) An offline presence is important for product promotion: Online game operators, such as Shanda, NetEase, and The9, frequently organize on-site informational sessions and contests at Internet cafes to promote their new games. Ctrip deploys its sales team at airports to promote its services.
We have found that when calling Ctrip we can talk to a contact person within five to ten seconds of the first phone ring. In service-intensive and highly competitive businesses, these offline efforts are crucial to enhancing user experience and reducing the churn rate of a customer base.
6) Labor costs are and may continue to be inexpensive given the abundant labor supply in China: We estimate average monthly compensation for customer service staff typically ranges between $200 to $300, or about 10% of US peers. Ctrip estimates that its labor cost per booking is around $0.36, only 8% of the commissions per airline ticket and 5% of the commission per room night. China Finance generated an operating margin of 65% for 2004. For the year, we estimate the companys net revenue per employee was around $67,000, 3.5 times higher than its operating cost per employee.
3) Offline channels are important for product distribution: Shanda currently distributes its physical prepaid cards, which gamers use to pay for online playing time, through over 180,000 retail sales points (and electronic prepaid cards through 130,000 e-sales systems). Dangdang has a well-known bicycle courier service to deliver many of its goods to buyers. Ctrip delivers air tickets to customers at airports and collect bills on delivery.
4) Close relationships with business partners often require a local offline presence: ISPs like Sina and TOM Online have deployed local sales teams in nearly all the 30 provinces to tighten their relationship with regional offices of mobile carriers. 51job uses its local sales forces to build rapport with local media partners in over 20 cities. Baidu, 3721.com, and Alibaba have nationwide sales networks to market their listing services to customers, mostly SMEs who may have never used Internet before.
We believe that it is important for investors to focus more on offline operations of Chinese Internet companies.
6 - State-Owned Enterprises (SOEs) - Referees and Players?
In China, the role of state-owned enterprises (SOEs) should not be underestimated. On most occasions, they are both rule-makers and major players. To date, the government is still the largest shareholder of a majority of listed Chinese companies and it is thus important for investors to understand that, like most SOEs in other nations, the service priorities for Chinas SOEs are as follows: Government, which appoints the heads of SOEs. Thus investors should pay close attention to the propaganda and policy swings of the Chinese government, such as the anti-pornography campaign in 2H2004 that led to the crackdown on MVAS. Consumers, who bring bread and butter to SOEs tables.
5) Offline services are pivotal for customer retention: Shanda reports that it has over 280 dedicated customer service representatives who account for around 20% of its total staff and handle an average 6,000 plus calls, 1,500 emails, 800 faxes, and 200 plus visitors per day. NetEase has one of the top 10 customer service centers in Chinas IT industry.
Page 18
Business partners, such as ISPs, who are important but have little bargaining power for the time being.
China Mobiles Mood Swing
China Mobile (0941.HK, Equal-weight; covered by Mark Shuper, Hani Abuali, Lina Choi and team) is the largest mobile carrier in the world in terms of mobile subscribers (around 220MM to date). We estimate MVAS contributed around 16% of its 2004 revenue. In 2004, China Mobile mandated a MVAS content cleanup, termination of inactive accounts, migration of short messaging service (SMS) to a new billing platform, and sanctions on billing malpractice by ISPs. Over 1H2005, there had been circulating rumors that China Mobile may raise its revenue share with ISPs, which partially explains the share price declines for MVAS operators. We do not believe China Mobiles intention is to handicap the MVAS industry, which has already become its most robust growth engine. While we are supportive of China Mobiles initiatives to discipline the malpractice of ISPs, we believe the company could handle the cases in a more transparent and more cooperative way to minimize the damage to the budding MVAS industry. In July 2005, China Mobile made a positive move to make its revenue sharing scheme with ISPs transparent, which removed a lot of uncertainties in the space. In fact, we estimate mobile carriers have collected a majority of MVAS revenue because the bulk of MVAS are person-to-person (P-to-P), which ISPs, such as TOM Online and Sina, cannot benefit from. In the long term, it will be in China Mobiles best interest to tighten the partnership with ISPs and incentivize them to work hard.
that the original founder of 3721 left Yahoo! China in August). Ourgame.com lost its market leadership to Tencent in concurrent game players within one year after its acquisition by NHN, a Korean Internet company. We believe there are several roadblocks that account for these shortfalls: 1) Regulatory barriers are high in China: Foreign companies are not allowed to directly operate online gaming in China. They thus need to partner with domestic players for online game offerings, which partially explains the alliance between NCsoft and Sina as well as Universal / Blizzard and The9. Foreign companies are restricted from providing news and online advertising by Chinese government. Even in traditional media business, no overseas TV company, except for Phoenix TV, offers any Chinese news programs. Google, a supposedly non-content provider, has been suspended several times in China for improper content in its search results.
2) Uncertainties in the operating environment, especially in terms of working relationships with SOEs, can deter foreign entry. For instance, until recently, we have not seen any foreign penetration into MVAS, which is by far the largest revenue contributor to Chinas Internet services. We attribute this abnormality to concerns related to regulatory crackdowns and the perceived difficulties in dealing with mobile carriers by foreign players. 3) Local content is another key factor. For instance, other than South Korean / Japanese online games that can have similar cultural origins, no other foreign games ranked within top 10 in China in 2004 (according to The News & Publishing Bureau of China). Other popular Internet applications, such as music content, also require local flavor. If history ever repeats itself, we believe foreign entrants, similar to the founders of the Qing Dynasty (which lasted from 1644 to 1911) need to transform themselves to fit Chinese local culture instead of the other way around. In traditional media business, MTV has achieved popularity in China in part because it produces half of its music programs using local content.
7 - Why Are Foreign Players Especially Challenged in China?
Despite the escalating presence of foreign interests in China, no foreign Internet company has thus far achieved runaway success in China. eBay EachNet has suffered market share loss to Alibaba / Taobao. Before its acquisition by Amazon in August 2004, Joyo was the traffic leader in online shopping but as of September 2005, its site traffic only accounts for 60% of Dangdangs (according to Alexa.com). Over the past six months, as per Alexa.com, Baidus traffic nearly doubled whilst 3721, which was purchased by Yahoo! China in November 2003, remained flattish (note
Page 19
4) Localized salesforces and distribution networks are key (See Why Can Click Plus Brick Outcompete?). 5) After selling their stakes to foreign partners, most local founders have achieved financial independence. It is conceivable that they lose their sense of urgency due to their financial comfort and status change from owners to employees. 6) The complex corporate structure can slows down the responsiveness of the foreign subsidiaries in China. It may take days or even weeks for the shuffling of correspondence between the local subsidiaries in China and their US headquarters while their domestic competitors need only a few hours to make a similar business decision. It is of note that the chiefs of foreign companies are typically not frequent visitors to China per year, which can limit their engagement and understanding of Chinas market. 7) When entering into China, most foreign players send their sales and marketing teams first and their R&D teams last. In a tech-heavy industry, such as the Internet, R&D support is critical to optimize products and services to make marketing and sales more effective. For instance, when using the Internet in China one typically finds noticeable latency (relative slowness) in the US-based sites (like Google) when compared with local rivals, such as Baidu and Sohu. 8) Few foreign Internet players seem to appreciate that they are competing against some of the brightest folks, who we consider the Yao Mings of the Internet.
Battle Between eBay EachNet and Alibaba / Taobao
In 2004, eBay completed its acquisition of EachNet, then the dominant leader in online auctions, for a total consideration estimated at $180MM. In 2005, eBay announced it would commit an additional $100MM investment (including marketing, promotion, R&D, capex etc.) in China for 2005. During 2004, we estimate eBay EachNets market share in gross merchandise value (GMV) shrank to around 50% in 4Q2004 from 90% in 1Q2004. Taobao reported its share expanded to over 40% in 4Q2004 from 9% in 1Q2004. For 1Q2005, eBay EachNet reported US$100MM in GMV while Taobao reported US$120MM. For 2Q2005, Taobao reported US$200MM in GMV while eBay EachNet did not disclose more details. That said, it is of note that Taobao's listings makeup is around 10% auctions, while eBay EachNet reports that its makeup is 40% auctions, 20% combined auction / fixed price, and 40% fixed price. With an early May listing fee reduction by eBay EachNet, we believe the market share shift has stabilized between the 2 ecommerce players and the gap in new listings between eBay and Alibaba may have narrowed.
Here are some (6) of our observations regarding eBays investment in China and its recent performance: 1) After the acquisition, one of EachNets two original Chinese founders, Haiyi Tan, retired. Bo Shao, another founder, became chairman in November 2004 (the COO James Zheng has been running the site since Bos move to chairman). We believe their managerial talents, savvy in customer service, and insights into competition were the key attributes for EachNets early success. eBay imported overseas managers to oversee its China operation, who, in our view, need to climb a steep learning curve before zooming ahead of the curve. However, of note on August 26, 2005, eBay EachNet announced that the former Chief Marketing Officer of Microsofts Greater China Region Wu Shixiong would become eBay EachNets new CEO. The position had been vacant since Bos change. 2) Unlike eBay EachNet, Taobao offers free listings and had 10MM listings, more than 10x eBay EachNet as of this writing. Taobao only had 20,000 listings two years ago and around 3MM half year ago. That said, it is important to note
Page 20
that Taobao listings have longer expiration date (Typically, sellers can choose between 7-14 day listing periods; at the expiry of the period, the listing is automatically extended for one more period, after which it is taken off), whereas a given eBay EachNet listing expires in 7-10 days. Furthermore, eBay EachNets conversion rate is running at roughly 7 days, so listings turn over much faster than with Taobao. We believe this improves the vibrancyof eBay EachNet, making the inventory for sale on eBay EachNet fresher than that of Taobao, in our view. We believe that Taobao may eventually convert to the eBay EachNet format. 3) Taobao offers a more user-friendly interface: Taobaos listings are organized into several major categories, such as Men, Women, and Home. In our view, such an arrangement makes it easier for buyers to find their preferred shopping destination and sellers to find their target customers. In contrast, eBay EachNet simply groups customers into Buyers. and Sellers. Note that Internet users only account for 8% of the total Chinese population. Most of the online shoppers are early adopters and any little help to them may determine who wins these customers. Taobaos listings appear to be more customer-centric whilst eBay EachNets appears to be more productcentric. For instance, in Taobaos Men section, a male customer may get the instant feel that Taobao really wants to make you look cool and stylish. Its listed goods are typically organized around such attention-grabbing topics as Elegant Men, Tasteful Men, and Fun-seeking Men. To us, by doing so, Taobao is trying to convey a strong message: You are my customer, I care about you. What can I do to better serve you? In contrast, eBay EachNet has a Mens Clothes and Accessories section with a long textual list of various mens products. To us, the message from eBay EachNet seems to be: We have
some great products here, so be our customers. Note that most of the Chinese Internet users are likely the only children of their families, and are accustomed to being the center of attention. Taobao emphasizes community experiences by allowing its customers to share knowledge through its IM service, Taobao Wang-Wang. We believe such services may create stickiness for Taobaos community and enhance the chances for online transactions as customers share their knowledge, buyers and sellers may get into each others mind in more depth (although some users are turned off by constantly being under pressure from users who want to buy off-site or want to negotiate a lower price). It is our experience that Taobaos web site tends to be graphically richer than that of eBay EachNet. Note that close to 70% of Chinese Internet users are under the age of 30. They are trendy, fun-seeking and are typically short in attention span. As you compare the web sites in Exhibits 14 and 15 (for an apples-toapples comparison, we selected the screen shots from the same book / audio / video channels from the two sites), it is not be difficult to figure out which site may be more eye-catching and more appealing to young Chinese online consumers. More importantly, it is less appealing to sellers if a web site has too much text and too few graphics as they most likely want to see their products standing out graphically rather than being buried within mundane text. Note that the front page placement of sellers / items on Taobao are not purchased, but chosen by Taobao, whereas eBay EachNet (similar to all eBay sites) doesnt assign one listing / seller top priority. Finally, the user satisfaction level was 77% for Taobao versus 62% for eBay EachNet for 2004 (according to iResearch).
Page 21
Exhibit 14
Screenshots for the Homepages of Taobao and eBay EachNet
Source: eBay EachNet, Taobao.

Exhibit 15
Taobaos Web Designs Tend to Be Graphically Richer Than Those of eBay EachNet
Source: eBay EachNet, Taobao (The screenshots were taken from the front pages of the book/ audio/ video channels from the websites for Taobao and eBay EachNet).
4) Taobao adopted AliPay, an escrow-based payment system. To date, Taobao reported 90%+ of its customers have accepted AliPay and 60%+ of the number of Taobaos transactions are done using the system. eBay EachNet also has an escrow-based payment system An Fu Tong which has been in place since October 2004. Additionally, eBay EachNet just rolled out the PayPal payment system. In our view, AliPay may continue its lead in the domestic payment market as it helps to resolve the settlement risks among buyers and sellers through its escrow mechanism,
whilst PayPal is a pure play online payment system (which may not resolve the settlement issues). However, in the long term, we believe PayPal may enable international trading as it could help Chinese online customers tap into overseas markets by leveraging PayPals large global user base (over 70MM users to date). PayPal China offers payments in the local currency and integration with 15 local banks and more than 20 different debit cards, and PayPal China also offers buyer protection off eBay, such as with the popular portals NetEase and TOM Online.
Page 22
5) The corporate reporting structure between eBay EachNet and its headquarters may, at the margin, slow down the decision-making process. An industry as dynamic as the Internet demands prompt responses to market changes. For instance, Taobao can typically turn around a decision within hours. 6) Alibaba, the parent company of Taobao and AliPay, is the largest B2B online marketplace operator in China. It also targets SMEs as Taobao. We believe Alibaba adds value to Taobao given its similar customer focus and extensive nationwide sales network. Additionally, the sellers on Taobao can source their products from wholesalers listed in the Alibaba marketplace. That said, it is of note that eBay has high standards for its operation. For instance, eBay EachNet focuses on a level playing field, similar to all eBay sites, and does not grant any one listing/seller top priority. In comparison, the front page placement of sellers/items on Taobao is not purchased but chosen by Taobao. We also note that on May 1, 2005, eBay EachNet lowered both its fixed price and auction listings fees while leaving final value fees unchanged, making the platform more attractive to users vis--vis Taobao free listings. (See Exhibit 16). And since that time, we believe the listings on eBay / EachNets web site have rebounded. Note, while not an apples-to apples comparison, we look at both eBay EachNet listings (P2P) and (as Taobao does not make its listings count available on its site) Alibabas listings (B2B) as rough indicators of the health of each platform with general directional significance. We believe that eBay EachNet listings have responded favorably since that price change (see Exhibits 17-18). We will continue to focus here as the China online auction market develops, and the initial signs of listings improvement for eBay EachNet could point to more than just the near-term impacts of the pricing change. More specifically, in the 50 days since eBay EachNet reduced its fees, we estimate that daily listings increased 42% to an average of 554K vs. 390K in the 50 days prior to the change.
Exhibit 16
eBay EachNet Fee Reduction in 2Q2005 (05/01) (Rmb) Original New Fixed Price Listing Fee Listing Fee 0.80 0.10 1.00 0.50 2.00 1.00 4.00 3.00 8.00 6.00 8.00 6.00 New Auction Listing Fee 0.05 0.25 0.50 1.50 3.00 3.00
Listing Price 0.01 - 1.00 1.01 - 99.99 100 - 499.99 500 - 1999.99 2,000 + Motor Vehicles
Source: eBay EachNet, Morgan Stanley
Exhibit 17
Weekly Listings in China eBay EachNet (P2P)

eBay EachNet (P2P)
600,000 500,000 400,000
Listings
300,000 200,000 100,000 0

11 /0 9/ 04 12 /0 9/ 04 02 /0 9/ 05 03 /0 9/ 05
trendline
eBay EachNet
06 /0 9/ 05
07 /0 9/ 05
07 /0 9/ 05
Source: eBay EachNet, Morgan Stanley Research

Exhibit 18
Weekly Listings in China Alibaba (B2B)

Alibaba Listings (B2B)
600,000 500,000 400,000
Listings
01 /0 9/ 05
04 /0 9/ 05
05 /0 9/ 05
trendline 300,000 200,000 100,000 0

11 /0 9/ 04 02 /0 9/ 05 03 /0 9/ 05 01 /0 9/ 05 12 /0 9/ 04 04 /0 9/ 05 06 /0 9/ 05 05 /0 9/ 05 08 /0 9/ 05
Alibaba
Source: Alibaba, Morgan Stanley Research; Taobao does not make its listings count available on its site
08 /0 9/ 05
Page 23
We believe the ongoing battle between Alibaba and eBay underscores the key challenges faced by foreign Internet companies in China, and reinforces our view that foreign companies that rely less on local elements or could better leverage local elements may have better odds for success in China: Over time, Google may have a strong chance to outperform in China because successful search results are especially technology and scale driven. While the majority of results for China searches today are endemic to China, this may evolve over time to become more global as Chinese people may increasingly want the best information instead of the best Chinese information. It is notable that Google spent $319MM in capital expenditures in 2004 (vs. $3MM for Baidu) in its quest to organize all the worlds information and make it universally accessible and useful and develop universal language translation on the fly and leverage its global base of distributed computers. eBay may be able to increase its market share if it tailors its global services toward the local needs rather than delocalizing its services to meet its global standard. In the long run, eBay may add value to Chinese consumers through its global trading platform (As of 1Q2005, 15% of eBays transactions are across-border).
In our view, Yahoo!s recent alliance with Alibaba should add several positives: 1) It may create cross-selling synergy among sponsored search (3721/Yisou), e-commerce (Taobao/1pai), and B2B marketplace (Alibaba) as they mostly target small and medium enterprises (SMEs); 2) AliPay may offer the much-needed payment channels for Yahoo!s China operation; 3) Yahoo! may help Alibaba (now the worlds largest online exporter) and Taobao to expand overseas trading opportunities; 4) Alibabas veteran management team, including Jack Ma and his crew, would be a huge plus for Yahoo! to penetrate Chinas local market given their managerial experience and local consumer knowledge. Despite these areas for upside, to date nearly all of the foreign Internet companies have lost momentum after they bought their ways into China. We need to emphasize here that integration would be the key for Alibaba and Yahoo! to unlock the hidden value, each with the help of the other. It may be challenging for other foreign companies to outcompete independently Chinese domestic rivals for local content, local services, or local logistics.
Page 24
China Internet Challenges / Risks

1 - Economic Hard landing or Soft Landing? 2 - Expect the Unexpected! 3 - SOE-Friendly versus Entrepreneur-Friendly 4 - Too Much Capital, Too Few Good Ideas 5 - To Buy or to Build? 6 - Long-Term Value versus Short-Term Profit 7 - Dearth of Innovation 8 - Founder Mentality
9 - Hollowing of Management Teams
1 - Economic Hard landing or Soft Landing?
We are concerned about the potential slowdown in Chinas economy because: Property bubbles in areas such as Shanghai may scare away investment fund inflows. In 2004, investment accounted for 44% of Chinas GDP (versus 20% for the US), the highest in the world and the highest in Chinas history.
Chinas domestic consumption is low, which contributed 42% of Chinas GDP in 2004, versus 70% for the US. In a time of economic downturn, consumption may likely be tighter. Non-performing loans (NPLs) could become a severe burden on Chinas economy and banking system. Morgan Stanley Asia/Pacific banking analysts estimate NPLs approximated 18% of Chinas GDP in 2004 after adding back NPL transfer, versus around 5% for Japan in 2004 and 8% in 2001, the highest level in a decade. If the economy decelerates, NPLs may accelerate.
Exhibit 19
Comparison of GDP Compositions between China And the US

As % of GDP China US 42% 70% 12% 15% 44% 20% 2% -5% 36% 7% 100% 100% Differential -28% -3% 24% 7% 29% 0%
Personal consumption Government consumption Investment Net export Export Total

Source: CEIC.
In our view, a market deceleration may have different impacts on various online businesses: Online advertising is cyclical and parallels economic development. For instance, over the past two years, Chinas automobile sector saw declining advertising sales following severe price wars and a sales draught. Basic MVAS, such as SMS, may be less affected by economic volatility due to their inexpensive price. We argue that more people may choose to select such low-cost communication means when they become pocket shy during an economic downturn. On the other hand, high-end offerings, such as some MMS and wireless application protocol (WAP) products, may be hurt as consumers may become more price-sensitive. Online Gaming may be the least affected by a downward spiral of the market due to the addictive nature of Online Gaming and their inexpensive price relative to other entertainment
Morgan Stanley analysts believe that baseline crude oil prices for 2006 may rise from $45 to $64. Morgan Stanley Global Economist Stephen Roach notes that China could be especially vulnerable in an energy-shocked climate. The energy intensity of Chinese GDP is double the global norm, and there is no other economy in the world as highly levered to the over-extended American consumer. The trade wars between China and the US/European Union may drag down Chinas exports, which approximated 36% of Chinas 2004 GDP (versus 7% for the US).
Page 25
formats (for instance, a ticket for a two-hour movie may cost five times more than playing two hours of a game online at an Internet cafe).
2 - Expect the Unexpected!
government is to minimize the unemployment rate and social unrest. The heads of SOEs are appointed by government and typically have better relationships with policy makers than entrepreneurs. Beijing Media Corp., the first overseas listed stateowned media company, has received a five-year tax exemption, versus two-year tax exemptions and three-year tax deductions (typically 50%) granted to private Internet companies. The company also secured the government endorsement for crossmedia ownership, which is typically restricted for private media companies.
In China, there have been relatively sudden policy and operational uncertainties that adversely affected company growth. For instance In 2H2004, owing to the sudden government mandates, the leaders of the top three telecom operators, including China Mobile, China Telecom, and China Unicom, shuffled among one another, which was equivalent to the exchanges of the CEOs among Verizon Wireless, Sprint-Nextel, and Cingular and unlikely to occur elsewhere in the world. Tencent announced that, starting from June 2005, China Mobile would cease to remit any revenue to Tencent for their joint venture, 161chat. In other words, China Mobile confiscated 161chat from mid-2005 onward. Cheating tools and private servers are widely publicized and offered online for popular massively multiplayer role-playing online games (MMORPGs). Illegal private servers caused The9s revenue to decline by 47% within a quarter in 1Q2004. MP3 search and music downloading, which may be perceived as a copyright violation in advanced markets, are among the most popular online applications in China. We estimate that MP3 accounts for around 20% of Baidus traffic. It is notable that NetEase terminated its MP3 search in August due to copyright concerns.
In our opinion, while selected areas, such as Zhong Guangcun high-tech village in Beijing and Zhangjiang hightech zone in Shanghai, offer a favorable local policy to the domiciled startup companies, the overall policy environment still needs improvement to expedite entrepreneurship and the shaping up of the Silicon Valley in China. That said, as noted in the following Exhibit (Exhibit 20), the Internet is one of the industries most opened up, and affected by the entry of China into the WTO.
3 - SOE-Friendly versus Entrepreneur-Friendly
We believe Chinas current business environment is more favorable to state-owned enterprises (SOEs) than to startups. Government is the largest shareholder of SOEs. We estimate government holdings account for 6070% of total shares outstanding for the listed SOEs. SOEs accounted for 44% of urban employment (CEIC 2003). The overarching priority for Chinese
Page 26
Exhibit 20
Effect of China's WTO Membership for Selected Industries Relative to Historical Degree of Protectionism
150
Dramatic
Chemicals
Internet Services
Banking
Distribution / Retailing
Effect of WTO Membership
100
Insurance Agriculture and Agribusiness
Moderate
Autos Securities 50 Pharmaceuticals Energy
Negligible
Electrical Equipment, Electronics 0 0
Processed Foods, Consumer Goods
Telecom Services
Low
50
Medium
100
High
150
Historical Degree of Protectionism
Source: United States-China Business Council, McKinsey Analysis.
4 - Too Much Capital, Too Few Good Ideas
In 2004, around 44% of Chinas GDP came from investment, the highest in the world and the highest in Chinese history. In our view, what happens in the pre-IPO market mirrors Chinas macro picture too much capital inflow, not enough good projects which may jeopardize the future returns for direct investment: Increasingly we have witnessed that: 1) Private equity or venture capital firms easily raise Chinarelated funds exceeding a couple hundred million dollars while Chinese start-ups seldom consume more than $1-2MM in funding per company in early rounds; 2) foreign investors who have rarely been to China have begun to pour money into China; 3) global executives who seldom visit China more than once or twice a year start to execute their China strategies. As the listed MVAS companies are trading at mid- to low-teens earnings multiples, the pre-IPO investment is converging with the public market valuation. When money becomes a commodity, prices rarely become cheap. New ventures take time to develop. The current Chinese Internet IPOs mostly began their
operations 4-5 years ago. Among the hundreds of MVAS providers, only six are publicly traded. Recently, companies that have not fully developed their core strengths have been pushed into the public markets with high expectations for impressive execution before they are ready. Baidus 4x share price jump on the first day of its IPO was a vivid example of a supply and demand imbalance for the shares.
Despite these challenges, we still see opportunities: Other than financing, venture capitalists may add more value through their industry insights, networks, and their abilities to coach the young entrepreneurs through ups and downs. Moreover, venture capitalists with solid track records also add credibility to the company. It is our observation that pure-play domestic startups still have difficulty raising money because they typically do not have returnees (who were born in China but trained overseas) as management and are not as polished in presenting their stories to investors. However, they may be close to the local
Page 27
market and offer services that consumers crave. It is thus important for venture capitalists to examine more closely what these entrepreneurs do instead of what they say in order to identify the investable opportunities. Industry leaders, such as TCL and Broad Air Conditioning, were self-financed in their early stage. Venture capitalists in Silicon Valley have the advantage of being close to new business models and new technologies both physically and intellectually. Their counterparts in China may not yet have that luxury. Thus being proactive in identifying future trends and being willing to endure extensive ground work would be crucial for long term investment returns in China.
Is there synergy between the target and our existing business, such as complementation in services or demographics, cross-selling and costsaving? What is the core competence of the target, R&D, marketing, or relationship? Will its leadership be sustainable? Is the management capable and trustworthy? Can we work with them seamlessly once we convert them from bosses to employees? Do we have the right incentive structure in place to keep the existing management motivated? Is the price for the acquisition right? If there is a premium to prior acquisitions or other comparables, does the target deserve it? Is the acquisition earnings-accretive?
5 - To Buy or to Build?
Chinese Internet companies are split into two camps in terms of growth through buying or building. One is represented by NetEase and Tencent, which have historically been focusing on building organic growth. The other is exemplified by Shanda (which bought Haofang, Bianfeng, and Digital Red, as well as stakes in Sina and Actoz), TOM Online (Puccini, Treasure Base, and Indiagame), Sohu (ChinaRen, 17173.com, and Focus.net), and Sina (Memestar and Crillion). In Chinas current business environment, we believe the biggest hurdle for acquisition is integration, during which problems may arise due to cultural conflicts and the ego issues of founders and key management. We believe it is key to focus on organic growth first. When making acquisitions, it is key for an acquirer to go through a Business School 101 checklist: Can we build the business on our own? If so, how long and how much money / time will it take? Can we achieve similar market leadership through self-development relative to potential acquisition? Will the cost be substantially different? If we have to buy, does our target address a sizable market with high growth potential? Is the business model irreplaceable, with high entry barriers? Will the acquisition help us to better service our customers?
Sinas M&A Savvy
Sina is a company built upon M&A (see our initiation report on Sina). In our view, its management team is among the most financially savvy in China. In 2003, Sina jumpstarted its MVAS business via the acquisition of Memestar, a leader in MVAS, for $24MM (including both cash and stocks). We estimate the face value for the acquisition was only 2-3X of Memestars 2003 operating profit and yet Memestar contributed 30% of Sinas total sales in 2003. In 2004, Sina purchased Crillion, a quasi-monopoly in wireless job search, for an initial consideration of $19MM plus two contingent payments of around 1.52.0X Crillions 2004 and 2005 earnings. We calculate the acquisition price was only around 3-4X of Crillions 2004 operating profit. Sina reported that Crillion and Memestar accounted for 30% of its 2004 sales. We note that integration and sustainable growth following acquisitions may remain a challenge for Sina.
6 - Long-Term Value versus Short-Term Profit
In our view, some Chinese Internet companies are too focused on delivering short-term profits, especially for the purpose of passing the quarterly earning test set by many on Wall Street. They may thus attempt to extract extra dollars from consumers, usually through irrational and erroneous sales and marketing tactics. This is a direct
Page 28
violation of our overarching investment thesis consumers always come first, share price follows.
ISPs Mishap in MVAS
New Internet applications, such as VoIP and blogs, are typically created elsewhere in the world and, after gaining popularity, are then adopted by Chinese companies. When a new online service becomes popular, you could expect the mushrooming of dozens or sometimes hundreds of Chinese copycats within a short timeframe.
In 2004, literally all the listed MVAS providers were sanctioned by mobile carriers for billing malpractice (such as erroneously charging customers who have been inactive), marketing manipulation (such as push marketing to sign on customers without their consent), and improper content (see Exhibit 21 below). Such behavior of ISPs in boosting short-term profits ended up hurting their share prices, which have to date not recovered to pre-sanction levels.
We believe Chinese Internet companies need to beef up their efforts in innovation to stay competitive and to sustain their leadership as foreign competition tries to enter Chinas market aggressively.
8 - Founder Mentality
Exhibit 21
Sanction List for MVAS Providers for 2004

Services Selected Companies That Were Sanctioned Reasons for Sanctions
Most Chinese Internet companies are founded by young entrepreneurs. They are talented, energetic, and resilient. However, we also observed that they appear to have some inherent limitations: Tight fists. Some founders tend not to delegate their responsibilities as they view the companies as their own creations. Their tight-fisted controls often create tension for professional managers hired to exercise their independent judgment and to add value with their own perspectives. Some founders also tend to empower their old aides, whose skill sets and managerial expertise may be limited as the companies reach a certain size. Limited vision. Some young entrepreneurs view IPOs as their endgame. Once they raise IPO proceeds, they are typically escalated to the richest class in China (our empirical formula is that if you multiply a Chinese persons wealth by a factor of 10, you will get someones personal wealth equivalent in the US). They may lose their sense of urgency and start to enjoy the fruits of wealth creation. Unfortunately, we believe if their vision is limited to IPO and personal gains, their business may fail to be a long-term winner. Lost focus. Unlike the fierce competition among US entrepreneurs Bill Gates (Microsoft), Steve Jobs (Apple) and Larry Ellison (Oracle), we seldom observe the fear factor for competition among Chinese Internet entrepreneurs. Given its robust expansion, Chinas Internet market could accommodate multiple winners at the present stage. However, the market may eventually slow down and may be consolidated by the most capable players.
SMS
Most listed ISPs Sina
Inactive accounts Billing malpractice Fortune-telling-related SMS Push marketing without mobile carriers' approval Billing malpractice Improper content
MMS
Sohu
WAP IVR
Linktone KongZhong/ Sina/ Tencent/ TOM Online
Source: Company data.
7 - Dearth of Innovation
We believe genuine innovation is still lagging in Chinas Internet industry: In the late 1990s, the first-generation Chinese Internet entrepreneurs saw what happened in the US and imported the US Internet models, such as portals, online auction, and paid search, into China. MVAS, which is the most popular Internet-related service in China, was pioneered by Japanese and South Korean Internet companies. In the online game space, Electronic Arts started the first MMORPG, Ultima Online, in the US in 1997 and NCSoft launched Lineage in South Korea in 1998. To date, other than a selected few, such as NetEase, most online game companies lack in-house development capability.
Page 29
The William Ding Premium
We consider William Ding, the founder of NetEase, to be one exceptional case among Chinese Internet entrepreneurs. He is focused on creating consumer value and willing to delegate his duties to professional managers (unlike other founders, Mr. Ding no longer assumes the CEO title). We believe he has the vision (the first to get into mobile messaging business, the first to steer away from it, and the first to focus on game development) and the leadership abilities to build his company to last. Moreover, Mr. Ding and his team are among the most innovative in the industry in R&D, marketing, and customer services, which altogether made NetEases Fantasy Westward Journey the most played MMORPG in China (see The William Ding Premium in our NetEase initiation report).
China Internet is still a budding industry with most companies having five years or less of history (considering that Sina and Sohu went public only in 2000). Experienced management is thus a scare resource. Given the fast pace of growth of Chinas Internet industry, a shallow management bench could inevitably derail companys future development.
Sohus Talent Exodus
Would Sohu be a different company if these people were still there? Over the years, Sohu saw an exodus of senior management, who in our view comprises an impressive All-Star cast.
Exhibit 22
Who Is Who Among the Ex-Sohu Executives

Name of Sohus Former Executive Victor Koo Sam Qian Lee Zhang Derek Palaschuk Yufan Zhou Positions at Sohu President Vice President Director CFO Founder of ChinaRen, now a Sohu subsidiary Vice President Positions After Sohu Plans to start own venture President, China Finance Co-founder, eLong CFO, eLong CEO, KongZhong
9 - Hollowing of Management Teams
It is a worrisome sign that some China Internet companies keep losing their key talent, who are either lured away by competitors or start their own ventures. We believe there are several negative implications: Their departure causes disruptions of company operation and their roles not easily replaceable. Their insights about corporate operations could help them to out-execute their former employers in their new competing roles.
Elaine Feng
EVP, TOM Online
Source: Company data, Morgan Stanley Research.
Page 30
China Internet Investment Framework Business / Competence / Price

Business Model - Large Market, Robust Growth, and High Barriers
When looking at business models, we prefer companies that target large markets with significant growth potential. This helps to explain why Shanda has become the No. 1 Chinese Internet company in market capitalization within four years and why there are more listed companies in MVAS than any other Internet sector. We prefer businesses that have achieved or may achieve clear market leadership, such as Sina in news content and Tencent in IM service. We believe their toll booth status may grant them pricing power (e.g., Sina raised its advertising price by around 20-30% per annum over the past few years except for 2005, while Tencent increased its advertising price by 50% in 1Q2005), which may enhance their bottom lines without incurring much incremental costs. We prefer businesses with high barriers to entry, which can be established through: Regulation - such as governmental restrictions on online news supply, which gives Sina a competitive edge. Cost - Tencent offers its basic IM service for free. Scale - TOM Online is now the largest MVAS content aggregator and Tencent has around three quarters of IM market share. Brand recognition - Sina and Sohu enjoy premium brand awareness among media buyers.
We prefer businesses with minimal regulatory interference. A contrary example is MVAS, which is subject to revenue sharing restructuring by mobile carriers, leading to increasing uncertainties and high discounts on share prices.
Core Competence - Management and Innovation
In our view, core competence is the corporate DNA that differentiates a company from its competition to produce a long-term winner. In its simplest form, core competence comprises the ownership of prime resources, such as news supply for Sina, online media assets for Sohu and IM community for Tencent. Moreover, we believe management and innovation are the two driving forces for the development of core competence. Regarding management, here are the top three questions we typically ask: Are they truly customer-focused? Are they capable in their own right and versus peers? Are they honest and trustworthy?
If the answer to all three questions is yes, we have the beginnings of an investment case. We view it as critical to invest in the right people because: Chinas Internet space is extremely fast-paced, with business models constantly subject to changes. Capable people with customer focus can figure out what is next and thrive on it. In the Internet space, you are competing against the brightest in China. Here are some good examples: the financially savvy team at Sina, the superior execution team at Ctrip, the competitive and consumer-oriented team at Alibaba/Taobao, and the innovative team at NetEase. We want to side with the best athletes who could win the proverbial championship for us. Make no mistake, the Internet space is volatile. You do not want to lose peace of mind by secondguessing managements incentives. Capable people
We prefer businesses with scalability. For instance, we estimate around 70-80% of Tencents content is selfgenerated by its IM users. NetEase and Shandas Online Gaming experienced continual margin expansion due to relatively fixed cost structures. We prefer businesses that enjoy network effects, which may increase the stickiness of an online community and the switching costs for its customers. For instance, more users for Online Gaming/IM service will typically attract more users due to interaction and community effect, thus forming a virtuous cycle for the expansion and the retention of customer base.
Page 31
may run fast but only those who are honest may share their rewards with you. If the key success factor for real estate is location, location, location, we believe the Holy Grail for Internet companies should be innovation, innovation, innovation. We prefer companies that are innovative and believe innovation may take different forms, including: In-house development capabilities - such as NetEases strength in gaming pipeline. Distribution network - such as Shandas coverage of over 300,000 retail sales points. Execution abilities - such as Ctrips outperformance in the service-intensive tourism business. Marketing strategies - such TOM Onlines alliance with handset manufactures (although it affects the companys margin in the short term but may help the company to gain long-term competitive advantages).
have shaky market leadership, with competition closing the gap focus more on ego and self-gratification than on consumer value and lag in innovation.
We prefer stocks that: trade at discounts to the growth potential of the underlying operations. For instance, our estimated 2006 PE ratios for our top picks, including Ctrip, NetEase, and Tencent, are currently at 05x- 07X of their forward operating profit growth rates. have margins of safety to our estimated discounted cash flow (DCF) values. For instance, our projected DCF value for Tencent is roughly HK$10.2 per share, versus the current share price of around HK$7.5-$8.0 (We assume a discount rate of 13%, versus an average of 10% for other listed Chinese stocks and 10-12% for the US Internet stocks under Morgan Stanley coverage, 11x exit free cash flow multiple, and a terminal growth rate of 4%). may offer higher value when each of the companys operations is treated as independent businesses. It is our observation that stock markets can sometimes underappreciate a companys asset quality due to complex structure and diverse business models, which may lead to more unknowns and thus heavier discounts. For instance, our estimated sum-of-the-parts valuation for NetEase is around $109 per share, versus the current share price of $70-75 (assuming 25x multiple for its online gaming operation, 22x for its online advertising business, 15x for its MVAS, plus net cash). have company managements with histories of outperformance (in other words, the share price may only reflect a low company guidance). For instance, Ctrip and NetEase both exceeded their own earning guidance by 20-25% for 2Q2005.
We see a direct correlation between innovation and consumer value creation (and hence share price performance). Taking the examples of Google and NetEase, it is evident that companies that are genuinely innovative are most likely ahead of the game. Those who are behind the innovation curve are likely underperformers.
Price - Margin of Safety Not an Alien Word
If you get your business model and core competence right, then you need to wait for Mr. Market to offer you the right price for the right stock. Regarding valuation, we tend to reward companies that are innovative and ethical and have achieved market leadership. On the other hand, we tend to assign discounts to businesses that: have heavy regulatory risks, such as MVAS experience margin erosion, such as some MVAS players whose margins are being squeeze by content providers and handset manufacturers undertake unethical or irrational behavior, such as business malpractice and ill-advised M&As
In addition, we also prefer to receive embedded options for free to further compensate for investment risks. Examples include:
Page 32
Tencents QQ Pets and Q-Zone operations NetEases email community, which is the largest in China re-rating of Sohus paid search business TOM Onlines opportunity in Skype.
Our Investment Case - Value Equals the Sum of Merits
In our investment framework, we assign 50% of the weight to business model, 30% to core competence, and 20% to price (for value investors, the order should be the other way around.). If a company has all three attributes (good business model, good core competence, and good price), it should prove to be a sound investment. If it misses all three, it will not be on our list
Page 33
Top China Internet Investment Ideas

1. Ctrip (Overweight-V)- A primary beneficiary of rising consumption power and travel growth in China. Ctrips sharp consumer focus, scalable platform and superior online and offline operational ability help make it a dominant player in the service-intensive travel industry. NetEase (Overweight-V) The leading innovator in the rapidly growing online gaming market in China. With its market savvy, focus on innovation, and strong management team, led by William Ding, we believe NetEase should be able to sustain market leadership. Tencent (Overweight-V) The powerhouse in Chinas rapidly expanding instant messaging market with over 60% of user share. Tencent has created a strong community with networking effects and high barriers to entry. The company has been successful at growing profits in Internet value-added services and MVAS and looks to continue to ramp monetization through Online Gaming and online advertising.
2.
3.
See our related reports on the merits and the investment risks for these three companies.
Page 34
China Internet Industry Analysis - Close to Consumers, Focus on Growth

Small Fish in a Big Pond or Big Fish in a Small Pond?
It is self-evident that market size is a pivotal indicator for revenue potential. For 2004, we estimate China Internet companies generated $700-800MM from MVAS, $350400MM from Online Gaming, and $200-250MM from online advertising, $100-120MM from Internet services (including ecommerce, avatar, online subscription services etc), and $50-80MM from paid search (please see the following chapters for detailed industry analysis). We place more emphasis on growth, which may stem from several fronts: Organic growth of the industry. We project the industry revenue CAGR for next 3 years at high30% for Online Gaming, mid-20% for online advertising, high-20% for MVAS, and 60% plus for paid search. New services. Examples include the deployment of 2.5G and 3G services in MVAS, which may significantly boost the user base for wireless application protocol (WAP), MMS, and K-Java applications (such as mobile games). We project such advanced MVAS expanding at CAGR of 4045% for the next three years, two to three times faster than existing 2G services, such as SMS. New products. Examples include 3D online games, such as Lineage 2 and The World of Warcraft. We believe these products may help to expand Chinas online game market due to their richer graphics and more interactive nature than 2D games. However, their revenue stream could be volatile due to their hit-driven nature, hardware requirement, and product life cycles. Moreover, these new products may cannibalize the user base of existing products. Outperformance due to market leadership. Industry leaders could grow their market share regardless of rain or shine. For instance, through better execution and raising barriers to entry, Ctrip has consistently outperformed its peers in the
travel sector while TOM Online has sustained its growth in a MVAS market which has been depressed by regulatory tightening. Industry consolidation. Sinas acquisition in MVAS, TOM Onlines buying spree, and Sohus purchase of prime online media assets may help to produce new growth opportunities due to crossselling synergies, potential cost saving and better services to consumers.
From a growth perspective, we rate Online Gaming and paid search first, MVAS (especially 2.5G) second, and online banner advertising last.
How Close Can the Companies Get to Consumers?
We believe there are several benefits of getting closer to consumers: better understanding of the market and more valueadded products to consumers lower discount to distribution channels faster turnover of customer feedback, thus improving service quality and customer retention more monetization potential and more bargaining power through direct access to consumers pockets (individual consumers or SMEs typically have the lowest bargaining power in the value chain in China nowadays)
Not all value chains are born equal in terms of consumer access (see the Exhibit 23). Online game operators can directly access gamers through Internet Cafes. MVAS providers, however, need to pay tolls, e.g, revenue sharing and policy adjustment, to mobile carriers before they could reach end users. Online advertising companies fall between online game operators and MVAS providers. They could approach their consumers but typically through advertising agencies/media buyers, who obviously have lower bargaining power than mobile carriers.
Page 35
Exhibit 23
Not All Value Chinas Are Created Equal Comparison Among Online Gaming, Online Advertising, and MVAS
Online Advertisers Advertising Agency
Online Game Developers
Online Game Operators
Internet Cafes
Consumers
Distributors
Content poviders
MVAS Operators
Mobile Carriers
(China Mobile & China Unicom)
Source: Morgan Stanley Research.
In terms of consumer access, we prefer Online Gaming over MVAS, with online advertising falling in between.
How Crowded Is the Space?
and TOM Online) and verticals (Chemease, IT168, and AsiaEC.com). Paid search providers also vie with traditional online media for market share. From the perspective of competition, we rate online gaming first, ahead of MVAS and online advertising.
Who Is the Boss in the Value Chain?
We prefer industries with a shrinking number of competitors. For instance: In MVAS, as a result of the regulatory crackdown, we saw higher barriers to entry and a declining number of competitors for new services. We estimate there were over 1,000 ISPs for SMS, around 200 for 2.5G, and only 100 for IVR. In online gaming, competition is also decreasing due to rising licensing costs and the need for an extensive sales and distribution network. However, the competitive landscape is still fragile as companies with one strong game could leapfrog the competition (for example, to date, NetEases Fantasy Westward Journey 2 surpassed Shandas Legend of Mir 2, which was the previous market leader, in the number of concurrent users). In online advertising, an oversupply of advertising inventory benefits up and coming portals (Tencent
The answer boils down to who sets prices and margins in the space: Game developers in online game business. Foreign game developers, such as Universal/Blizzard and Actoz, are commanding higher revenue shares. Game distributors are having their margins squeezed. iResearch estimates that channel discount, or the percentage of revenue cut for channel distributors, declined to 17% in 2004, down from 22% in 2003 and 30% in 2002. Integrated online game operators with selfdevelopment capabilities are enjoying margin expansion. NetEase now boasts one of the highest margins among Internet peers with gross margin at 90% and operating margin north of 60%.
Page 36
Leading portals in online advertising business. We estimate Sina, Sohu, and NetEase raised advertising pricing by 15-30% annually in the past. However, they appeared to lose their pricing power as they kept their advertising rate unchanged for 1Q2005 due to competition from up and coming portals, such as TOM Online (30% price increase for 1Q2005) and Tencent (50%), online verticals and paid search listings. Content providers and mobile carriers in MVAS (See Content Becoming the King). However, we believe MVAS leaders may start to regain pricing power as they reach critical scale. For instance, TOM Online saw its operating margin hit a trough in 3Q2004 and recover in 4Q2004.
handsets, and the likely slowdown in property markets due to macro tightening. Competing or disruptive products in online games, such as The World of Warcraft. Shift of revenue-sharing schemes for MVAS providers by mobile carriers. SOEs, which are integral players of the value chain that cannot be ignored. Whether you like it or not, they sometimes could be your best enemies (as they are easy to compete against, such as those in the online traveling business) but worst partners (such as mobile carriers in MVAS in some cases).
Industry Risks Double-Edged Sword
There are industry-specific risks to be mindful of: Economic or sector slowdown, which may hurt the cyclical online advertising business. For instance, leading portals already feel the adverse impact of weakening auto prices, competition in mobile
However, we believe market leaders may be able to play industry risks to their advantages. For instance, in MVAS, TOM Online leveraged its rapport with local mobile carriers to achieve its market leadership. KongZhong and Hurray! secured their leading positions for WAP through tight relationship with mobile carriers. Our top-down view is that online game is attractive, paid search robust, brand advertising challenging, and MVAS bottoming out.
Exhibit 24
Industry Scoreboard for MVAS, Online Brand Advertising, Online Gaming, and Paid Search
Industry Growth Consumer Access Entry Barriers Industry Risks Total Score Industry Ranking
MVAS 2 1 2 1 6 4
Online Brand Advertising 2 2 2 2 8 3
Online Gaming 3 3 3 2 11 1
Paid Search 3 2 2 3 10 2
Notes For Scores (3: Sales CAGR for next 3 years >30%, 2: 20%< but <30%, 1: <20%) (3: Good direct access; 2: Some direct access; 1: Access all through third party) (3: High entry barriers; 2: Some barriers; 1: Low barriers) (3: Low risks; 2: Medium risks; 1: High risks)
Page 37
Mobile Value Added Services (MVAS) - Light at the End of the Tunnel
MVAS Outlook - Overlooked Investment Opportunity
We believe MVAS may represent an overlooked investment opportunity: 1) The growth of the MVAS industry is robust: China ranks No. 1 in mobile phone users, which is more than those of the next three nations, the US, Japan, and Germany, combined. We estimate MVAS users may climb to roughly 30% of Chinese population by the end of 2005, up from 26% a year ago. MVAS revenue may expand at high-20% CAGR for the next three years, with new services, such as 2.5G, growing at around 40% CAGR.
3) Is there enough margin of safety? We believe yes. MVAS companies are trading at low- to mid-teens forward earnings, or around a 30-40% discount to Online Gaming and online advertising companies. Using the three-year forward industry CAGR, the PEG for MVAS companies as a group is around 0.5. 4) MVAS companies appear to be bottoming out, with signs of recovery. For instance, Sohus MVAS revenue recorded 28% Q/Q growth in 1Q2005 after five quarters sequential decline. TOM Onlines MVAS revenue jumped 22% sequentially in 2Q2005 after a modest 5% growth in 1Q2005. In our view, most listed MVAS companies may turn the corner in the coming quarters. 5) What other shoes may drop? Over the past year, we observed mobile carriers imposing regulatory tightening on literally every single MVAS business line, including platform migration for SMS, penalties on push marketing for WAP and on billing malpractice for MMS, as well as a content crackdown on IVR. After platform migration for MMS and termination of idle WAP accounts (which may last two 2 to three quarters), mobile carriers may exhaust their regulatory measures. 6) As a result of rising entry barriers in MVAS, competition may decline and it is more feasible than ever to pick the leaders. Our top stock picks in this sector include: Tencent. Its MVAS center on the interactions among its IM users, which constitute one of the most sticky online communities in China. This allows the company to enjoy lower content costs (as the bulk of its MVAS content is usergenerated) and fewer idle MVAS accounts (as most of IM users tend to stay active in communicating with their online buddies) than peers. TOM Online. The company is now the revenue leader in the MVAS space, with one of the most balanced portfolios of MVAS products. It owns differentiating content (such as the mobile games from Indiagame), differentiating services (TomSkype), and differentiating distribution channels (such as distribution partnership with 90 TV
Exhibit 25
Forecast for Commercial MVAS Growth in China

(CY2003- CY2010E)
2500 MVAS Sales Growth Rate 2000 MVAS Sales (US$MM) 120%
100%
80% 1500 60% 1000 40% 500 % Growth y/y
20%
0
20 04 20 05 E 20 06 E 20 07 E 20 08 E 20 09 E 20 10 E 20 03
0%
E = Morgan Stanley Research estimates Source: Morgan Stanley Research (The commercial MVAS refers to the computer-to-person (C-to-P) MVAS).
2) Other players have also experienced strong growth in MVAS. Foreign media companies, such as Viacom/MTV (China), expect that the MVAS contribution in China could be larger than their pay TV revenues in several years.
Page 38
stations, 300 radio stations, 50 newspapers, 45 magazines, and close to 40 video/audio distributors.).
Who Says There Is No Entry Barrier In MVAS?
One investment concern about MVAS is that the business appears to have low barriers to entry. However, over the past 12 months, we have observed a range of entry barriers shaping up: Cost barriers. To date, China Mobile has kept its revenue sharing with ISPs at 15% for bill collection. However, it will charge an additional 10-20% more if ISPs need to use China Mobiles marketing channels or needs to rely on China Mobile for customer services. For smaller ISPs which do not have sufficient capacity for distribution and customer services, gross margins may automatically be 10-20 percentage points lower than those of their larger counterparts. The costs are thus prohibitively high for new entrants who typically have shallow pockets. Regulatory barriers. Mobile carriers have installed a series of new regulatory measures, including content clean-up, termination of delinquent accounts, and new billing protocols, which significantly raise the bar for new entrants. Relationship barriers. We view China Mobile as a conglomerate of 30 local mobile carriers. These regional branches are responsible for promoting MVAS products and adjusting revenue sharing schemes with ISPs. We consider good relationships with local mobile carriers, which take time to develop, a key factor to achieve dominance in local markets. Distribution barriers. Some MVAS leaders pioneered innovative distribution channels, such as alliances with mobile handset manufacturers or TV channels to distribute new MVAS products. These deals are typically exclusive (we acknowledge that these arrangements may hurt margins in the near term but may help challenge competition in the long term). Content acquisition barriers. Acquiring quality MVAS content is now a difficult task. Whoever gets great content may draw large crowds and
increase the switching costs of their customers. For instance, KongZhong secured exclusive wireless distribution rights for Kung Fu Hustle and A World Without Thieves, two leading blockbuster movies in China for 2005. TOM Online has an exclusive deal for music content with Jay Chou, the hottest Chinese pop star at the moment. Scale barriers. This is a self-evident competitive advantage for large players. For instance, TOM Online has distribution alliances with over 90 TV stations, 300 radio stations, 50 newspapers, and 40 plus magazines. Would content providers and MVAS users prefer larger players for their scale in content and distribution over small players? We believe so.
We believe the market dynamics have validated these existing barriers: Other than the purchase of Skyinfo ($80MM in March 2005) by Japan-based Index, there has not been any meaningful foreign entry into MVAS. There are 1,000 ISPs that used to provide SMS. The number drops to 200 for 2.5G and 100 for IVR.
What Does It Take to Be a Winner In MVAS?
For the most part, MVAS providers often offer interchangeable services and products. We believe the key to success is to differentiate from competitors by offering what your rivals cannot, such as the following: Unique business models. Sinas Crillion Unit achieved a leadership position via exclusive contracts with regional job centers to provide jobrelated MVAS services. Treasure Base, a wholly owned subsidiary of TOM Online, has exclusive relationship with China Central Television (CCTV) to offer MVAS to TV viewers. These edgy models offer Sina and TOM Online growth opportunities even in a weak MVAS market. Unique services. Tencents MVAS centers on its instant message community and is communicationdriven rather than content-driven. As a result, Tencents MVAS has fewer inactive accounts and is stickier than peers. Hurray! focuses on WAP, which contributed over 50% of its 2004 revenue
Page 39
and is the most value-added service among all the MVAS, in our view. Unique products. In an industry replete with commodity content, grasp of edgy (and yet ethical) content may put one ahead of the competition. Examples include KongZhongs wireless distribution rights of leading blockbuster movies and TOM Onlines mobile game offering from Indiagame.
Exhibit 26
League Table for MVAS Providers in China

Company Names Sina TOM Online Tencent Linktone KongZhong Hurray! Sohu NetEase Total C-to-P MVAS MVAS Revenue ($MM) 124 113 77 50 48 43 38 17 770 % market share* 16% 15% 10% 7% 6% 6% 5% 2% 100%
What Shall We Bet On - Portals or MVAS Pure Play?
We believe each camp has its own advantages: Relative to MVAS pure plays, portals have lower customer acquisition costs, more cross-selling opportunities between Internet users and MVAS users, alternative non-mobile carrier-dependent distribution channels, and more diversified revenue streams. Players, such as Sina and TOM Online, leverage their portals as self-owned free distribution channels for MVAS products. Given the large traffic on their portals, they could more easily convert Internet users to MVAS users with relatively lower acquisition costs. For instance, we believe portals have advantages in SMS and MMS as they could display and promote these products on their heavily trafficked website. Relative to portals, MVAS pure plays are less distracted by competition in other areas, such as online advertising or Online Gaming. They could thus be more focused on promoting their MVAS products. Interestingly enough, pure plays, such as KongZhong and Linktone, have realized the importance of portals as a distribution platform and are endeavoring to develop their own portals.
Source: Company data, Morgan Stanley Research (Data are based on the 2004 results. *: Refers to the market share in computer-to-person, or C-to-P, MVAS market).
What Models Shall We Focus On ISP or ICP?
We believe MVAS providers may gradually transform into ICPs (Internet content providers) from ISP (Internet service providers) because: Mobile carriers are seizing technical platform and billing practices into their own hands. The partnership between China Mobile and MVAS providers may evolve into a similar relationship between cable operators, such as Comcast, and content aggregators, such as MTV and HBO. MVAS players will likely place increasing focus on content as the key to drive subscriber and sales growth.
Which MVAS Product May Flourish?
First, voice-based MVAS, including IVR and ringback tone (RBT). We project voice-based MVAS revenue growing 68% Y/Y in 2005 and at a CAGR of 22% for next three years.
Exhibit 27
Wireless Messaging Product Mix

Product SMS MMS WAP IVR RBT Description Basic text messanging Picture/Graphic messanging Web browsing (including search) Voice-based access to information and entertainment Tone that callers hear when they dial another person
Page 40
Exhibit 28
Growth Comparison among MVAS Products*

Revenue Annual Sales Growth (For 2004, $MM) (For 2005E) Voice-based MVAS IVR/ RBT SMS 2.5G MMS WAP 86 484 68% 23% Sales CAGR (2006E-2008E) 22% 13%
94 108
53% 51%
29% 32%
A handful of early movers in IVR, including TOM Online, can operate their own IVR technical platforms and typically share around 15% of revenue with China Mobile. However, latecomers are required to use mobile carriers intermediary IVR platform, now offered through Hi-Sun (or Gao Yang in Chinese and a Hong Kong-listed company), and thus pay an extra 15% of revenue to mobile carriers.
E = Morgan Stanley Research estimates Source: Morgan Stanley Research (*: Forecasts do not include MVAS sales collected solely by mobile carriers, such as the person-to-person (P-to-P) MVAS sales, bandwidth and transmission charges).
Exhibit 29
Growth Projection for Voice-based MVAS in China

(CY2004- CY2010E)
350
IVR/ RBT Sales
Voice-based MVAS services, especially RBT, are now the fastest (though still small) growth drivers for mobile carriers. China Mobile now operates its own RBT services and may continue seizing market share from MVAS providers due to its massive distribution platform and bargain power over content providers. In May 2005, China Mobile and MTV inked a deal to provide RBT to China Mobiles 220MM mobile users. China Mobile started to allow prepaid users, a majority of its mobile subscriber base, to access IVR service from 2Q2005, boosting ISPs revenue in IVR.
80%
Growth Rate
Voice-based MVAS Sales (US$MM)
300 250 200
70% 60% % Growth y/y 50% 40%
150 30% 100 50 0

E 04 E E E E 07 06 20 05 08 09 20 20 20 20 20 20 10 E
20% 10% 0%
E = Morgan Stanley Research estimates Source: Morgan Stanley Research (Forecasts do not include bandwidth and transmission charges by mobile carriers.).
Second, data-based MVAS, such as SMS. We project three year revenue CAGR of 17% for SMS. We view SMS as a cash cow business due to its broad user base (around 196MM or 65% of total mobile users by the end of 2004 by our estimates) and slower growth relative to other MVAS.
These services have the following attributes: IVR is a high-margin business as most of the voice-based content is pre-recorded and could be played repeatedly. TOM Online estimates that, before revenue sharing, mobile carriers transmission and information fee amount to 60% of entire IVR/RBT revenue, giving mobile carriers high incentives to promote voicebased MVAS service. Voice-based MVAS service, such as music downloading and pre-recorded audio content, requires less product innovation and thus distribution becomes the key differentiating factor.
Page 41
Exhibit 30
Exhibit 31
Growth Projection for SMS in China

(CY2003- CY2010E)
1200
SMS Sales Growth Rate
Growth Projection for 2.5G MVAS (MMS) in China

(CY2004- CY2010E)
120%
450 400 MMS Sales Growth Rate 50% 350 60%
1000 SMS Sales (US$MM)
100%
MMS Sales (US$MM) 300 250
% Growth y/y
800
80%
40% % Growth y/y
600
60%
30% 200 150 100 20%
400
40%
200
20%
50
10%
0
E E E E E 20 03 20 04 E 20 05 20 06 20 07 20 08 20 09 20 10
0%
0
04 E E 06 E E E 20 05 07 08 09 20 20 20 20 20 20 10 E
0%
E = Morgan Stanley Research estimates Source: Morgan Stanley Research (Forecasts do not include P-to-P SMS sales).
E = Morgan Stanley Research estimates Source: Morgan Stanley Research.
MMS is a low margin business because: MMS is typically based on daily push with ISPs sending out SMS to promote their MMS products, thus incurring high marketing-related expenses. MMS is graphics-rich and it is harder for users to self-generate MMS than SMS. MMS may thus incur higher content development costs than SMS, especially as MMS providers source content from external parties. MMS is more data-intensive than SMS, resulting in higher transmission costs.
SMS could bring several benefits to MVAS players: The cheap price for SMS (typically $0.01 to $0.05 per message) makes it an ideal tool for mass communication. SMS is text-based and incurs fewer bandwidth problems than other services, such as MMS. SMS can produce more user-generated content (UGC) than other MVAS, thus lowering content cost for SMS operators.
Third, 2.5G, including multimedia services such as MMS and WAP. We project a three-year revenue CAGR of 38% for MMS from 2005 onward.
Unlike SMS, which generates four to five times more P-to-P traffic (person-to-person traffic, or the SMS traffic among mobile users) than C-to-P traffic (computer-to-person traffic, or the traffic for downloading SMS from ISPs websites to users mobile phones), we believe it is difficult to send P-toP MMS due to its high bandwidth requirement (thus high failure rate) and multimedia nature. MMS is hence a less lucrative business for mobile carriers as they need to rely mostly on revenue sharing with ISPs on C-to-P traffic. Consequently, mobile carriers may have lower incentives to promote MMS than IVR/ RBT or WAP. That said, portals may have advantages in MMS operation as they have selfowned distribution platforms and could tap into their large Internet user base to lower customer acquisition costs.
Page 42
We view WAP as the most appealing service among all the existing MVAS because: We forecast WAP may expand at three year sales CAGR of 43% from 2005 onward, higher than any other MVAS products.
WAP is based on pull rather than push as WAP users take initiatives to conduct wireless browsing, thus resulting in fewer inactive accounts. (That said, we also acknowledge the existence of inactive WAP accounts, which is the target for mobile carriers new regulatory initiatives.) Unlike SMS and MMS, WAP service does not have platform migration problem as it is already on mobile carriers new billing platform. As a matter of fact, these new billing platforms, such as China Mobiles Misc platform, were initially designed for WAP services.
Exhibit 32
Growth Projection for 2.5G MVAS (WAP) in China

(CY2004- CY2010E)
600
WAP Sales Growth Rate
60%
500
50%
400
40% % Growth y/y
300
30%
200
20%
100
10%
We believe the biggest challenge for WAP providers is to sustain their menu positions on mobile carriers WAP portals, which partially depend on objective metrics, such as traffic, and partially depend on mobile carriers arbitrary assessment, such as service quality and customer satisfaction. As WAP users typically focus on the first page of their mobile screen for WAP search, a decline in ranks (and the slip to the second page of mobile screen) could mean serious defection of eyeballs and material loss in revenue. Last, but not least, we believe it is important for MVAS providers to build a balanced portfolio of MVAS products because: MVAS providers may then seize different growth opportunities. They may offset weakness in one area, such as that caused by regulatory measures, with strength in others. They may have more opportunities to cross-sell their content.
WAP Sales (US$MM)
0
8E 9E 5E 20 04 6E 7E 20 0 20 0 20 0 20 0 20 0 20 1 0E
0%
E = Morgan Stanley Research estimates
WAP is a high-margin business as it incurs minimal content costs. Internet content in HTML language cannot be directly used as WAP content, which requires XML language. WAP providers could thus add more value to and have higher bargaining power over Internet content providers. TOM Online estimates that mobile carriers may get as high as 80% of total WAP revenue, likely the highest among all MVAS products, thus giving mobile carriers high incentives to promote WAP services.
Among the listed MVAS operators, TOM Online appears to have the most balanced MVAS portfolio, with a nearly even sales split between IVR/RBT, SMS, and 2.5G (MMS/WAP).
Page 43
Exhibit 33
Breakdown of MVAS Revenues for Selected Chinese Internet Companies

Company Names Voice-based MVAS (IVR/RBT) Sales ($MM) as % of Total MVAS SMS Sales ($MM) as % of Total MVAS 2.5G (MMS/WAP) Sales ($MM) as % of Total MVAS Total MVAS Sales ($MM) as % of Total MVAS Tom Online 29 26% 55 49% 29 26% 113 100% Sina 5 4% 103 83% 16 13% 124 100% Tencent 6 8% 65 83% 7 8% 77 100% Linktone KongZhong 3 8 5% 16% 40 7.8 80% 16% 8 32.5 15% 68% 50 48 100% 100% Sohu 1 2% 30 79% 7 19% 38 100% Netease 0 1% 14 85% 2.3 14% 17 100% Hurray! 3 7% 14.9 35% 25 59% 43 100%
Source: Company data, Morgan Stanley Research (Data are based on 2004 results.).
Risks - Regulation, Revenue Sharing, and Content Costs
In the near term, we see continual regulatory crackdowns by mobile carriers. Since early 2005, China Mobile required MVAS providers to migrate their MMS to the new billing platform and terminate charges on MMS delivered to but not accepted by customers. Similar to what happened to SMS migration, the whole process of migrating MMS may slow new MMS subscriber growth (due to traffic loss and the difficulty to sign up customers during transition) and increase the churn of existing users (due to easier disconnection). China Mobile also mandated the termination of idle accounts and double confirmation for WAP services. While we support such initiatives due to their long-term benefits to consumers, we believe the transition may cause sales volatility, especially in the 2.5G area. In the intermediate term, it is likely that mobile carriers may raise revenue sharing with MVAS providers. On the headline figures, China Mobiles revenue sharing (15%: 85%) is still lower than what DoCoMo charges for Japanese ISPs. However, note that: Revenue sharing change is more likely applicable to new MVAS products, such as 2.5G, as they are the primary growth drivers. Larger MVAS providers may likely get more favorable terms than their smaller counterparts due
to their revenue contribution to and thus higher bargaining power over mobile carriers. In July 2005, for the first time ever, China Mobile publicly announced that it would stay with the current 15:85 revenue sharing scheme (15% of revenue to China Mobile and 85% to ISPs) but may split the revenue 50:50 if ISPs need marketing services and 30:70 for customer services.
In the long term, MVAS providers may need to secure quality content (See Content Becoming the King) and distribution (with handset manufacturers, for instance) at the expense of margins. That said, we believe large players, especially those with broad content aggregation and large distribution platforms, may increasingly gain cost advantages over their smaller rivals. Judging from the latest developments, we believe mobile carriers may open their own MVAS shops. For instance, China Mobile has been rapidly expanding its in-house ringtone (RB)/RBT service. In May 2005, China Mobile and MTV inked a deal to provide RB/RBT to China Mobiles 220MM mobile users. The company has also developed its own WAP portal. However, in our view, the large MVAS content providers may outrival mobile carriers due to their content expertise, innovative capability, and superior customer services.
Page 44
Why Do We Believe Mobile Carriers May Show Mercy Toward the Large MVAS Players?
First, MVAS contributed 16% of China Mobiles 2004 revenue. We estimate the top five MVAS players accounted for over 50% of the C-to-P (computer-toperson) MVAS sales. Based on the 20/80 rule, it would be more cost-efficient for China Mobile to focus on the top 20% of its MVAS partners who may contribute 80% of its MVAS revenue. Second, mobile carriers already get lions share of MVAS revenue: We estimate P-to-P (person-to-person) MVAS, such as those SMS sent among mobile users, from which mobile carriers keep all the revenue, accounts for 7080% of total MVAS traffic. TOM Online calculates that, before revenue sharing, transmission and information fees charged by mobile carriers total 80% of WAP revenue and 60% of IVR revenue. Simply put, MVAS providers have literally been working for mobile carriers for free most of the time. Third, the new head of China Mobile came from China Unicom in later 2004. As the new leader on the block, he may need to milk the relationship with MVAS players, especially the large ones, to deliver growth. Moreover, China Unicom has historically been more linear toward ISPs than China Mobile. Fourth, 3G license may be granted to other carriers, such as China Netcom and China Telecom in the coming future. This may intensify the competition among telecom carriers, thus forcing them to give more favorable terms to MVAS providers. Fifth, the large MVAS players are gaining bargaining power due to their unique content/service offerings and broad customer bases.
Deployment of 3G phones takes time. We estimate around 9MM Chinese consumers, or 3-4% of total mobile users, may use 3G mobile phones in 2006, up from zero in 2005. 3G phones typically cost 1.5-2 times more than existing mobile phone models. The high cost structure may limit 3G services only to the most well-to-do consumers in China. Content needs to catch up for 3G services to take off. For instance, bandwidth limitation restricts the existing mobile games to single-player games instead of multi-player games, which may be a killer application to sign on 3G subscribers. The battery life for existing models can typically last only half an hour if operated continuously, which is insufficient for users to engage in interaction-intensive 3G services.
That said, we believe mobile games may see a boom in the future as these bottlenecks are alleviated. IDC estimates that mobile game sales in South Korea expanded at a CAGR of 56% over the past two years and may climb 30% Y/Y to $350MM in 2005, nearly equivalent to the entire online game market size in China for 2004. More recently, Nokia, Motorola, and Sony-Ericsson are striving to improve the functionality of mobile phones to suit mobile game operation. Second, we forecast growth in Mobile-commerce or Mcommerce, which utilizes mobile phones as the platform for commercial transactions. We view it a natural evolution for mobile users to use a mobile payment system to purchase merchandises or air tickets. We believe M-commerce may take off in the long run because: There are three times more mobile users than Internet users in China. Mobile carriers may endorse the M-commerce initiative, which offers them another growth leg. The key limiting factors, including security and settlement, may eventually be resolved. For 2004, only 1.7% of Chinese Internet users used mobile phones as payment channels (according to iResearch).
New Opportunities on the Wireless Horizon
First , 3G MVAS 3G services, such as K-Java-enabled mobile games may fire imaginations. However, we believe such services may not be an instant top-line contributor due to the following bottlenecks:
Page 45
The fast deployment of WAP services make it feasible for SMEs to list their merchandises on mobile networks and reach a critical mass.
traffic due to the proliferation of wireless information and the hunger for mobile users to find the right information. Heavy traffic should attract advertisers and merchandisers. Advertisers will go where the eyeballs are and merchandisers will go where the consumers are. Mobile phone screens are more suitable for advertising than PC screens as they can only accommodate one advertisement per screen (thus resulting in a lower degree of distraction by other content) and mobile users cannot skip the advertising pages before reaching their desired sites.
Third, mobile search/mobile advertising is another promising area given that: The fast penetration of WAP leads to the proliferation of wireless websites, and vice versa. China Mobile targets over 90MM WAP users by the end of 2005, up from 50MM in 2004. Similar to Google for Internet, mobile search engines provide a tool to sort out the ocean of wireless data and serves as a gateway to access wireless sites. Mobile search may thus draw heavy
Page 46
Online Advertising - A Race for Eyeballs and Performance

Online Brand Advertising- Secular Uptrend, but Not Performance-Based Yet
Exhibit 34
League Table for Chinas Online Advertisers

Company Names Online Advertising Sales (US$) 65 56 19 8 7 220 % Of Market Share 29.7% 25.3% 8.6% 3.4% 3.0% 100.0%
China will host the 2008 Olympic Games in Beijing. We expect such sports events may be more appealing to online populations, around 70% of which are under the age of 30, than the general public. We factor in a 30% Y/Y growth for online advertising sales for 2008. We project online brand advertising (excluding paid search) to grow 28% Y/Y to $281MM in 2005 and may climb at a CAGR of 25% to $550MM in 2008.
Sina Sohu NetEase TOM Online Tencent Total online advertising
Source: Company data, Morgan Stanley Research (Data are based on FY2004 results).
Exhibit 35
Forecast for Online Brand Advertising Growth in China

(CY2003- CY2010E)
800 1.6
Online Brand Adv Sales Growth Rate
Advertising in China has the following attributes: Chinas advertising market has been the fastest growing in the world with a sales CAGR of 19% over the past 10 years and is the second largest in Asia, second only to that of Japan (according to ZenithOptimedia). ZenithOptimedia estimates that advertising spending accounted for 0.54% of Chinas GDP in 2004, below 1.34% for the US, 0.82% for Japan, and 1.21% for South Korea. Advertising market in China is extremely concentrated with the top four cities contributing roughly 50% of the market share (also see Regional Focus Can Vary). Multinational companies, such as Proter & Gamble and Unilever, were five out of the top ten advertising spenders in China in 2004. We estimate Chinas online advertising sales quadrupled over the past two years, faster than any other advertising subsector, including TV and print media. Online advertising accounted for 1.6% of Chinas advertising spending, below the 3-4% for the US and Japan. Unlike popular practice in the US, Chinas online advertising sales are billed on the basis of cost per period rather than cost per impression (CPM or cost per click through).
Online Advertising Sales (US$MM)
700 600 500 400 300 200 100 0

03 04
1.4 1.2 1 0.8 0.6 0.4 0.2 0
05 E
06 E
07 E
08 E
09 E
20
20
20
20
20
20
20
In the online advertising space, we prefer Internet companies that have: brand recognition edgy content or services large user base targetable demographics and pricing power.
20
10 E
% Growth y/y
Page 47
In our view, online advertising leaders may face challenges due to competition, scaling-back of advertising spending in selected sectors (such as auto and property), and the shift to performance-centric business model. Among the companies under our coverage, we believe that: Tencent may enjoy substantial upside in advertising (despite its current small contribution) given its well recognized QQ brand, its large market share in IM (77% for 2004 in user frequency), its sticky community (easy for targeting), and its higher-than-peers pricing power (50% advertising price increase in 1Q2005 versus 30% for TOM Online and nil for Sina, Sohu, and NetEase). Note that Tencents advertising revenue jumped 65% Q/Q in 2Q2005 to 8% of revenue. Sohu may have multiple growth engines due to its mix of brand advertising and sponsored listing. Note that: 1) Sohu now has five out of the top 50 most-trafficked websites in China. 2) Its 2004 listing revenue was 75% of Baidus whereas its search traffic was only 5% of Baidus. Presently, SoGous search traffic climbs to one third of Baidus. Sina may continue to be the preferred advertising destination due to its content leadership and premium user demographics.
Brand awareness. It is always easier for media buyers to allocate advertising dollars to branded sites than more obscure sites.
We consider online advertising a high-quality revenue model because: In emerging economies, advertising sales have historically outpaced GDP growth by several percentage points per annum and online advertising has typically outperformed the general advertising market. Online media that have reached critical mass may enjoy pricing power. They could raise advertising prices periodically, a practice inconceivable by online game or MVAS operators. Online media could thus expand their top lines without incurring much additional costs, resulting in sustainable margin expansion.
Who Will Be the Long-Term Winner?
We believe three key factors may dictate the long-term success of online advertising: Content. Sinas partnership with over 1,000 news supply as well as Sohus content alliance with the NBA and Disney put them ahead of the competition in brand advertising. Demographics. Sina attracts more upscale white collar professionals than its peers due to its news-driven content nature. (Who has access to the Internet in the morning to check out daily news? Most likely, white-collared professionals.) ACNielsen estimates that around 40% of Sinas users have college degrees or above, versus 31% for general Internet users (according to CNNIC) and less than 5% for the general population. Around 52% of Sinas users have an annual income of Rmb25,000 or above, versus around 19% for general Internet users. On the other hand, NetEase and Tencent are more focused on online community, thus making them a more ideal platform for product advertising. Effectiveness of advertising. Presently, online advertising payment in China is not based on performance, such as click through rate (CTR), CPM, or purchase conversion rate, but instead on the duration of the advertising campaign. We feel that media buyers will inevitably focus on advertising effectiveness, which may be achieved through the following:
Online Advertising High Barriers, High Quality
Online advertising has high entry barriers due to: Regulation. Online content is regulated in China, although to a lesser degree relative to traditional media. Websites, including Google, have been sanctioned for improper content. Heavy regulation thus deters foreign entry into online media/content. Content. Differentiated content in China is still a scare resource. Content advantages for Sina (news) and Sohu (entertainment) are hard to reproduce. Scale. Sina has 30MM unique readers per day for its portal content. Sohu is top-ranked in web traffic in real estate (Focus.cn), online game information (17173.com), and alumni network (ChinaRen), according to Alexa.com. Brand advertisers prefer to be where the eyeballs are.
Page 48
Online verticals, which may help advertisers to target a desirable online community. Examples include Sohus Focus.cn (property portal) and 17173.com (online game) as well as pure-play verticals such as Chemease (chemicals) and AsiaEC.com (IT and stationery supplies). Paid search, which is based on the pull instead of the push model. Search users take initiatives to seek for their desired content/products. As a result, the effectiveness of paid listing, which is tailored toward search users, is typically higher than that of push advertising. Googles outgrowth over Yahoo! in advertising sales illustrates the power of paid listing. In China, leaders in search engines include Baidu, Google, Yahoo!/3721.com, Sohu/SoGou, and Zhongsou. Performance-driven advertising. Allyes, which is the largest online advertising agency in China, now rolls out its affiliated marketing product, through which it sells advertising inventory for other websites. Payment for the service is based on performance, such as CTR. Community-based targeting, such as NetEases email community and Tencents IM community, both of which are the largest in their categories in China. The log-in page and the special interest IM groups, such as music and sports clubs, may offer ways to enhance advertising effectiveness. Narrowad, which matches advertising with the key words in online texts and delivers advertising to readers who are interested in particular key words. For instance, to market a new mobile handset model, narrowad providers could first search for popular online articles with the key word mobile phone, for instance, and then deliver the mobile phone advertisement to the spare space next to these articles. It is conceivable that readers of these articles may have higher interests and thus higher click through rates on the advertisement for new mobile phone models. Tixa.com, formerly known as VeryE, is a leader in narrowad and has received venture capital funding from Sumitomo. Email-directed advertising. Allyes.com estimates that the reach ratio for email-directed advertising was 83% in 2004, up 10 percentage points from
2002. For opened email advertising, the click through rate increased to 21% in 2004 from 17% in 2002. Email advertising could help advertisers to target specific consumer segments based on their personal profile. However, such marketing-related emails are often perceived as spamming and thus filtered out by email operators. Among the above formats, we believe: Paid search offers the best targeting for online advertising and yet the lowest stickiness for users. Verticals provide some kind of targeting with some kind of stickiness. Email service has the lowest targeting but the highest stickiness. (Email users typically do not change their accounts frequently due to high switching costs, such as loss of contact and stored information in old accounts.)
The Tale of Two Formats Online Advertising Versus Traditional Media
Despite its small size, online advertising sales have gained market share over traditional media in the past few years. We estimate the share of online advertising in the total advertising market climbed from 0.7% in 2002 to 1.6% in 2004 and may expand to 2% in 2007. We believe the new and old media both have their own pros and cons: First, relative to traditional media, online media could reach a broader market. For instance, Sinas daily unique readers are more than the combined total of Chinas top 10 newspapers) provide better targeting, and offer more accurate measurements of advertising performance. Second, compared to online media, traditional media, such as print, have competitive edges in local markets because they generally have a better understanding of local culture and local demands. They have better relationships with local government and local sales channels. While the number of Internet users are growing rapidly, they only accounted for 7% of Chinas total population at the end of 2004. The bulk of Chinese population still uses TV and print media as their primary source of information. Moreover, over 70% of Chinas Internet users are under the age of 30. For advertisers who target a more aged population, such as health care and pension fund providers, traditional media may be their preferred choices.
Page 49
Third, competition between online media and traditional media is heating up: Foreign brands, including Procter & Gamble and Colgate, accounted for five out of the top ten advertising spenders in China for 2004. They have also been allocating more marketing budgets to online advertising to target young professionals who have more spending power than the general public. The9.com has secured co-marketing sponsorship from Coca Cola for World of Warcraft. We estimate Coca-Cola may commit around $30MM to $40MM toward marketing expenses, which could otherwise go to traditional advertising channels. Pepsi also plans to team up with Shanda to promote Shandas new games. eBay EachNet and Alibaba/Taobao, by our estimates, may combine for 10-15% of Chinas total online advertising sales. Despite their .com origins, they also compete heavily for advertising slots on traditional media, such as TV entertainment programs produced by Enlight Media, the largest private TV entertainment program producer in China. Sohus stated its 17173.com subsidiary, the leading game portal in China, has not benefited much from the rollout of new online games as game operators allocate a bulk of their marketing budget to the offline advertising format. 51job reported that the sales growth of its online job posting underperformed offline recruiting for 2004. We believe a majority of job seekers may not have regular Internet access, which is still considered an upscale luxury in China, and would thus rely more on newspapers for job information. Focus Media, the leading outdoor advertiser in liquid crystal display (LCD) TV advertising, has advertising presence (mostly exclusive) in over 10,000 commercial locations. Such a massive office network makes Focus a cost-effective platform to target white collar professionals relative to online media. For instance, Ctrip chose Focus as a primary marketing channel to promote
its airline and hotel booking services among the high-end consumers.
Paid Search and Its Unique Story in China
We believe paid search in China may leapfrog traditional online advertising in revenue growth. Googles outperformance of Yahoo! and NHNs fast expansion in paid search suggest to us the power of the online search business model. We estimate Chinas paid search market sized at around $70MM-80MM in 2004. We project paid search sales expanding at a CAGR of 67% for next three years.
Exhibit 36
Projected Paid Search Industry Growth in China

(CY2003- CY2010E)
1000 900 800 120% Paid Search Sales ($MM) 700 % Growth y/y 600 500 400 300 40% 200 100 0
6E 5E 7E 9E 20 0 20 0 8E 20 0 20 0 20 0 20 0 20 1 20 0 0E 3 4
160% Paid Search Sales Growth rate
140%
100%
80%
60%
20%
0%
We believe paid search has several appealing attributes: First, for search users, paid listings provide product information that may be about their best interests. Relative to random online shoppers, search users may be more interested in goods related to their search key words and thus more willing to convert their page view into buying decisions. Second, for paid listing operators, paid search allows them to enjoy network benefits. High search page views will induce more paid listing. More paid listings increase the
Page 50
revenue for search operators, who may allocate more resources to better their search technology and services. As a result, more search users may use the services and so do the paid listing customers. For 2004, Google had 50% of search queries but 70% of the search revenue in the US. Third, paid search offers customers cost-effective means to enhance advertising effectiveness through targeting and pay-for-performance models. Dynamic pricing in bidding listing also provides them with instant insights into the competition and the popularity of their listed products. Last but not least, among all the online businesses, we believe search is the most R&D-intensive. Factors such as the comprehensiveness of search results, relevance, and updating speed are the keys to draw search users. If you have good search technology, the chances are that you perhaps need little marketing. For instance, Google achieved its dominance in the search market with minimal marketing effort. However, paid search operations in China has clear distinctions relative to those in other markets: Search results are regulated in China. Googles search service has been suspended multiple times when the government perceived its search results to be improper. Sohu has been filtering out the improper content on the first pages of the search results for politically sensitive key words. Paid search operators need to deploy a nationwide sales force to solicit paid listing from SMEs, who may have never used the Internet or a PC before. We view such sales networks as a hurdle for foreign players, such as Google, to monetize their search operation in China because they typically operate a self-serve model (in which case SMEs could directly register and bid for search key words online) in the overseas market. MP3 search is among the most popular search service in China. We estimate MP3s account for
roughly 20% of Baidus traffic. And yet foreign competitors, such as Google, cannot operate such service due to copyright concerns. Note that NetEase terminated its MP3 search services in August. The competitive landscape of Chinas search market is intriguing. Note that: Since Baidus IPO, Google has accelerated its pace of expansion in China. It hired Dr. Kai-Fu Lee from Microsoft to head its China operation and struck deals with three distributors within one week to sell its search key words in China. Google plans to commit $700MM capex for 2005 on technology development (vs. Baidus capex of $3MM in 2004A), including universal language translation (namely, translating foreign language search results into the local language). Going forward, we believe Chinese Internet users may care more about the best information instead of the best Chinese information (note that 70% of Chinese Internet users are under the age of 30, with some knowledge of English). Googles advantage in the global search market and its technology strength may help it to gain a foothold in local markets such as China. Sohu had only 5% of Baidus search traffic but 75% of Baidus search revenue in 2004. SoGou, Sohus new search product with unique features including comparison shopping and local map search, saw its traffic climbing to one third of Baidus two quarters after its commercial launch. It is of note that SoGou currently ranks among the top 15 most-trafficked sites among all the web sites in China. Yahoo! China folded its search operation (3721/Yisou) into Alibaba, whose strong nationwide sales network and close rapport with SMEs may help Yahoo! (China) to accelerate the monetization of its search traffic.
Page 51
Zhongsou and Its NetPIG
Zhongsou is an up and coming search provider and a wholly owned subsidiary of HC International, a Hong Kong-listed company and the leader in yellow page listings in China. In May 2005, Zhongsou rolled out its advanced search product, net personal information gate, or NetPIG, with the following features: Like a combination between My Yahoo! and Google, NetPIG provides personalized search service. Once search users choose preferred key words or preferred subjects, NetPIG can automatically search and deliver the desired results to users personal portal periodically. NetPIG can be downloaded to the desktop, thus improving convenience by enabling users to directly conduct searches on their desktops without the need to navigate through search bars. NetPIG has a user-friendly interface that allows search users to operate its function easily.
Industry Risks- Sector Volatility, Competition, Shift to Performance-Centric Advertising Model
In the short term, continual weakness in selected sectors may adversely affect online advertising spending: A price war in the auto segment with 10-20% annual price decline, hurt automakers budgets. Competition and oversupply in the mobile phone sectors made handset manufacturers more pocketshy for marketing campaigns. Macro tightening in the real estate segment is intensifying. With the governments firm resolution to combat the overheated property market and the tightening monetary policy, we believe the property market may be bound for a correction sooner rather than later. In areas such as Shanghai, our economics team estimates the property price versus average per capita income is 4-5X higher than in New York, 3X than in Tokyo, and 2X than in Hong Kong.
Exhibit 37
Unlike existing search services, NetPIG can extend search functions from Internet to Intranet and the hardware of users PCs. NetPIG can automate news searches among 6,000 news sources with the help of only a handful of editorial staff. Relative to other online news suppliers, who heavily rely on editors, NetPIG may be able to provide more comprehensive news in a more timely fashion. Compared to existing search engines, Zhongsous NetPIG, via personalized search, appears to be more customer-centric and thus more sticky. That said, will NetPIG attract only the high-end Internet users (as the service may be too advanced for average Chinese Internet users)? Will Zhongsou be able to ramp up its user base fast enough to compete against the industry leaders (as it needs to distribute the client software for NetPIG)? Will the company be able to monetize on its search business? For a budding company that has innovative products, we are hopeful of its future and eager to find out.
Urban Property Price Comparison

Square Meter Price/Monthly Disposable Per Capita Income 4.8 Shanghai 3.8 Beijing 3.7 Tianjing 2.7 Hong Kong 2.3 Chongqing 1.0 Tokyo 1.0 New York Source: CEIC and Morgan Stanley Note: (1) Average prices for Chinese cities are from CEIC except Shanghai. The price and income refer to the average figures for the first 7 months of 2005 (2) Shanghai's average price is estimated at Rmb 7,000/square meter. (3) Hong Kong's disposable per capita income is estimated at $18,000 per annum. City
Chinas domestic A-share stock market has been hovering around multi-year lows. China Finance, the leading financial information provider, which positions itself as the equivalent of Bloomberg in China, warned that the underperforming stock market may impact its business.
In the medium term, we see intensifying competition for advertising dollars among portals, paid search providers, pure play verticals, and traditional media:
Page 52
There is an oversupply of advertising inventory, especially from the up and coming portals such as TOM Online and Tencent. Sina and Sohu, as the prime online media in China, have more or less exhausted their advertising inventory on their home pages and prime channels, which leads to a dilemma for media buyers: Do you want to advertise on the second-tier channels on Sina and Sohu or on the prime channels on TOM Online, NetEase, and Tencent? Competition for brand advertising is also heating up. Alibaba/Taobao inked exclusive advertising deals with Sohu while eBay EachNet partnered with NetEase and TOM Online. Sina gets none due to the perceived potential conflict of interest from 1pai, Sinas online auction joint venture with Yahoo! China. Competition also forces online media to provide better user experiences. For instance, popup advertising is often perceived as visual pollution by readers. Alexa.com estimates that 45-55% of the sessions for Sina and Sohu contain pop-ups. Cleaning up of the pop-ups may hurt the advertising sales of these portals.
Competition also weakens the pricing power of the online media leaders. Sina, Sohu, and NetEase did not raise advertising price in 1Q2005, while historically they consistently jacked up advertising prices by 10-20% at the beginning of each year. We estimate paid search sales may expand 60% plus Y/Y for next three years. We believe the power of targeting and the pay-for-performance model make paid search more appealing to SMEs than traditional portal advertising. Competition from offline media formats may also erode online advertising budgets (also see The Tale of Two Formats Online Advertising Versus Traditional Media).
In the long term, media buyers cravings for advertising performance may force online media operators to move away from the current pay-per-period model to a pay-forperformance model. Presently, media buyers spend their online advertising budget with little knowledge of advertising performance. We believe such transitions may prove to be painful for the Sinas and the Sohus in China as they may need to compete for performance against paid search and pure play verticals more effectively.
Page 53
Online Gaming - Pipeline and Distribution Fuel Hypergrowth

Online Game Market In China Sizable and Robust
Exhibit 39
League Table For Chinese Online Game Operators

Company Names Shanda Netease The9 Total Online Game Sales Online Game Revenue ($MM) % Market Share 154 39% 72 18% 4 1% 391 100%
There are several trends for Chinas Online Gaming business: The online game industry came into full existence in China only 3-4 year ago but now ranks No. 2 in terms of industry revenue among all online businesses. We estimate the online game market to be $390MM for 2004, larger than online advertising ($220MM) and second only to MVAS ($770 for C-to-P MVAS). We project online game business (37% revenue CAGR over next three years) outgrowing MVAS (high-20%) and online advertising (mid-20%).
Source: Company data, Morgan Stanley Research (Data are based on 2004 results).
In the online game space, we prefer integrated online game operators with core expertise in both game development and game distribution. Our top stock pick in the space is NetEase. By developing games in house, NetEase eliminates licensing fees and could expedite the game upgrading cycle. By controlling distribution channels, the company could minimize channel discounts and respond to customer demand more efficiently, in our view.
Ten Reasons Why Online Gaming Should Do Well in China
Exhibit 38
Growth Forecast for Online Game Industry in China

(CY2003- CY20010E)
2000 1800 Online Game Sales (US$MM) 1600 1400 1200 1000 800 600 400 200 0
20 05 E 20 06 E 20 07 E 20 08 E 20 09 E 20 10 E 20 03 20 04
Online Game Sales Growth Rate
1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 % Growth y/y
For 2004, we estimate Chinas online game revenue expanded 90% Y/Y, substantially higher than the 78% for online advertising and the 12% for SMS (we left out the Y/Y comparison for 2.5G as they did not exist until late2003). We believe there are several factors contributing to the strong growth: 1) Online gaming helps to eliminate piracy issues, which, in effect, have kept the development of PC games and console games hostage in China. Online gamers must download the client software of their favorite games and yet have no access to the source code, which are typically stored on game operators servers. 2) Online games operate on a service-model (customers get billed on hourly or monthly basis for services) while PC games and console games operate in a retail-model (customers purchase game copies or game consoles). The revenue streams for Online Gaming are thus recurring while sales for PC games and console games are often one-off. 3) Unlike MVAS providers, online game operators control the last mile or the billing relationship with customers, which enables them to better service customers and minimize margin squeeze by distribution channels.
Chinas online game market is highly concentrated with top 2 players, Shanda and NetEase, contributing 58% of the entire online game sales in 2004.
Page 54
4) As online gamers prepay for un-rendered services, game operators thus has negative working capital requirement and substantial financial leverage. 5) We believe online gaming fills in entertainment blanks in China, where quality entertainment is still a scarcity, especially in the second-tier areas. Shanda and NetEase both reported that the bulk of their revenues come from non-top-tier regions. 6) Online games, especially the MMORPGs, help to engage group interaction among young people in China, who are typically the only children of their families and hungry for social interactions. 7) Online gaming is inexpensive even by Chinese standards. Online games typically cost: around $0.04-$0.05 per hour for playing time plus $0.3-$0.7 per hour for Internet cafe expenses or $0.04-$0.05 per hour for playing time plus $15 to $20 monthly Internet access fees at home, which are usually subsidized by parents who use the Internet to access online information. For two-hour entertainment, we estimate it costs around $1 to play games at Internet cafes versus $5-10 for movies and $10-15 for Karaoke.
IDC estimates that desktop PCs shipped for home usage accounted for 39% of the total 2004 shipment volume, more than those to any other sector. iResearch estimates that 62% of gamers played online games at home versus 29% at Internet cafes for 2004.
We believe such migration bodes well for the development of casual games (more suitable for family entertainment than MMORPGs) and 3D games (as PCs at home have better-equipped hardware, such as 3D display cards, than those at Internet cafes). This explains why Shanda has shifted its focus to home-based users and why The9 comarkets with Coca-Cola to focus on traditional consumers.
Barriers for Online Gaming The Most Well-Rounded Thrive
Online game industry has high entry barriers due to: Game development capabilities. To date, only selected companies, such as NetEase and Kingsoft, have self-developed games with hit potential. Most other game operators license their games, thus resulting in high royalty fees, slow upgrades due to lack of access to source code, and sometimes even litigation. Distribution network. Shanda has over 300,000 retail sales points, which helped to convert Mir 2, which was perceived as a non-top-tier game in South Korea, into the most profitable online game in China for 2004. Local content and cultural barriers. Thus far, no foreign game with western content has achieved financial success in China. In 2004, the top 10 most popular games in China are all of Asian origins. Most recently, domestic games, such as NetEases Westward Journey and Kingsofts FS Online, are gaining increasing popularity, which speaks for the importance of local game content (Westward Journey and FS online are based on two of the most famous Chinese classical novels, Westward Journey and Fengshen Bang). Regulatory restrictions. Presently, no foreign company is allowed to operate online games in China. That is why foreign game companies, such as Universal / Blizzard, NCSoft, and Webzen, need to collaborate with domestic partners for game operation.
8) According to IDC estimates, for every $1 in online game revenue in 2003, the revenue multiplier (or the revenue that other players in the industry foodchain may generate) was around 11X (telecommunication industry - 6X, IT industry3X, and media and publishing- 2X). Such magnification effects provided compelling incentives for business partners to rally behind online game operators. 9) By our estimates, broadband penetration in China is still low (3% of total population for 2004) but may trend upward at a CAGR of 37% for the next three years. In other more advanced countries, such as South Korea, broadband penetration is a leading indicator for online game growth. 10) We believe there is a significant migration of online gamers from Internet cafes to homes: According to a CNNIC survey in January 2005, 68% of Chinas Internet users choose home as their preferred place for Internet access versus 41% for working places and 25% for Internet cafes.
Page 55
Exhibit 40
Screen Shots from NetEases MMORPGs
Source: NetEase.
Game Community Sticky, Self-enhancing, and Scalable
First, online gaming is perhaps stickier than all other online businesses because: Online gamers can often get addicted to game content. Online game is community-based and can enjoy networking effects. Gamers tend to go where their online buddies are.
Movie hits have short life cycles. While cinema operators rarely show a blockbuster movie for more than a year, online game providers can operate the same game for 3-5 years, such as Lineage in South Korea, Ultima Online in the US, and Legend of Mir 2 in China. The popularity of movies usually declines over time while popular online games can gain popularity. For instance, the average concurrent users for Fantasy Westward Journey have grown by an average of 54% per quarter over the past five quarters. Would you expect more viewers for last years blockbuster movie this year? Online gaming revenue is based on playing time and subscription, thus recurring. Box office sales are typically one-off with little to do with the length of the movies. Chinese people tend to love Hollywood but none of the Western online games ranked within the top 10 in China in 2004.
Second, like the movie business, online gaming is hit-driven. However, there are clear distinctions between movie and online game industries: Movies are finished products once released while online games can be works in progress, which can be upgraded periodically. Movie goers passively accept the content while online gamers actively participate in the content. Interactivity is typically disallowed in cinemas but is part of the draw for online games. While you meet online friends all the time while playing online games, you typically do not make new friends while watching movies. You adore stars in movies while you will be adored if you become a star in the virtual game community. While you hold back your ego in cinema, you can unleash or even inflate it in the virtual world.
Third, the online game business is scalable due to its low capex and generally fixed-cost structure. For instance, Shandas operating margin increased by an average of 500 basis points per quarter during 2004.
Page 56
Exhibit 41
Screen Shots from Shandas MMORPGs
Source: Shanda.
Online Game Products Four Pairs of Comparison
1) Domestic games versus foreign games: No online game with Western culture content has ever achieved much success in China. Among the top 10 most popular online games for 2004 as surveyed by The News and Publishing Bureau, none of them is based on Western culture content.
were developed domestically. The domestic games have gained traction through improvement in game design, storylines, and technology. In 2Q2005, Fantasy Westward Journey recorded over 700, 000 PCUs and became the No. 1 mostplayed MMORPG in China. FS Online, Kingsofts new game on top of JX Online, became a potential blockbuster by reporting 130,000 PCUs during its beta-testing in early 2005. Game operators need to pay upfront licensing fee plus revenue sharing (typically between 20-30%) to foreign licensed games while they could keep all the revenue for self-developed domestic games. Foreign game providers are restricted by the government from operating online games in China and thus need domestic partners to distribute their games. It takes substantially longer periods of time for foreign games to pass government censorship than domestic games. For instance, it took six months for NC-Sinas Lineage 2 to get government approval but only a couple of weeks for NetEases Fantasy Westward Journey. It is easier to prevent hacking and cheating issues for domestic games than foreign-licensed games due to game operators access to source codes.
Exhibit 42
Top 10 Most Popular Online Games In China (2004)

Rank 1 2 3 4 5 6 7 8 9 10 Name of The Online Games Lineage 2 Mir 2 Mir 3 World of Legend Fantasy Westward Journey Mu JX Online Magic Baby Westward Journey Ragnarok Online (RO) Developers Operators NCsoft NC-Sina Actoz Shanda Actoz Optisp Shanda Shanda NetEase NetEase Universal/Blizzard The9 Kingsoft Kingsoft Square Enix Joypark NetEase NetEase Gravity Gameflier
Source: The News & Publishing Bureau of China.
We believe domestic games may gain increasing popularity primarily due to cultural acceptance. For instance, Chinas TV drama industry used to be dominated by TV dramas produced in Japan, Taiwan, and Hong Kong a decade ago. However, domestic productions have improved substantially and now rake in the bulk of the top ratings. For 2004, four out of the top 10 most popular games
Page 57
Exhibit 43
Screen Shots from 3D Online Games Lineage II
Source: NC-Sina.
2) 2D games versus 3D games: 3D games are graphically richer than 2D games, thus more appealing to fun-seeking online gamers. Like what happened in the movie industry, once people get used to color movies, it will be hard to lure them back to the black-and-white ones. In South Korea, which is arguably the most advanced nation in Online Gaming, most popular online games are typically three-dimensional. 3D games have higher hardware requirement, such as 3D graphics card. IDC estimates the average selling price for a desktop PC in China was only $570 in 2004. We estimate that a 3D display card may cost around $150-200, 25-35% of the average PC cost or the equivalent of the average monthly salary per capita in top cities, such as Shanghai or
Beijing. The9 and NetEase stated that three quarters of Internet cafes could support World of Warcraft and Fly for Fun using their existing hardware. NetEase also targets to cover over 50% Internet cafes with its new 3D game, which may be launched in 2H2005. We believe the increasing PC shipment to home users may bode well for 3D games as home PCs are typically better equipped than those at Internet cafes. 3D games usually cost more to play than 2D games. For instance, World of Warcraft and Lineage 2 cost around $0.06 per hour to play versus $0.04-0.05 for other popular 2D games. Since Online gamers, a bulk of who live in second-tier and third-tier cities, are typically price-sensitive. Higher price may deter their enthusiasm for new games.
Page 58
Exhibit 44
Screen Shots from Tencents Casual Games
Source: Tencent.
3) Casual games versus MMORPGs: The success of MMORPG is more product-driven with great game content winning in the long run. Casual game is more distribution-driven as people have been playing similar games, such as Mahjong and pokes, for infinity. Often, it is where you are rather than what you play that drives casual game growth. MMORPG is interactivity-centric, with hundreds of thousands people playing the same game simultaneously. Casual gaming is entertainmentcentric as only less than a handful of people play the same game against each other at the same time (people could also choose to be lone rangers to compete against themselves). Gamers play MMORPG for hours (according to IDC, 87% of Chinese gamers play more than three hours per day) while people typically spend several minutes to half hour on casual games.
Players for MMORPGs pay for playing time or subscription. Casual games do not charge for playing time but instead for game items, such as avatars, cyber cash, and virtual weapons. We believe pure plays, such as Shanda, may have advantages in MMORPGs as they are more focused on resources, product development, and distribution. Portals may outperform in casual games due to demographic similarity between portal users and casual game population and to self-owned free distribution channels. For instance, Tencent surpassed Ourgame.com, the former pureplay leader in casual games, in concurrent users (1.4MM for Tencent in 2Q2005 versus around 0.6MM for Ourgames.com) in four quarters. Sina partnered with NetMarble in 2004 to launch the casual game portal. Sohu plans to debut its casual game operation through its 17173.com portal and NetEase may roll out 9 casual games altogether for 2005.
Page 59
Exhibit 45
Screen Shots from Mobile Games
Source: TOM Online/Indiagames.
4) Mobile games versus online games: China is more mobile-centric than PC-centric; there were 3.6 times more mobile users than Internet users in China at the end of 2004 (see China Internet- Mobile-centric Rather Than PCcentric). Our Morgan Stanley Technology team forecasts that there will be around 9MM 3G phones, or around 4% of total mobile phones, in China in 2006. IDC estimates that the South Korean mobile game market has expanded at a CAGR of 56% since 2002 to $350MM in 2005. Mobile users may play interactive mobile games on a continuous basis, leading to high transmission fees for mobile carriers, who may thus be more incentivized to promote mobile games than other MVAS. Unlike online games, mobile gamers may have short playing time (likely less than half hour due to short battery life) and lack of interactive features due to bandwidth restrictions (literally all the existing mobile games are single-player games instead of multi-player games, which are the killer application for online games).
Risks Competition, Pipeline, and Regulation
In the short term, The9s World of Warcraft may cannibalize the market share of existing online games because: World of Warcraft has achieved success in the US and South Korea. It has surpassed Everquest to be the most popular MMORPG in the US with 200,000 peak concurrent users (PCUs). According to the company press, World of Warcraft is the most popular game ever during beta-testing phase in South Korea. World of Warcraft has seen phenomenal growth in many countries because: 1) Its PC game version has already prempted the gamer market; 2) It is easy for casual players to understand and get involved; 3) It also offers enough sophistications to engage serious gamers, who may 3+ hours per day playing the game. For World of Warcraft, The9 committed to pay a minimum of $51MM licensing fee against a 22% revenue sharing on the face value of prepaid cards and prepaid online points sold. By comparison, we estimate Westward Journey, NetEases leading game in 2004, netted around $47MM in sales. With such tight revenue-sharing commitment, we believe The9 may do anything possible to seize monetization opportunities. The9 has committed $13MM minimal marketing expenses over the licensing period to promote World of Warcraft. By comparison, Shanda spent $11MM on sales and marketing for its entire game portfolio in 2004.
Page 60
The9 co-markets World of Warcraft with Coca Cola. Coca-Cola now places the icons of the virtual characters of World of Warcraft on iCoke bottles. We believe the two parties have marketing synergies as the demographics for online gamers and for Coke drinkers are quite similar. Similarly, Shanda plans to comarket its new MMORPGs with Pepsi Coca. World of Warcraft places more emphasis on team work and less on violence, such as PK (people killing), thus likely receiving more positive press coverage and less scrutiny from parents. Two months after its commercial launch, World of Warcraft already recorded 1.5MM registered paid users, making it the fastest growing online game in China ever.
distributors. iResearch estimates that revenue share for game distribution channels declined to 17% in 2004, down from 35% in 2001. In the long term, we view cheating and private servers as credible threats to online game business in China: According to an IDC survey, 53% of Chinas online gamers used plug-ins or cheating tools and 30% used illegal game servers in 2003. The9s revenue to Webzen from MU dropped 47% Q/Q in 1Q2004 due to cannibalization by illegal game servers. Shanda stated that it has been cleaning up BOTS since 4Q2004. These automatic cheating programs can play games and accumulate points nonstop, which could occupy excessive server capacity, lower the ARPU for online gamers (Shandas online game revenue is mostly subscription-based), and hurt the morale of the online community due to unfair competition. Sina stated that the ACUs for Lineage 2 climbed 43% Q/Q to 93,000 in 1Q2005, largely due to the release of new expansion packs winning gamers over from private servers. However, the company expected that private servers may crawl back to erode its customer base shortly.
We believe the current market leader in online game sales may face challenges on multiple fronts: For MMORPGs, NetEases leading game, Fantasy Westward Journey, recorded over 700,000 PCUs in 2Q2005 and exceeded Shandas Mir 2 to be the No. 1 online game in China. Other than NetEase, Kingsoft has also demonstrated respectable selfdevelopment capability. Its two in-house games, FS Online and JX Online, all ranked within top 10 most popular games as surveyed by iResearch in March 2005. For casual games, Tencent recorded 1.4MM PCUs in 2Q2005 and ranks No. 1 in the category only around 1.5 years after the launch of its game portal. The9s World of Warcraft may adversely impact Shandas game sales as both are located in the same cities and likely overlap in sales, marketing, and distribution channels, which may result in head-to-head competition.
We believe governmental regulations may come into place as there is increasing public outcry for saving kids from spending excessive amounts of time playing online games. We consider three to five hours game playing time per day a serious time commitment, which may lead to adverse effects on school children. Constructively, we believe the following approaches may help alleviate the issue: Learning from Koreas experiences, Chinese government could implement a rating system on games to segment game population, thus preventing underaged kids from getting addicted to online games. Game operators should provide more healthy content. For instance, Kingsofts new game, Fantasy Spring and Fall (or Fantasy Chun Qiu), may embed 365 Chinese classical stories into the
In the medium term, we are concerned that pipelines for domestic game operators are still thin. Only NetEase and Kingsoft have produced two self-developed games with PCUs over 100,000. Most online game companies still rely heavily on the licensed foreign games. Foreign game developers increasingly demand high revenue sharing, such as the case for The9s licensing contract with Blizzard. On the other hand, we observe deteriorating power of
Page 61
game design, making the game more educational and more informative than the current PK games. Game operators could make games more familyoriented. For instance, Linktone licensed Mr. Hammer, one of the most popular casual games in Korea. The game has a limited three-minute playing time, thus having less impact on players time. Since it combines entertainment, combat and strategy, families could thus play together.
In August 2005, the Chinese Ministries of Information Industry and Culture released a series of regulations on online game operation, including: implementation of technical measures to restrict excess playing time. prohibition of under-aged people from playing PK games. use of real names for online gamers.
Page 62
Online Commerce Solid Outlook

We believe Internet commerce should proliferate in China because: Chinese people are natural-born entrepreneurs. There are around 25-30MM SMEs in China, or roughly 1 SME for every 20 working-aged people. Internet services, such as paid search and online auction, offer innovative means for SMEs to mushroom. Chinese people love community. Especially for the younger generation, who are typically the only children of the family, the Internet allows them to fulfill two basic needs at the same time make friends and make a living online. China is fragmented in terms of geography and ethnicity. There are 56 ethnic groups in China with different cultures and different dialects, albeit unified by the teaching of Mandarin as the official language in government schools. The Internet commerce helps to eliminate information barriers and facilitate trades among different groups of people. Chinese people tend to be bargain hunters. We view this a positive as we ourselves are also bargain hunters in the investment field. The online marketplace, such as online traveling services, provides a competitive pricing environment. China and has developed diversified payment channels with customers. In this section, we highlight some online operations, including IM, sponsored search, online auction, online traveling, online recruiting, online shopping and online blogging, which in aggregate may help to expand Chinas online commerce market. Our favorite stock picks in the space include Ctrip (online traveling) and Tencent (IM).
IM A Building Block for Internet Value-Added Service
We estimate that there were around 176MM active IM accounts in China by the end of 2004. Assuming that one IM user has an average of 3 IM accounts, we estimate that IM users stood at around 59MM, or 62% of Chinas total Internet population in 2004, and may expand at 15% CAGR to 100MM in 2008. Note that: iResearch estimates Tencent dominated the IM market with 77% of share in use frequency in 2004, up from 74% in 2003. Tencent reported 135MM active accounts for its QQ IM services, representing the largest online community in China.
Exhibit 46
Forecast For IM User Growth in China

(CY2003- CY20010E)
120 80%
Average IM Users Growth Rate
Average IM Users (MM)
We believe that 2005 may turn out to be a watershed year for e-commerce providers to develop payment, credit and logistics systems before ecommerce break out in China in 2006. Our top stock picks in the space include: Ctrip, which combines online technical expertise with offline customer services. Thanks to its click plus brick model, we believe Ctrip is more efficient than the brick-and-mortar players due to its online platform and superior to online rivals due to its offline distribution. Tencent, which may evolve into one of the most dynamic online marketplaces as it has the largest and one of the most sticky online communities in
70% 60% 50% % Growth y/y
100
80
60
40% 30%
40 20% 20 10% 0
20 03 20 04 20 05 E 20 06 E 20 07 E 20 08 E 20 09 E 20 10 E
0%
Page 63
We believe IM services have several unique attributes: 1) Entry barriers for IM services are high because: There is no interoperability among different IM communities as they are surrounded by layers of firewalls. Competitors can find themselves blocked from entry into other IM communities. Basic IM services are free, thus preventing price competition. IM community is sticky. IM users typically communicate among online buddies and enjoy networking effects. Once you are in the IM community, it is challenging for rivals to lure you away from your online buddies.
communication tool (QQ is perceived more as a tool for entertainment and online connectivity). In May 2005, Microsoft launched MSN China, which integrates information, communication, and content services. That said, we believe most IM users may choose both MSN (for more professional usage) and QQ services (for its larger community and entertainment). We believe Skype, a voice-based IM services, is emerging as a credible contestant. Skype recorded 51MM subscribers worldwide in August and constitutes the fastest-growing online community ever. In China, Skype formed its only joint venture in the world with TOM Online to expand Skype services in China. To date, China already ranks within top three around the global in new Skype user growth.
2) While basic IM services are free, IM operators monetize value-added services, such as avatars (virtual characters one could choose as his/her online identify), premium IM clubs (such as online dating clubs), and MVAS (such as forwarding IM to mobile phones). 3) Scale is critical in monetizing IM services. We assess that only 6% of Tencents active accounts were paying accounts for 2004. We believe small IM operators will have a hard time surviving as they may not get enough valueadded service customers to subsidize their free basic IM services. 4) Unlike portal businesses, IM services focus on interaction and communication. A majority of IM content is generated by IM users, thus lowering the content costs for IM operators. 5) IM operators offer alternative payment mechanisms independent of mobile carriers. For instance, Tencent reported close to 45% of its revenues were collected from non-mobile channels, such as prepaid cards, e-sales, and credit card sales. Despite Tencents dominance, competition is intensifying: Relative to Tencent, MSN is more competitive in the high-end market, such as white collar young professionals, due to its natural advantages as an embedded feature of Microsoft Window, its better protection for security and privacy, and its perceived image as a more professional
We believe IM communities are a natural fit for Internet value-added services (IVAS), such as online dating clubs and avatars, as they are sticky (high customer retention), fragmented (easy to target), and youth-oriented (thus higher possibility for new IVAS). In the long run, similar to what happened in the US, we believe a few winners may split the IM market.
Paid Search Core of Online Business for SMEs
We believe paid search is revolutionizing the business models for Chinese SMEs: We estimate there are around 25-30MM SMEs in China, which account for 99% of total registered companies and yet are under-serviced for online business. In fact, most SMEs have never used the Internet before. SMEs are constrained in their marketing budgets and thus need more effective marketing channels. Relative to the more expensive online banner advertising, sponsored search offers SMEs a more cost efficient way to promote their products. SMEs are deficient in information about their market and customers. It is thus difficult for SMEs to target their customers using the traditional push approach. Instead, sponsored search helps SMEs to pull customers.
Page 64
SMEs typically focus on selected local markets. Sponsored search services allow them to identify business opportunities in remote markets. SMEs have limited resources. Sponsored search connects SMEs with high potential customers, who take incentives to search for their services/products, thus improving success rate and labor efficiency for SMEs.
that Taobaos GMV expanded to $200MM, up % 67% Q/Q, whereas eBay EachNet did not disclose further details on its GMV. We believe there are lots of lessons to be learned from the contest between eBay EachNet and Alibaba/Taobao (also see the section in this report titled Why Are Foreign Players especially Challenged in China? and our case study on The Battle between eBay EachNet and Alibaba/ Taobao). In our view, such intense competition may accelerate the expansion of the entire ecommerce market in China and bring the following benefits to Chinese Internet users: Customer education about online commerce. eBay plans to commit up to $100MM in China for 2005 (By comparison, we calculated the 2004 total costs, or cost of sales plus operating costs, were only $95MM for Shanda, the No. 1 Internet company in China in market capitalization in 2004 and $131MM or Sina, the No. 1 Internet company in China in revenue in 2004). We believe that part of the cost may go into marketing campaigns and customer education. In our view, the combined marketing campaign by Taobao and eBay EachNet may substantially raise customer awareness of online commerce and convert more Chinese offline shoppers into online shoppers. Benefits to online sellers. P2P online retailing offers Chinese SMEs alternative sales channels. Presently, Taobao offers listing services for free, thus incentivizing merchant sellers, most of them still early adopters of online auctions, to try this alternative sales channel. We estimate Taobao has around 10MM current listings, No. 1 in China, versus 5MM for eBay EachNet in CQ2, 1MM for 1pai, and 187MM for eBay (US). It is noteworthy that Taobao only had 20,000 listings two years ago and around 3MM half year ago. That said, it is important to note that Taobao listings have longer expiry date (Typically, sellers can choose between 7-14 day listing periods; at the expiry of the period, the listing is automatically extended for one more period, after which it is taken off), whereas a given eBay EachNet listing expires in 7-10 days. Furthermore, eBay EachNets conversion rate is running at roughly 7 days, so listings turn over much faster than with Taobao. We believe this improves the vibrancyof eBay EachNet, making the inventory for sale on eBay EachNet fresher than that of Taobao, in our view. We believe that Taobao may eventually convert to the eBay
We believe sponsored search operators have competitive edges in China as they are more customer-centric than traditional media, which are mostly state-owned, and more cost-efficient and more targeted than online brand advertising. However, it is important to note the significant differences between search services in China and those in other markets, especially in areas such as ground sales force requirement, content regulation, and the range of service offerings (please see Paid Search and Its Unique Story in China in this report).
Online P2P Retail The Battle between eBay and Alibaba May Expand the Whole Market
We believe Chinas P2P online market has significant growth potential: According to iResearch, the gross merchandise value (GMV) for online auctions in China totaled around $410MM for 2004 and may expand at a CAGR of 84% for next three years. Note that iResearchs definition broadly includes person-toperson (P2P) trading sites that have fixed price listings, so that sites like Taobao are included, but sites like Joyo or DangDang that do not have multiple sellers are excluded. There are around 12MM registered users for these sites, with one third of them having completed successful online auction transactions during 2004 (iResearch).
Competition in the field is intense eBay EachNet was the online auction leader in China in 2004 with $307MM of GMV or around 70% of the market share in terms of GMV in 2004. However, Taobao has potentially surpassed eBay EachNet in headline GMV. The company reported its GMV exceeded Rmb1B ($120MM) for 1Q2005, versus $100MM for eBay EachNet and around $10-30MM for 1pai, the online auction joint venture between Sina and Yahoo! (1pai recorded around Rmb3MM per day in April 2005). Note
Page 65
EachNet format. Also, on May 1, eBay EachNet lowered its fees across the board. The listing fees were reduced by 2550%. Since the price change, more sellers have listed their goods on eBay EachNet. Benefits to online buyers. In our view, P2P online retailing offers Chinese consumers opportunities for: 1) bargain hunting as Chinas living costs vary substantially from region to region, online customers may take advantage of the price gaps in different regions more efficiently than what they could achieve offline; 2) social interaction as a lot of the P2P sites are community-based, users, especially those who are the only children of the families, could thus make friends while doing online shopping; and 3) fun-seeking wouldnt it be fun to fish for your favorite items from the 10MM listed goods on Taobao to make you look really cool? You bet. Payment systems. In our view, the push by Alibaba / Taobao and eBay EachNet for online payment systems may help to resolve one of the key bottlenecks for Chinas online commerce development. Alibaba launched AliPay in late 2003 and over 90% of Taobaos customers have adopted the system. AliPay serves as an independent third party between online buyers and sellers. It holds the cash from buyers in escrow accounts and releases it to sellers only after buyers confirm the settlement of their orders. And eBay EachNet also has an escrow-based payment system An Fu Tong which has been in place since October 2004. In our view, such escrow systems help to minimize the settlement risks in China, which still lacks a credit system to date. On the other hand, eBay debuted its PayPal China system in July. PayPal is the global leader in online payment with over 70MM existing users. As a foreign entrant into Chinas heavily regulated financial services industry, PayPal may face resistance from local policy makers. But eBay was proactive in partnering PayPal with leading domestic banks, including Industry and Commerce Bank of China, the No. 1 state-owned commercial bank in China, and China Merchant Bank,
the leading online bank in China. PayPal China offers payments in the local currency and integration with 15 local banks and more than 20 different debit cards, and PayPal China also offers buyer protection off eBay, such as with the popular portals, NetEase and TOM Online. More dynamic marketplace and more service choices. In August, Yahoo! invested $1B in cash plus the contribution of its Yahoo! China business to become the largest shareholder of Alibaba, with a 40% stake and 35% voting rights. We believe the combined entity, which was valued at $4.3B at the time of transaction, may lift Yahoo! Chinas operation given the cross-selling synergy among Alibaba, Taobao, 3721, Yisou, and 1Pai (as they all mostly target SMEs), AliPays payment channels, and the managerial expertise of Jack Ma and his crew. Moreover, other players are also trying to seize the massive opportunities in P2P online retailing. For instance, Tencent plans to launch its own online auction in 2H2005. We believe Tencent has advantages in P2P market due to its large and sticky IM community, demographic match for online auction, communityoriented nature (thus easy for targeting), and alternative non-mobile dependent payment system (45% of 2Q2005 total sales). We believe these elevated levels of market activities may benefit Chinese customers in the long run by offering them more online service and product options. Cross-border trading opportunities. In our view, despite the current government restrictions on Rmb leaving China, eBays PayPal may offer a potential trading platform between the 100MM Chinese Internet users and the 70MM PayPal users, which may help global customers to benefit from the low-priced goods produced in China and Chinese SMEs to venture into international marketplace. After its alliance with Alibaba, Yahoo! may help Taobao and Alibaba expand their overseas markets (Note that Yahoo! has outcompeted eBay in Japan and Taiwan).
Page 66
Exhibit 47
China P2P Online Retail Competitive Landscape

(US$MM, Metrics in Thousands)
GMV, 1Q2005E Listings Listings Expiry? Listing Fees? Listings Format Top 5 Selling Categories (Listings)
eBay EachNet $100 1,000+ Yes Yes 40% auctions, 20% combined auction / fixed price, and 40% fixed price ------
Alibaba Alibaba -580 Yes Yes Buyer Inquiry / Seller Negotiation Home Supplies (58) Industrial Supplies (51) Apparel & Fashion (48) Electronics & Electrical (42) Construction & Real Estate (37)
Taobao $120 10,000+ Yes No > 10% auctions ------
Source: Alibaba/Taobao, eBay EachNet, Morgan Stanley Research.
Page 67
Online Travel and Online Recruiting The Incarnation of Click Plus Brick
Exhibit 48
Projections for Online Travel Industry in China

(CY2003-CY2010E)
400 120%
Online Traveling Sales Growth Rate
Chinas online travel market presents compelling growth potential: Forrester Research estimates the global online travel market at $53B in 2004 and may grow at a sales CAGR of 18% for next three years. The China National Tourism Administration estimates Chinese tourism sales climbed 37% Y/Y to $57B in 2004, accounting for 4% of Chinas 2004 GDP. In 2004, there were roughly 1.1B people traveling (including double counts of people traveling multiple times during the year) in China with average spending per traveler per visit approximating a modest $50, implying substantial growth upside. We assess that frequent flyers, who fly once a month, account for only 1% of total population. The World Tourism Organization estimates 10% of traveling transactions will be online in China in the next few years and we estimate less than 5% of people now transact online. For 1H2005, Ctrip, the market leader in online travel services, recorded that its hotel booking business (68% of its sales) expanded 32% Y/Y and air ticket booking (28% of its sales) climbed 1.8x Y/Y. We forecast online travel revenue grew 96% Y/Y to $62MM in 2004 and may climb at three-year sales CAGR of 53% to $222MM in 2008.
350 Online Traveling Sales (US$MM) 300
100%
200 150
60%
40% 100 20% 50 0

20 03 20 04 20 05 E 20 06 E 20 07 E 20 08 E 20 09 E 20 10 E
0%
E = Morgan Stanley Research estimates Source: Company data, Morgan Stanley Research
An attractive market also attracts competition. The Xinhua News Agency estimates there are over 5,000 Chinese websites offering traveling information and 300 of them specializing in traveling/ tourism. Ctrip is currently the leader in online traveling, with eLong as the second. We believe online traveling companies could build entry barriers through: Exclusive arrangement with suppliers for hotel rooms and air tickets such as guaranteed hotel room allotment. If you have guaranteed room allotment, customers will come to you because you always have what they want. Otherwise, you may have a hard time meeting customer demand during the high season and they may not come back again after several trials. Ctrip reports that one third of its contracted hotels offer guaranteed room allotment, which contributes two thirds of its hotel room sales. Customer base which could generate a networking effect. More customers would induce more business allocation from hotels and airlines, which in return boosts customer transactions.
% Growth y/y
80% 250
Page 68
Online Recruiting Sales (US$MM)
Scale such as total number of hotels and total number of airline routes covered. For instance, Ctrip has a national coverage of over 3,000 hotels in China and ticket capabilities in roughly 40 cities. Local distribution air tickets and hotel rooms are generally commodities (except for peak seasons) and distribution thus becomes one of the key differentiating factors. Whoever controls the last mile or the delivery/ billing relationship with customers, will gain advantages in the long run. Optimized customer services at the end of the day, it is the customer service that makes all the difference, especially in a service-intensive business like online traveling. Ctrip has installed six sigma standard to stress on the consistence and the quality of customer services. As a result, around 60% of its customers are repeating customers and the company adds an average of one new customer per minute (based on 4Q2004 result).
400 300 200 100 0

20 04 20 05 E 20 06 E 20 07 E 20 08 E 20 09 E 20 10 E 20 03
40% 30% 20% 10% 0%
Net-Temps reported that 78% of companies in the US hired people through online channels in 2004. By comparison, iResearch estimates online recruiting only accounted for 13% of the entire 2004 recruiting service revenue in China. In 2004, there were around 64,000 online job openings for information technology (IT), nearly 2.5x more than the No. 2-ranked manufacturing industry. We believe IT industry is among the fastest growing industry in China and is tailormade for online recruiting because IT professionals are typically more adaptable to online environment than people in other sectors. News Evening reported that 26 online recruiting companies hosted a united online recruiting in Shanghai for two months in 2004, which attracted 1,400x more visitors (online) per job opening than what all the 555 job fairs in Shanghai achieved offline from February to September 2004. We assess that 51job generated around 30-40% of its sales from Shanghai and Beijing; two cities that have the highest numbers of Internet users in China.
We believe online recruiting has also entered a robust growth stage: IDC projects that global online recruiting sales expanded 41% Y/Y to $7.9B in 2004 and may trend up 70% Y/Y to $13.4B in 2005. Euromonitor estimates Chinas employment services market may grow 21% per annum from $1.2B in 2003 to $2.7B in 2007. As compared with its global peers, Chinas online recruiting is still relatively small but is expanding rapidly. We project online recruiting revenue in China climbed 71% Y/Y to $66MM in 2004 and may climb at a CAGR of 53% to $400MM in 2008.
Similar to online traveling business, we believe online recruiters could develop competitive advantages through scale (such as coverage in multiple markets), brand image,
% Growth y/y
Brand image It is an easier decision for consumers to choose the online traveling agency they could recognize than those unfamiliar ones. For instance, around 30% of Ctrips new customers come from the companys general branding efforts. Hotels and airlines would also prefer doing business with branded partners than with those more obscure ones due to perceived higher credentials.
Exhibit 49
Projection for Online Recruiting Industry in China

(CY2003- CY20010E)
800
Online Recruiting Sales
80%
Growth Rate
700 600 500
70% 60% 50%
Page 69
local sales and distribution, and networking effects (potential employees and employers always want to aggregate together).
Cross Examination between CTrip and 51job
2Q2005. We believe the following factors may contribute to such bifurcation: Business model. Ctrip operates an agency model and holds no inventory. It extracts commissions regardless of the price swing. 51job operates a merchant model by owning advertising inventory, resulting in inventory risks and the need to offer promotional discounts for new customer acquisition. As a result, 51jobs revenue per advertising page declined at a compounded rate of 5% Q/Q over the four quarters of 2004. Scalability. Ctrips revenue is transaction-based and is driven by demand, which is more scalable as the company could leverage its transaction platform. As a result, Ctrips average commission per air ticket increased at a compounded rate of 7% per quarter over the past 4 quarters. 51jobs revenue is contract-based and is driven by sales. It needs to deploy marketing and sales force in every city to solicit and retain contracts, which partially explains its margin weakness. Customers. Ctrip directly sells to retail consumers. 51job sells to SMEs (Around 75% of 51jobs customers are SMEs with fewer than 50 people). SMEs typically have more bargaining power and more service requirements than individual consumers, thus impacting 51jobs margins. Business partners. Ctrip secured guaranteed room allotment from one third of its hotel suppliers, which resulted in two thirds of its transaction volume. Hotel operators tend to side with partners which can maximize the customer order flow. 51job has exclusive contractual relationships with local newspapers for distribution. These contracts are typically short-term-based with one-off payment (typically without revenue sharing). Upon contract renewal, competitors may kick in with high bids. Newspapers probably care more about who pays them more rather than who has more customers, creating margin pressure for 51job. Operational efficiency. The average opex per employee was roughly $8,000 for Ctrip and $15,000 for 51job in 2004.
Both companies are market leaders in their own space: Ctrips revenue was 2.6x larger than the No. 2-ranked eLong in 2004. According to iResearch survey, 61% of online travelers booked air tickets or hotels through Ctrip, nearly 2x more than eLong and 7x more than the next competitor. We estimate the revenue size for 51job is 5-10x larger than that of the closest competitors (it is hard to gauge the exact differences as these competitors are unlisted). According to Alexa.com, the total daily reach of Internet users for 51job.com equals that of the next three competitors combined, including ChinaHR, Zhaopin, and Cjol.
We believe Ctrip and 51job are the embodiment of the click plus brick model in China: Both have material online operation. For 2Q2005, Ctrip generates around 30% of sales from online transactions and 51job around 26% from online recruiting. Both have sizable offline presence. Ctrip has local sales and distribution teams in around 40 cities. 51job has local sales forces in 24 cities. Ctrip has over 1,300 call center service staff and 51job has 1,000 ground sales people. Both are active in high growth markets, with huge customer demand (Please see our text above). Both are customer service-intensive. Unlike other Internet businesses which are content-centric, such as news portals and Online Gaming, the growth in online recruiting and online traveling is driven by customer services.
While both companies expanded their top line (61% Y/Y for Ctrip and 26% Y/Y for 51job in 2Q2005), the operating margin remained at mid-40% for Ctrip, flattish Y/Y, but dipped to 13% for 51job, down 10 basis points Y/Y, in
Page 70
Online Shopping Budding and Hopeful
There are encouraging signs for Chinas online shopping business: iResearch estimates that online shopping revenue grew 165% Y/Y to $540MM in 2004 and may climb 93% Y/Y to $1B in 2005. For 2004, IT products accounted for 53% of total market share, 8x more than the No. 2 (groceries and home electronics) and more than everything else combined. Heavy online spending in IT sector, which is among the fastest growing industry, bodes well for the growth prospects of online shopping business in China. CNNIC reported 40% of China Internet users did online shopping at least once in 2004 and 42% used online payment. According to iResearch, over half of Internet users visited Joyo and Dangdang, two of the largest B-to-C websites in China in 2004. Around a quarter to onethird of the visitors bought books/ products from Joyo or Dangdang, which are more than Sina, Sohu, and NetEase combined. The import and export GMV transacted through Alibaba reached $1.5B in 2004. We estimate that there are approximately 6MM SMEs registered with Alibaba, which recorded a GAAP revenue of $46MM and nonGAAP cash receipts of $70MM in 2004. Cncard.com, the leader for online prepaid card sales with roughly 1/4 of online transaction share, reported that its peak transaction volume is around 60,000 per day or roughly 1 transaction per 1.5 second.
delivery system. However, online merchants control the customer relationship and typically enjoy bargaining power over suppliers. Like what happens to Wal-Mart, the large online merchants may enjoy scale benefits. They may command vendors to finance their inventory and thus have negative working capital requirement, such as the case for Dangdang. We view comparison shopping as another area of growth. For instance, Googles Froogle and Zhongsous comparison shopping services offer online customers value for their money and suits the bargain hunting genes embedded in many Chinese shoppers (One of the writers, Richard Ji, is included). In our view, Chinas e-tailors need to resolve the following bottlenecks to make their business materially profitable: Payment is still mostly on COD basis. Improvement in payment and settlement is needed to extend e-tailors reach beyond the local markets. Sourcing requires the knowledge of local consumer demand and extensive network with local vendors. Dangdang estimates that 99% of its products are endemic. Foreign companies, which have price advantages in their home market, may need to develop their local savvy to regain its cost leadership. Logistics makes or breaks e-tailors. Amazon.com achieved market dominance in the US through its superior logistic management. In China, which lags in logistic infrastructure, efficient inventory management and timely delivery are the keys to lower costs and win customers over competition. For instance, Dangdang, the current market leader in online shopping, innovated a bicycle team for local delivery.
There are currently two online commerce models: Platform or agency model - which has lower inventory risks, low working capital requirement, and high margin as it entails minimal inventory and logistics support. However, such a model has lower entry barriers with online agencies subject to margin squeeze by both merchandise suppliers and customers unless they reach critical scale. Merchant/ principal model - such model has high inventory risks and lower margins as online merchants need to operate their own logistics and
Blogs From Passive to Interactive Online Media
Blogs, in their simplest form, are online journals composed by independent online writers on selected subjects that are typically link to other websites/online postings and are posted periodically. Blogs operators encourage readers to comment on the original posts and to become involved in content creation. We believe blogs will become the next growth phenomenon in Chinas Internet space because:
Page 71
Blogs enjoys the networking effect as people of similar interests tend to aggregate together, lowering user acquisition costs. Most blog content is user-generated. Like eBay in online auction and Tencent in IM, blogs operators may enjoy low content costs. Blog communities are highly interactive. Unlike the users of the traditional online portals, who typically are passive readers of portal content, blog users can actively participate in content creation and feedback, thus enhancing the stickiness of the blog community.
Blog communities are highly targetable for online commerce and advertising due to the aggregation of similar interest groups and their higher incentives to purchase goods related to their interests than average Internet users.
Blog operators have demonstrated robust growth. For instance, the web traffic for BlogChina, the current leader in Chinese blog space, jumped to top 30 among all Chinese websites from 1,500 a year ago. The companys traffic has nearly doubled every quarter over the past. It now has 2MM registered users with over 20MM page views per day. As a result, blogs recently become the darlings of venture capitalists. Softbank financed BlogChina.com, IDG backed ChinaBlog.net, and UCI funded BlogBus.com.
Page 72
Exhibit 50
Global Internet Companies Price Performance

Price Information
(US$ in Millions, Except per Share Data)
Company US - Attractive Google eBay Yahoo! Electronic Arts Amazon.com InterActive VeriSign Navteq Activision WebMD Monster.com (4) CNET EarthLink aQuantive GSI Commerce InfoSpace Overstock.com Shopping.com Blue Nile JAMDAT Europe - Cautious T-Online (3) Tiscali Japan - Not Rated Yahoo! Japan (3) Softbank (3) Rakuten (3) China - Attractive Baidu NetEase Shanda Tencent (3) Sina Ctrip Tom Online Sohu The9 51job KongZhong chinadotcom Linktone eLong China Finance Korea - In-Line NHN (3) NCsoft (3) Daum (3) Webzen (3) Nasdaq S&P 500 Footnotes
Ticker GOOG EBAY YHOO ERTS AMZN IACI VRSN NVT ATVI HLTH MNST CNET ELNK AQNT GSIC INSP OSTK SHOP NILE JMDT TOIGn.DE TIS.MI 4689.T 9984.T 4755.Q BIDU NTES SNDA 700.HK SINA CTRP TOMO SOHU NCTY JOBS KONG CHINA LTON LONG CHFI 035420.KS 036570.KS 035720.KS 069080.KQ COMP SPX
Rating O-V O-V O-V O O-V -E-V -O -O-V E-V --E -----E-V E-V ----O-V -O-V E-V O-V E-V E-V -E-V -----E-V E-V U-V ----
9/8/05 Price $295 39 33 57 43 25 23 47 18 11 31 14 10 19 19 25 41 21 35 25 $11 3 $2,362 52 772 $84 76 31 1 27 60 15 17 19 15 11 3 9 12 0 $141 70 27 17 $2,166 1,721
Diluted Shares 303 1,420 1,501 328 429 335 271 91 219 345 122 157 135 64 50 33 19 30 18 24 1,224 397 15 351 12 32 32 70 1,774 53 16 53 36 24 28 34 111 26 11 58 16 20 15 13 ---
Equity Value (1) $89,364 55,265 50,035 18,638 18,439 8,523 6,258 4,270 4,012 3,743 3,738 2,156 1,339 1,210 962 820 758 624 606 591 $13,521 1,204 $35,672 18,258 9,117 $2,700 2,439 2,168 1,828 1,447 955 761 632 457 404 369 327 237 129 24 $2,186 1,418 408 223 ---
2002 -1 (8) (17) 75 (16) (79) -(44) 21 (74) (70) (55) -(82) (59) ----(42%) (50) 60% (29) 51 -1,675 --311 --433 ---(6) ----(27) 6 -(32%) (22)
Total Return % (2) 2003 2004 (5) -91 175 92 179 48 103 -87 5 100 152 83 -167 173 ----133% 56 331% 168 478 -222 --419 --367 ---185 ----155 68 -50% 29 127% 80 67 29 (16) (19) 106 111 66 (9) 53 65 15 (13) 82 106 247 57 35 29 3% (47) 43% 59 159 -43 286 26 (5) 35 (2) (41) 39 271 (4) (43) (40) 38 4,450 39% 46 (50) (37) 9% 11
52-Week 2005 YTD 53% (33) (12) (8) (3) (13) (31) 1 21 33 (9) 22 (14) 110 7 (48) (41) (26) 25 20 (17%) (18) 8% 21 (32) 209% 43 (27) 84 (15) 31 (5) (2) (20) (72) 12 (36) 10 (35) (81) 75% (13) 19 (24) (0%) 3 High $318 59 40 71 47 32 36 49 23 12 34 14 12 20 20 58 77 36 36 35 $14 4 $2,716 53 1,220 $154 81 45 1 40 60 17 24 29 56 13 5 11 26 4 $142 98 30 27 $2,220 1,246 Low $101 31 30 43 31 21 18 30 9 6 21 8 8 7 8 24 30 12 23 15 $10 3 $1,930 34 602 $60 33 21 0 19 30 10 14 15 12 6 2 6 8 0 $74 64 16 14 $1,849 1,090
(1) Equity Value equal to diluted shares multiplied by current price. (2) Total Return computed as the sum of the return on capital gains and dividend yield for individual stocks, and S&P500. Computed as return on (2) capital gains for NASDAQ. (3) Based on Conversion Rates as of 9/08/05, 1 US$ = 0.81 Euro, 1 US$ = 1027.50 KRW, 1 US$ = 110.51 JPY, and 1 US$ = 7.77 HKD. (4) Monster.com is covered by Douglas Arthur, in the Publishing Group. Rating is relative to the Publishing Group, which is rated in-line. (5) 2004 performance since IPO for Google, Navteq, eCost, Shopping.com, JAMDAT, Blue Nile, Overstock.com, Shanda, 51job, Tencent, (5) eLong, KongZhong and Webzen. 2005 performance since IPO for Baidu.
Page 73
Summary Slide Presentation
Page 74
Creating Consumer Value In Digital China

September 2005
Richard-Weidong Ji richard.ji@morganstanley.com Mary Meeker mary.meeker@morganstanley.com
All of Morgan Stanleys technology equity research reports are available on the Internet through Client Link at https://secure.ms.com. If you wish to receive this service please contact your institutional sales representative. This presentation can also be downloaded from http://www.morganstanley.com/techresearch/
Morgan Stanley does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. Please see our research reports for detail on companies under coverage. Customers can access this independent research at www.morganstanley.com/equityresearch or can call 800-624-2063 to request a copy of this research.
1
Contributors
Andy.Xie@morganstanley.com Economics China Mark.Shuper@morganstanley.com Telecommunications China Hani.Abuali@morganstanley.com Telecommunications China Lina.Choi@morganstanley.com Telecommunications China Viktor.Ma@morganstanley.com Technology China
2
Jenny.Wu@morganstanley.com Internet/Media China Brian.Fitzgerald@morganstanley.com Internet US Ramji.Srinivasan@morganstanley.com Internet US
Page 75
Outline

Perspective Focus on Consumer Value will Drive Shareholder Value 10 China Internet Trends Mysteries Challenges Industry Analysis Consumer Focus and Innovation are Key Mobile Value-added Services Online Advertising Online Gaming Online Commerce
Investment Framework - Business / Competence / Price Investment Ideas Ctrip, NetEase, Tencent
3
Risks
In addition to the risks outlined throughout this document:
It is key to remember that investing in emerging markets (the Internet) in emerging markets (like China) is always fraught with risks and challenges, and a portfolio approach to investing is crucial
4
Page 76
Perspective
5
Internet Emerging Sweet Spot of Chinas Consumer Ecosystem
China
Most consumers
(1.3B)
Internet +
Potential to reach more Chinese consumers than other industries
= Size
Fastest growing economy

(19x GDP expansion over past 20 years)
Fastest growing industry in China
= Growth
6
Page 77
Most Trafficked Chinese Sites in China

Traffic Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Company Name Baidu Sina NetEase Sohu QQ / Tencent 3721 (Yahoo!) / Alibaba Allyes TOM Online Taobao (Alibaba) Google Yisou (Yahoo!) / Alibaba ChinaRen / Sohu SoGou / Sohu China.com 21CN eBay EachNet Major Business Paid search Portal (MVAS and online advertising) Portal (online games) Portal (online advertising) Instant messaging Paid search Online advertising Portal (MVAS) Online auction Paid search Paid search Online community Paid search Portal Portal Online auction
Sources: Alexa.com, Morgan Stanley Research (The traffic rank is based on three months of aggregated historical traffic data from Alexa Toolbar users and is a combined measure of page views and users (reach) prior to September 6, 2005. Alexa data may be somewhat skewed to the Asian countries given that the results are a function of opt-in downloads of the Alexa toolbar.)
8
Page 78
Most Trafficked Global Sites

Traffic Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Company Name Yahoo! MSN Google Yahoo! Japan Baidu Sina Passport.net eBay 163.com (NetEase.com) Sohu Microsoft QQ / Tencent Amazon.com 3271 (Yahoo!) / Alibaba Naver.com Myspace.com Country USA USA USA Japan China China USA USA China China USA China USA China Korea USA
Sources: Alexa.com, Morgan Stanley Research (The traffic rank is based on three months of aggregated historical traffic data from Alexa Toolbar users and is a combined measure of page views and users (reach) prior to September 7, 2005. Alexa data may be somewhat skewed to the Asian countries given that the results are a function of opt-in downloads of the Alexa toolbar.)
9
10 China Internet Trends
10
Page 79
Trend
Growing Faster than Other Marketswith More Potential

China leads the world in number of mobile subscribers China ranks No. 2 in Internet users with more Internet users under the age of 30 than any other country Growth drivers for Chinese Internet have been / are robust
By the end of 2004 CAGR for next 3 years Volume 13% Internet users 94MM (18% yoy; < 7% of population) 32% Broadband users 43MM (146% yoy; <3% of population) 12% Mobile users 335MM (24% yoy; 26% of population) ARPU Internet users $5-$6 per month Broadband users $10-$15 subscription fee per month Industry revenue MVAS $770MM (89% yoy) Online Gaming $390MM (90% yoy) Online Advertising $220MM (78% yoy)
28% 37% 24%
CNNIC, Morgan Stanley Research; Notes: MVAS- mobile value-added services; ARPU- average revenue per user
11
Trend 2 Unlike the Rest of the World, China Internet is Mobile-centric Rather Than PC-centric
Mobile Users (MM) 363 177 88 69 54 54 37 Internet Users Mobile Phone to (MM) Internet User Ratio 100 3.6:1 211 0.8:1 78 1.1:1 51 1.4:1 37 1.5:1 32 1.7:1 32 1.2:1 Installed PCs (MM) 53 207 55 39 26 16 27
Country China US Japan Germany UK Italy S. Korea
Euromonitor, CNNIC, World Bank, Morgan Stanley Research (July 2005)
China has 360MM mobile users, 3.6x more than its Internet users and amounting to the level of the next 3 nations combined MVAS contributed $700-800MM in revenue to Chinese Internet firms in 2004, 2x > than Online Gaming and 3x > than online advertising Indicating what users do SMS
12
Page 80
Trend 3 Internet Brings about Cultural Evolution

Pre-Internet Age Status Information Gaping holes Regulated (Top 10 Newspapers all state-owned) Only 8 model movies during the Cultural Revolution Deficient due to one-child policy Post-Internet Age Filling in holes Portals (Sina, Sohu)- alternative news sources New entertainment formats, such as Online Gaming (Shanda, NetEase) Instant messaging (Tencent) and MVAS (TOM Online, Sina) offer connectivity
Entertainment
Interaction
13
Trend 4 Foreign Interests Accelerating

Foreign Entrants Local Targets EachNet 3721.Com Alibaba Joyo eLong ChinaHR Baidu Dates Transaction Type Market Position of Target
eBay Yahoo!
7/03 11/03 8/05 8/04 7/04 2/05 6/04
Acquisition
No. 1 in online auction
Acquisition No. 2 in paid search Bought 40% Stake No. 1 in B2B marketplace Acquisition Bought 30% Stake Bought 40% Stake Bought 3% Stake Leading ecommerce store No. 2 in online traveling No. 2 in online recruiting No. 1 in paid search
Amazon.com InterActive Monster.com Google
Source: Morgan Stanley Research (Market positions were based on our estimates on the dates of transactions)
14
Page 81
Trend 5 Hotspot for Mergers and Acquisitions

Vertical integration Hostile share purchase Shanda vs. Sina TOM Online bought Puccini Sina acquired Memestar & Crillion Sohu acquired ChinaRen, Focus.net, and 17173.com China Internet Shanda purchased PC game/ mobile game/ casual game companies Horizontal integration Poison Pill Sina vs. Shanda
15
Trend 6 A League of Big Players

Industry Online Advertising Companies Sina Sohu Industry Rank No. 1 No. 2 Total share = Online Gaming Shanda NetEase No. 1 No. 2 Total share = MVAS Sina Tom Online Tencent No.1 No. 2 No. 3 Total share = No. 1 Market Share 30% 25% 55% of industry rev 39% 18% 58% of industry rev 16% 15% 10% 41% of industry rev 77% of IM users 65% plus 29% plus 94% plus of GMV
Instant Messaging (IM) Tencent Online Auction
eBay EachNet No. 1 Alibaba/ Taobao No. 2 Total share =
Sources: Company date (2004), IResearch, Morgan Stanley Research; GMV- gross merchandise value
16
Page 82
Trend
Leadership Positions may not Necessarily be Secure

Sales ($MM) MVAS Sina TOM Online as % of Sina Online Gaming Shanda NetEase as % of Shanda Online Advertising Sina Sohu as % of Sina 2Q2003 14 2Q2004 31 29 93% 2Q2005 23 41 180%
17 4 24%
35 16 46%
56 40 70%
Narrowing gaps or Change in leadership
10 7 72%
16 13 86%
20 17 83%
Sources: Company date, Morgan Stanley Research
17
Trend
Purer Players Tend to Have Higher Market Share

Industry Business-line as % of companys revenue
Industry
Companies
Rank in the Industry
Online Gaming
Shanda NetEase
No. 1 No. 2 No. 2
87% 83 66
Online Advertising
Sohu
MVAS
TOM Online
No. 1
95
Company date, Morgan Stanley Research (Estimates are based on 2Q2005 results)
18
Page 83
Trend 9 Regional Focus Can Vary

Distribution of Chinese Internet Users
Geographical concentration
2.9% Beijing 4.3 % 1.3% 3.6 % 0.3% 0.2% 0.1 % 5.6% 1.0% 2.2% 3.0% 12.6% 1.3% 2.8% 2.2 % 3.2% 4.6% 3.3% Tianjin 2.1 % 9.0% 7.0 % 2.6 % 5.7% 1.7 % 3.5% Shanghai 4.7 % 1.9% 3.4%
Chinese Internet users are concentrated in coastal areas (56% of total, 2004, CNNIC) Top 4 cities contribute 50% plus of advertising sales Top 5 cities accounted for 60% of Ctrips hotel sales and 80% of its air ticket sales volume, 2004
1.0 %
West China 19%
Geographical dispersion
Coastal China 56%
0.5%
Source: CNNIC
IDC estimates that 80%+ of online gamers came from non-top tier cities (2003)
19
Please see analyst certification and other important disclosures starting on page 75
Trend 10 Content Becoming King
Content providers gaining pricing power

TOM Online paying out 40-50% of revenue to music companies for distribution rights KongZhong paid 5x more for wireless distribution rights for leading blockbuster movie vs. 2 years ago Foreign game developers, such as Actoz and Blizzard, command multiple fold higher licensing payment vs. several years ago
Rollout of new wireless and broadband services creates demand for quality content
China Mobiles color ringtone users have expanded at quarterly compound rate of 90% over past 1.5 years Most popular song, Mouse Loves Rice, downloaded 5MM times in 2004 with potential profits exceeding the best-selling CDs
Content providers, such as MTV, may produce more revenues from MVAS than from pay TV in a few years
20
Page 84
Mysteries
21
Mystery
Chinese Internet Players at Valuation Discounts to Global Peers
Top 5 Internet Companies China US

Average P/E multiples Average earnings growth Average operating margin 15-25x 25-40% 30-40% 40-50x 25-30% 15-20%
Our diagnosis
Macro uncertainty Economic concerns create market issues Policy risks - revenue sharing changes, content clean up, etc. Competition - Industry leadership changes faster in China than elsewhere High margins may not be sustainable Operational risks Management playbooks still developing
Company date, I/B/E/S, Morgan Stanley Research (Top 5 Internet firms are selected based on their market capitalization; Excluding stock-based share compensation costs, the average operating margin for the US companies should be around 20%-30%)
Business models still unfamiliar to global investors
22
Page 85
Mystery
Real Revenue Perhaps Significantly Higher than Headline Revenue

Industry revenue
($B, CY2004) Headline figures MVAS operators Online Gaming Online Advertising Total After adjusting for PPP Total 0.8 0.4 0.2 Amazon Yahoo eBay Google
US company revenue
($B, CY2004) 6.9 3.6 3.3 3.2
1.4
6.2
17.0
After adjusting for PPP & industry multiplier MVAS-related revenue 21.1 Online Gaming-related revenue 19.4 Online Advertising-related revenue Total 1.1
41.4
Sources: Company date, Morgan Stanley Research (PPP- purchase power parity)
23
Mystery
Chinese Internet Players Have Much Higher Margins than Global Peers
Company Name China China Finance NetEase KongZhong Tencent Shanda Ctrip Sina Sohu Linktone TOM Online 51job Operating Margin Company Name (For CY2004) US 65% eBay 50% Google 41% Yahoo! 40% InterActive 39% Verisign 38% Amazon.com 35% 33% 25% 25% 20% Operating Margin (For CY2004) 32% 20% 19% 17% 11% 6%
Average Margin
37%
18%
Sources: FactSet, Morgan Stanley Research, Company date (based on 2004 reported data; Excluding stock-based compensation, the average operating margins for these Chinese Internet companies and the US companies are 39% and 27%, respectively.)
+: China has abundant labor supply and low living expenses -: There may be cost pressure due to competition, rising property price, content costs, R&D, and marketing costs 24
Page 86
Mystery 4 ecommerce Ready to Take Off in China?

Pillars for ecommerce Development in China GDP per capita Payments $1,270 for China in 2004, 3% of that for the US but expanding rapidly eBay EachNet debuted PayPal in July 2005; Au Fu Tong (escrow service) in place since October Payment volume for Alibaba's Alipay increased 64% per month during 2004 Tencent collects 45% of its sales from non-mobile payment channels in 2Q2005 Shanda and NetEase have direct billing relationships with end customers Logistics Merchant model (Dangdang), which integrates inventory and distribution systems Agency model (Ctrip), which controls delivery without owing inventory
Credit/ Settlement
'Online signature law', released in April, grants online and printed signatures the same authorities Alipay could help to alleviate settlement risks through the escrow system
We believe 2005 will be an important year in the development of payment, credit and logistics systems before ecommerce truly breaks out in 2006 25
Mystery 5 Why can Click Plus Brick Outcompete?

Chinese Internet companies mostly have online plus offline operations For online companies, offline presence is instrumental for:
customer acquisition - Ctrip has over 1,300 call center staff to sell air tickets and hotel room nights product promotion - Paid search operators, such as Baidu, have ground sales forces to market their services to small and medium enterprises product distribution - Shanda distributes its physical prepaid cards through 300,000 retail sales points local partnership - Sina and TOM Online have deployed local forces to tighten their relationship with regional mobile carriers
26
Page 87
Mystery 6 SOEs - Referees and Players?
State-owned Enterprises (SOEs) typically rule-makers
China Mobile has ordered content clean-up, cancellation of inactive accounts, and billing platform migration for MVAS (since 1Q2004).
SOEs also players
MVAS already accounted for ~16% of China Mobile revenue (2004)
Challenges rise when SOEs are key to value chain

Unlike online advertising and online gaming, mobile carriers lie between MVAS providers and consumers Potential rise in revenue sharing by mobile carriers may hurt margins
27
Mystery
Why are Foreign Players Especially Challenged in China?

Domestic players
Regulatory barriers Reporting structure Local management Relationship with SOEs Customer responsiveness Local content Local service Low Lean In charge Good Quick
Foreign entrants
High
Struggles of foreign entrants

Fat (often multi-layer) Mostly left after acquisitions O.K. Slow
eBay EachNet experienced market share erosion in ecommerce to Alibaba / Taobao Amazon/ Joyo yielded its traffic leadership to Dangdang Yahoo! China lost the founder of 3721 NHN/ Ourgame.com outrivaled by Tencent in casual games
+ +
+/+/-
Financial pocket
Source: Morgan Stanley Research
Not so deep
Very deep
28
Page 88
Case StudyBattle between Alibaba / Taobao and eBay EachNet

Alibaba / Taobao
GMV Total no. of listings Listing fees Listing standard $120MM in 1Q2005; $200MM in 2Q2005 10MM No Front page placement of sellers / items are not purchased but chosen by Taobao; longer listing period (1 to 2 weeks plus an automatic extension for 1 more period) Has an early lead by launching Alipay (escrow-based payment) in October 2003 Lean and can typically turn around a decision within hours Alibaba has similar customer demographics (SMEs) and nationwide sales network; Taobao's sellers may sources their products from wholesalers listed on Alibaba
eBay EachNet
$100MM in 1Q2005; not disclosed in 2Q2005 1MM Yes, but reduced by 25%- 50% since May 2005 Focuses on a level playing field and do not allow any one seller / listing top priority; shorter listing period (7 to 10 days), likely making listed products 'fresher' Launched An Fu Tong (escrow-based) in October 2004 and PayPal (online payment) in August 2005 Global corporate structure may slow down decision-making May enable cross-border trading with eBay's large global user base over the long term
Payment system
Reporting Structure Rapport with parent company
Source: Alibaba / Taobao, eBay EachNet, Morgan Stanley Research
29
Case StudyAlibaba / Taobao vs. eBay EachNet

Women Men Buying Selling
Listings Web Design
More customer-centric, with categories like Men, Women and Home Graphics-rich, more appealing to online customers
More product-centric, with categories for Buying and Selling Text-rich, less appealing to both buyers and sellers
Source: Alibaba / Taobao, eBay EachNet, Morgan Stanley Research
30
Page 89
Challenges
31
Challenge 1 Economic Hard Landing or Soft Landing?

As % of GDP China US Differential 42% 70% -28% 12% 15% -3% 44% 20% 24% 2% -5% 7% 36% 7% 29% 100% 100% 0% China 18% Japan 5% 29%
Chinas economy relies heavily on: investment, which is at historical high but may be affected by the burst of property bubbles and overcapacity export, which may be impacted by trade wars and Rmb revaluation Online advertising and advanced MVAS (WAP and MMS) are procyclical Online gaming and basic MVAS (SMS) may be counter-cyclical
Personal consumption Government consumption Investment Net export Export Total Non-performing Loans
Sources: CEIC, Morgan Stanley Research (Economic data were based on 2004 results)
32
Page 90
Challenge 2 Expect the Unexpected!
Regulatory / government-related uncertainties

Leaders of 3 telecom carriers shuffled among each other in late 2004 Changes in revenue sharing by mobile carriers Google has been suspended multiple times due to improper content
MP3 search is among the most popular search applications
It accounts for 20% of Baidus search traffic
Cheating tools and private servers for Online Gaming are widely publicized in China
The9s online game sales declined 47% within a quarter in 1Q2004 due to illegal private servers
33
Challenge 3 Long-term Value vs. Short-term Profit

Short-term behaviors could lead to value destruction ! Case Study- The mishap of MVAS operators in 2004
Services Selected Sanction List SMS Most listed ISPs Sina MMS Sohu Reasons For Sanctions Inactive accounts Billing malpractice Fortune-telling related SMS Push marketing without mobile carriers' approval Billing malpractice Improper content
WAP IVR
Linktone KongZhong/ Sina/ Tencent/ TOM Online
Sources: Company data, Morgan Stanley Research
34
Page 91
Challenge 4 Hollowing of Management

Most Chinese Internet firms have experienced high levels of management turnover Case study- Sohus talent exodus
Name of Sohu's Former Executives Victor Koo Sam Qian Lee Zhang Derek Palaschuk Yufan Zhou Positions At Sohu President Vice President Director, marketing CFO Founder of ChinaRen, now a wholly owned unit of Sohu Vice President Positions Post-Sohu Plans to start up his own venture President, China Finance Co-founder, eLong CFO, eLong CEO, KongZhong
Elaine Feng
Senior Vice President, TOM Online
Sources: Company date, Morgan Stanley Research
35
Other Challenges
Chinas current business environment is more favorable to SOEs than start-ups, especially in areas like regulation and financing Dearth of innovation
lots of me-too products, services, and content
Too much capital, too few good ideas To buy or to build?
Most Internet companies are facing the dilemma on whether to expand through organic growth or acquisitions
Founder mentality

Start-up founders may sometimes lose momentum once becoming rich William Ding is one good example who keeps his focus on consumer value, innovation and building a long-lasting organization
36
Page 92
Industry Analysis Consumer Focus and Innovation are Keys
37
How do We Rate an Industry
What is important
Small fish in a big pond or big fish in a small pond? How close can we get to consumers? How crowded is the space? Who is the boss in the value chain? Industry risks
What we are looking for Growth really matters The closer, the better The less competitive, the better We want bargaining powers Returns should be risk-adjusted
38
Page 93
Industry Analysis Not All Value Chains were Created Equal

Online advertising
Online Advertisers Advertising Agency
Online gaming
Online Game Developers
Online Game Operators
Internet Cafes
Consumers
Distributors
MVAS
Content poviders
MVAS Operators
Mobile Carriers
(China Mobile & China Unicom)
Distance from service providers to consumers: Online Gaming < Online Advertising < MVAS 39
Scoreboard for Chinese Internet Industries

MVAS Industry Growth Consumer Access Entry Barriers Industry Risks Total Score Industry Ranking 2 1 2 1 6 4 Online Brand Advertising 2 2 2 2 8 3 Online Gaming 3 3 3 2 11 1 Paid Search 3 2 2 3 10 2 Notes For The Scores (3: Sales CAGR for next 3 years >30%, 2: 20%< but <30%, 1: <20%) (3: Good direct access; 2: Some direct access; 1: Access all through third party) (3: High entry barriers; 2: Some barriers; 1: Low barriers) (3: Low risks; 2: Medium risks; 1: High risks)
Sources: Morgan Stanley Research
Our top-down view

Online game attractive, paid search robust, MVAS bottoming out, and online brand advertising challenging
40
Page 94
Chinese Online Gaming IndustryGrowth and Scale are the Name of the Game
Chinese Online Gaming Industry Sales
$390MM sales for 2004 and 37% CAGR for next 3 years
Large players lead the space
2000 1800 Online Game Sales (US$ mn) 1600 1400 1200 1000 800 600 400 200 0
E E E E E 03 04 05 06 07 08 09 20 20 20 20 20 20 20 20 10 E
Online Game Sales Growth Rate
100% 90% 80% 70% % Growth yoy 60% 50% 40% 30% 20% 10% 0%
Company Names Shanda NetEase The9
Online Game Revenue ($MM) Market Share $154 72 4 $391 39% 18 1 100%
Total Online Game Revenue
41
Beauty of Online Gaming

Online Gaming is:
self-enhancing with networking effect - NetEases most popular game has an average concurrent users of around 300,000-400,000, 5x more than US leaders sticky - average user playing time > 3 hrs per day (IDC) scalable - Shandas operating margin on average expanded 5% per quarter in 2004
Sources: NC-Sina, NetEase, Shanda
42
Page 95
Whats Unique about Chinese Online Gaming Industry
Revenue multiplier for Chinese Online Gaming sales is around 11x (IDC estimates) 6x for telecom industry, 3x for IT industry, and 2x for media and publishing industry Online Gaming is inexpensive entertainment even by Chinese standard For two-hour entertainment, it costs around $1 playing games at Internet cafes versus $5-$10 for movies and $10-$15 for Karaoke
Games of Asian culture origins dominate the top 10 list
Rank 1 2 3 4 5 6 7 8 9 10
Name of The Online Games Lineage 2 Mir 2 Mir 3 World of Legend Fantasy Westward Journey Mu JX Online Magic Baby Westward Journey Ragnarok Online (RO)
Developers Ncsoft Actoz Actoz Shanda Netease Universal/Blizzard Kingsoft Square Enix Netease Gravity
Operators NC-Sina Shanda Optisp Shanda Netease The9 Kingsoft Joypark Netease Gameflier
Sources: The News & Publishing Bureau of China (2004)
43
Comparing Online Gaming with Movies All Hit-driven, but with Different Life Cycles
Movies
Content Star Effect Life Cycle Popularity User Experience Interactivity Revenue Model Style Preference Finished products once released Big movie stars draw crowd Even blockbuster movies typically do not last over 1 year Mostly decline over time for movies Movie goers passively accept the content Disallowed in cinema Box office sales, typically one-off, and sales of relalted items Chinese people love Hollywood
MMORPGs
Work-in-progress that can be upgraded periodically Online gamers can become self-made stars Leading games can last over 5 years May gain traction for popular games Online gamers actively participate in content creation Part of the 'killer application' for online games Based on playing time or subscription, thus recuring None of the Western style games ranked within top 10 in 2004
Sources: Morgan Stanley Research
44
Page 96
Profiling Different Game Products

3D MMORPG* 3D MMORPG Casual Game Mobile Game
Scale
Multiplayers, typically thousands to hundred thousands
Single player to a few players Game item sales
Single player to a few players Subscription
Revenue Model Subscription, hourly charge, game item sales Playing Time Several hrs per day
Several minutes to half hr per day Several minutes to half hr per day
Source: NetEase, NC-Sina, Tencent, TOM Online, Morgan Stanley Research (MMORPG- massively multiplayer online role-playing game) * First Person Game
45
Our Investment Focus in Online Gaming - Integrated Game Operators

Game content is king
Foreign game developers command 5x-10x higher game licensing fees now than a few years ago NetEase and Kingsoft are leaders in self game development
Channel margin for game distributors declined by 50% over past 3 years (iResearch)
Shanda is the front runner in game distribution
Game Developers
Game Distributors
Internet Cafes
Consumers
Integrated Game Operators

46
Page 97
MVAS Largest Revenue Contributor for China Internet

Chinese commercial MVAS industry sales
$770MM in 2004 with high-20% CAGR for next 3 years
MVAS industry is more fragmented than Online Gaming and Online Advertising
2500 MVAS Sales Growth Rate
120% 100% 80%

Company Names Sina TOM Online Tencent Linktone KongZhong Hurray! Sohu NetEase Total commercial MVAS
2000 MVAS Sales ($MM)
1500 60% 1000 40% 500 20% 0%

20 04 20 05 E 20 06 E 20 07 E 20 08 E 20 09 E 20 10 E
MVAS Revenue ($MM, CY2004) 124 113 77 50 48 43 38 17 510
% Market Share 16% 15% 10% 7% 6% 6% 5% 2% 66%
0
20 03
Sources: Company data (2004), Morgan Stanley Research (MVAS- mobile value-added services, commercial MVAS- computer-to-person MVAS)
% Growth Y/Y
47
MVAS Light at the End of the Tunnel
New Trends
Entry barriers are rising
Regulation, relationship, scale, and costs
Our investment focus

Market leaders with edgy business model, edgy content, and edgy distribution
Tencent TOM Online
Content becoming hot

TOM Online pays out 40%-50% of revenue to leading content providers
China Mobile becoming more transparent with its policies

Data services contributed 16% of the companys 2004 sales
Listed MVAS providers mostly trade at low-teen multiples of forward earnings
48
Page 98
What does It Take to be a Winner in a Choppy Industry?
MVAS has more risks than other Internet sectors
Winners DNA
Unique business model - Crillion, now wholly owned by Sina, enjoys exclusive relationship with job centers Unique services - TOM-Skype Unique products - KongZhongs wireless distribution rights of leading blockbuster movies Unique distribution - Portals (Sina & TOM Online) as self-owned distribution channels; Treasury Bases exclusive alliance with TV channels 49
Short term - migration to new billing platform Medium term - Revenue sharing change with mobile carriers Long term - rising content costs
MVAS - Which Product has the Best Prospects?
MVAS Products IVR/ RBT
Product Revenue (CY2004, $MM) 86
Annual Sales Growth (CY2005) 68%
Sales CAGR (CY2006-CY2008) 22%
SMS a cash cow 2.5G robust IVR/ RBT attractive
SMS 2.5G MMS WAP
484
23%
13%
94 108
53% 51%
29% 32%
Sources: Morgan Stanley Research (IVR- Interactive voice response, MMS- multimedia messaging, RBT- ringback tone, SMS- short messaging services, WAP- wireless application protocol)
50
Page 99
Art of Managing MVAS Portfolio
Company Names IVR/RBT Sales ($MM) as % of Total MVAS SMS Sales ($MM) as % of Total MVAS 2.5G (MMS/WAP) Sales ($MM) as % of Total MVAS Total MVAS Sales ($MM) as % of Total MVAS
Sina 5 4% 103 83% 16 13% 124 100%
Tencent Tom Online Linktone KongZhong Sohu Netease Hurray! 6 29 3 8 1 0 3 8% 26% 5% 16% 2% 1% 7% 65 55 40 8 30 14 14.9 83% 49% 80% 16% 79% 85% 35% 7 29 8 33 7 2 25 8% 26% 15% 68% 19% 14% 59% 77 113 50 48 38 17 43 100% 100% 100% 100% 100% 100% 100%
Sources: Company data (2004), Morgan Stanley Research
Sina leads in SMS KongZhong leads in 2.5G TOM Online has the most balanced MVAS portfolio
51
Online Advertising Race for Eyeballs and Performance

Chinese Online Advertising Industry Sales
$220MM in 2004 with mid-20% CAGR for next 3 years
800 Online Advertising Sales (US$ mn) 700 600 500 400 300 200 100 0
20 03 20 04 20 05 E 20 06 E 20 07 E 20 08 E 20 09 E 20 10 E
Online Brand Adv Sales Growth Rate
A League for the Big Guys
160% 140% 120% % Growth yoy 100% 80% 60% 40% 20% 0%
Company Names Sina Sohu NetEase TOM Online Tencent Total Ad Revenue
Online Ad Revenue ($MM) Market Share $65 56 19 8 7 $220 29.7% 25.3 8.6 3.4 3.0 100%
Sources: Company data (2004), Morgan Stanley Research
52
Page 100
Our Investment FocusContent, Demographics and Performance
What stands out in China

Chinese advertising market - fastest growing in world and second largest in Asia Chinese online advertising sales quadrupled over past 2 years, faster than any other advertising subsector Online advertising sales are billed on the basis of cost per period rather than cost per impression in China Risks - sector volatility, competition & shift to performance-centric model
Winners attributes:
Content
Sinas partnership with 1,000 plus content providers; Sohus content alliance with NBA and Disney
Demographics
Sina attracts upscale Internet users (40% with college degree versus less than 5% for general public)
Advertising effectiveness
May be improved by online verticals, community-based targeting, search, email advertising & narrowads
53
Paid Search and its Unique Story in China

Chinese paid search industry sales
$74MM in 2004 with 60%+ CAGR E for next 3 yrs
1000 900 800 Paid Search Sales Growth rate
160% 140% 120%
Top 3 players accounted for 86% of search traffic in 2004 (iResearch)Baidu (33%), Yahoo/Yisou/3721 (30%), Google (22%) Search results are regulated in China - Google has been suspended for improper content Ground salesforce, YES; direct sales, NO!- Sohu had 5% of Baidus search traffic but 70% of its 2004 search revenue due to effective sales MP3 is among the most popular search services in China and accounts for 20% of Baidus traffic
Paid Search Sales ($MM)
700 600 500 400 300 40% 200 100 0

20 06 E 20 07 E 20 08 E 20 05 E 20 09 E 20 10 E 20 03 20 04
80% 60%
20% 0%
Sources: iResearch, Morgan Stanley Research
% Growth Y?Y
100%
54
Page 101
Online Commerce - Road to the Promised Land

Our View
Instant Messaging (IM) A building block for Internet value-added service
Features
Tencent has 3 quarters of IM market share as China's largest online community Yahoo! bought 40% stake in Alibaba / Taobao in August eBay EachNet launched PayPal to counter Alipay Chinese travelers spent a modest $50 per person per trip in 2004; frequently flyer only 1% of population Online recruting contributed 13% of China's 2004 recruting service revenue (In comparison, 70% plus US companies hired people online in 2004) CNNIC reported that 40% of China Internet users did online shopping at least once in 2004 Close to 30% of US Internet users read Blogs, among the fastest growing online applications; lower content costs but higher stickness than existing online media
Current Leaders
Tencent, MSN
P2P Online Retailing
The combat between eBay EachNet and Taobao may expand the whole market The incarnation of 'click plus brick'
eBay EachNet Alibaba / Taobao Ctrip, eLong
Online Traveling
Online Recruiting
Riding the employment boom of SMEs
51job, ChinaHR Zhaopin
Online Shopping
Budding and hopeful
Dangdang Amazon.com/Joyo ChinaBlog, BlogChina
Blogging
From passive to interactive online media
Sources: CNNIC, China National Tourism Administration, company data, iResearch, Net-Temps, Morgan Stanley Research
55
Investment Framework - Business / Competence / Price
56
Page 102
Our Way of Picking Stocks

Large market Robust growth High barriers
Good Business Model
+
Managementcapable, reliable & consumer-centric Innovation
Good Core Competence
Our Stock Picks
+
Margins of safety Free embedded options
Good Price
57
Case Study- Business Model Matters Microscopic Cross-Examination Between Ctrip and 51job
Ctrip
Business models Agency model with little inventory risks Commission-based Scalable platform- could expand new business on existing platform
51job
Merchant model with advertising inventory Contract-based, entailing promotional discount for customer acquisition Needs to deploy marketing and sales force in each new city, leading to margin erosion Mostly SMEs, who enjoy some bargaining power Contract-based (no revenue sharing) relationship with local newspapers, which care more about who pays them more than who brings in more customers, creating margin pressure for 51job Opex- $15,000 per head for 2004 13%, down 10 ppts year-on-year
Scalability
Customers Business Partners
Retail customers with low bargaining power Revenue-sharing-based relationship with hotel or air ticket suppliers, who preferring siding with whoever could maximizes their customer traffic
Sales Efficiency
Opex- $8,000 per head for 2004
Operating Margin 44%, flat year-on-year

Sources: Company data, Morgan Stanley Research (Margin estimates were based on 2Q2005 results)
58
Page 103
Case StudyCompetence Make Differences Between Ctrip and eLong

Differential Between Ctrip vs. eLong Rank In Travel Consolidation Industry Revenue Hotel Room Nights Sold Air Tickets Sold Net Profit Revenue Per Head Revenue Per Contracted Hotels Opex Per Head Customer Acquisition Costs
No. 1 vs. No. 2 2.6x 7x 2x $16MM vs. loss of -$2.6MM 1.4x 1.8x 0.6x 0.5x
59
Case StudyOur Price Discipline: Focus on Growth and Bargains

Valuation Criteria Discount relative to potential growth Examples Our projected PEGs for Ctrip, NetEase, and Tencent are at around 0.5x (relative to their operating income growth) Ctrip and NetEase exceeded their own earning guidance by 25% for 2Q2005 Our top 3 picks trade at discounts relative to our assumed DCF values (DCF values: Ctrip- $71 per share; NetEase: $114, Tencent: HK$10.2) Our estimated sum-of-parts value for NetEase is $107 per share - Re-rating of Sohu's search business - NetEase's email community, the largest in China - Tencent's Q-Zone and QQ Pets - TOM Online's opportunity in Skype
Upside from company's outexecution Margin of safety' versus DCF value
Hidden value in sum-of-parts Free 'embedded options'
60
Page 104
Investment Ideas
61
Our Top Stock Ideas in China Internet Ctrip (CTRP, Overweight-V) - Setting an operational standard rarely seen in China NetEase (NTES, Overweight-V) - Best-in-class in online gaming Tencent (0700.HK, Overweight-V) - Leader of virtual community
62
Page 105
Ctrip (Overweight-V) High Flyer from the First to the Last Mile
Investment Positives
One of fastest growing companies in China - Sales up 94% Y/Y in 2004, versus 75% for Sina, 67% for NetEase, and 28% for Sohu; widened its gap against No. 2-ranked eLong Higher barriers to entry - 1/3 of contracted hotels offer guaranteed room allotment, contributing 2/3 of total hotel room sales in 2004; 80% more sales per contracted hotel than eLong
Service platform that enjoys the best of all:

Scalability - operating margin of 41% in 2004, up from 34% in 2003 and 23% in 2002 Networking benefits Broad coverage - 3,000 plus hotels and air ticketing capability in 39 cities Agency instead of principal - low inventory risks 6-Sigma service quality - high customer loyalty with 70% plus of transactions from repeat users Labor arbitrage through click plus brick - 70% of sales from offline with online type of growth and margin
63
Ctrip (Overweight-V) Setting an operational standard rarely seen in China

Ctrip's Operating Efficiency
Ctrips Cost Leadership
Ctrip eLong NetEase S&P companies Microsoft
Opex Per Head (US$) 8,000 13,000 27,000 73,000 300,000
Ctrip's total revenue (CY2004) Air ticket sales Hotel room sales Operating revenue per head Operating revenue per hotel Ctrip's commission per booking
2.6X larger than eLong 2X 7X 1.4X 1.8X 13X cost per air ticket booking 22X cost per hotel room booking
Investment Risks
Double exposure to economic shakeout Business travel - 88% of Ctrips 2004 business Top 5 cities- 70% of Ctrips air ticket sales May expose to inventory risk in packaging business and margin pressure in leisure travel: Chinese travelers averaged only $50 per person per travel in 2004, 3% that of US travelers
Sources: FactSet, Company data, Morgan Stanley Research
64
Page 106
NetEase (Overweight-V) In a Class by Itself

NetEase generates 80%+ revenue from online game sales, which have grown by 32% per quarter over the past 8 quarters The peak concurrent users for NetEases two leading Online Gaming, Fantasy Westward Journey and Westward Journey 2, climbed to 710,000 and 480,000 in 2Q2005
Sources: NetEase (Screenshots were taken from NetEases massively multi-player online role-playing game, Fantasy Westward Journey)
65
NetEase (Overweight-V) Best-in-Class in Online Gaming

NetEase is the China Internet leader in innovation and was the first Internet company to: Tap into MVAS, thus becoming the first profitable Internet company in China Shift its focus away from MVAS due to regulatory uncertainties Focus on in-house game development, producing 2 out of the top 10 Online Gaming and the most popular MMORPG (Fantasy Westward Journey) in China Recorded the fastest growth in online game sales in China - compound growth rate of 32% per quarter over past 8 quarters Embedded option value in portal, online community and casual games NetEases portal ranks No. 4 in web traffic and its online advertising ranks No. 3 Owns the largest email user base in China with 47% of market share (iResearch) Plans to launch 9 casual games in 2H2005 and could leverage its portal as a distribution platform The William Ding premium - As Gates is a valuable asset for Microsoft, so is Ding for NetEase
Sources: iResearch, Company data, Morgan Stanley Research
66
Page 107
NetEase (Overweight-V) Competitive Advantages, Threats

Our View
NetEase vs. Shanda - Closing The Gap

NetEase vs. Other Portals More nimble, more creative Deeper pipeline, deeper pocket More local flavor, better local distribution
Netease
Online game sales as % of Shanda's in 2Q2005 as % of Shanda's in 2Q2004 Compounded growth over past 6 Qs 70% 46% 28%
Shanda
100% 100% 18% 1 3-tiers 14,000 2.5MM $0.027 45% $30,000 to $32,000
NetEase vs. Domestic Game Operators NetEase vs. Foreign Rivals
Investment Risks
No. of self-developed games within top 10 2 Distribution channels More direct sales No. of servers 300 to 400 No. of peak concurrent users 1MM plus ARPU for online games $0.035 Operating margin (2Q2005) Annual opex per staff (2004) 63% $27,000
Competition - Kingsoft, Shanda and The9 Regulation of online addiction - Money has color Average online game playing time increased to 6 hours per day in 2004, up 1 hour Y/Y (iResearch) Marginalization of non-game business Online Gaming now 80% plus of NetEases revenue
Sources: FactSet, Company data, Morgan Stanley Research (ARPU- average revenue per online game user)
67
Tencent (Overweight-V) Reigning Champ of Virtual Community

Online Advertising & Avatars
MVAS
Virtual Items
Online Gaming
Sources: Tencent
68
Page 108
Tencent (Overweight) Monetizing Online Communication and Entertainment

Dominating instant messaging (IM) space with authority 3 quarters of IM users frequented Tencents IM in 2004, whose gap against MSN was widened to 5x 300MM downloads of QQ client software, 16MM voice-over-IP (VoIP) calls per day 50% + of Chinese Internet users use QQ IM service High entry barriers - insulated by firewalls; core IM services for free; scale of QQ community Fellowship of IM community and its networking benefits QQ identity, in combination with IM communication, enhances stickiness Low content costs - 70% to 80% of content is user generated Fewer inactive MVAS accounts due to the interactivity among IM users No. 1 casual game operator in user volume in China - 1.4MM PCUs and 7% of Tencents total sales Accelerated monetization - QQ Pet/Q-Zone/auction/non-mobile payment channel/advertising price rise (50% for 2005)
Investment risks- battle against giants (MSN, TOM-Skype), MVAS exposure

Sources: iResearch, Company data, Morgan Stanley Research
69
Sina (Equal-weight-V) Executing, but on Increasingly Competitive Fronts

Leading online content provider
Overall web site, homepage, news, sports, entertainment, and lifestyle channels all top-rated in traffic in China Premium demographics - Higher spending power than general Internet users Up-market adv. customers - Average spending 1.5x more than Sohus in 2004 Recovery of MVAS (60%-plus of Sinas sales) in progress Management among most financially capable in China - Memestar & Crillion acquisitions highly earning-accretive
% of revenue Online Advertising 42% MVAS 52% Others 6% Total Sales 100%
Industry Rank No. 1 No. 2 No. 1
Investment Risks
Stretched Rivalry -TOM Online in MVAS, Sohu in advertising, NetEase, Shanda & The9 in Online Gaming, Tencent in IM, eBay EachNet & Alibaba / Tao in Online Auction Advertising volatility & loss of pricing power Face-off with Shanda in anti-takeover battle
Sources: Alexa.com, CNNIC, Company data, Morgan Stanley Research (Industry ranks are based on the company earnings for 2Q2005)
70
Page 109
Case Study
What we Think about a Potential Sina / Shanda Combination
The Bull Case
- Sina's content leadership may fuel the growth of Shanda's home entertainment strategy - Sina's management depth may help Shanda to venture into new businesses - The consumer demographics of the two parties may complement each other - Shanda may broaden the distribution of Lineage 2 and NC-Sina may add some potential hits to Shanda's pipeline - The transaction may transform Shanda's image into a digital media leader from an online game operator
The Bear Case

- Shanda's core competence center on online games instead of MVAS or online advertising- Shanda thus needs Sina's management buy-in but the 'Poison Pill' suggests the opposite - Low cost-saving synergy due to little business overlaps - Shanda's risk profile may escalate due to MVAS exposure - Shanda is undergoing major transformation. In its core gaming business, rivals, such as NetEase and The9, are closing the gap. Its interactive TV initiatives may prove thorny rather than rosyIt may repeat what happened to ISPs in MVAS as it needs to collaborate with state-owned content producers/ telecom carriers
- Culture clash may be conceivable as Sina is the oldest Internet company in China with a professional management team taking charge of a diverse spectrum of businesses whilst Shanda is a new entity on the block with relatively simpler operations
71
Sohu (Equal-weight-V) The Purest Advertising Portal

Web Property Nature Of The Property Web portal Alumni network Sponsored search Online game information protal Real estate portal Traffic Ranks No. 4 No. 12 No. 13 No. 31 No. 44
Sohu's Portal ChinaRen.com SoGou
Owns 5 of top 50 sites in China, second to none Ranks No. 2 in online advertising sales; narrowing its gap with Sina Sponsored search sales only behind Baidu and Yahoo!/3721 12% of total sales for 2Q2005 SoGou- among the top 15 sites overall in China MVAS - Minimal downside, rebound in sight
17173.com Focus.net
Investment Risks
Rivalry - Sina in brand advertising, Baidu / Google in search, verticals and rising portals Management turnover Advertising - sector volatility, secular shift toward performance-centric model
Sources: Alexa.com, Company data, Morgan Stanley Research (Traffic ranks are based on user reach or the number of visitors per 1 mn Internet users on September 6, 2005)
72
Page 110
TOM Online (Equal-weight-V) A Portal by Origin; A Wireless Leader by Choice

Examples Of TOM Online's Edgy Practice Content Content alliance with Universal, Warner, BMG, EMI etc Acquired Indiagame, the leading mobile game operators in India Exclusive operator of Skype in China
Market leader in MVAS sales Only company that ranks within top 3 in every MVAS business Most balanced MVAS portfolio in the industry with IVR/RBT, SMS, and 2.5G evenly contributing to top line Edgy content - One of the largest wireless content aggregators Edgy services - TOM-Skype Edgy distribution- alliance with radio, TV and telecom operators; portal as a selfowned MVAS distribution channel
Services
Distribution Treasure Base has exclusive distribution alliance with CCTV channels Distribution alliance with over 90 TV stations, 300 radio stations 50 newspapers, 45 magazines, 40 video/audio distributors
Investment Risks
Policy risks - billing platform migration, idle accounts, content crackdown, and revenue sharing changes Margin pressure due to content and distribution alliance Under-monetized portal traffic
73
51job (Equal-weight-V) A Job Still to Be Done

Operating margin
Scale benefits: We estimate 51jobs revenue size is 5-10x larger than closest rivals Daily reach for 51job.com equal to next 3 players combined (Alexa.com) Brand recognition - Both Career Post Weekly and 51job.com among the leading brands in recruiting services Broad national coverage - Operations in 23 cities, 2x more than the runner up Networking effect - Unique employers jumped 52% Y/Y in 2004; 5MM online resumes nearly 2x more than ChinaHR
3Q2004 26%
4Q2004 16%
1Q2005 2Q2005 6% 10%
Revenue growth Career Post Weekly Online Differential
2004 (Y/Y) 1Q2005 (Q/Q) 2Q2005 (Q/Q) 65% 27% -5% 45% 7% 5% 20% 20% -10%
Investment Risks
Scalability in question Op margin dipped 10 ppts over past 1 year Revenue per advertising page dropped at a compounded rate of 5% per quarter in 2004 Competition - tougher, smarter, and stronger
Sources: Alexa.com, Company data, Morgan Stanley Research (Traffic ranks are based on user reach or the number of visitors per 1 mn Internet users on July 2005)
74
Page 111
Disclaimer www.morganstanley.com/research
Analyst Certification The following analysts hereby certify that their views about the com panies and their securities discussed in this report are accurately expressed and that they have not received and will not receive direct or indirect com pensation in exchange for expressing specific recommendations or views in this report: Richard Ji, Mary Meeker. Unless otherwise stated, the individuals listed on the cover page of this report are research analysts. Important US Regulatory Disclosures on Subject Companies The inform ation and opinions in this report were prepared or are dissem inated by Morgan Stanley Dean W itter Asia Lim ited and/or Morgan Stanley Dean W itter Asia (Singapore) Pte.(Registration num ber 199206298Z) and/or Morgan Stanley Asia (Singapore) Securities Pte Ltd (Registration number 200008434H) and/or Morgan Stanley & Co. International Limited, Taipei Branch and/or Morgan Stanley & Co International Limited, Seoul Branch, and/or Morgan Stanley Dean W itter Australia Limited (A.B.N. 67 003 734 576, holder of Australian financial services licence No. 233742, which accepts responsibility for its contents), and/or JM Morgan Stanley Securities Private Limited and their affiliates (collectively, "Morgan Stanley"). Certain disclosures listed above are also for com pliance with applicable regulations in non-US jurisdictions. As of July 29, 2005, Morgan Stanley beneficially owned 1% or m ore of a class of common equity securities of the following com panies covered in this report: SINA CORP COM SHS, SOHU COM INC COM. W ithin the last 12 m onths, Morgan Stanley managed or co-m anaged a public offering of securities of 51job, Inc. W ithin the last 12 m onths, Morgan Stanley has received com pensation for investm ent banking services from 51job, Inc. In the next 3 months, Morgan Stanley expects to receive or intends to seek com pensation for investm ent banking services from 51job, Inc, Sina Corp. W ithin the last 12 m onths, Morgan Stanley has provided or is providing investm ent banking services to, or has an investment banking client relationship with, the following com panies covered in this report: 51job, Inc, Sina Corp. The research analysts, strategists, or research associates principally responsible for the preparation of this research report have received com pensation based upon various factors, including quality of research, investor client feedback, stock picking, com petitive factors, firm revenues and overall investm ent banking revenues. Morgan Stanley & Co. Incorporated m akes a market in the securities of Tom Online, 51job, Inc, SINA CORPORATION, SOHU.COM INC, NETEASE.COM INC.
75
Disclaimer www.morganstanley.com/research
Stock Ratings Different securities firms use a variety of rating terms as well as different rating systems to describe their recommendations. For example, Morgan Stanley uses a relative rating system including terms such as Overweight, Equal-weight or Underweight (see definitions below). A rating system using terms such as buy, hold and sell is not equivalent to our rating system. Investors should carefully read the definitions of all ratings used in each research report. In addition, since the research report contains more complete information concerning the analysts views, investors should carefully read the entire research report and not infer its contents from the rating alone. In any case, ratings (or research) should not be used or relied upon as investment advice. An investors decision to buy or sell a stock should depend on individual circumstances (such as the investors existing holdings) and other considerations. Global Stock Ratings Distribution (as of August 31, 2005)
Coverage Universe S tock Rating Category Count % of Total Investment B anking Clients (IBC) Count % of Total IBC % of Rating Category
Overw eight/Buy Equal-w eight/Hold Underweight/Sell Total
689 902 379 1,970
35% 46% 19%
265 312 93 670
40% 47% 14%
38% 35% 25%
Data include common stock and ADRs currently assigned ratings. For disclosure purposes (in accordance with NASD and NYSE requirements), we note that Overweight, our most positive stock rating, most closely corresponds to a buy recommendation; Equal-weight and Underweight most closely correspond to neutral and sell recommendations, respectively. However, Overweight, Equalweight, and Underweight are not the equivalent of buy, neutral, and sell but represent recommended relative weightings (see definitions below). An investor's decision to buy or sell a stock should depend on individual circumstances (such as the investor's existing holdings) and other considerations. Investment Banking Clients are companies from whom Morgan Stanley or an affiliate received investment banking compensation in the last 12 months. ANALYST S TOCK RATINGS Overweight (O). The stocks total return is expected to exceed the total return of the relevant country MSCI index, on a risk-adjusted basis, over the next 12-18 months. Equal-weight (E). The stocks total return is expected to be in line with the total return of the relevant country MSCI index, on a risk-adjusted basis, over the next 12-18 months. Underweight (U). The stocks total return is expected to be below the total return of the relevant country MSCI index, on a risk-adjusted basis, over the next 12-18 months. More volatile (V). We estimate that this stock has more than a 25% chance of a price move (up or down) of more than 25% in a month, based on a quantitative assessment of historical data, or in the analysts view, it is likely to become materially more volatile over the next 1-12 months compared with the past three years. Stocks with less than one year of trading history are automatically rated as more volatile (unless otherwise noted). We note that securities that we do not currently consider "more volatile" can still perform in that manner. Unless otherwise specified, the time frame for price targets included in this report is 12 to 18 months. ANALYST INDUSTRY VIEWS Attractive (A): The analyst expects the performance of his or her industry coverage universe over the next 12-18 months to be attractive vs. the relevant broad market benchmark, as indicated below. In-Line (I): The analyst expects the performance of his or her industry coverage universe over the next 12-18 months to be in line with the relevant broad market benchmark, as indicated below. Cautious (C): The analyst views the performance of his or her industry coverage universe over the next 12-18 months with caution vs. the relevant broad market benchmark, as indicated below. Benchmarks for each region are as follows: North America - S &P 500; Latin America - relevant MSCI country index; Europe - MSCI Europe; Japan - TO PIX; Asia/Pacific - relevant MSCI country index. Stock price charts and rating histories for companies discussed in this report are available at www.morganstanley.com/companycharts or from your local investment representative. You may also request this information by writing to Morgan Stanley at 1585 Broadway, (Attention: Equity Research Management), New York, NY, 10036 USA.
76
Page 112
Other Disclosures
Other Important Disclosures This research report has been published in accordance with our conflict managem ent policy, which is available at www.m organstanley.com/institutional/research/conflictpolicies. For a discussion, if applicable, of the valuation m ethods used to determine the price targets included in this summary and the risks related to achieving these targets, please refer to the latest relevant published research on these stocks. Research is available through your sales representative or on Client Link at www.m organstanley.com and other electronic systems. This report does not provide individually tailored investm ent advice. It has been prepared without regard to the individual financial circumstances and objectives of persons who receive it. The securities discussed in this report may not be suitable for all investors. Morgan Stanley recomm ends that investors independently evaluate particular investm ents and strategies, and encourages investors to seek the advice of a financial adviser. The appropriateness of a particular investm ent or strategy will depend on an investors individual circumstances and objectives. This report is not an offer to buy or sell or the solicitation of an offer to buy or sell any security or to participate in any particular trading strategy. The "Important US Regulatory Disclosures on Subject Companies" section lists all companies m entioned in this report where Morgan Stanley owns 1% or more of a class of common securities of the companies. For all other com panies m entioned in this report, Morgan Stanley may have an investm ent of less than 1% in securities or derivatives of securities of companies m entioned in this report, and may trade them in ways different from those discussed in this report. Em ployees of Morgan Stanley not involved in the preparation of this report may have investm ents in securities or derivatives of securities of companies m entioned in this report, and may trade them in ways different from those discussed in this report. Derivatives may be issued by Morgan Stanley or associated persons. Morgan Stanley & Co. Incorporated and its affiliate companies do business that relates to companies covered in its research reports, including market making and specialized trading, risk arbitrage and other proprietary trading, fund managem ent, investm ent services and investm ent banking. Morgan Stanley sells to and buys from custom ers the equity securities of companies covered in its research reports on a principal basis. W ith the exception of information regarding Morgan Stanley, reports prepared by Morgan Stanley research personnel are based on public information. Morgan Stanley makes every effort to use reliable, comprehensive inform ation, but we m ake no representation that it is accurate or complete. W e have no obligation to tell you when opinions or information in this report change apart from when we intend to discontinue research coverage of a subject company. Facts and views presented in this report have not been reviewed by, and may not reflect information known to, professionals in other Morgan Stanley business areas, including investm ent banking personnel. Morgan Stanley research personnel conduct site visits from tim e to tim e but are prohibited from accepting paym ent or reimbursem ent by the company of travel expenses for such visits. The value of and incom e from your investm ents may vary because of changes in interest rates or foreign exchange rates, securities prices or market indexes, operational or financial conditions of companies or other factors. There m ay be tim e limitations on the exercise of options or other rights in your securities transactions. Past performance is not necessarily a guide to future performance. Estimates of future performance are based on assumptions that may not be realized. Unless otherwise stated, the cover page provides the closing price on the prim ary exchange for the subject companys securities. To our readers in Taiwan: Information on securities that trade in Taiwan is distributed by Morgan Stanley & Co. International Limited, Taipei Branch (the "Branch"). Such information is for your reference only. The reader should independently evaluate the investm ent risks and is solely responsible for their investm ent decisions. This publication may not be distributed to the public m edia or quoted or used by the public m edia without the express written consent of Morgan Stanley. Information on securities that do not trade in Taiwan is for informational purposes only and is not to be construed as a recomm endation or a solicitation to trade in such securities. The Branch may not execute transactions for clients in these securities.
(continued on next page)
77
Other Disclosures
Other Important Disclosures To our readers in Hong Kong: Information is distributed in Hong Kong by and on behalf of, and is attributable to, Morgan Stanley Dean Witter Asia Limited as part of its regulated activities in Hong Kong. If you have any queries concerning this publication, please contact our Hong Kong sales representatives. Certain information in this report was sourced by employees of the Shanghai Representative Office of Morgan Stanley Dean Witter Asia Limited for the use of Morgan Stanley Dean Witter Asia Limited. This publication is disseminated in Japan by Morgan Stanley Japan Limited; in Hong Kong by Morgan Stanley Dean Witter Asia Limited (which accepts responsibility for its contents); in Singapore by Morgan Stanley Dean Witter Asia (Singapore) Pte. (Registration number 199206298Z) and/or Morgan Stanley Asia (Singapore) Securities Pte Ltd (Registration number 200008434H), regulated by the Monetary Authority of Singapore, which accepts responsibility for its contents; in Australia by Morgan Stanley Dean Witter Australia Limited A.B.N. 67 003 734 576, holder of Australian financial services licence No. 233742, which accepts responsibility for its contents; in Korea by Morgan Stanley & Co International Limited, Seoul Branch; in India by JM Morgan Stanley Securities Private Limited; in Canada by Morgan Stanley Canada Limited, which has approved of, and has agreed to take responsibility for, the contents of this publication in Canada; in Germany by Morgan Stanley Bank AG, Frankfurt am Main, regulated by Bundesanstalt fuer Finanzdienstleistungsaufsicht (BaFin); in Spain by Morgan Stanley, S.V., S.A., a Morgan Stanley group company, which is supervised by the Spanish Securities Markets Commission (CNMV) and states that this document has been written and distributed in accordance with the rules of conduct applicable to financial research as established under Spanish regulations; in the United States by Morgan Stanley & Co. Incorporated and Morgan Stanley DW Inc., which accept responsibility for its contents. Morgan Stanley & Co. International Limited, authorized and regulated by Financial Services Authority, disseminates in the UK research that it has prepared, and approves solely for the purposes of section 21 of the Financial Services and Markets Act 2000, research which has been prepared by any of its affiliates. Private U.K. investors should obtain the advice of their Morgan Stanley & Co. International Limited representative about the investments concerned. In Australia, this report, and any access to it, is intended only for "wholesale clients" within the meaning of the Australian Corporations Act. The trademarks and service marks contained herein are the property of their respective owners. Third-party data providers make no warranties or representations of any kind relating to the accuracy, completeness, or timeliness of the data they provide and shall not have liability for any damages of any kind relating to such data. The Global Industry Classification Standard ("GICS") was developed by and is the exclusive property of MSCI and S&P. Morgan Stanley has based its projections, opinions, forecasts and trading strategies regarding the MSCI Country Index Series solely on publicly available information. MSCI has not reviewed, approved or endorsed the projections, opinions, forecasts and trading strategies contained herein. Morgan Stanley has no influence on or control over MSCIs index compilation decisions. This report or any portion hereof may not be reprinted, sold or redistributed without the written consent of Morgan Stanley. Morgan Stanley research is disseminated and available primarily electronically, and, in some cases, in printed form. Additional information on recommended securities is available on request.
78
Page 113
ModelWare is a proprietary framework for financial analysis created by Morgan Stanley Research. This new framework rests on the principles of comparability, transparency, and flexibility, and aims to provide investors with better tools to view the anticipated performance of an enterprise. The result of an 18-month global effort, ModelWare harmonizes the underlying data and calculations in Morgan Stanley models with a broad set of consistently defined financial metrics. Our analysts have populated the database with over 2.5 million data points, based on an extensive taxonomy of more than 3,500 unique metrics and more than 400 Morgan Stanley calculations. The ModelWare framework will also have the flexibility to allow analysts and investors to quickly customize their own analytical approach. What makes the ModelWare architecture distinctive lies in the separation of data from calculations. Its transparency will permit users to see every component of every calculation, to choose elements or recombine them as they wish without laborious adjustments or recalculations. When choices must be made in defining standard or industry-specific measures, ModelWare defaults to economic logic, rather than favoring one accounting rule over another. This discipline facilitates comparability across sectors and regions. Underlying the ModelWare data is a new set of systems that check the internal consistency of forecast data in each of our analysts models. ModelWare EPS illustrates the approach taken. It represents ModelWare net income divided by average fully diluted shares outstanding. ModelWare net income sums net operating profit after tax (NOPAT), net financial income or expense (NFE), and other income or expense. ModelWare adjusts reported net income to improve comparability across companies, sectors, and regions. These adjustments include the following: We exclude goodwill amortization and items deemed by analysts to be one-time events; we capitalize operating leases where their use is significant (e.g., in transportation and retail); and we convert inventory to FIFO accounting when LIFO costing is used. For more information on these adjustments and others, as well as additional background, please see Morgan Stanley ModelWare (ver. 1.0): A Road Map for Investors, by Trevor Harris and team, August 2, 2004.
Page 114
Analyst Certification
The following analysts hereby certify that their views about the companies and their securities discussed in this report are accurately expressed and that they have not received and will not receive direct or indirect compensation in exchange for expressing specific recommendations or views in this report: Richard Ji, Mary Meeker. Unless otherwise stated, the individuals listed on the cover page of this report are research analysts.
Important US Regulatory Disclosures on Subject Companies

The information and opinions in this report were prepared or are disseminated by Morgan Stanley Dean Witter Asia Limited (which accepts the responsibility for its contents) and/or Morgan Stanley Dean Witter Asia (Singapore) Pte. (Registration number 199206298Z, regulated by the Monetary Authority of Singapore, which accepts the responsibility for its contents), and/or Morgan Stanley Asia (Singapore) Securities Pte Ltd (Registration number 200008434H, regulated by the Monetary Authority of Singapore, which accepts the responsibility for its contents), and/or Morgan Stanley & Co. International Limited, Taipei Branch and/or Morgan Stanley & Co International Limited, Seoul Branch, and/or Morgan Stanley Dean Witter Australia Limited (A.B.N. 67 003 734 576, holder of Australian financial services licence No. 233742, which accepts responsibility for its contents), and/or JM Morgan Stanley Securities Private Limited and their affiliates (collectively, "Morgan Stanley"). Certain disclosures listed below are also for compliance with applicable regulations in non-US jurisdictions. As of July 29, 2005, Morgan Stanley beneficially owned 1% or more of a class of common equity securities of the following companies covered in this report: SINA CORP COM SHS, SOHU COM INC COM. Within the last 12 months, Morgan Stanley managed or co-managed a public offering of securities of 51job, Inc. Within the last 12 months, Morgan Stanley has received compensation for investment banking services from 51job, Inc. In the next 3 months, Morgan Stanley expects to receive or intends to seek compensation for investment banking services from 51job, inc, Sina Corp. Within the last 12 months, Morgan Stanley has provided or is providing investment banking services to, or has an investment banking client relationship with, the following companies covered in this report: 51job, Inc, Sina Corp. The research analysts, strategists, or research associates principally responsible for the preparation of this research report have received compensation based upon various factors, including quality of research, investor client feedback, stock picking, competitive factors, firm revenues and overall investment banking revenues. Morgan Stanley & Co. Incorporated makes a market in the securities of Tom Online, 51job, Inc, SINA CORPORATION, SOHU.COM INC, NETEASE.COM INC.
Page 115
Stock Ratings Different securities firms use a variety of rating terms as well as different rating systems to describe their recommendations. For example, Morgan Stanley uses a relative rating system including terms such as Overweight, Equal-weight or Underweight (see definitions below). A rating system using terms such as buy, hold and sell is not equivalent to our rating system. Investors should carefully read the definitions of all ratings used in each research report. In addition, since the research report contains more complete information concerning the analysts views, investors should carefully read the entire research report and not infer its contents from the rating alone. In any case, ratings (or research) should not be used or relied upon as investment advice. An investors decision to buy or sell a stock should depend on individual circumstances (such as the investors existing holdings) and other considerations. Global Stock Ratings Distribution
(as of August 31, 2005)
Coverage Universe Stock Rating Category Count % of Total Investment Banking Clients (IBC) Count % of Total IBC % of Rating Category
Overweight/Buy 689 35% 265 40% 38% Equal-weight/Hold 902 46% 312 47% 35% Underweight/Sell 379 19% 93 14% 25% Total 1,970 670 Data include common stock and ADRs currently assigned ratings. For disclosure purposes (in accordance with NASD and NYSE requirements), we note that Overweight, our most positive stock rating, most closely corresponds to a buy recommendation; Equal-weight and Underweight most closely correspond to neutral and sell recommendations, respectively. However, Overweight, Equal-weight, and Underweight are not the equivalent of buy, neutral, and sell but represent recommended relative weightings (see definitions below). An investor's decision to buy or sell a stock should depend on individual circumstances (such as the investor's existing holdings) and other considerations. Investment Banking Clients are companies from whom Morgan Stanley or an affiliate received investment banking compensation in the last 12 months.
ANALYST STOCK RATINGS
Overweight (O). The stocks total return is expected to exceed the total return of the relevant country MSCI index, on a riskadjusted basis, over the next 12-18 months. Equal-weight (E). The stocks total return is expected to be in line with the total return of the relevant country MSCI index, on a risk-adjusted basis, over the next 12-18 months. Underweight (U). The stocks total return is expected to be below the total return of the relevant country MSCI index, on a risk-adjusted basis, over the next 12-18 months. More volatile (V). We estimate that this stock has more than a 25% chance of a price move (up or down) of more than 25% in a month, based on a quantitative assessment of historical data, or in the analysts view, it is likely to become materially more volatile over the next 1-12 months compared with the past three years. Stocks with less than one year of trading history are automatically rated as more volatile (unless otherwise noted). We note that securities that we do not currently consider "more volatile" can still perform in that manner. Unless otherwise specified, the time frame for price targets included in this report is 12 to 18 months.
ANALYST INDUSTRY VIEWS
Attractive (A): The analyst expects the performance of his or her industry coverage universe over the next 12-18 months to be attractive vs. the relevant broad market benchmark, as indicated below. In-Line (I): The analyst expects the performance of his or her industry coverage universe over the next 12-18 months to be in line with the relevant broad market benchmark, as indicated below. Cautious (C): The analyst views the performance of his or her industry coverage universe over the next 12-18 months with caution vs. the relevant broad market benchmark, as indicated below. Benchmarks for each region are as follows: North America - S&P 500; Latin America - relevant MSCI country index; Europe - MSCI Europe; Japan - TOPIX; Asia/Pacific - relevant MSCI country index.
Stock price charts and rating histories for companies discussed in this report are available at www.morganstanley.com/companycharts or from your local investment representative. You may also request this information by writing to Morgan Stanley at 1585 Broadway, (Attention: Equity Research Management), New York, NY, 10036 USA.
Page 116
Other Important Disclosures

This research report has been published in accordance with our conflict management policy, which is available at www.morganstanley.com/institutional/research/conflictpolicies. For a discussion, if applicable, of the valuation methods used to determine the price targets included in this summary and the risks related to achieving these targets, please refer to the latest relevant published research on these stocks. Research is available through your sales representative or on Client Link at www.morganstanley.com and other electronic systems. This report does not provide individually tailored investment advice. It has been prepared without regard to the individual financial circumstances and objectives of persons who receive it. The securities discussed in this report may not be suitable for all investors. Morgan Stanley recommends that investors independently evaluate particular investments and strategies, and encourages investors to seek the advice of a financial adviser. The appropriateness of a particular investment or strategy will depend on an investors individual circumstances and objectives. This report is not an offer to buy or sell any security or to participate in any trading strategy. In addition to any holdings disclosed in the section entitled "Important US Regulatory Disclosures on Subject Companies", Morgan Stanley and/or its employees not involved in the preparation of this report may have investments in securities or derivatives of securities of companies mentioned in this report, and may trade them in ways different from those discussed in this report. Derivatives may be issued by Morgan Stanley or associated persons. Morgan Stanley & Co. Incorporated and its affiliate companies do business that relates to companies covered in its research reports, including market making and specialized trading, risk arbitrage and other proprietary trading, fund management, investment services and investment banking. Morgan Stanley sells to and buys from customers the equity securities of companies covered in its research reports on a principal basis. Morgan Stanley makes every effort to use reliable, comprehensive information, but we make no representation that it is accurate or complete. We have no obligation to tell you when opinions or information in this report change apart from when we intend to discontinue research coverage of a subject company. With the exception of information regarding Morgan Stanley, reports prepared by Morgan Stanley research personnel are based on public information. Facts and views presented in this report have not been reviewed by, and may not reflect information known to, professionals in other Morgan Stanley business areas, including investment banking personnel. Morgan Stanley research personnel conduct site visits from time to time but are prohibited from accepting payment or reimbursement by the company of travel expenses for such visits. The value of and income from your investments may vary because of changes in interest rates or foreign exchange rates, securities prices or market indexes, operational or financial conditions of companies or other factors. There may be time limitations on the exercise of options or other rights in your securities transactions. Past performance is not necessarily a guide to future performance. Estimates of future performance are based on assumptions that may not be realized. To our readers in Taiwan: Information on securities that trade in Taiwan is distributed by Morgan Stanley & Co. International Limited, Taipei Branch (the "Branch"). Such information is for your reference only. The reader should independently evaluate the investment risks and is solely responsible for their investment decisions. This publication may not be distributed to the public media or quoted or used by the public media without the express written consent of Morgan Stanley. 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. The Branch may not execute transactions for clients in these securities. Certain information in this report was sourced by employees of the Shanghai Representative Office of Morgan Stanley Dean Witter Asia Limited for the use of Morgan Stanley Dean Witter Asia Limited.
Page 117
This publication is disseminated in Japan by Morgan Stanley Japan Limited; in Hong Kong by Morgan Stanley Dean Witter Asia Limited; in Singapore by Morgan Stanley Dean Witter Asia (Singapore) Pte. (Registration number 199206298Z) and/or Morgan Stanley Asia (Singapore) Securities Pte Ltd (Registration number 200008434H), regulated by the Monetary Authority of Singapore, which accepts responsibility for its contents; in Australia by Morgan Stanley Dean Witter Australia Limited A.B.N. 67 003 734 576, holder of Australian financial services licence No. 233742, which accepts responsibility for its contents; in Taiwan by Morgan Stanley & Co. International Limited, Taipei Branch; in Korea by Morgan Stanley & Co International Limited, Seoul Branch; in India by JM Morgan Stanley Securities Private Limited; in Canada by Morgan Stanley Canada Limited, which has approved of, and has agreed to take responsibility for, the contents of this publication in Canada; in Spain by Morgan Stanley, S.V., S.A., a Morgan Stanley group company, which is supervised by the Spanish Securities Markets Commission (CNMV) and states that this document has been written and distributed in accordance with the rules of conduct applicable to financial research as established under Spanish regulations; in the United States by Morgan Stanley & Co. Incorporated and Morgan Stanley DW Inc., which accept responsibility for its contents. Morgan Stanley & Co. International Limited, authorized and regulated by Financial Services Authority, disseminates in the UK research that it has prepared, and approves solely for the purposes of section 21 of the Financial Services and Markets Act 2000, research which has been prepared by any of its affiliates. Private U.K. investors should obtain the advice of their Morgan Stanley & Co. International Limited representative about the investments concerned. In Australia, this report, and any access to it, is intended only for "wholesale clients" within the meaning of the Australian Corporations Act. The trademarks and service marks contained herein are the property of their respective owners. Third-party data providers make no warranties or representations of any kind relating to the accuracy, completeness, or timeliness of the data they provide and shall not have liability for any damages of any kind relating to such data. The Global Industry Classification Standard ("GICS") was developed by and is the exclusive property of MSCI and S&P. Morgan Stanley has based its projections, opinions, forecasts and trading strategies regarding the MSCI Country Index Series solely on publicly available information. MSCI has not reviewed, approved or endorsed the projections, opinions, forecasts and trading strategies contained herein. Morgan Stanley has no influence on or control over MSCIs index compilation decisions. This report or any portion hereof may not be reprinted, sold or redistributed without the written consent of Morgan Stanley. Morgan Stanley research is disseminated and available primarily electronically, and, in some cases, in printed form. Additional information on recommended securities is available on request.
The Americas
Europe
Japan
Asia/Pacific
1585 Broadway New York, NY 10036-8293 United States Tel: +1 (1)212 761 4000
25 Cabot Square, Canary Wharf London E14 4QA United Kingdom Tel: +44 (0)20 7425 8000
20-3, Ebisu 4-chome Shibuya-ku, Tokyo 150-6008, Japan Tel: +81 (0)3 5424 5000
Three Exchange Square Central Hong Kong Tel: +852 2848 5200
2005 Morgan Stanley

China Internet 091205

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

China Internet 091205

Uploaded by

Copyright:

Available Formats

Comment

Equity Research Global

Richard Ji (Internet / Media - China)

September 12, 2005

Mary Meeker (Internet - US / Global)

Creating Consumer Value in Digital China

Table of Contents Report Roadmap

Investment Thesis - Focus on Consumer Value will Drive Shareholder Value

China Internet September 12, 2005

China Internet Trends

1 - Growing Faster Than Other Marketswith More Potential

Major Drivers for Chinese Internet Are Trending Up

China US Japan Germany UK Italy S. Korea

3.6:1 0.8:1 1.1:1 1.4:1 1.5:1 1.7:1 1.2:1

Source: Euromonitor, CNNIC, Morgan Stanley Research (July 2005).

28% 37% 24%

China Internet September 12, 2005

China Internet User Age Demographics

Ages 25-30 17%

71% under 30 54% under 24

Examples of Foreign Entries into Chinas Internet Market

4 - Foreign Interests Accelerating

China Internet September 12, 2005

Alexa Global Web Site Traffic Rankings

Alexa Chinese Web Site Traffic Rankings

5 - Hotspot for Mergers and Acquisitions

China Internet September 12, 2005

Bull Case for Combination of Shanda and Sina

Bear Case for Combination of Shanda and Sina

The transaction could transform Shandas public image from

Sina and Shanda have different business cultures. Sina is the

6 - A League of Big Players

Online Leaders Have Lions Share of Chinese Market

eBay EachNet Alibaba/ Taobao

8 - Purer Plays Tend to Have Higher Market Share

7 - Leadership Position May Not Necessarily be Secure

Pure Plays Tend to Dominate the Internet Market

leadership in MVAS to TOM Online, a much purer MVAS player.

China Internet September 12, 2005

West China 19%

Coastal China 56%

10 - Content Becoming King

China Internet September 12, 2005

China Most Frequently Used Online Services

China Internet September 12, 2005

China Internet Mysteries

1 - Chinese Internet Players at Valuation Discounts to Global Internet Peers

Price Discount of China Internet Firms versus US Peers

2 - Real Revenue Perhaps Significantly Higher than the Headline Revenue

FY2004 Reported Operating Margin

China Internet September 12, 2005

4 - Ecommerce Ready to Take Off?

China Internet September 12, 2005

China Internet September 12, 2005

China Mobiles Mood Swing

7 - Why Are Foreign Players Especially Challenged in China?

Battle Between eBay EachNet and Alibaba / Taobao

China Internet September 12, 2005

China Internet September 12, 2005

Screenshots for the Homepages of Taobao and eBay EachNet

Source: eBay EachNet, Taobao.

Source: eBay EachNet, Morgan Stanley

Weekly Listings in China eBay EachNet (P2P)