41 Global Equity Research 05 Januar y 2009 Imran Khan (1-212) 622-6693 imran.t.khan@jpmorgan.com For our international forecast, we expect two main drivers to impact eCommerce growth: (a) a very weak consumer outlook, with historically low consumer confidence in parts of the world, including the UK, and (b) continued shift of wallet share from offline retailers to online sellers. Both in the US and worldwide, we expect the eCommerce to benefit from several drivers, some region-specific and some general. Of primary importance among these is the growing adoption of broadband, which is much more conducive to eCommerce growth than slower forms of Internet access. Other factors include: (1) continued rises in online shopping penetration in Western Europe, (2) continued investments by online retailers in broadening selection, (3) improvements in shipping infrastructure, (4) improved payment systems and (5) better fraud protection. Table 18: Global eCommerce Forecast (Excluding Travel) $ in millions Global eCommerce Forecast 2004 2005 2006 2007 2008E 2009E 2010E 2011E ’08E – ’11E CAGR US 88,780 106,851 128,708 151,443 160,563 168,270 187,958 207,505 8.9% Europe 52,430 72,690 96,631 123,698 152,537 140,740 170,942 198,200 9.1% Asia 31,972 43,054 55,556 69,705 90,585 112,573 127,949 139,617 15.5% ROW 9,440 13,216 18,502 25,903 34,970 41,963 55,811 73,113 27.9% Total 182,622 235,811 299,397 370,749 438,655 463,546 542,660 618,435 12.1% Y/Y Growth 29.1% 27.0% 23.8% 18.3% 5.7% 17.1% 14.0% Source: Department of Commerce, Internet WorldStats, UK eStats, Forrester Research, Iresearch, Korea National Statistics Office, Japanese Statistics Bureau, eMarketer, PhuCusWright, TIA.org, Jupiter, company reports, J.P. Morgan estimates Increasingly, It’s Amazon vs. Everybody Else Whereas 2007 saw the large-cap eCommerce companies grow revenue at a faster pace than the US market, in 2008 it was Amazon maintaining its outperformance, while eBay fell somewhat behind, with decelerating sales and GMV growth as the Auctions business appears to be fairly mature; eBay continues to benefit from strong PayPal revenue growth. Table 19: eCommerce Industry Comparable Table -- Small and Large Cap Revenue Growth Rates Market cap in $ millions Company Ticker Rating Price Mkt Cap 12/30/2008 12/30/2008 ’05/’06 ’06/’07 ’07/’08E Amazon.com Inc. AMZN OW 50.76 21,776.0 26% 39% 26% eDiets.com Inc. DIET NR 3.40 85.5 -9% -39% -17% drugstore.com inc. DSCM NR 1.04 100.4 4% 7% -10% eBay Inc. EBAY N 13.96 17,824.7 31% 29% 13% 1-800-Flowers.com FLWS NR 3.55 225.6 17% 1% 5% MercadoLibre Corp. MELI OW 15.69 695.0 84% 64% 62% Netflix Inc. NFLX NR 28.66 1,694.4 46% 21% 13% Blue Nile Inc. NILE UW 23.67 343.7 24% 27% -3% Overstock.com Inc. OSTK NR 11.51 261.8 -2% -4% 17% Shutterfly Inc. SFLY OW 6.75 169.2 47% 51% 12% Stamps.com Inc. STMP NR 9.23 177.1 37% 1% -2% Group Average 28% 18% 11% US E-Commerce 21% 18% 6% Source: Department of Commerce, Company reports, Reuters, J.P. Morgan est. for rated companies, FactSet estimates for other companies. J.P. Morgan ratings: OW = Overweight; N = Neutral; UW = Underweight 42 Global Equity Research 05 Januar y 2009 Imran Khan (1-212) 622-6693 imran.t.khan@jpmorgan.com Takeaways from the J.P. Morgan Internet Team’s 2008 Consumer Survey In the second half of November, we conducted a proprietary survey of US consumers. The survey had 766 participants, residents of the United States aged 18+. In addition to asking consumers about their shopping plans for the 2008 holiday season, we asked the respondents a variety of questions about their online shopping habits in general, and the key broad takeaways are summarized below: Price Is Biggest Factor in Choosing a Site Nearly half of shoppers indicated that price was their top factor when it comes to choosing an online store, with selection also important. Of the options included in our survey, the least popular one, by a wide margin, was, “recommendations from friends and relatives.” Table 20: Price Is Paramount: Factors that Influence Buyers’ Choices % among respondents who indicated they shop online Factor #1 factor #2 factor In top 5 Price 48.8% 19.5% 91.0% Selection 12.0% 25.3% 79.1% Familiarity/experience with store 12.0% 8.7% 57.3% Payment options 5.0% 10.2% 50.9% Customer service 5.0% 7.9% 50.0% Promotions / advertisements 3.7% 9.9% 42.8% Access to customer reviews / product info 2.7% 5.9% 39.0% Ability to purchase multiple items 2.7% 4.9% 33.4% Name recognition 5.8% 5.0% 32.5% Recommendations from friends / relatives 2.1% 2.6% 21.0% Source: J.P. Morgan Internet Team 2008 Consumer Survey Looking at the data on a site-by-site basis does not change the top-line conclusion: price remains the #1 consideration for most shoppers regardless of site, with 46-50% of shoppers at each of Amazon, eBay, Walmart.com, Target.com and Sears.com listing price as their top factor in choosing a site. Higher-income shoppers look for slightly different things in a site. More shoppers with incomes over $100K listed “Familiarity/experience with store” as one of their top five factors than did those earning less than $100K. Even for the >$100K-income shoppers, price is the biggest factor, but notably fewer respondents chose it as their #1 factor: 38% for the higher-income group, vs. 50% for those with incomes below $100K. Table 21: Familiarity Matters More to Higher-Income Shoppers % of online shoppers who indicated factor was one of their top 5 factors in choosing a site Income <$100K Income >$100K Difference Price 92.3% 82.4% 9.9% Selection 79.7% 75.3% 4.4% Familiarity/experience with store 56.7% 61.2% -4.4% Payment options 51.3% 48.2% 3.1% Customer service 49.7% 51.8% -2.0% Promotions / advertisements 42.0% 48.2% -6.2% Access to customer reviews / product info 38.4% 43.5% -5.2% Ability to purchase multiple items 33.8% 30.6% 3.2% Name recognition 32.4% 32.9% -0.5% Recommendations from friends / relatives 20.7% 23.5% -2.9% Source: J.P. Morgan Internet Team 2008 Consumer Survey 43 Global Equity Research 05 Januar y 2009 Imran Khan (1-212) 622-6693 imran.t.khan@jpmorgan.com Walmart.com, Target.com Catching Up to Large Online Sites Third-party metrics indicate eBay and Amazon are the two largest online-only stores in terms of unique users, whereas Walmart.com and Target.com are the two biggest sites belonging to brick-and-mortar retailers. In ’07, Amazon and eBay drew business from more online shoppers than any other sites we asked about. For the ’08 holidays, our survey reveals no significant change in Amazon and eBay’s reach, compared to double-digit increases in the number of shoppers planning to use Walmart.com and Target.com; the only other site in our survey with a statistically significant Y/Y increase, Sears.com, also belongs to a brick-and-mortar retailer. Table 22: Online Shoppers Who Used (’07) or Plan to Use (’08) Specific Sites for Holiday Shopping Site 2007 2008 Y/Y difference significant? Amazon 50.9% 48.3% No eBay 33.6% 32.4% No Walmart.com 31.3% 34.7% Yes, up 11% Y/Y Bestbuy.com 22.1% 22.4% No Target.com 21.7% 26.8% Yes, up 24% Y/Y Overstock.com 16.4% 18.2% No Circuitcity.com 12.0% 13.6% No Sears.com 9.0% 12.3% Yes, up 36% Y/Y QVC.com 7.3% 7.0% No HSN.com 5.1% 5.1% No Source: J.P. Morgan Internet Team 2008 Consumer Survey. Statistical significance at a p<.05 level. Media, Apparel Top Categories Books, music and video comprised the top category, with ~65% of online shoppers reporting they had bought such items online within the past year; the number was in line with our survey result a year ago. Apparel and accessories, at 59% penetration, was second, tilted predominantly toward female shoppers. Unsurprisingly, men were more likely to shop in both the Computers and videogames category and the Electronics category. The results were very similar to those in our 2007 survey, with the top four categories among men unchanged (Toys replaced Health and Beauty at #5). For women, the five most popular categories remained the same, though Apparel leapfrogged over Media into first place, and Toys edged out Computers and videogames for #4. Table 23: Media Items Popular with Both Genders: Top Five Categories among Men and Women %, among respondents who indicated they shop online, that indicated they had made purchases in category Men Women Books, music, videos and other media 61.5% Apparel and accessories (includes shoes, handbags and jewelry) 69.0% Computer and videogames software and hardware 50.5% Books, music, videos and other media 68.5% Apparel and accessories (includes shoes, handbags and jewelry) 46.2% Health & Beauty 41.1% Electronics (includes television and stereo equipment) 33.2% Toys 39.7% Toys 29.6% Computer and videogames software and hardware 37.5% Source: J.P. Morgan Internet Team 2008 Consumer Survey 44 Global Equity Research 05 Januar y 2009 Imran Khan (1-212) 622-6693 imran.t.khan@jpmorgan.com Younger Users Shop Online More, Use Classifieds More Among online shoppers, 40% of those aged 18-41 expect to do more than 40% of their holiday shopping online this year, compared to 27% of shoppers aged 42+. Nearly half of the older group plans to do 20% or less of their holiday shopping online. At the same time, the 42+ bracket included almost twice as high a percentage of people who planned to do almost all (over 80%) of their shopping online, 8% vs. 4%. Figure 20: Expectation of Portion of Holiday Shopping to Be Done Online, Holiday Season ’08 % among respondents who indicated they shop online 34% 26% 23% 13% 4% 47% 25% 11% 9% 8% 0% 10% 20% 30% 40% 50% 0-20% 21-40% 41-60% 61-80% 81-100% Age 18-41 Age 42+ Source: J.P. Morgan Internet Team 2008 Consumer Survey Younger shoppers also appear to be more open to using classified sites, with significantly higher usage rates. The discrepancy is even more extreme at the edges of the age distribution: 48% of online shoppers 18-25 reported having bought something on a classified site, compared to only 17% of those aged 58 and older. Likewise, more than 41% in the youngest group reported having sold something on a classified site, compared to only 16% of respondents aged 58 and older. Figure 21: Online Classified Sites Attract a Younger Demographic % of online shoppers saying they had ever bought/sold on a classified site (such as craigslist) 40% 34% 21% 17% 0% 10% 20% 30% 40% 50% Bought on a classified site Sold on a classified site Age 18-41 Age 42+ Source: J.P. Morgan Internet Team 2008 Consumer Survey Internet Sales Tax: In US, the Action Will Be in the States In 2008, the state of New York began requiring online retailers to collect sales taxes even if they had no physical presence in the state, so long as members of their affiliate networks (who get paid a commission for driving sales traffic to the site) are present in the state. 45 Global Equity Research 05 Januar y 2009 Imran Khan (1-212) 622-6693 imran.t.khan@jpmorgan.com Several companies, notably Amazon, have started collecting the NY sales tax while appealing the ruling. Overstock.com, on the other hand, has chosen to drop its NY- based affiliates and not charge the tax. Given that many states are facing significant revenue shortfalls this year, we expect to see continued efforts by states to collect tax revenue on online sales. While the implementation of state sales taxes can be a slight negative, we do not expect this trend to have a material impact on US eCommerce. Catalysts for International Growth We believe the rising tide of increased Internet use across the world is likely to help lift eCommerce globally. However, we see three key challenges to overcome for eCommerce to fulfill its potential: • Improvement of shipping infrastructure. Postal and parcel service in many parts of the world can be unreliable, and a reliable distribution channel is an essential prerequisite for the growth of eCommerce. • Improved payment systems. This is not a world-wide challenge, but rather a slew of country-specific challenges related to the idiosyncrasies of different countries’ banking systems and conventions. Even in more developed countries, significant differences emerge: e.g., Germany and Austria have historically seen lower rates of PayPal use than other eBay geographies due to the prevalence of bank transfers as a mode of payment there. Table 24: PayPal Penetration on Select eBay Country Sites % of listings that include PayPal Country % Penetration US 99.3% UK 98.8% Canada 98.6% Australia 94.2% Switzerland 89.4% Italy 87.6% Germany 82.2% France 81.6% Spain 78.9% Belgium 42.2% Source: eBay.com (and country sites), J.P. Morgan estimates, data collected December, 2008 • Better fraud protection. The promise of eCommerce has been one of lower prices and/or better selection, with the trade-off that many purchases must be made sight-unseen. The threat of fraud remains present, and structures that insure buyers against fraud should help smooth operations in an eCommerce environment that is not yet fully mature. (An example is eBay’s PayPal, which has, over the course of the past year, augmented the level of fraud protection it provides in the US, from $2K per transaction to as much as the full purchase price of the transaction. PayPal uses such protection to create buyer confidence, with transaction loss rates for the PayPal unit of 24-33 bps, as denominated by total payment volume). 46 Global Equity Research 05 Januar y 2009 Imran Khan (1-212) 622-6693 imran.t.khan@jpmorgan.com We note that the above is not intended to be an exhaustive list of catalysts for international growth – many specific markets can present unique challenges – such as governmental ones – for the operation of eCommerce companies. 47 Global Equity Research 05 Januar y 2009 Imran Khan (1-212) 622-6693 imran.t.khan@jpmorgan.com 2009 Online Travel Outlook In 2008, online travel agency companies in the US performed relatively well compared to offline agencies, primarily due to some counter-cyclical effects of the weak macroeconomic environment. The companies in our coverage universe moved to push their lowest price or discount offerings and began a variety of promotions to woo travelers to their sites. However, growth in international markets began a significant deceleration, pressured not only by the weak economy but also by weaker currency exchange rates. Furthermore, any positive counter-cyclical effects seen in the US in the first half of the year were overpowered by the negative macroeconomic environment going into the second half of the year. In 2009, we expect to see the macroeconomic environment and difficult comps have a large impact on both domestic and international growth. We are lowering our estimates from those published in the 2008 Nothing But Net. We are now expecting F’09 total global travel gross bookings growth of (4.7%) Y/Y. We are expecting global online gross bookings growth of 3.9% Y/Y vs. our prior estimate of 15.4%. Table 25: Global Online Travel Market Forecast $ in millions 2004 2005 2006 2007 2008E 2009E 2010E 2011E 2008-2011E CAGR Total Travel Spend 671,000.0 701,000.0 742,040.0 787,524.4 813,340.6 774,985.4 811,714.9 859,557.7 1.8% % online 19.1% 22.8% 28.1% 34.5% 40.5% 44.2% 49.9% 53.3% 9.5% Online Travel Spend 127,960.0 159,830.0 208,228.8 271,927.8 329,233.7 342,199.5 405,074.0 457,955.0 11.5% Total Travel Spend Growth 4.5% 5.9% 6.1% 3.3% -4.7% 4.7% 5.9% Online Travel Spend Growth 24.9% 30.3% 30.6% 21.1% 3.9% 18.4% 13.1% Source: Company reports, PhoCusWright, and J.P. Morgan estimates Domestic Growth Expected to Slow We expect a very challenging macroeconomic environment to impede both customer volume growth and pricing in 2H’08 and F’09. Furthermore, we believe online penetration increases will be less effective in offsetting total travel market weakness as penetration increases begin to flatten. We are now looking for F’09 US online travel market growth of 2% Y/Y, well below our prior estimate of 8% growth. We think the total US travel market will decline 5% Y/Y in F’09. This is the slowest growth rate since 2003, when the market declined 8% following the effects of the Sept. 11, 2001 terrorist attacks. Table 26: US Online Travel Market Forecast $ in millions 2004 2005 2006 2007 2008E 2009E 2010E 2011E 2008-2011E CAGR Total Travel Spend 233,000.0 253,000.0 267,000.0 282,000.0 290,460.0 275,937.0 289,733.9 304,220.5 1.5% % online 38.0% 41.0% 46.0% 51.0% 54.0% 58.0% 61.0% 63.0% 5.2% Online Travel Spend 88,540.0 103,730.0 122,820.0 143,820.0 156,848.4 160,043.5 176,737.6 191,658.9 6.8% Total Travel Spend Growth 8.6% 5.5% 5.6% 3.0% -5.0% 5.0% 5.0% Online Travel Spend Growth 17.2% 18.4% 17.1% 9.1% 2.0% 10.4% 8.4% Source: Company reports, PhoCusWright, and J.P. Morgan estimates 48 Global Equity Research 05 Januar y 2009 Imran Khan (1-212) 622-6693 imran.t.khan@jpmorgan.com OTAs Make US Volume Gains but Pay the Price in Discounts According to PhoCusWright data, online travel agencies (OTAs) started to benefit from a limited counter-cyclical lift in the second half of 2007 as consumers started to shop for online bargains as airlines continued to raise fares and reduce capacity. According to PhoCusWright’s CTTS10, 47% of online travelers cited online travel agencies as their primary shopping method in 2007 vs. 40% in 2006. However, we think these traffic volume gains made from supplier websites are mostly being offset by pricing discounts. Orbitz, Priceline, and Expedia have all offered promotions and discounts to attract new traffic and conversions. In fact, PhoCusWright data shows that OTAs are attracting more infrequent leisure travelers, who may take just a few trips per year and are more price sensitive. Meanwhile, suppliers are still the preferred destination for more frequent loyal leisure/unmanaged business travelers who may be less concerned with price and more interested in schedule, convenience, and booking with specific suppliers. Figure 22: Suppliers Accounted for 61% of the Online Market in 2007 Supplier Websites, 61% Online Travel Agencies, 39% Source: PhoCusWright data Advertising Revenues Are Becoming More Prominent With US competition from suppliers for gross bookings, we think many OTAs are turning to advertising as an additional means of monetizing the traffic coming to their sites. Expedia has launched several new online ad initiatives, and Orbitz and CheapTickets began using Kayak-powered search ads on their sites. Expedia launched PasswordAds, an advertising network that attempts to capitalize on downstream traffic. Expedia purchases advertising space on websites that consumers go to after they visit Expedia and serve ads that click directly to the advertiser's website. Orbitz links directly to suppliers and competitors from its search results display by offering a sponsored-link alongside requested travel search results. It appears that focus is shifting from transactions to more of an emphasis on the broader shopping and decision-making process. In our opinion, there is a large market for this. According to PhoCusWright, more than 80% of travelers who visited at least one website when planning a leisure trip read a traveler review, and 95% of those indicated the reviews were influential. Furthermore, we think travelers are even more likely to spend more time shopping and less time booking in the current economic environment, thus making this revenue stream even more attractive. 49 Global Equity Research 05 Januar y 2009 Imran Khan (1-212) 622-6693 imran.t.khan@jpmorgan.com Figure 23: Use and Influence of Social Media, 2008 Survey 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100 % Read a review of a product or serv ice Read a blog Posted a review Found reviews by trav elers to be somewhat influential, influential, or v ery influential Source: PhoCusWright Consumer Technology Survey Second Edition The International Opportunity Despite macroeconomic weakness impacting total travel spend, we think international markets will continue to benefit from increased penetration trends. According to recent PhoCusWright data, Europe has only a 32% online penetration for leisure/unmanaged business and corporate travel, which is still well below the US rate of 51%. This is a significant growth opportunity for online travel agencies, as the EU27 is more than 60% larger than the US in population size. Additionally, the Asia-Pacific region is only in the very early stages of online travel adoption with an estimated 15% online penetration rate. Thus, we think macroeconomic headwinds will be slightly offset by online adoption forces. Table 27: Europe Online Travel Market Forecast $ in millions 2004 2005 2006 2007 2008E 2009E 2010E 2011E 2008-2011E CAGR Total Travel Spend 260,000.0 263,000.0 284,040.0 307,524.4 322,900.6 297,068.6 311,922.0 336,875.8 1.4% % online 9.0% 15.0% 22.0% 32.0% 41.0% 45.0% 53.0% 57.0% 11.5% Online Travel Spend 23,400.0 39,450.0 62,488.8 98,407.8 132,389.3 133,680.9 165,318.7 192,019.2 13.1% Total Travel Spend Growth 1.2% 8.0% 8.3% 5.0% -8.0% 5.0% 8.0% Online Travel Spend Growth 68.6% 58.4% 57.5% 34.5% 1.0% 23.7% 16.2% Source: Company reports, PhoCusWright, and J.P. Morgan estimates Table 28: Asia-Pacific Online Travel Market Forecast $ in millions 2004 2005 2006 2007 2008E 2009E 2010E 2011E 2008-2011E CAGR Total Travel Spend 178,000.0 185,000.0 191,000.0 198,000.0 199,980.0 201,979.8 210,059.0 218,461.4 3.0% % online 9.0% 9.0% 12.0% 15.0% 20.0% 24.0% 30.0% 34.0% 19.1% Online Travel Spend 16,020.0 16,650.0 22,920.0 29,700.0 39,996.0 48,475.2 63,017.7 74,276.9 22.7% Total Travel Spend Growth 3.9% 3.2% 3.7% 1.0% 1.0% 4.0% 4.0% Online Travel Spend Growth 3.9% 37.7% 29.6% 34.7% 21.2% 30.0% 17.9% Source: Company reports, PhoCusWright, and J.P. Morgan estimates 50 Global Equity Research 05 Januar y 2009 Imran Khan (1-212) 622-6693 imran.t.khan@jpmorgan.com For F’09, we are projecting European online gross bookings growth of 1% Y/Y due to macroeconomic weakness and an estimated 6 points of negative impact from foreign currency weakness. We are projecting F’09 APAC online travel gross bookings growth of 21% Y/Y. Western Europe Growth Flattens Out… The UK was the largest total and online travel market in Europe in 2007, accounting for over 50B euros in gross bookings. It also had the highest online penetration rate of 40% (excludes corporate travel), according to PhoCusWright data. France and Germany were the second and third largest travel markets in Europe with online penetration rates of 23% and 20%, respectively. We think these markets will be less of a growth driver moving forward as penetration gains slow and euro weakness impacts earnings. Figure 24: Online Penetration of Key European Countries Leisure/Unmanaged Business Gross Bookings, 2007 billions euros 0 102030405060 Ireland Scandinavia Italy Spain Germany France UK Total Online Gross Bookings Total Gross Bookings 58% penetration 37% penetration 13% penetration 17% penetration 20% penetration 23% penetration 40% penetration Source: PhoCusWright reports and J.P. Morgan estimates …But Eastern Europe Has Attractive Penetration Growth Eastern Europe is at a much earlier stage of online adoption than its Western counterparts. We think countries in this region have the potential to post strong online travel growth despite macroeconomic headwinds. According to PhoCusWright data, online travel gross bookings increased 52% Y/Y in 2001 in the US, despite an overall travel market decline of 8% due to recession, terrorist attacks and travel fears. As such, we think Eastern European gross bookings may also continue to grow given similarly low penetration rates. We believe online travel agencies with a larger exposure to these regions will outperform the competition. . 31,972 43, 054 55 ,55 6 69,7 05 90 ,58 5 112 ,57 3 127,949 139,617 15. 5% ROW 9,440 13,216 18 ,50 2 25, 903 34,970 41,963 55 ,811 73,113 27.9% Total 182,622 2 35, 811 299,397. 106, 851 128,708 151 ,443 160 ,56 3 168,270 187, 958 207 ,50 5 8.9% Europe 52 ,430 72,690 96,631 123,698 152 ,53 7 140,740 170,942 198,200 9.1% Asia 31,972 43, 054 55 ,55 6