1、Cross Asset Research 18 July 2019 FOCUS Barclays Capital Inc.and/or one of its affiliates 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.
2、Investors should consider this report as only a single factor in making their investment decision.FOR ANALYST CERTIFICATION(S)PLEASE SEE PAGE 36.FOR IMPORTANT FIXED INCOME RESEARCH DISCLOSURES,PLEASE SEE PAGE 36.FOR IMPORTANT EQUITY RESEARCH DISCLOSURES,PLEASE SEE PAGE 37.Restricted-Internal U.S.Int
3、ernet Breaking Up Is Hard to Do Anti-trust bark louder than bite,but regulatory risks are real:Calls to break up the largest internet companies have become louder recently,amid a general rise in concerns about the potential market power these firms may have accumulated.We do not believe that attempt
4、s to break up these companies would be successful,nor would they mitigate the detrimental effects of market power.Given current antitrust law,the bar to prove consumer harm is quite high,and at a macro level,the business practices and organizational structure of companies such as Google,Amazon and F
5、acebook would be difficult to challenge,at least in the US.Any cases will likely take a decade or longer to play out;meanwhile,the companies would continue to operate.Further,the risk of targeting the wrong companies is real.Both media and retail have been singled out as industries becoming dominate
6、d by internet giants.Indeed,our Barclays Competitiveness Indicator(BCI)suggests that competition in media has fallen,likely due to the rise of digital media.However,according to our BCI,competition in retail is healthy,suggesting that economies of scale are driving consumer surplus and that breaking
7、 up Amazon would only further entrench other large retailers.Data portability is a risk:New regulations targeting technology companies are also being discussed.Approaches designed to keep dominant companies from abusing their positions could actually work in favor of incumbents by raising new barrie
8、rs to entry.Changing the ownership structure of consumer data could allow platforms to retain network effects,economies of scale and monetization opportunities while allowing competitors to leverage data collected on a given platform at de minimis cost.Much like the PSD2 regulations in banking,such
9、a move could minimize unintended consequences,with a higher chance of levelling the playing field and may present a challenge to todays winners.Where we could be wrong:The FTCs$5bn Facebook fine was a“slap on the wrist”relative to the companys$585bn market cap.Googles$9bn in EU fines similarly pale
10、in comparison to its$800bn market cap.However,similar to what occurred in banking from 2009 to today,at some stage the fines could become material enough to matter for stock prices.Banks have paid an aggregate$150bn in fines worldwide since 2009,and this has weighed meaningfully on certain names.BAC
11、 and JPM have paid 15-20%of their market cap in fines,respectively.If regulators were to fine large tech in that neighborhood as a percentage of market cap,investors would likely feel it.Who benefits from big tech regulation?If Google were required to roll back some of the bias on SERP,the primary b
12、eneficiaries would be TripAdvisor,B,Kayak,Skyscanner(CTRP),and Expedia in the travel categories.In mobile advertising,data portability would likely close the competitive advantage ARPU between Google and Facebooks smaller advertising/mobile companies.We see SNAP,TWTR and PINS as primary beneficiarie
13、s on data portability.INDUSTRY UPDATE Equity Research Ross Sandler+1 415 263 4470 BCI,US Deepak Mathivanan+1 415 274 5351 BCI,US FICC Research Jeffrey Meli*+1 212 412 2127 BCI,US James K Martin*+1 212 412 2345 BCI,US *This author is a member of the Fixed Income,Currencies and Commodities Research de
14、partment and is not an equity research analyst.Barclays|U.S.Internet 18 July 2019 2 CONTENTS FICC RESEARCH.3 ANTITRUST IS UNLIKELY,BUT REGULATORS HAVE OPTIONS.3 EQUITY RESEARCH.7 ASSESSING THE REGULATORY RISKS.7 Alphabet(GOOGL,OW,PT:$1,315).8 Possible Changes to Business Practices.9 Googles Anti-Com
15、petitive Behaviour to Date(according to the EU).12 Google US Market Share Material for Search Only.13 Amazon(AMZN,OW,PT:$2,050).15 Possible Business Model Changes.16 Facebook(FB,OW,PT:$240).19 Possible Business Model Changes.19 Potential Beneficiaries of Antitrust Enforcements.22 What If Politicians
16、 Get Their Way?The“Break Up”Scenario.27 GOOGLs Sum-of-the-Parts Valuation Suggests+18%Upside.27 AMZNs Sum-of-the-Parts Valuation Suggests Minimal Upside.31 FBs Sum-of-the-Parts Valuation Suggests Potential Downside.32 Barclays|U.S.Internet 18 July 2019 3 FICC RESEARCH Antitrust is unlikely,but regul
17、ators have options Concerns that the largest technology companies have accumulated too much market power have continued to gather steam,with both policy makers and politicians offering commentary on the subject recently.Assistant Attorney General Makan Delrahim sounded a more hawkish tone about anti
18、trust in the digital age at the Antitrust New Frontiers Conference.“While antitrust is not a panacea for every policy challenge presented by the digital market,the Antitrust Division will not shrink from the critical work of investigating and challenging anticompetitive conduct and transactions wher
19、e justified,”he said.1 The pace of hearings about the influence of technology companies has increased,with concerns being raised from both sides of the political aisle.We expect the focus on this issue to accelerate,driven in part by the political cycle.Several aspiring presidential candidates have
20、released commentary or proposals,most notably advocating breaking up large technology companies,reversing past acquisitions,or prohibiting certain business practices from being performed by the same company.2 More importantly for investors,equity markets have begun to take notice.Earlier this year,t
21、he prices of FB,AMZN,and GOOGL fell 5-8%on news that various federal agencies were intensifying their scrutiny of market power.Although stocks were fast to recover these losses,investors are clearly alert to signs that companies will come under pressure.Of the FAANG equities,only Amazon has outperfo
22、rmed the S&P over the past twelve months in price terms(Apple performed roughly in-line),after several consecutive years of meaningful outperformance from these companies.As we detailed in Assessing the Elephants in the Room,the presence of large powerful firms does not always mean that an industry
23、is non-competitive.Depending on which sector you focus on,the emergence of these large technology companies could be the result of a highly competitive industry where more of the rewards go to the most efficient producer(“winner-take-all”).Our Barclays Competitiveness Indicator(BCI)suggests that in
24、the retail sector,the accumulation of market share by Amazon(as well as WalMart)has not been detrimental to the competitive dynamics in the industry(Figure 1).Despite higher concentration and the rise of e-commerce,companies continue to invest in employees and new projects;in turn,employees continue
25、 to take on new jobs and leave old ones at the same pace as they did 15 years ago.Simply because Amazon is putting some retailers out of business does not mean that competition is dead in that sector.While this may not shield Amazon from regulatory scrutiny,it at least provides a defense for its bus
26、iness model.Conversely,the BCI of the media sector suggests the opposite that the level of competition has declined as concentration has risen(Figure 2).Although we consider the entire media landscape,we see digital media,which has taken share from traditional media,as the driver of anticompetitive
27、forces.Traditional media outlets appear competitive,with new entrants forcing incumbents to innovate their business models.But this is a shrinking portion of the media pie.Digital media is characterized by a small number of large firms with significant market share.This growing part of the market is
28、 more likely responsible for our aggregate findings about media.1 https:/www.justice.gov/opa/speech/assistant-attorney-general-makan-delrahim-delivers-remarks-antitrust-new-frontiers 2 For example,Elizabeth Warren has proposed prohibiting platform companies from owning both the platforms and partici
29、pants on that platform:https:/ Jeffrey Meli+1 212 412 2127 BCI,US James K Martin+1 212 412 2345 BCI,US Barclays|U.S.Internet 18 July 2019 4 FIGURE 1 Retail HHI versus Barclays Competitiveness Indicator FIGURE 2 Media HHI versus Barclays Competitiveness Indicator Source:Compustat,Barclays Research So
30、urce:Compustat,Barclays Research The concerns about the potential ways digital media companies can abuse their position are wide-ranging,but several themes are emerging:1.Cannibalization of old-line media revenues,along with dominance in their own space,per the recent hearing on Google and Facebook
31、news revenue.For example,GOOGL generates substantial ad revenue when users search for content,which potentially reduces the revenue available for the content creators themselves.2.Anti-competitive behavior and category abuse by GOOGL by trying to capture certain verticals such as shopping,travel,loc
32、al advertising,and jobs:Companies such TRIP,YELP,BKNG,ANGI,were affected when GOOGL took over the search engine optimization(SEO)space.3.FB copying features from SNAP,PINS,MTCH,etc.and scaling quickly using cross-promotion to its user base.Regardless of these concerns in particular,or of the larger
33、allegations of market power being leveled at the largest tech companies,it is important to consider the ramifications they will face in their businesses should any of the proposed remedies to market power become a reality.We examine three possible approaches;for each we assess the macro and micro im
34、plications for market power.1.Traditional application of antitrust regulation,with a particular emphasis on breaking up(or attempting to break up)the large technology firms.2.Legislation designed to keep dominant companies from abusing their positions.3.Pro-competitive regulation designed to alter t
35、he ownership structure of consumer digital data.Scenario 1:Break em up Breaking up the large tech companies has been one of the most commonly cited remedies to perceived market power,and proposals along these lines have received notable airtime on the campaign trail in the US.For example,in her Marc
36、h Medium post,Elizabeth Warren specifically cited Amazons merger with Whole Foods and Zappos,Facebooks WhatsApp and Instagram acquisitions,and Googles acquisitions of Waze,Nest,and DoubleClick as potential unwinds.In an even more extreme scenario,the list of possible separations could also be expand
37、ed to include YouTube and AWS,as well as other entities.200220240260280300320340360380400-2.5-2.0-1.5-1.0-0.50.00.51.01.52.02.51998199920002001200220032004200520062007200820092010201120122013201420152016Index(10-10,000)Indicator UnitsRetail BCIRetail HHI(RHS)3004005006007008009001,000-2.5-2.0-1.5-1.
38、0-0.50.00.51.01.52.02.51998199920002001200220032004200520062007200820092010201120122013201420152016Index(10-10,000)Indicator UnitsMedia BCIMedia HHI(RHS)Barclays|U.S.Internet 18 July 2019 5 This represents a fairly traditional antitrust approach to addressing market power:breaking up companies and/o
39、r constraining future acquisitions limits companies operational reach,and therefore their ability to establish or maintain a dominant position.That said,in our view,this approach is unlikely to resolve the underlying issue,for several reasons.First,in big picture terms,the business practices of thes
40、e companies could be difficult to rule against under current US antitrust law/interpretation.For example,case law supports activities such as bundling products and services.3 Vertical integration has been largely accepted in the US for decades and is very unlikely to be successfully challenged,parti
41、cularly retroactively.This is not to say that at a more micro level the large tech companies have never engaged in activity that has raised antitrust concerns.There are such examples we review the notable ones in the company specific sections below some of which have resulted in fines and/or agreeme
42、nts to modify behavior.Instead,we are suggesting that the types of consumer harm being generated by these companies,if they exist,are unlikely to fit in the typical lens currently used by antitrust regulators,which is dominant firms raising prices and restricting output to maximize profits.Our BCI c
43、aptures more aggregated effects,such as declining business dynamism and lower investment,which imposes long-term costs for consumers,for example,through reduced innovation or less choice.But these costs are difficult or impossible to measure and to ascribe to any one company.Further,any company targ
44、eted for breakup would certainly challenge this action.Given the full coffers and sizeable influence of these firms,such an attempt would consume a substantial amount of time and resources,and there is no guarantee of eventual success.In the meantime,these companies will be able to continue to opera
45、te under the status quo,with new entrants discouraged and investment and labor share remaining low.Another concern is the risk that authorities might target the wrong companies.As we demonstrated in Assessing the Elephants in the Room,we believe that the increased market share of Amazon in the retai
46、l sector has not led to a decline in competition.Breaking up Amazon could serve no other purpose than entrenching other incumbents,such as WalMart.If the underlying drivers of higher concentration in a particular sector are economies of scale and network effects that are positively correlated with l
47、arger firms,then breaking up the largest firms would actually be counterproductive for consumers,to whom at least some of these benefits accrue.This risk is accentuated since firms that rapidly acquire market share at the expense of competitors may make for expedient targets.Even in sectors such as
48、media,where we find competition has deteriorated,there could be negative unintended consequences to breaking up large companies.Part of the reason that competition has remained so strong in retail is that smaller newcomers have leveraged low cost marketing opportunities provided to them by social me
49、dia platforms.Limiting the reach of these platforms or their ability to identify potential customers could inhibit competition in the retail space.In general,the potential for unintended consequences from such targeted action is high.As a result,we believe that breaking up the largest firms is both
50、difficult in practice and conceptually unlikely to be an effective remedy for the concerns about market power.We are therefore less concerned about this outcome in terms of risks to the underlying equity valuations.That is not to say that“break em up”will not continue to generate headlines it will b