Loading...
The URL can be used to link to this page
Your browser does not support the video tag.
Home
My WebLink
About
Supplemental - 5 Drivers of 5G Value received from Jonathan Waller
Morgan Stanley I RESEARCH • November 18, 2018 09:00 PM GMT Telecoms Primer 5 Drivers of 5G Value GLOBAL FOUNDATION r .l7 I_ nr I' C>t �; <- -t Ko ; -. .a «n i a�r � , n4\/ u'C'e r f E't l'��'s ��. _ C. (.T;'� .'r .,.l. , t.. t.l -3I If . ...,f:) .Y. t } �e.'. ,Yc Ule IC. /PI ti_ !?.;s C, 4.'IJi =y 0 C iii- E >'(': "i •Morgan Stanley does and seeks to do business with companies covered in Morgan Stanley Research. Asa result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of Morgan Stanley Research. Investors should consider Morgan Stanley Research as only a single factor in making their Investment decision. For analyst certification and other important disclosures, refer to the Disclosure Section, located at the end of this report. += Analysts employed by non U.S. affiliates are not registered with F1NRA, may not be associated persons of the member and may not be subject to NASD /NYSE restrictions on communications with a subject company, public appearances and trading securities held by a research analyst account. Morgan Stanley I RESEARCH • i ontri utors MGRGAN STANLEY ASIA LIMITED. Gary Yu Equhy Analyst +85228186918 Gary.VU@mwganshnky.cwn MGRGAN STANLEY 6 W. LLC Simon Flannery E9ulty Analyst 81212761-6432 Simon.Flanmy @mwganslanky.rom MORGAN STANLEY MIffGSEWRITIESCO.. LTD,. MORGAN STANLEY ASIA LIMITED. Tetsuro Tsusaka, CFA Yunchen Tsai 11 Ew NnNst Ean M.Nat .81368368412 88522848 5636 TnwoTSUSake @mwgenzlaNeyrnuf8larn Yunchen Tsai @mwgmstanky.mm E VoRGAN STANLEY 6 Cn LLC Spencer Gantsoudes, CFA Resaemh ASSwiate +1212 761 5374 Spennw.GaniswEes @ mwganslanley.cwn GLOBAL FOUNDATION kGiUASIMIEY8c NI A1pNNtiffGA3 C� Seyon Park EwIN k-w 88223994936 Sry F%xk@mxganstanley.mn WMM STANLEY ASIA LIMREG+ Yang Liu Ep Awls, +8522239-1911 Yef14L1U@mMganstBnEy cdn W%R STANLEY 8 CO LLC Alexis R Roper R.z Assatlate +12127614590 NniaROpA@mwganslanley.com ' i STMLEYMIALPMITEO Sara Wang Resevch Aeaoclate to +85222391230 Swa Wang@mwgwstarky.cwn WRGAN STANLEY 8 W. LLC Landon Park ReseaMAss«'ute .1212761fi363 LandwRa�k@morganstanl icon E VoRGAN STANLEY 6 Cn LLC Spencer Gantsoudes, CFA Resaemh ASSwiate +1212 761 5374 Spennw.GaniswEes @ mwganslanley.cwn GLOBAL FOUNDATION kGiUASIMIEY8c NI A1pNNtiffGA3 C� Seyon Park EwIN k-w 88223994936 Sry F%xk@mxganstanley.mn WMM STANLEY ASIA LIMREG+ Yang Liu Ep Awls, +8522239-1911 Yef14L1U@mMganstBnEy cdn W%R STANLEY 8 CO LLC Alexis R Roper R.z Assatlate +12127614590 NniaROpA@mwganslanley.com Morgan Stanley I RESEARCH ® Telecoms Primer 5 Drivers of 5G Value GLOBAL FOUNDATION 7C IS cd e :`- 7f''E ent de[ ;e r tOiC )S Ir' C-li 1, f i .; ?.: : :1 .cF 1i1 We r I dprlt_ir hve :,he(' t .. ,�.- d ive v ki ", a c �i", ( - -e'X 'v�, ac- 'd -�,� r; lal° �:1� .oncr_r r s, ��NC mE,t t., slt -c' rev._, ue t,.T=.;or.:�:litv tc,. r _aSPc haP_ !.; SI r 10 F-i D ' . I R P P U 5G will be the main share price driver for global telcos over the next decade: Unlike the 2G/3G/4G cycles, SG will create new use cases (fixed wireless and the Internet of Things), open up new mar- kets (enterprise /industrial), trigger major changes in market struc- ture (consolidation, new entrants, network sharing), and offer different pricing structures (tiered by service /performance vs. ®usage). 5G will likely take longer to reach maturity and will emerge as the most important value driver for global telcos over the next decade. We project SG capex of US$872bn — 1.7x 46 capex — driven by China: We project SG capex of US$421bn in China, US$265bn in the US, US$1291bn inJapan, and US$58bn in Korea in 2019- 30.These esti- mates imply that 5G capex will be 1.3 -14x 4G capex in the US, Korea and Japan but 2.2x 4G capex in China. However, this will be spread over a longer period of 10 -12 years for 5G vs. only 6 -8 years for 4G. In China, we expect high SG capex to deteriorate ROIC in the early years, particularly for China Mobile, and the first 56 capex guidance, potentially in March 2019, could be a negative catalyst. In the US, we estimate the Big 4 carriers could spend US$261an on spectrum pur- chases over the next three years. Incremental annual revenue of US$156bn for seven 5G use cases by 2030,40% of current mobile service revenue. Including fixed wireless, cloud gaming, autonomous vehicles, manufacturing auto- mation, connected surveillance cameras, connected drones, and • O industryview Greater China Telecoms and Media - In -Line S. Korea Telecoms, Media & Internet - In -Line Japan - Telecommunications - In -Line North America - Telecom Services - In -Line remote healthcare services, we estimate incrementa L annua l revenue of US$156bn for seven SG use cases in the four markets by 2030, which represents 40% of current mobile service revenue. In Asia, we believe operators with dominant enterprise /industrial market share are weLLpositioned totap these sizeable incremental revenue oppor- tunities, including China Telecom and China Unicorn in China, KTCorp in Korea, and Docomo and KDDI in Japan. Return of pricing power —an opportunity for telcos that is not in the price: GLobaI telcos have genera Llyfailed to monetize 3G /4G, but we see 5G as a rare opportunity for the industry to regain pricing power given: l) Less regulatory and competitive pressure in the enter- prise /industrial segment vs. consumer; 2) better market structure with fewer competing networks as a result of network sharing or even mergers; and 3) network slicing, enabling operators to charge premium prices for superior, tailored services. With MSCI World Telcos trading at a 2019e EV /EBITDA of 6.Sx (vs. an average 6.7x since 2013), we believe such potential upside is not in the price. Morgan Stanley I RESEARCH i Team List LATIN AMERICA TELECOM ® Cesar Medina Ernesto Gonzalez GLOBAL FOUNDATION ASIA PACIFIC TELECOM NORTH AMERICA TELECOM Greater China Simon Flannery Gary Yu Spencer Gantsoudes Yang Liu Lisa Lam Sara Wang Landon Park South Korea Alexis Roper Seyon Park ASEAN EUROPE TELECOM Mark Goodridge Emmet Kelly ® Da Wei Lee Luis Prota India Terence Tsui Parag Gupta Bhavish Patel Amruta Pabalkar Nicole Florack Australia EEMEA TELECOM Andrew McLeod Madhvendra Singh Elise Kennedy Andrea Ferraz Joseph Michael David Gardner JAPAN TELECOM Miriam Adisa Tetsuro Tsusaka LATIN AMERICA TELECOM ® Cesar Medina Ernesto Gonzalez GLOBAL FOUNDATION Morgan Stanley I RESEARCH GLOBAL FOUNDATION rn L-A • 4 Team List 6 Key Charts 9 Executive Summary 11 Theme 1: 5G Capex is 1.7x 4G, Driven by China 18 Theme 2: Incremental Annual Revenue of US $1561bn for Seven SG Use Cases by 2030 25 Theme 3: 5G Could Trigger Major Changes in Market Structure 28 Theme 4: Return of Pricing Power — A Rare Opportunity for Telcos 35 Theme 5: Passive Infrastructure Remains Key 40 Global Telco Comps 41 Appendix 1 — China 45 Appendix 2 — US 49 Appendix 3 — South Korea 50 Appendix 4 —Japan 53 Appendix 5 — Spectrum L Morgan Stanley I easeeeae Key Charts Exhibit 1: US. 4,95OMHz of mmWave spectrum to come to market in 2018 -19 MHz Offered 3.500 _... ..._. ._._.3400 _... 3,000 2.500 2.000 1,500 1.000 500 0 tlov to 'C's 1H 2099 2H2019 Paand cal =s _,._... 21, GH_ 3139S�GNZ i27.42GHz : "5.53Hz: Source. Morgan Statle,Reeearoh estlrtates. FCC ExhlbR 3: 2030 incremental annual revenue from seven 513 use cases vs. 2018 mobile service revenue USD bn 2030E SG incremental annual revenue a 2010E mobile seN e`avenue Y China us Smith Korea Japan Source; Canna, Morgan Stanley Research estimates GLOBAL FOUNDATION Exhibit 2: Capex comparison: 4G investment through 2018 vs, our projections for 5G (including spectrum) USDbn ■4G 4 S a50 421 Ora 350 SO0 zao acG 15o ;C4 50 0 Chma Us So,h Korea Japan Space Morgan Stanley R.s innn,vtes, company data Hdtibiit 4: 2030 incremental annual revenue by application AV S1bn Drone Mn 5% 3% USD bn Cv3O 40 450 300 150 Q Source'. Morgan Stanley Research inat. Korea Japan ry r _ ry O i Sf A d h q' 3 a 3 a 3 a 2n5 ¢E [ C E o • r n E _ lV p S N � 3 Source Company daa Morgan Stanley Research earnates ry r _ ry O i Sf A d h q' 3 a 3 a 3 a 2n5 ¢E [ n E o r re ry r _ ry O i Sf A d h q' 3 a 3 a 3 a 2n5 ¢E [ n E o Morgan Stanley I aa4a.a =a • Exhibit 6: 5G- related infrastructure plays GLOBAL FOUNDATION Source Morgan Santry Research Exhibit 7: 5G timetable with key events in the four markets 1016 Category Tower /small cells China Tower China Data centers 21 Vianet, Sinnet, GDS Construction China Communication Services Tower /small cells American Tower, Crown Castle, SBA, UNIT Fiber Zayo, Boingo US applia0on Data centers Equinix, Digital Realty, CyrusOne, CTS, CoreSite Spectrum holder Dish Network, Intelsat, Globalstar System Huawei (Private), ZTE, FiberHOme, Ericsson, Nokia, Juniper, Ciena • Accelink, Innolight, Broadcom, Finisar, Lumentum, Accelink, Oclaro, Global Transceiver Truelight, Luxnet, APAC, PCI, O -net, Applied, Oplink Equipment aAOabon service taurrpr, Corning, Furukawa, Sumitomo, Fujikura, YOFC, HengTong, TongDing, Fiber Zhongtian Source Morgan Santry Research Exhibit 7: 5G timetable with key events in the four markets 1016 2010 3GPP Re ase 15 ($A venronl NeMak and 3018 applia0on hat • Spectrum Fired- raless SpecWmaudpn 4019 aAOabon service taurrpr, 101(23GHr) 5G aQ.ianant Spectrum auction 1019 system b Gtr 102 (24GHa) pn� mmnardial 5G ikansin9 188q 2019 iwpyl SG handsel C-band FCC b be ready Report and Order 3019 Spectrum eutlion 4019 3GPP Rebore 10 Mc-19 (37, 39, OGH., 1020 Sp nn -union (CSRS) 2020 Cmmwrciel �dt 3020 4020 Source Morgan Santry Research • Pre-Mndand and O Wnier Gpmpk Spectrum A..t. Commmcial lnal of SG Carvnertlai Iaunen 5G NeM * and apphwion e!perimeMa Spagrun aaoca0on (3512, 4 H.9, 27.529.5 G" Commerpa @anon of 5G, eMBB sernce launch Showcase SG b Summer 0ympc Morgan Stanley I .r.. - • ExhibR 8: Master 5G comparison table for the four markets Mid band 3.33.6 & 4.8- 5.OGHz; maybe 2.6GHz Spectrum mmWave 24.75 -27.5; 3742.SGHz Low band NA Spectrum allocation timing Spectrum cost(USSbn) 6G capes (USSbn) Expected service launch Focused application in early stage Potential market structure changes Infrastructure plays Potential Incremental revenue (US$ b i) • as % telecom service revenue as % mobOe service revenue ..curre Morgan STone R?,rct. • Late 2018 NA 421 2020 IoT Network sharing: CU and CT Tower, Data Centers. Conslrucbon Vendor 60.2 32% 45% 3.54.2GHz, maybe 2.5GHz', 24, 28, 37, 39, 47GHz 600MHz mm Wave in 4018 and 19; Mid trend will be later 26.0 239 4018 Fixed- wireless Merge: T- Mobile and Sprint Tower, Data Centers, Fiber, Spectrum Holders 76.3 23% 42% 3.42- 3.7GHz 26.5- 28.9GHz NA Jun -18 74 50 1019 IoT NA NA 6.8 21% 32% GLOBALFOUNDATION u 3.64.2, 4.4- 4.9GHz 27.6- 29.5GHz NA Ils1E NA 129 4019 IoT New entrant Rakuten NA 12.2 12% 21% Related Reports: Global Technology: Learning to Ride a 5G Cycle (Oct 15, 2017) Auto & Shared Mobitity. Global Telcos. Tech, SRI: Autonomous Cars & Telcas: We're eonna need a New (5G) Pipe (May 7, 2018 North America Tetecom Services & CabLe/SateLLlte: SG Spectrum Primer (Nov 12.2018) China Mobile: A Painful SG )ourney Ahead Out 4, 2017) Verizon Communications; Lessons from Day 1 of the 5G Home Launch (Oct 2,201a) Morgan Stanley I AE.E. 40 Executive Summary SG is a key investment debate for global telcos, particularly for leading markets Like China, the US, Korea, and japan.In this SG primer, we identify five themes that will likely emerge as key share price drivers for telcos, both around major events (e.g. spectrum alloca- tion/auctions, licensing, and provision of capex guidance) in the near term and as the 515 cycle evolves in the next decade. To address a major market concern, we try to quantify SG capex, including a bull - bear range based on different scenarios. We take our first attempt to size the revenue opportunity for 5G, focusing only on seven (out of many more) use cases. We also highlight the potential return of pricing power as a rare opportunity for telcos that is not yet in the price. Theme is 56 apex of Us$872bn —1.7x 4G apes —sd0 be driven by China. including spectrum, base transceiver stations (BTS), trans- mission and towers, we project SG capex of US$421bn in China, US$265bn in the US, US$129bn in Japan, and US$58bn in Korea in •2019 -30. These estimates imply that SG capex will be 1.11.4x 4G capex in the US, Korea and Japan but 2.2x 4G capex in China, although this will be spread over a Longer period of 10 -12 years for SG vs. only 6-8 years for 4G. In China, market concern regarding SG capex risk appears to have subsided after recent newsflow around 2.6GHz spectrum, but we expect high 5G capex to deteriorate ROIC in early years, particularly for China Mobile, and the first 5G capex guidance, which may come in March 2019, could be a negative catalyst. In the US, we estimate that the Big4 carriers could spend US$26bn on spec- trum purchases over the next three years. Given uncertainties around spectrum, coverage, timing, technology, network sharing, and use cases, our scenario analysis derives a buLL -bear range of US$322 -576bn for China, US$201 -340bn for the US, US$89 -162bn for Japan, and US$51 -69bn for Korea. Theme 2: Incrementalannual revenue of US$156bn for seven SG use aces by 2030,40% of current mobile service revenue. These include fixed wireless (US$18bn), cloud gaming (US$9bn), autono- mous vehicles (US$7bn), manufacturing automation (US$64bn), connected surveillance cameras (US$20bn), connected drones (US$4bn), and remote healthcare services (US$32bn). In the US, fixed wireless will likely be the early use case for 5G, and Verizon esti- mates incremental revenue of US$81bn for fixed wireless (44% of •current mobile service revenues of US$183bn). In Asia, operators will Likely focus on the Internet of Things (IoT) for the enterpriserndus- trial segment. We believe operators with dominant enterprise/indus- trial market share are well positioned to tap these sizeable GLOBAL FOUNDATION incremental revenue opportunities, including China Telecom and China Unicom in China, KT Corp in Korea, and Docomo and KDDI in Japan. Theme 3: SG could trigger mayor changes in market structure. In China, Bloomberg reported in September that the government is considering merging CU and CT to spur SG development. Even without a merger, we expect more extensive active network sharing between the two companies in the SG era. In the US, T- Mobile and Sprint combined have the strongest 5G spectrum portfolio, with a good combination of high/mid/low band, although whether competi- tion will ease after the merger remains uncertain. In Japan, new entrant Rakuten Mobile Networks was allocated the 1.7GHz spec- trum in April 2018 for 4G service (carrier service expected to com- mence in October 2019) and has officially entered the telecom carrier business (changing the market landscape from a three MNO market to four). With SG spectrum allocation expected at end -March 2019 for Japan, there could be more competition around 515 services if Rakuten is allocated SG bandwidth. In Korea, we expect the three - player market to be maintained in the SG era, with all operators rolling out separate networks. Theme 4: Return of pricing power — a rare opportunity for telcos. 5G provides a rare opportunity for a potential return of power pricing to the telcos. A bulk of the 5G incremental revenue opportunity Lies in the enterprise segment, and the 82B market tends to be less sensi- tive to price competition and faces less regulatory risk than B2C, par- ticularly in China and Korea. We also expect fewer competing networks in China and the US due to mergers and network sharing. In contrast to the pre -515 era, when net neutrality made it structurally difficult to package data, 5G will allow telcos to sell 'slices' of the network, at different parameters, customized for specific end appb- cations. This is a fundamental shift away from net neutrality-based economics which we believe has potential to create completely new ways to price data service. Theme 5: Passive Infrastructure remains key. At a high level, we expect SG to Leverage many of the same passive infrastructure assets that comprise a modern 413 network: towers, small cells, fiber, and data centers. Overall, we do expect the mix of spend to shift at the margin with a greater emphasis on fiber and small cells, particularly in denser urban areas. However, the broad infrastructure Landscape should see an overall benefit as many 5G use cases require significant coverage in suburban and rural markets as well. Morgan Stanley I • Exhibit 9: 5 Drivers of 5G Value • • Source. Company data, Gantt. E = Morgan Sanley Researtlt estnnates USD M -4G 450 s2`. 400 350 300 26.5 250 200 ..90 ,0 -,W 29 i 100 Sit 50 4 o Ghma U ^a Swum Kxea Japan I, ♦vim. bent. sun sx s% use bn oeTX1° a% 600 44 45C AYalrkiYNY1 3:If: lasaxrML MYW a� Theme 3, F AY Ict iDaut • 4•Keu St.Q Ui R: Sof16ank - Poe wlxtikl perry 1 VYatel ne1wvik m e ¢Mm We see the bulk of the 5G sacramental revenue opportunity lies in the enterprise Theme 4 segment, and the B20 market lends to be less seretkve to price competition and Return ofpflcing power, a rare also faces less regulatory risk than 132C, particularly in China and Korea. We also opportunity for telcos erect fewer competing networks in China and the US due to network sharing or mergers. giving more leverage to the operators. Towmismall calls China Tower China OMs cement Mars( Simret GOS Construction China COmmunkaaon Samna Towarturall tale Arroncmt Tovar, Crest Cexe, SBA, UNIT Theme 5 us Fiber Zayo, Bortgo Passive infrastructure key Data centers , Digiel Reny, GTB, LoraSks remains Did, N Ghealstur Spedmm Adder Diah NeMOrk, Intalsat, Gtobaleer System Huawn(Privae FiW (Private), rHOrre. Ericsson, Nokia. Junipaf, Clans GbWI Transceiver Acceknk, Innokght, Bmadcom, Feisar, Lumentum, ACCM KOC4m. Epgment TmahghtLuanet. APAG PCI. O-wt. Appled, Oplink Fiber Coming, Furukav a Sumitomo. Fullkwa, YOFC, HegTo q. TongDing, .......... . .. .._.___._.__._._. Aorptian _ .................. _._._.._ -- _.____.._ ... _. Source. Company data, Gantt. E = Morgan Sanley Researtlt estnnates Morgan Stanley I - -..... 0 Theme 1: 5G Capex is Wx 4G, Driven • LJ by China Market view: Concerns regarding China's ultra -high capex burden started subsiding at the end of August 2018, after market reports that China Mobile may be able to use the 2.6GHz band to deploy SG to save some capex and accelerate 5G development. Our view. We forecast that total SG capex will be 1.3 -1.4x 4G capex in the US, Korea and Japan but 2.2x 4G capex in China, although this will be spread over a Longer period of 10 -12 years for SG vs. only 6 -8 years for 4G. In China, market concern regarding SG capex risk appears to have subsided after recent newsflow around 2.6GHz spectrum, but we expect high SG capex to negatively affect ROTC in early years, particularly for China Mobile, and the first 5G capex guidance, which may come in March 2019, could be a negative catalyst. Our proprietary analysis: Our bottom -up SG capex model includes spectrum, base transceiver stations (BTS), transmission and tower spending in 2019.30. Under our base case, we project SG capex of US$421bn (20% of mobile revenue) for China, US$265bn (10% of mobile revenue) for US, US$58bn (17% of mobile revenue) for Korea and US$129bn (15% of mobile revenue) for Japan. Given uncertainties around spectrum, coverage, timing, equipment cost, network sharing, and use cases, our scenario analysis derives a bull -bear range of US$322- 576bn for China, US$2O1.340bn for the us, u5$51 -69bn for Korea, and US$89.162bn for Japan. Spectrum Spectrum is a key determinant of SG capex, both because of the likely high costto acquire such a scarce resource globally and the implications for network deployment. In order to fulfill the eight performance requirements identified by the International GLOBAL FOUNDATION Telecommunication Union ( Exhibit 10 ), 5G requires much wider fre- quency spectrum than 26/36/46, and most industry participants suggest a minimum of 1OOMHz contiguous band per operator. This creates a dilemma: there is insufficient unoccupied spectrum in the low frequency bands, but the abundant idle spectrum in the high fre- quency bands suffers from narrow radio coverage range and poor indoor propagation. Global harmonization of 5G spectrum: To promote the global har- monization of spectrum, the ITU, the United Nations agency for infor- mation and communication technologies (ICTs), organizes the World Radiocommunication Conference (WRC) every three or four years to review the Radio Regulations, the international treaty governing the use of spectrum. The 3rd Generation Partnership Project (3GPP), an international industry group uniting seven telecoms standard devel- opment organizations from Asia, Europe and North America, defines 5G standards via a series of releases, notably Release 15 frozen injure 2018 and Release 16 expected to be frozen in late 2019. Both the WRC held in 2015 (WRC -15) and 3GPP support the use of 3.3- 3.8GHz band (C -band) for 5G, and the next W RC to be held in 2019 (WRC -19) aims to identify spectrum above 24GHz band (mm Wave) for SG. Mutt! -Layer spectrum to meet different requirements: In November 2017, Huawei published a White Paper: SG Spectrum Public Policy Position, to promote a globally harmonized spectrum framework for 5G. Specifically, Huawei proposed a multi -layer spec- trum approach to address a wide range of usage scenarios and requirements: • Coverage and capacity Layer using 2-6GHz offers the best compro- mise between capacity and coverage • Coverage layer using below 2GHz provides wide -area and deep indoor coverage • Super data layer using above 6GHz addresses specific use cases requiring extremely high data rates Morgan Stanley I wE.-- • Exhibit 10: 5G network performance requirements Peak data rate ?Gbitls) Area traffic capacity WVT s/'O Network enerEy efficiency Connection density (devices/ km') So M Internatmnai TeiecanmunicatiW Union Exhibit 11: • Huawel: Mufti -layer frequencies approach for 5G usage scenarios „nn��e r+nawe� • High Frequencies Super Data Layer Addressing specific use cases requiring extremely high data rates Medium Frequencies Coverage & Capacity Layer Best compromise between capacity and coverage Low Frequencies Coverage Layer Wide-area and deep indoor coverage User expericrced data rate(Mbit/s) 00 Spectrum efficiency LRtencytrns) Above 6 GHz 800 MHz assignments (contiguous) 2 -6GHz 100 MHz assignments (contiguous) Below 2 GHz (up to 20 MHz paired ! unpaired) GLOBAL FOUNDATION Morgan Stanley I .r,e•.= • Exhibit 12: Carrier bandwidth: Increasing demand for spectrum • 213 200kHz 1G 25130kHz Source HKT. H.,, Exhibit 13: General considerations of spectrum for 5G 4G !OMHz 51131 100MHz in <6GHz/ 80OMHz in mm Wave Low - frequency bands below 6GHz are 2015 High - frequency bandswithin 6- 1GGGHz 2019 always necessary for IMT ,/ might be introduced in 2019 and beyond scarce iTu Exhibit 14: 5G will require spectrum across different frequencies t g 8 Source T -MO ,ie Reliable Low - Latency Comms •Localized, reliable high capacity service Enhanced Mobile Broadband -Consistent capacity and speed experience ® N Massive IoT • Nationwide coverage reliability GLOBAL FOUNDATION Morgan Stanley I aeaa. GLGBALFOUNDATION • Exhibit 15: Coverage range and spectrum frequency 650 MHz Z fi. °,J aq in 700 MNZ 3,400 so; m / 2.4 GHz lUnlicensWl 12 e4, me r f 2GRz sq in,YYSj ^...,. 5.9 GMz 500 sQ m � (Unilcensad) 2.5 GRz(BRI EBa) 300511 en j � 6RiNlruhrWarS 3.5 GHz lti9tMf t.iei J l� swo Coll 30 sq. no ? � 003 sq. as ,. Source :Vantage%m%Specnum v.,ange comparison (Example at parminnl p pa FCC rulea assumed pa band) Exhibit 16: 5G spectrum preparation and timetable Sub-2 GHz NA "NA 700MHz is used by analog N 3.3�3.6GHz; Allocation is expected in 2018; 3.3-3.4GHz for indoor only; regulator indicates a is China is 2-e GHz 4.BS.OGHZ 3.63.7GHz likely to allocate 3.63.7GFtr l0 5G Above 6 GHz NA 24.75- 27.5GHzi Finished consultation; low priority In allocation 37.042.5GHz Sub-2 GHz T Mobile will potentially deploy 5G on its existing 600MHz spectrum 2-6 GHz Sprint will potentially deploy 53 on its existing 2.5GHz spectrum U.S 26GHz and 24 GHz auction will be In November 2018 (confirmed) 37 GHz+ 39 ADOVe 6 GHz 28, 24 GHz 37 GHz + 39 GHz+ 47 GHz GHz + 47 GHz auction is expected to happen in 2019 Sub-2 GHz NA NA South 2-6 GW 3.42- 3.7GFR NA Allocation completed by auction in July 2018 Korea K Above 8 GHz 26. 5-26.9GHZ NA Allocation completed by auction in Jury 2018 Sub-2 GHz NA NA Japan 2-6 GHz NA 3.6GHz 4 +2GHz, 4.4GHZJ.9GHZ To be Confirmed and allocated by end of March 2019 Atom 6 Gtr NA 27.5GHz- 29.5GHz To be confirmed and allocated by end of March 2019 .,ource: Morgan Stanley researai, China WIT, Japan MIC. US FCC Network rollout Cowrap (habitotsys. natiom ride):This is the single biggestvari- able to watch to forecast 5G capex. Nearly all previous mobile net- works were built starting from business -rich places and then expanded, and it was up to operators to decide network coverage, Influenced by expected investment returns and regulations. But many 515 applications will have significantly different requirements for network coverage and will provide different monetization oppor- tunities. We present a detailed discussion of investment retums in the 5G use case section. In China, the regulator previously set specific timetables for coverage requirements, and it is possible this practice will be continued under 5G. Regulators in other countries tend to use market - driven methods, mainly in the form of competition. It is also likely that there will be a minimum base station commitment when telcos receive spectrum licenses. Timing: The primary reason that the timing of network rollouts makes a big difference to capex is declining equipment prices as the upgrade cycle matures. Equipment prices have tended to fall every year in previous technology cycles, driven by economies of scale on the equipment vendors' side, and we expect this phenomenon will continue in the SG era. The rate of decline is expected to be particu- larly high in the first two years after product launch. So globally the early movers will suffer from high equipment costs and late movers will enjoy the benefits of a mature value chain from a capex perspec- tive. Morgan Stanley I - A key potential change in 5G is that China, the world's biggest tel- ecom market, will change from a follower to a Leader, rolling out its SG network together with traditional Leaders the US, Japan and Korea, which could help to lower initiaL5G equipment pricing through scale. We think global teLcos could see a win -win situation if they har- monize their SG investment plans. We currently expect large -scale 5G investments to start from 2020 for these leading SG markets, with some Limited -scale trial investments or fixed wireless capex in 2019. • China: ALL three operators aim to Launch 515 service in 2020, and we expect 5G investment to start from 21-119. But neither the regu- lator nor the operators have specified the scale of the network at Launch. • US: Verizon has already Launched SG service with a focus on fixed wireless applications in selective markets. Verizon sees a 25 -30% residential broadband market (30mn households) addressable by 5G as an initial SG opportunity. • South Korea: Operators target to Launch SG service in 1Q19, and we expect the capex upcycLe to kick off from early 2019. •Exhibit 17: Timetable of 3GPP standards .te GLOBAL FOUNDAt1GN • Japan: Operators target SG commercialization in 2020, with SG coverage only available in major cities in the beginning, and utti- mately around 2022 to have nationwide SG service coverage. Standalone vs. non - standalone: These are two broad deployment scenarios for SG networks. 3GPP froze the non - standalone NR tech- nology in December2017, followed by standalone technology injune 2018. Standalone would be a newLy -built network, including new base stations with SG equipment, backhaul. Links and core network- A non - standalone network would operate as a hybrid model on existing infrastructure supplemented by targeted small -cell deploy- ment in hot spots, allowing 4G and SG services to run in paraLLeL Standalone offers larger scale economies and high performance as well as Less complexity from legacy 4G integration, but it requires more capex in the early commercial stage. By contrast, non -stan- datone offers a quicker route to market, but its suitability is arguably greater for hyper - Localized deployments than national rollouts, and additional complexity needs to be introduced to realize interopera- bitity with existing 4G networks. 2017 2018 2019 Release14 I ROM S13 E.WflliOh Rat- t95pgat °,`. ; • Source: 3GPP 1) EDIMM771 Morgan Stanley I aF•F• • Availability of hrfrastructure :5G networks can be built using some of the infrastructure of existing LTE networks, basically towers and fiber. There is no doubt that a comprehensive LTE infrastructure will help save SG- related capex and accelerate the network rollout. A tower- sharing scheme would be helpful to save capex considering that the numberof towers in SG networks could be much higherthan LTE and it is becoming more difficult to find physical sites for future towers. In addition, fiber is a far better SG backhaul solution than microwave in terms of capacity and data speeds, but at a much higher cost. In fact, we believe microwave will be not able to meet most of the real SG capacity requirements, so a denser existing fiber network will also reduce incremental capex for 5G. Macro vs. small alb: Small cells are low -power cell sites that are easy to install. Macro and small cells are primarily complementary to each other in terms of network capacity and coverage. We see rising importance for small cells going into the SG era given the higher net- work performance requirements and the poorer propagation of 513 spectrum. In fact, small cells could be the only viable option for sev- eraL key 5G employment scenarios: •e 6GHz spectrum, where propagation Limits cell sizes e Radios using shared and License - exempt spectrum, which gener- ally mandates Lower power e Areas of hyperdense traffic demand So teLcos' smaLL -cell rollout plans, small -ceLL value chain develop- ment, and cost dynamics will be key variables for SG capex. ALSO, we need to watch teLcos' communications with regulators in terms of potential small -cell sites like streetlights and sides of buildings, given the famous complaint from US teLcos: it takes a year to get approval to deploy a small cell and only one hour to install it. Edge computing- Edge computing refers to infrastructure that ena- bLes data processing as close to the source as possible, which allows for faster processing of data, reducing latency and improving the cus- tomer experience. This type of computing wilt require micro -data centers close to SG EITS. The edge computing discussion is still in the very early stages, as potential applications like autonomous driving are still several years away. We also note that micro -data center investment could be shared by teLcos (network), data center opera- tors (physical environment) and information service providers (com- puting capacity), and future deveLopmentwiLL be tied up with specific applications. • GLOBAL FOUNDATION Exhibit 18: Macro cell and small cells t...r. fmx ew Tali. Visually prominent. Discreet. Lower to the ground. Source: Gavin Castle Exhibit 19: Small cells source E,r.. Morgan Stanley I safe• ^ \" • Exhibit 20: Mobile edge computing: VR use case � y fi�YL YWn„ .F r�r WI 4iabiii W�.+�r•• % owwW a+u4 source: TechNanve 5G capex forecast Our proprietary bottom -up model projects total 5G capex of USS672bn: Including spectrum, base transceiver stations (BITS), transmission and towers, we forecast totaL 5G capex of US$872bn for all. four markets combined in 2019 -30, 1.7x total 413 capex of •US$522bn in 2011 -18. Our scenario analysis suggests US$872bn under our base case and US$1,148bn under our bear case. In China, market concern regarding 5G capex risk appears to have subsided after recent newsflow around the 2.6GHz spectrum, but we expect high SG capex to negatively affect ROIC in earLyyears, particularlyfor China Mobile, and the first 515 capex guidance, which may come in March 2019, could be a negative catalyst. We believe US teLcos will spend US$26bn on SG spectrum overthe next few years in the base case, with a wide bull.-bear range of US$10 -42bn. Morgan Stanley $ Co. LLC ("Morgan Stonley) is acting as financial advisorto Deutsche Telekom AG ("Dn and is providing financing ser- vices in relation to T- Mobile USA, Inc. ("T- Mobile") for the proposed • GLOBAL rOUNDAPON Exhibit 21: Mobile edge computing: AV use case aooe MobH*NatNdk viNirfyq lNYM uu la.rtbn. ir4 .L�rn r�uwf [mlrol Ihixrrprp the Leaf 4r M GrWpM CM4 and tAFC �W�ova�q YWMa+ ✓aMr� �aonlom The \ea0w o- 0 0 0 o 0 0 o O O O O O O Q `tID a $ppce TecbNO��'F acquisition by T- Mobile of Sprint Communications, Inc ("Sprint') as announced on 29th April 2078. The transaction is subject to approval by T- Mobile and Sprint shareholders and other customary closing conditions. DT and T- Mobile have agreed to pay fees to Morgan Stonleyforitsfinancialservicesthatarecontingentupontheconsum- mation of the transaction. Please refer to the notes at the end of the report. Exhibit 22: spectrum and capex forecasts and bull -bear range Sowce: Morgan Stanley Resea¢s Bug 322,315 China Base NA 420,672 Bear 576,201 Bull 10,000 190,996 us Base 26,000 236,748 Bear 42,000 298,435 Bull 6,429 44,800 South Korea Base 7,393 50,303 Bear 9,000 60,363 80 89,201 Japan Base NA 128,731 Bear 161,948 Sowce: Morgan Stanley Resea¢s Morgan Stanley I - GLOBAL FOUNDATION 40 Theme 2: Incremental Annual Revenue of US $156bn for Seven 5G Use Cases by 2030 Market view: Although 5G promises significantly faster speeds, Lower Latency, higher reliability, and more device connections, there is limited visibility on the monetization of applications that require these features, suggesting that revenue opportunities for 5G remain uncertain. Our view: In the US, fixed wireless will likely be the early use case for 513, and Verizon estimates a total addressabLe market of US$81bn for fixed wireless (vs. US$180bn for mobile currently). In China, Korea and Japan, operators will likely focus on IoT use cases for industrial markets. We estimate incremental annual revenue of US$156bn for the top seven 513 use cases, 40% of current mobile service revenues in these markets. We believe operators with dominant market share in the enterprise segment are well positioned to tap these sizeable incremental revenue opportunities, including CT and CU in China, KT in Korea, and Docomo and KDDI in Japan. Our proprietary analysis: We attempt to size the revenue opportunities for the top seven 5G use cases based on key revenue drivers. We estimate incremental annual revenue of US$156bn, including US$18bn for fixed wireless, US$64bn for manufacturing automation, US$91bn for cloud gaming. US$7bn for autonomous vehicles, US$20bn for surveillance, US$4bn for drones, and US$32bn for healthcare. • SG will offer unprecedented bandwidth, Low Latency, fast mobility, and high capacity, which will support a set of brand new applications on top of a better experience using traditional smartphone connections. Based on the network capability requirement, SG applications can be classified into three key categories: enhance mobile broadband (eMBB), ultra reliable Low Latency communications (URLLC) and massive machine type communications (mMTC). • Enhance mobile broadband (eMBB): enhanced indoor and outdoor broadband, enterprise collaboration, aug- mented and virtual reality • Ultra reliable low Latency communications (URLLC): autonomous vehicles, smart grids, remote patient moni- toring and teLehealth, industrial automation • Massive machine type communications (mMTC): IoT, asset tracking, smart agriculture, smart cities, energy monitoring, smart homes, remote monitoring eMEB is expected to be She primary use case for SG in its •early stage. eMBB will bring high-speed mobile broadband to hotspots, enable high -speed streaming and will allow enterprise collaboration services to evolve. Some opera- Exhibit 23: 5G usage case Enhanced Mobile Broadband ;ga:p'ss in a seconn ,.� .itlec. U:J YtBni Smsrt nsma!- ::amng rye Horan Gey '..� Massive Machine Type Communications oa«e iro Future IMT Yac�k e1, Nay m the no r twyerented raat3y laausuy a�eomatmc FAissian cHliral armNCatwn tsar ar-.n; ca, Ultra- reriable and low latency Communications Morgan Stanley I • tors consider eMBB as the fast -mile solution in areas Lacking fiber connections to homes, particularly for US mobile operators. Telsos are more excited about the revenue opportunities in the industrial market than the consumer market. Based on the experi- ence of 3G/4G migrations, and the current global trend of offering unlimited data amid ARPU pressure, telcos doubt the willingness of consumers to pay more and telcos' pricing power to charge more going into 5G. Most of 5G's network capabilities are designed to meet industrial demands, and this market represents an incremental rev- enue opportunity for telcos. Network slicing features could equip telcos with pricing power when offering customized service to enter- prises. Use cases Smartphones: The 5G smartphone business modelwilt be quite sim- iLar to 3G and 4G, where telcos charge monthly fees by offering much faster network speeds. According to chipset and handset vendors' progress, we expect the first batch of SG handsets will hit the market •in mid or Late 2019- TeLcos might charge higher monthly fees when upgrading customers to a faster network, but this will be subject to competitive dynamics. From the customer's perspective, most online content demand can be met by LTE networks, with the exception of HD WAR, which we discuss below. From the telcos' perspective, 5G technology will be helpful to expand network capacity in hotspots like stadiums, exhibition centers, and business districts. Fixed wireless: Time will tell whether 5G fixed wireless is a cost -ef- fective substitute for FTTH (fiber to the home), but both Verizon and T- Mobile have suggested it is viable. On 11 September 2018, Verizon announced details of its SG fixed wireless offering — 5G Home — in parts of four initial markets: Sacramento, Los Angeles, Indianapolis, and Houston. We assume the pre- standards rollout will be limited to perhaps 30 -50k households. Verizon is initially deployed with 28GHz millimeter wave spectrum and will require a technician to install ser- vice. It hopes to move to self - installation over time, which would improve economics, while using Lower band spectrum such as 2.5/3.5GHz would also make self - installation more practical. The service is free for the first three months and then priced at US$50 per month for Verizon wireless customers and US$70 for others, and will include three months of free YouTube TV (US$40 per • month thereafter). These prices are competitive, especially for GLOBAL FOUNDATION Exhibit 25: Verizon fixed wireless terminals Line of Sight Low -E Glass Verizon customers, who may represent 30 %+ of targeted house- holds. On speeds, the typical300Mbps is competitive for most cus- tomer needs, particularly with no data caps and peak speeds of nearly 1Gbps. The service was launched on 1 October 2018. Related research: Verizon Communications: Lessons from Day 1 of the 5G Home Launch. Morgan Stanley I wE.a - • CloudgaminaandVR/ AR: Given their data-intensive and interactive nature, good VR and AR experiences set high requirements for net- work speed and latency, as well as storage and computing capabili- ties in the cloud. Though an average network speed of 100Mbps might be possible over existing 4G networks, some premium VR/AR applications will demand higher speeds and lower latency. KT u MIN 26: VR /AR connectivity requirements and evolutionary stages GLOBAL FOUNDATION showed a 360 - degree VR live application at the 2018 Winter Olympic Games via a pre- standard network (see Video: KT Showcases its 5G Trial Network at the 2018 Winter Olympics (27 Feb 2018)). VR/AR content now is highly skewed to gaming and video (including sports and education), sowe expect the business modelto be similarto cur- rent online video platforms (advertising revenue plus subscription fees for platforms). The busi- ness opportunity for telecom operators mainly lies in the potential ARPU uplift when migrating subscribers to faster S rre HUawei Wheless x t abs Exhibit 27: VR /AR business models User pays tc e!ray'A portent -. A&' Company pays vR platfa by CPM A Oempary 4 CPM based Payment vR Ptstfcrm L'R Pants P cdu ar = rsms^t waI R Subscrha* ..puma Hua WvNessx Labs �� User P&ys'uR ciatform. VR P''arlcr sharestf -a payrtu twith theconcertproducer __ Suas(criber_ ! wMmetwc VR Plntfertn.'Apo l� as>merf VR Gonxe*t Producer User my$ VP sto'e: PPV *Ray par downtoad: Platfo*m Shares payment wi#, W t pmdr 5ubscrbe, `33 byueage; °aY per vexWaredaea 'JRApp Stare; ftaticrm Payment Vf, fantam Producer Stage pit Stage Stale 3/4 speed networks on the con - sumer side, and setting band- VR `. width on the corporate side. ARpticationa &Technical Garning. 36T ,,deo. Immerswecort�t. Gaming i>r 'qg Features s:muaeion etlwat3on irrc actwesimulai iw' pr�enpenente rauahzat[onldesgn 'rtc:ar w' rw,lnyem �r ine s 3ar rza, vs^aaeae r. r:: tuw r. Orq'£PM FII �,vL'ec Y.Ti g � .w.ea o x c wb:.mcG sroar R,FXcav IO SI G a^Ff= m�'anc,'En �C 9•SarT 2 AR Applications &Technical Assembly instructions, gam, rg, Hcl0graphicvim :Mat clad- baaedmi and Features mca[iw- based,re.Tza w^ wiR , creasinguniversesize. merged reality - applicauma asuatizeUw for retafVna7set ng '+ghli connaM:Rd public Increased user density arts sMetuAR appications ronrrect:v ty - ncM'.m�a xmcsa v.- .rts�, �^�sic.+4t atl ucua -ce�a �a mesa seam:; Mna[e P-a :zau»c r arsre mracr� -a� Connectivity Pnmany X -F Gand W, -Fi Stray `rgrA 2S: Mbps 56G Btv 'Pgto 46 Mhos BtrenmengS tOO Mbps Requirement Oonnecrt-,' K ms Late 20 M Latency -8 GGbps ."( 2 -1C ms iat ' S rre HUawei Wheless x t abs Exhibit 27: VR /AR business models User pays tc e!ray'A portent -. A&' Company pays vR platfa by CPM A Oempary 4 CPM based Payment vR Ptstfcrm L'R Pants P cdu ar = rsms^t waI R Subscrha* ..puma Hua WvNessx Labs �� User P&ys'uR ciatform. VR P''arlcr sharestf -a payrtu twith theconcertproducer __ Suas(criber_ ! wMmetwc VR Plntfertn.'Apo l� as>merf VR Gonxe*t Producer User my$ VP sto'e: PPV *Ray par downtoad: Platfo*m Shares payment wi#, W t pmdr 5ubscrbe, `33 byueage; °aY per vexWaredaea 'JRApp Stare; ftaticrm Payment Vf, fantam Producer Morgan Stanley I • Exhibit 28: Different levels of autonomous driving IL IF Driver Automated Combined Hands off the Automated No steering Controis all subsystems automated steering wheel but driver can wheel aspects of subsystems conditionalty intervene driving source: TOmiom. Exhibit 29: Data comparison: Today's iPhone vs. a 2050 autonomous vehicle r t 3,300 Years -44 1 Hmir -t2 GB J Phoe Usage • -3 �a FaWm� o oo c8!ricam °eo sa0x iPtwne usage -37 EB -S,GW EB Today '7 IFe sa+asrtuas�n 1 • Source Mwgan San"Research,Shutterstoct Exhibit 38: We estimate that level 5 penetration will reach 1 % of global miles traveled by 2030 Global Autonomous Vehicle Penetration toil 75% ,6% 25% 0% lot 2015 2020 2025 2030 2035 2040 2045 2050 ■ Level 0 R Level 1 f 2 ■ Level 3 '.° level 4 ■ Level 5 Swr.. Magan Stanley Research estimetea GLOBAL FOUNDATION Autonomous driving: This is one of the mission - critical IoT applications, with super -high requirements for seamless cov- erage. The Low network Latency and good mobility features of 5G are the key enablers of autonomous driving. Autonomous vehicle data could increase global wireless traffic by 40x over current levels, delivering a boost for telcos comparable to the Phone in 2007. The business model is still in the very early stage, but we size up the opportunities in our gLobaL collaboration report with the Autos team: Autonomous Cars &Telcos: We're gonna need a bigger (513) pipe (7 May 2018). Morgan Stanley I wE,E• • Exhibit 31: Autonomous vehicle /telco /technology business model Car controtlirg Collision v navigatior avoidance RemoT -e Cortent a service appLicatior T -map related API services Regular con nectm:y - V2X Network Car sharing insurance Charging& mairtenance CarPar✓ing 12XAPl sc'vices. .. Beyond connectivity, - Priority transmission Cloud Da-,a aentre Tetecem `, B26 partnership •Smart manulacturinr. The linection of future manufacturing level, opments includes higher productivity (Leaner), safety, efficiency, and flexibility. Though wifi and Bluetooth have started to replace wired technology for manufacturers, the massive number of connections and Low latency of SG will enable smart manufacturers to evolve to cloud -based controLs for robotics. In particular, the incorporation of Al or machine learning into smart manufacturing will need cloud - based control. Based on different demands from manufacturing, all three types of 5G will be applicable in this field: eMBB (cameras, sen- sors, remote control, cloud/edge computing), mMTC (tracking/moni- toring), and uRLLC (synchronized robotics). Smart manufacturing may Just require seamless network coverage in a certain area, Like a factory ora logistics center, in the early stage, but this will expand to wider coverage when it evolves to end -to -end solutions or product lifecycles. Telcos, factories and value chains (Like software devel- opers, system integrators) will need to explore business models case by case given customized demands. • GLOBAL FOUNDATION API fee so[ution fee Comer, fee Apo fee Revenue shar,..g with V parties Pdverrseme�t e;gdata In-app purchase eunotingsates 5G networkfee Priority trersrnissior fee Cther nc;wc %k fee �touc service fee IDC fee Remote healthcare: 5G will enable remote healthcare. Services Like remote diagnosis (wireless endoscopy and uLtrasound), real -time health tracking, and medical robotics depend on interaction between the device terminal and the patient. The sensitivity of force- feedback requires a Lower Latency network as weLL as uninterrupted connec- tivity, which can only be met by 5G technology. We think teLcos'role in remote healthcare may not only be restricted to providing connec- tions but is likely to expand to cloud -based big data or Al analytics services. Drones: 5G will help drones transmit real -time HD video when moving quickly. When combined with anaLytics tooLs or industrial vertical sensors, drone services will be highly value -added in the security, construction, energy, agriculture, transport, and content production field. Some easy applications include site/property sur- veillance or monitoring, data collection and analytics. The business opportunity fortelcos could vary from a pure connection service pro- vider to integrated solutions, or the so- called drone -as -a- service. Morgan Stanley I ^E.E. • Exhibit 32: Demand- supply business model for drones GLOBAL FOUNDATION • S.-, Nww, . SPG L . Smart cities and surveillance: There is no clear boundary to what municipal activity needs to be included in a smart city management system, and aspects most frequently discussed are video surveil- lance (social security), transportation, utility meters, and weather and environmental monitoring. The future development of smart cities will be wireless, cloud -based and analyticaLly rich, all setting higher requirements than current LTE and NB -loT networks can deliver. Smart cities may also need network slicing owing to cyberse- curity requirements. For teLcos, cooperation with the final customer - municipal governments - will be essentiaL. Enersy: Energy companies need low Latency SG networks to build intelligent distributed feeder systems at a Low communication cost to control grids accurately and quickly. Demand is driven by the increasing contribution of renewable energy (soLar power, wind and hydro- electric), which lacks stability. A distributed feeder system can respond to an outage, calculate, and implement the fault Location and isolation quickly, which is essential to retaining high power suppLy reliabiLity and saving costs at the same time. Smart grids need dedicated networks, and the network slicing of SG will play an impor- • tant role here. Incremental annual revenue of US$156bn We have attempted to size the revenue opportunities for top 5G use cases. We estimate a incrementalannual rwenue of US$156bn for the top seven use cases in China, the US, Korea, and Japan by 2030. • Fixed wireless (US$18.Sbn): 22mn household subscribers x ARPU of US$70 /month; this appbcation only applies to the US market • Automation (US$64.tbn): US$0.03 per GB data pricing x 21ZB data traffic on SG network from manufacturing sensors • Gaming (US$8.7bn): US$8.7bn total gaming revenue x 60% share for cloud gaming x 10% revenue pie for telcos • Autonomous vehicles (US$7.3bn): Total. 494bn miles from LeveLs 4 & 5 autonomous vehicles x US$0.01 -0.03 price per mile • SurveiLlance (smart cities) (US$20.5bn): Total 1.3bn cameras x 50% with SG connections x ARPU at US$1 -10 /month • Drones (US$4.lbn): Total 143mn drones x 50% with SG connec- tions x ARPU at US$1 -10 /month • Healthcare (US$32.4bn): Total. heaLthcare service revenue US$3.4tn x 0.1 -1% spending on SG networks Droroe Partnor Companies Sensor Partner "« o s . - Lomwnlas f EnaaGngtew^itrotogl! 7 � - "' compani" n �. , " {aausucw„ I'� x �11-1'JLO� Air Spec* Precision Flight :. 94Z Data Analytic• $DP 'r 9 Slicing Str ices Air Traffic Autonomous ReaE Tlmv DeCwaloxl En sne g . > Third Party data seta e.g. weathac, landratry, Property e*v'nsraNP, Mda�nagaxrnent F€ t a crop yY*l4t. crop paces Piegistratan Emarg ®nry ' E "g RN�e Traekreg Senses cop"t ' * Storage Data Proc•ssmg ; Y' ✓ r UAV Camrot plat €arm Enhanced Corimmointy ,. Exietnal development ecv,., :, • S.-, Nww, . SPG L . Smart cities and surveillance: There is no clear boundary to what municipal activity needs to be included in a smart city management system, and aspects most frequently discussed are video surveil- lance (social security), transportation, utility meters, and weather and environmental monitoring. The future development of smart cities will be wireless, cloud -based and analyticaLly rich, all setting higher requirements than current LTE and NB -loT networks can deliver. Smart cities may also need network slicing owing to cyberse- curity requirements. For teLcos, cooperation with the final customer - municipal governments - will be essentiaL. Enersy: Energy companies need low Latency SG networks to build intelligent distributed feeder systems at a Low communication cost to control grids accurately and quickly. Demand is driven by the increasing contribution of renewable energy (soLar power, wind and hydro- electric), which lacks stability. A distributed feeder system can respond to an outage, calculate, and implement the fault Location and isolation quickly, which is essential to retaining high power suppLy reliabiLity and saving costs at the same time. Smart grids need dedicated networks, and the network slicing of SG will play an impor- • tant role here. Incremental annual revenue of US$156bn We have attempted to size the revenue opportunities for top 5G use cases. We estimate a incrementalannual rwenue of US$156bn for the top seven use cases in China, the US, Korea, and Japan by 2030. • Fixed wireless (US$18.Sbn): 22mn household subscribers x ARPU of US$70 /month; this appbcation only applies to the US market • Automation (US$64.tbn): US$0.03 per GB data pricing x 21ZB data traffic on SG network from manufacturing sensors • Gaming (US$8.7bn): US$8.7bn total gaming revenue x 60% share for cloud gaming x 10% revenue pie for telcos • Autonomous vehicles (US$7.3bn): Total. 494bn miles from LeveLs 4 & 5 autonomous vehicles x US$0.01 -0.03 price per mile • SurveiLlance (smart cities) (US$20.5bn): Total 1.3bn cameras x 50% with SG connections x ARPU at US$1 -10 /month • Drones (US$4.lbn): Total 143mn drones x 50% with SG connec- tions x ARPU at US$1 -10 /month • Healthcare (US$32.4bn): Total. heaLthcare service revenue US$3.4tn x 0.1 -1% spending on SG networks Morgan Stanley I area. ... • Exhibit 33: C J • 5G incremental annual revenue by application and country USD bn X Fixed wireless NA 18.5 NA NA = Subscribers (mn) 22 76"4.1 x ARPU (USD /mth) 70 Automation 30.3 20.4 3.6 918 = Data usage by manufacturing sensor (EB) 10,110 6,791 1,196 3,280 x % 5G 10% 10% 10% 10% x Data pricina (USDIGB) 0.03 0.03 0.03 0.03 Cloud gaming 4.9 1.6 0.7 1.5 8.7 = Mobile gaming revenue (USD bn) 81.8 25.8 11.5 25.2 x % cloud -based 60% 60% 60% 60% x 5G % of revenue 10% 10% 10% 10% Autonomous vehicle 2.2 4.6 0.4 0.2 7.3 = No. of AV miles (bn) 337 140 11 6 x 5G revenue (USD /mile) 0.01 0.03 0.03 0.03 Surveillance 1&0 7.0 NA 0.4 20.5 = No. of camera (mn) 1,000 235 50 x %5G 50% 50% 30% x ARPU USD /month 2.2 5.0 2.0 Drone 2.6 1A NA 0.1 4.1 = No. of drones (mn) 100 23 20 x % 5G 50% 50% 50% x ARPU USD /month 4.3 10.0 1.0 Healthcare 7.2 22.9 2.2 0.2 32.4 = Healthcare revenue (USD Ion) 1,074 7,057 215 253 x % service 67% 32% 100% 67% x 5G % of revenue 1.0% 1.0% 1.0% (1.1%1 2030 5G incremental annual recenue of 7 fi.Ids (USD bw .r 2018E telecom service revenue 190.5 325.0 33.0 102.3 650.9 5G incremental revenue as % of 32% 23% 21% 12% 24% 2018E mobile service revenue 133.6 1 183.0 21.5 57.3 395.4 5G=remen1al revenue as %ol 45% 42% 32% 21% 39% Source : Gartner, Ragan Stanley Research E - Magen Stanley Research estimates Morgan Stanley I ......- GLOBAL FOUNDATION 0 Theme 3: SG Could Trigger Major Changes in Market Structure Market view: Changes in market structure, such as industry consolidation (positive) or new entrants (negative), are unpredictable binary events. The potential merger between T- MobiLe and Sprint in the US and new entrant Rakuten in Japan are known to the market, while Bloomberg reported in September that the Chinese government is considering a merger between CU and CT to spur 5G development. Our view. Major changes in market structure often happen around new technology cycles, so we do not rule out the possibility that SG will trigger more unexpected events. Even without a merger, CU and CT should benefit from capex savings via SG network sharing, which is not yet in the price given a 10 -20% discount on 2018 EV /EBITDA vs. CM. Our proprietary analysis: Under our base case, we assume CU -CT network sharing will focus on selective areas such as tourist spots, highways, and high -speed railways, and we estimate combined SG capex of USS182bn (vs. US$192bn for CM). Under our bull case we assume even more extensive network sharing, including in key hotspot areas, which would effectively mean one single network competing with CM, and we estimate a tower combined 5G capex of US$134bn. Active network sharing. Operators not only share passive infra- structure like towers and electricity equipment, but also base station equipment. Network sharing among smaller operators could become a new trend in the SG era given the high capex requirement and uncertain investment returns. Compared with industry consoLi- dation, network sharing is more Likely to win approval from reguLa- tors as Long as operators run businesses separately and the market remains competitive. Operators of similar size and 5G rollout timeta- bles, using adjacent spectrum, can generate more capex savings by working together. The economics between partners could be very simple, such as CU and CT in the 4G era, purely splitting capex in remote areas, or could it be complicated, including setting up a net- work JV and details of roaming agreements. In- market consolidation. We believe the two primary reasons for telco consolidation — spectrum and capex— will be magnified in the SG era compared with the previous cycle. TeLcos need a block of 20MHz spectrum to achieve the theoretical 413 speed, which increases to 100MHz under 5G, but the limitations of spectrum supply remain unchanged. Faced with Limited resources to acquire spectrum and build expensive networks, smaller operators will either have to seek M &A opportunities or stay out of the 5G market. It looks Like three or four is a globally accepted number of independ- ently running mobile operators in one country, as most five -to -four •consolidations have won regulatory approval while three -to -two consolidations have faced regulatory hurdles. Most governments are preparing enough spectrum for three or four operators. Fixed- mobile convergence. The potentially huge volume of 5G data traffic and denser 5G BTS both require comprehensive fiber network coverage. Pure mobile pLayers' profitabiLity will be pressured by the increasing backhauL cost going into SG, which compares to the syn- ergy from sharing fiber networks, marketing bundling service, and exploring new 8213 use cases for integrated players. So we think SG will become an incremental new driver for global fixed - mobile con- vergence. Some fixed - mobile integrated operators have shifted more capex to fiber network investments following the peak of 4G cov- erage expansion and before the massive 515 network rollout. Exhibit 34: Potential changes in 513 market structure c. AT &T Rt Y= ®L.G U+ SoftBank Sprint i Poterxial co -build Potential Poterxlal network new merge entrant ..omce Mogan Sanley Res rch Morgan Stanley I "r,E. • Country 2G Potarltbilnwentrant"5Gopensnew 4G markets such as industrial LOT applica- China tions. Some SG use cases require net- 1 CM work coverage in a specific Location (as opposed to very broad coverage for 2 CU traditional consumer teLecom ser- 2 CU vices). This may attract new entrants to the industry. In the US, while the 3 CT wireless carriers are clearly interested in leveraging SG for manufacturing 1 CSL automation use cases, this is an area 1 HKT (CSL PCCW merged) where other vendors could leverage private SG networks, perhaps using 2 3HK unlicensed spectrum, to provide end - to -end solutions to industrial I 3 SMT Writers. Communications equipment 3 HKT vendors could provide end -to -end pri- vate 515 networks with necessary ser- 4 MKT vice level agreements connecting m various elements of the production 5 New World Telecom line, robots in warehouses, distribu- tion, etc. The tech companies could also play a rote here, with Amazon and Taiwan GoogLe showing a strong interest In 1 CHT the connected home via ALexa, Nest O and other products. It is not hard to see 2 TWM their interest expanding to the enter- 2 TWM prise market. In Japan, Rakuten received 4G spectrum in 2018 and 3 FET plans to spend US$Sbn (through end - 2028), aimingfor 96% nationwide cov- 4 KGT erage (by end - 2025), 10mn 4 APT subscribers (by end - 2028), and posi- tive net profit by 2023. As the govern- 5 VIBO ment's stated policy is to promote competition, we think the Ministry of 6 Mobdal Communication (MIC) is Likely to be receptive to Rakuten's planned appik Japan cation for SG spectrum. In China, China 1 NTT Docomo Broadcasting Network Corporation (CBN), the user of 70OMHz spectrum, 2 KDDI is often viewed as a potential new 2 KDDI entrant as a result of the government's 'three networks convergence" initia- tive GLOBAL FOUNDATION Exhibit 35: Market structure changes in different technology cycles Region Country 2G 3G 4G China 1 CM 1 CM 1 CM 2 CU 2 CU 2 CU 3 CT 3 CT Hong Kong 1 CSL i CSL 1 HKT (CSL PCCW merged) �. 2 3HK 2 3HK 2 SMT 3 SMT 3 S&TT 3 HKT v 4 CR Peoples 4 MKT 4 CMHK m 5 New World Telecom 5 CMHK (formerly Peoples) Taiwan 1 CHT 1 CHT 1 CHT O 2 TWM 2 TWM 2 TWM 3 FET 3 FET 3 FET 4 KGT 4 APT 4 APT 5 TransAsle 5 VIBO 5 TStar 6 Mobdal 6 Fire) Japan 1 NTT Docomo 1 NTT Docomo t NTT Docomo 2 KDDI 2 KODI 2 KDDI 3 Sofibank 3 Sollbank 3 SoBbank 4 Rakuten g. Korea 1 SKT i SKT i SKT o 2 KT 2 KT 2 KT x. 3 LGU 3 LGU 3 LOU USA 1 AT &T 1 AT &T 1 AT &T 2 Verzon 2 Vernon 2 Verizon 3 T- Mobile 3 T -Moble 3 T- Mobile 4 Sprint 4 Sprint 4 Sprint 5 U.S. Cellular 5 U.S. Cellular 5 U.S. Cellular 6 Leap 6 Leap 6 Leap 7 MetroPCS 7 MetmPCS 7 MetroPCS 8 Clearvrire 8 Clearxire 8 Cleaiwire 9 Cincinnati Bell 9 Cinannaa Bell 10 Nemel 11 Axel UK 1 TEF 02 1 TEF 02 1 TEF 02 2 EE 2 EE 2 EE 3 Vadamne 3 Vodafone 3 Vodafone 4 3 4 3 4 3 Germany 1 DT 1 DT 1 DT 2 Vodatone 2 Vodafone 2 Vodafone 3 E-Plus 3 E -Plus 3 TEF D (E -Plus) 4 TEF D 4 TEF D France t Orange 1 Orange 1 Orange 2 SFR 2 SFR 2 SFR 3 Bouygues Tel 3 Bouygues Tel 3 Bouygues Tel g 4 Ifatl stared 2012 wItaly 1 TIM 1 TIM i TIM 2 Vodafone 2 Vadafone 2 vocatone 3 wind 3 Wind 3 W,nd13 (merged in 2017) 4 Blu a I t u ca 2002 4 3 leunMed 2004 Spain 1 TEF t TEF 1 TEF 2 Vodafone 2 Vocations 2 Vodafone 3 Orange 3 Orange 3 Orange 4 yo' o 4 yo' o 4 Masmovil i. India 1 Artel 1 Airtel f Anal 2 Vodafone 2 Vodafone 2 Vodafone Idea 3 Idea 3 Idea 3 Reliance Jio 4 BSNL 4 BSNL 5 MTNL 5 MTNL Indonesia 1 Telkomsel t Telkomsel 1 Telkomsel 2 Indosat 2 Indosat 2 Indosat 3 x1 Axiata 3 xL Amata 3 XL Axiata 4 Hutch 4 Hutch 4 Hutch i 5 Smart Telecom 5 Smarlhen 5 Smamren y 8 Mobile -8 B A>as S 7 Axis 7 Balshe Telecom 8 Bakrie Telecom 9 Telkomflexi 10 Sampoema Telecom Swrce. Morgan Same, Research Morgan Stanley I ^F .... • Japan With the three mobile carriers injapan conducting 5G use -case exper- iments since 2016, we believe they are all likely to launch SG net- works. However, much will depend on SG spectrum allocation, which is scheduled for end -March 2019. The 3.7GHz , 4.SGHz, and 28GHz spectrums are seen as the top candidates for 5G bandwidth alloca- tion, and with the emergence of Rakuten as a fourth telecom carrier in Japan, competition could be fierce. A new entrant: Rakuten Mobile Networks was allocated the 1.7GHz spectrum (4G) in April 2018 and has officially entered the telecom carrier business. However, the Ministry of Internal Affairs and Communications has imposed 'additional conditions' on Rakuten, including the need for Rakuten to build its own network capabilities/ infrastructure. Rakuten plans to commence its telecom carrier ser- vice in October 2019 and has announced plans to invest up to V600bn by 2025. However, this V600bn of capex is approximately the same amount NTT Docomo invests in a singLe year. With SG spec- trum allocation expected by end -March 2019, market participants will be closely watching whether Rakuten is allocated SG spectrum • before commencement of its 4G service. China Our base case is for a stable three mobile fixed -line integrated player market in China going into 5G. We expect to see active net- work sharing between CU and CT,which willvary on spectrum alloca- tion and negotiations, but operationally there will be three competing operators in the market. CU and CT did have active net- work sharing in the late 4G cycle on a relatively small scale (40k macro BTS mainly in rural areas and 40k indoor systems), saving Rmb3bn capex and Rmb1bn opex. Both CU and CT have mentioned that they are open to discussing the co- building of a 513 network, and emphasized that having adjacent SG spectrum will be key to network sharing. Potential CU-CT merger is an unpredictable andbinaryewrk: We have no knowledge as to whether such a transaction could occur, but we see both pros and cons to such an arrangement. A merger could result in significant 5G capex savings with better returns, encour- aging operators to accelerate SG network rollouts. It could also • improve the overall profitability of the combined entity given bene- fits from scale and synergies. However, it would create a duopoLy between CM and CU- CTthat could significantly reduce the competi- tive pressure on telecom tariffs. It could also affect CU's role as a pilot for the ongoing mixed - ownership reform. Now entrant In China? China Broadcasting Network Corporation (CBN), the user of 700MHz spectrum, is often viewed as a potential new entrant as a result of the government's'three networks conver- gence' initiative. We see a Low Likelihood of this happening, as CBN cannot get a mobile telecom License and Lacks the capital to build out a network. Having said that, we do not rule out the possibility of a restructuring between CBN and the three telcos. Korea We expect the three. player market to be maintained in the 5G era. All three operators have acquired spectrum in the 3.SGHz and 28GHz bands and are expected to roll out separate networks. The Ministry of Science has announced policies aimed at sharing passive infrastructure as a way to lower the capex burden. The bulk of these assets are owned byl(Tand the regutatoris currently negotiating the pricing for access. The regulator expects aggregate savings in the range of W400bn to W1tn as a result of the shared infrastructure policy. That said, none of the companies have proposed sharing of active infrastructure and we expect three full -scale nationwide SG networks to be rolled out over the next 3 -5 years. US Sprint & T -Mobile seeking to mere. The US is currently a four - player wireless market in most locations, with Verizon and AT &T the Largest players. The regulator has blocked four -to -three consolida- tion in the past, but Sprint and T- Mobile argue that their merger would help accelerate 513 deployment, whereas they would have less financial capacity to deploy robust SG networks on their own. Cable companies have recently entered the wireless market via MVNO agreements, and may consider acquiring spectrum to build out net- works in high- traffic areas. Dish is also planning to build a 5G IoT net- work but will need a partner to help fund the estimated US$10bn build cost. Morgan Stanley I - GLOBAL FOUNDATION 0 Theme 4- Return of Pricing Power — A Rare Opportunity for Telcos Market view: With a few exceptions, global telcos have generally failed to monetize the 3G and 4G cycles due to a lack of pricing power as a result of regulatory and competitive pressure. With rising adoption of unlimited data plans in most markets, operators are unlikely to have pricing power even with booming data usage in the SG era. Our view: SG provides a rare opportunity to return pricing power to the telcos. We believe the bulk of the SG incremental revenue opportunity lies in the enterprise segment, and the B2B market tends to be Less sensitive to price competition and also faces less regulatory risk than B2C, particularly in China and Korea. We also expect fewer competing networks in China and the US due to network sharing or mergers, giving more leverage to the operators. Our proprietary analysis: From our deep dive into the concept and potential use cases for network slicing we see significant potential for telcos to exert higher pricing power. In contrast to the pre -SG era, where net neutrality made it structurally difficult to package data, SG will allow telcos to sell "slices" of the network, at different parameters, customized for specific end applications. This is a fundamental shift away from net neutrality -based economics, and we believe it has the potential to create completely new ways of pricing data services. We believe pricing power will remain a function of demand, competition and regulations and is likely to differ by country. Markets where the number of effective players could shrink, such as the US, should see operators regain some pricing power. • For B2C service, we believe select markets such as Korea and Japan may be able to price SG at a premium, but it will depend on whether flagship handsets are available to target early adopters. We believe 828 — with its potential for higher pricing and new concepts such as network slicing — is likely to see new pricing models in which operators charge more for superior performance. Unlike retail pricing, where regulations play a part in the overall price- finding process, B2B pricing is decided purely on commercial terms. This should be good for markets Like China, Japan and Korea, where politics is Likely to have a smaller impact and operators can focus on projects which offer high returns on capital. How much pricing power telecom operators have under SG will be a factor of 1) demand, 2) competition, and 3) regulations. We believe pricing will be a moving variable, depending on how the three factors mentioned above evolve, ultimately deciding what return on invest- ment operators will be able to generate. Given the many unknowns (we are currently not aware of any official announcement yet on 5G • pricing), we have written this section using a question & answer approach, providing our views on what we believe will be key ques- tions investors will need to consider. As 5G is expected to signifi- cantly expand the market for B28 services, we have structured our questions separately for consumers and for corporate services. Morgan Stanley I PeSE^ • Consumer services • • 1. How will SG services be priced as a mobile service? With unlimited data service having already become the tariff plan of choice for most countries where LTE service is readily available, we will require, or be willing to pay a premium for, download speeds in excess of 10oMbps. For now we believe video streaming is likely to be one of the few services where 5G's faster speeds wilL allow for an enhanced experience — ultra- high- definition without the buffering. However, we believe applications are Likely to emerge to take advan- tage of the higher speeds and lower latency, allowing operators to charge a premium— service concepts such as cloud gaming would be believe SG unlimited data is Likelyto be the standard service offering. an example, as it would require much shorter latency times to meet The higher data throughput fol coupled with the larger spectrum allocations, imply that the marginal cost to offering higher amounts of data will be much Lower, further supporting the case for unlimited data offerings. Hence, we believe the tiering of tariff plans could end up being based on speed rather than total data usage. This is because not all users E1ddb@ 36: Comparison of unlimited LTE data plans among the four countries the response times required for a ubiquitous gaming experience. Other forms of consumer service might be conducted on devices other than smartphones, such as mobile ARNR helmets and drones. We think pricing schemes could be similarto when tablets and Apple Watches were introduced to the market, with a monthly fee per new Sim. Country Operator China Mobile Lowest unlimited price 58 plan 8 Notes In select provinces. No speed cap for first20G8 local, then Imbps China Mobile 78 11 No speed cap for first 1050, then Imbps China Unicom 99 15 No speed cop forfirst20GB, then Imbps China CU Big King (with Tencent) 49 7 1GB per day. Unlimited data for TencentApp China Telecom 49' 7 In select provinces. No speed capforfirst 10511, then Imbps China Telecom 99 15 No speed capforfirst10G11, then Imbps Softball 5,480 49 No speed cap for first SOGB Japan KDDI 5,500 50 No Speed capforfirst 25GB SK Telecom 51,7501 46 No speed cap forfirst 100G8, then Smbps South LG Telecom 51,750 46 Daily SGB Korea" Korea Telecom 51,750 46 Na speed cap for first 10050, then Smbps United AT &T i 40 40 No speed cap, throttling starts at 22GB x Verizon 40 40 No speed cap, throttling starts at 22GB States T Mobile 35 35 Unlimited data, throttling starts at SOGB Nate: `Effective cost to customer after 25 % tariff disi eN inclunwe & VAT 41604140 for 4lines, sppleabk 4knoPay and paperless billing are enabled source -. Company data Magi, Stanley ResearM Morgan Stanley I aaaa. • 2. Will operators be able to price 513 at a premium? • • For certain markets we think this could be possible. Premium pricing strategies have tended to work in markets such as Korea and Japan, and we believe the 5G novelty factor, coupled with a mix of flagship smartphones, should be able to get early adopters to pay a premium. GLOBAL FOUNDATION 3. Won't competition work to bring down pricing? Yes, but this could differ by market. Given the large amount of capex and spectrum costs required for 5G, we believe the number of effec- tive players could potentially shrink in select markets such as the US, where SG spending is a key driver behind the proposed Sprint/ While critics may say that SG's faster throughput will only be margin- T- Mobile merger. In such cases, we do expect operators with SG net- ally felt by consumers, from what we were able to experience first- hand with trial networks, the rapid response times do give the impression that 5G is much faster, even if the actual time saved was less significant. Past experience in China indicates that, at least in the early stage of service launch, telcos can price the newer and faster network connections at a premium. We also believe operators are likely to be creative with marketing to position SG as a premium service, such as offering a bundle with an OTT subscription. Video streaming is expected to be one of the few applications where subscribers can take full advantage of 5G, and hence offers a valid value proposition to upgrade subs to higher priced plans. Exhibit 37: Unlimited data plans bundled with OTT subscriptions works to have more pricing powerthan 4G operators, enabtingthem to enjoy higher overall pricing. In Korea, a market renowned for being hyper - competitive, the three network operators (SKT, KT and LGU)were not able to reap the bene- fits of higher ARPU when 3G was launched, despite being one of the first countries to commercialize the technology. The companies did see a material improvement when migrating to 4G, however, as com- petition was focused more on subsidies than tariffs. For 5G, none of the operators is showing an intent to aggressively roll out the neM- generation network as a way to gain a competitive advantage. Given regulatory pressure which led to visible cuts to 4G tariffs, we believe operators are not as eagerto compete aggressively on price from the start. A stronger- than - expected take -up by consumers once the infrastructure is in place would set the stage for increased competi- tive pressure, in the form of either subsidies or tariffs. In China, the concentrated mobile market share towards China Mobile (65% for CM and 17 -18% each for CU /CT) is likely to provide a strong incentive for the smaller operators to use price as a weapon to gain market share, especially if the consumer market telecom ser- vice focuses on pure connection (which is hard to differentiate). Should the regulator set a difficult tariff reduction target, similar to 2015 and 2018, telcos may have to further bring down prices. Country Operator price plan South SK Telecom Not Avaitabie Korea- LG Uplus Unlimited 88 66,000 59 Access to U* movies KT Carp Data ON premium 66,750. 602,54 1doyfor mobile IPN; w11,000 warth points to usefor VoO Unlimited &More AT &T Premium 40' 40 Any 1 of 7- HBO, Cinemax, Showtime, Star, VBV, Amazon Music, or Pandora United go / beyond / above Verizon unlimited 40, 40 Apple Music subscription at no extra cost, no video service States' Unlimited Basic /Plus Sprint /Premium 40. 401 Hulu limited Commercials plan per account T- Mobile I T-Mobile ONE 35! 35 Nerflix 2 -screen subscription at no extra cost on 2- qualifying lines Nae'EffWnr, cost tocustamr after 25% tariff diaoaan and mclui of VAT 44=11JONW40 for 6 hnes, aWlicade if A,nnPay and paprless baling are enabled Source. Ca,npany di Morgan Stanley Research M Morgan Stanley I aE --- Morgan Stanley I wr.,aara • 4. For fixed wireless access, is cable pricing expected to be the benchmark? Yes, we think so. Verizon has pitched its ini- tial 5G Home fixed broadband pricing at US$70 per month standatone, or US$50 per month bundled with wireless. The ser- vice is indicated to provide at least 300MBps with speeds peaking at over 1GBPs. These prices are broadly competi- tive with cable prices, and we would note that in some cases cable companies have introduced more aggressive pricing in response to the entrance of Verizon. B2B market 1. What are some of the 62B • opportunities 5G can offer? 0 Given the full scale network virtualization 5G is offering, the ability to support mas- sive machine - to-machine (M2M) communi- cations, and the ultra -low latency required for mission - critical applications, we believe the business opportunities for 5G are likely to be exponentially bigger than previous generation networks. Many upcoming industries related to the 4th industrial rev- olution (Al, cloud computing, drones, AR/ VR) wit[ require the enhanced functionabty 5G can offer, and hence white the imme- diate opportunities may be limited to a small number of areas, we believe revenue opportunities will emerge as technologies such as cloud -based gaming, autonomous vehicles, and remote surgery become mainstream- sz GLOBAL FOUNDATION Exhibit 39: Verizon versus peers: Price plans in four launch markets for 5G broadband service Location Provider Speed (Mbps) Price Details Sacramento Verizon 1 300 $501 Existing wireless customers Verizon 300 $70 non - Verizon customers Comcast 50 $50 No Agreement Comcast 60 $30 First 12 months; lyr agreement Houston Verizon 300 $50 Existing wireless customers Verizon 300 $70 non - Verizon customers Comcast 150 $50 First 12 months; 1yr agreement Comcast 15 $30 Los Angeles Verizon 300 $50 Existing wireless customers Verizon 300 $70 non - Verizon customers Cher art 100 $45 First 12 months; lyr agreement Indianapolis Verizon 300 $50 Existing wireless customers Verizon 300 $70 non - Verizon customers Charter 300 $45 Comcast 250 $60 First 12 months; lyr agreement Comcast 150 $50 First 12 months; lyr agreement comcast 25 $30 First 12 months; lyr agreement S....: Company data. Morgan Stanl, Re.,d Exhibit 40: Potential 5G use cases Delay AlllOtglnpl5 Cd y 1acbip lms driving. dy =Internet Virtual geau,y toms I'dL Disaster Peal time Multrpenon JW alert gaming video can Automotive BFdimctiowl r<4 remowconwlir v Oaths Device It remote • First responder controlling connectMry 1PxxrS Persona,• �WirWM loud MonBwlrg do d y bawd Mice wnsor networks stwu ing SP -V that Can be delmwed • Fmd $ Npmddx, i OntlAgo bliegM MnYorks ❑ServKWINtcouldbe Whlccnrxctwfty enabled by SG Sowce Gsw Bandwx th throughput • • • Morgan Stanley I aE,E. 2. What is network slicing? Network slicing is the concept of creating and operating multiple virtually independent 'Logical networks on a shared or common physical infrastructure. As service requirements of business verticals as well as individual customers differ (in terms of bandwidth, Latency, data security, coverage, etc), network slicing can enable a network operator to customize the design of individual Logical networks or'network slices' to deliver dedicated services to specific verticals or customers. According to GSMA, "network operators could deploy a single network slice type that satisfies the needs of multiple verticals, as well as multiple network slices of different types that are packaged as a single product targeted towards business customers (a business bundle) who have multiple and diverse requirements." This can enable the operator to achieve economies of scale through efficient utilization of network infrastructure while enhancing user experience on the dedicated network slice. Exhibit 41: Visualization of network slicing GLOSAL FOUNDATION 3. What kind of business models can network slicing bring to the market? Similar to how cloud services have allowed corporates to access computing power on demand, we believe network slicing is Likely to open a whole new market for data services by enabling businesses to lease bandwidth varying on parameters such as speed, Latency, pri- ority, and throughput. In particular, we believe the value proposition network slicing can offer will be of significant value for mission -crit- icaL applications such as autonomous driving, remote surgery, mili- tary communications, and emergency networks. While most of these industries will take time to roll out, the growth opportunity should become much clearer over the next 5-10 years. In contrast to the previous generation networks, where the net neu- trality framework and technological challenges made it difficult for operators to price their networks by application, in the SG era we expect operators to have both the Legal framework and the tech- nology to price the network by slice. From another perspective, net- work slicing is expected to change the current standardized connection service model to a customized service model. This would be a major paradigm shift in how telecom services are priced, and the much enhanced flexibility with service offerings should in theory lead to greater pricing power. We believe this is a key area which the market could be underes- timating. SG networks need to serve Customers with very different needs utilltie5 {g}� Do Automotive R r (gad r Manufacturing 5G Network Sr networks subdivided into virtual networks each optimised for one business case -- Utilities k r YM Automotive YY RY r O ii Y li Manufacturing 5G Network ,al let 0. xara ,ind". f Lay Iat.r"viw S ,. GSMA In a May 2018 report on the SG opportunity from autonomous vehicles, we estimated that autonomous vehicles will require 3 gigabytes of data per hour (GB/hr) for Level 1 vehicles and up to 50 Gl for level 5, making access to the cloud critical. As con- stant connectivity will be crucial to provide real -time information to navigate vehicles safely, securing sufficient network capacity with the required bandwidth, Latency and speed will be key to offering self - driving functionality. This is a clear example of just how important network slicing will be, and what kind of revenue opportunity it presents for the tetco operators. At the 2018 GSMA Shanghai conference, potential 5G enterprise customers such as PetroChina and China Southern Power Grid mentioned that they would require a dedicated layer of the 5G network for their specific demands to guarantee reliability. Morgan Stanley I aE,r. • Exhibit 42: Connectivity platform for AVs on a network slice • ._ MEC Servers installed at base station Swore Morgan Stanley Res Network shC with guaranteed ultra-low latency and high throughput sold at premtom pncirrg 4. What kind of pricing structure can we expect with B2B? Unlike retail. pricing, where regulations play a part in the overall price - finding process, B2B pricing is decided purely on commercial terms. This is good for markets like China, Japan, and Korea, where politics is Likely to have a smaller impact and operators can focus on projects which offer high returns on capital. The wiLLingness of enterprise cus- tomers to pay will depend on the value they can create from 5G ser- vices, and we see strong opportunities from customers that will reap major cost savings, such as in manufacturing automation. With 5G allowing networks to be 'managed', we believe B2B solu- tions will most LikeLy be packaged as a total solution. Some of the more complex B28 projects on the current 413 networks are already •sold to consortiums consisting of network access, hardware and sys- tems software. For SG, the ability to manage network parameters, NeFwork sl=ce 1 GLOBAL FOUNOAP,ON Sfandard hest etron ne*Wo sc along with emerging technologies such as edge computing, is LikeLy to add another layer of sophistication. We believe bidding through a consortium structure is Likely, to be the norm, which could allow operators to charge higher pricing than for plain vanilla leased Lines. In theory, services Like network slicing should command a premium, as operators are committing bandwidth to the client, implying a higher opportunity cost. That said, as corporate accounts tend to be stickier, especially for customized services, we believe competition could be fierce forfLagship accounts li.e. a scenario where operators compete to be the primary network service provider forTesla's self - driving service). We also believe that for mission - critical accounts, a dual vendor policy could be the norm, as service providers will require an alternative option in case of emergency, and leverage this as a way to keep prices in check. Morgan Stanley I Pe$Exxaa GLOBAL FOUNDATION 0 Theme 5: Passive Infrastructure Remains Key Market view: Similar to the 3G and 415 cycles, telecom infrastructure companies (towers, fiber, data centers) shouLd see significant benefits, but there is a risk that spending will move away from some traditional assets Like towers. Our view: At a high Levet, we expect SG to leverage many of the same passive infrastructure assets that comprise a modern 4G network: towers, small cells, fiber, and data centers. Overall, we do expect the mix of spending to shift at the margin with a greater emphasis on fiber and small cells, particularly in denser urban areas. However, the broad infrastructure Landscape should see an overall benefit as many SG use cases require significant coverage in suburban and rural markets as well. Potential infrastructure winners: Tower companies (Al CCI and SBAC in the US; China Tower in China), fiber companies QAYO in the US), data centers (DLR and CONE in the US; GDS, VNET and Sinnet in China), and network construction companies (CCS in China). We expect 5G to Leverage many of the same passive infrastructure assets that comprise a modern 4G network: towers, small CELLS, fiber and data centers. In the US, we have a strong preference for communication infrastructure assets that benefit from 5G network investments. These incLudes the tower companies (AMT, CCI, SBAC, fiber companies (ZAYO), and data centers (DLR, CONE). We think small Cells will become increasingly important in a 5G world, and note that Crown Castle is the best way to gain exposure. In China, we think China Tower and China Communication Service (CCS) are well positioned to benefit from denser 5G networks and high investment costs. 0 What does SG need? Similar infrastructure, different mix: To meet the higher capacity and Lower Latency demands of a 5G network, we expect the focus/mix of investment to change along with the ultimate density and Location of infrastructure assets. Carriers will tailor their network deployments to individual areas, as noted by AT &T: 'Our 5G deploy- ment strategy will include using millimeter wave spectrum to deploy SG in pockets of dense areas —where demand on our network is high and extra capacity and coverage is needed most. In other parts of urban areas and in suburban and rural areas, we plan to deploy 5G on our mid- and Low -band spec- trum holdings.' We highlight the following global public SG players for different types of infrastructure: I, ,I u Exhibit 43: Global public 5G infrastructure players Crown Castle CCI t® ® 11D 411111 Rai Way SBA Comm. SBAC 40 4D El Towers Ell Zayo ZAYO Interxion INRN Europe Unit, UNIT r SIEMENS India Mingo WIFI TBIG.IK Indonesia all EOIx Indonesia - Netlink NBN Digital Realty DLR Telesites SITESB1.Mx Cyrul CONE Source Marian Shelay Research Note: AMT. SBAC, ZAYO. UNIT. MIX. DLR, CONE each have expos,xe onside the US, AMT a SBAC ass, have small exposure to Small Cells Is DAS, Male AMT owns fir assets in some mtenia ltermories OTS GETS Celeste Cox Switch SWCH China China Tower 0788. HK 21V,anet VNET GDS GDS Inwt INW.MI Italy ® 11D 411111 Rai Way RWAY.MI Italy 40 4D El Towers Ell Italy 4 Interxion INRN Europe Ill Infratel SIEMENS India Tower Bersama TBIG.IK Indonesia Saran Menara TOWRJK Indonesia - Netlink NBN NETLSI Singapore Telesites SITESB1.Mx Mexico 40 Source Marian Shelay Research Note: AMT. SBAC, ZAYO. UNIT. MIX. DLR, CONE each have expos,xe onside the US, AMT a SBAC ass, have small exposure to Small Cells Is DAS, Male AMT owns fir assets in some mtenia ltermories Morgan Stanley I ^rrE• • Exhibit 44: Macro towers are typically 100+ ft structures that serve as the edge of wireless networks Types of Towers a Monopole s 100 -200 feet f Typical use. telephony • ..OJICe. AmP,ie'Sn 9wP Lattice t 200 - 400 feet I Also called self - support > Typical use telephony Guyed s 200 - 2,000 feet f Typical use televivon and radio broadcasting. paging and telephony Macro towers: A wireless tower is a vertical structure built to accommodate telecom communications equipment, incLuding base stations. The tower serves as the edge of a wireless network as the first point of connection for end -user devices and remains the most cost - efficient manner in which to deploy wireless spectrum. The structures are unlikely to change in a 5G environment though the equipment deployed to them will evolve to make use of the Latest spectrum bands and developments related to things such as carrier aggregation, massive MIMO and beamforming. Carriers are Likely to deploy mid- and low -band spectrum to macro sites in order to pro- vide coverage for 5G use cases such as IoT and autonomous driving. Small cells/DAS: Similar to towers, small cells and DAS (distributed antenna systems) serve as the edge of a wireless network in higher - demand environments such as dense urban areas and sports venues. They operate as a way to optimize network performance and offload demand from overburdened macro sites. The antennas and other equipment are typically deployed to existing urban 'furniture' such as Light or utility poles with antennas operating in high -band or mmWave spectrum. Current 1+G deployments are somewhat Limited •and geared towards providing additional capacity in areas where towers maybe overloaded as carriers seek to densify their networks. However, the expectation is for a significant growth in depLoyment over the next 10 years focused on dense urban environments. Per the Stealth s Range in sure I Generally used to maintain aesthetic quality of area s particularly useful In areal sdh strict zoning regulations CTIA in the US, S&P Global Market Inte ftence projects that more than 800k smaLL cells will. be deployed by 2026 vs. around 100k today. In China, small -cell deployments did not begin until this year, but China Tower has expressed optimism over potential growth in a 5G environment and has already secured over 10mn sites for small cells (mainLy municipal Light poles). At a high Level, smaLL cells can provide coverage for 100+ users in about a 1,000+ ft. radius vs. a macro tower site that can support 1,000+ users in a 1-2+ mite radius (these numbers can vary meaning- fully based on specific depLoyment environments and we note that small cells and towers are more typically viewed as complements than substitutes). Cost also remains an important factor with Crown Castle in the US notingthat anchor builds can cost US$100k per node on a gross basis and US$70k after carrier capital. contributions with the bulk of cost being fiber investment, while noting deployment typically takes 18 -24 months. A recent report from the ITU outlined muLtiple deployment scenarios with average capex per site in the US$50 -60k range, comprising 50% implementation costs (design, permitting, etc.) and 25% each for RAN equipment and fiber (see RU — Setting the Scene for 5G). By comparison a multi - tenant macro tower costs US$250 -300k on average in the US and can support over 10x the number of users. MorganStantey I ^E.E. GLOBAL FOUNDATION • Exhibit 45: Small cells combine fiber assets with existing urban infrastructure to provide network capacity What Are Small Cells? Fiber fed small cells enable wireless carriers to add much needed coverage and capacity to relieve congestion on their networks at ark€ sr 3.�saarn�. ME Fiber: Fiber optic cable provides the backhaul from the network edge (towers, small cells) to carrier aggregation points/central offices and uttimately the mobile network core. For small cells, fiber represents a significant portion of the capex discussed above. In a 5G environment a higher density of fiber, as indicated by fiber strand counts, is needed as additional. edge nodes/equipment are deployed to serve higher demand and meet the network capacity requirements of 5G standards. Crown Castle has discussed deploying denser strand counts of 288+ given that each smaLL -ceLL node requires two strands of fiber. Last year, Verizon discussed placing 1,700 strand cables on main feeder routes in areas such as Boston, which followed a US$11bn contract with Corning to purchase up to 12.5mn miles of optical. fiber annuaLly from 2018 to 2020. 40 viii. ^i I-ina Plic+s K t )4 ifi d h e� .I el3'R' altitie2 t" j,r eo-+a`bi R4Y{gn. a Data centers (including edge data centers): Co- location data cen- ters are real. estate faciLities where servers, used to process data, are stored. There are fiber optic cabLes, as part of the broader tel.ecom- munications network, that connect to these facilities, to allow data to move to and from the relevant servers. In a SG environment, more data and processing requirements may to move closer to the net- work edge as carriers seek to fulfill. ultra Low Latency needs. This could drive increased investment in edge data centers Located in cen- tral offices or at the base of towers. Existing data centers architec- tures will remain important as the primary Locations for data processing. With the Larger amounts of data created as a result of 56, data centers provide central. Locations for things such as data Lakes, for such use cases as AI and autonomous driving. Morgan Stanley I axsl ^apa • Exhibit 46: 5G networks leverage 4G components but layer in higher density and new equipment technologies • Source'. EMF Elam (5G Explained- How 56 Worksl EAlbft 47: 5G networks will utilize macro towers and small cells Wireless dntrastucture: A Heterogeneous Network Ms a MR40 LTE -Rio and W ars^aw /!l it MMm Cell lt` RRH pOr Q OMldpor Small CNI fir f,' Soufrn porno (Small Cell Network and the Evolohoo of 5G) • Mott 9i6W OWdaor Oa5 ,{r J 4r, J fff oas Hua rm meool ons � � k ors s s NwtrEnd 1 .z Oi0110erve, Mc GLOBAL FOUNDATION Morgan Stanley i e•a•a•ae • Exhibit 48: 5G has evolved through densification and new equipmenUtechnology • • The Evolution of 56 Massive MIMO and Macro base station Denal ication digitat beamforming • Increasing frequency • Small cells • Up to b GHz beyond 2.7 GHz to 6 GHz •DAS •Array antenna • increasing bandwidth • LTE -U • increasing efficiency q0r" Seeroe_ Orem (Smell Cell Nearork arM fie Evolution of 5E Exhibit 49: 5G cycle expected to be weaker than past cycles given elongated catalysts t5 ti} 5 0 200, sless Infrastructure Spend ($bns) s o ; °et' GeC10 - 1 iG MGR DG g- 5G LAG0.119 _'T51`1x- 2005 2008 2013 2017 2021 2025 Source. Dell Oro Morgan Stanley Research estimates a 43 s Exhibit 50: Investment to be slower as use cases for 5G still years away from being realistic 60 eveaGlPa eo nabre OM W 14446 2ozo Massive lot nzsd MAR Autonomous Car vaniou 9K Video source Dell' Oro. Morgan Slanley Research 90 to 95 GHz 2.9 GHz available 70 to 85 GHz 10 GHz available 38 GHz 4 GHz available 2e GHz 2 GHz available Millimeter waves • Ultra -wide bandwidth • Ultra -high ffrroughput 020170erve. ln: When does 5G infrastructure need to be in place? As highlighted in our Global Insight published last year, the SG cycle should be longer than 4G with investment starting now but not hitting a full run rate until closer to 2023 (see GlobalTechnolo : Learning to Ride aSGCycle (15 Oct 201»): Our research indicates that the ramp -up of 5G will be more arduous than 413 due to: costs that need to come down 60 -70% (explored in our proprietary model), permitting issues, spectrum allocation, regulatory buy -in, and technical challenges. We do not expect these issues to be resolved by 2020, when sizable spend is expected to begin. Consequently, we believe the SG cycle will be more elongated with smaller initial wireless spend compared to previous cycles. This is at odds with the view provided by some that SG could have a synchronized ramp in more regions than for the 4G launch. Instead of traditional blan- keted network rollouts, we expect SG spend to start with fixed wireless use cases in the US and Europe, whereas use cases in China may be focused more on IoT. If we are wrong and the SG cycle proves faster than expected, itwoutd likely be driven by quicker rollouts in China spurring a global com- petitive reaction (given deployment and cost issues are not as large). Morgan Stanley I a .... I- GlobaLTelco Comps Exhibit 51: Global telco comps .�..�,.. CNU. yar 1. 151 IIA Ily 331 la BR 69Y 1. 1)% la DJ%I'. IsreX[vn CM 4,111 qy2 XYD 1]T; IIN! 31 am DR 3A1 Bm Om 13% I' GIw UOmn v.zv Xlm RZ ¢W 11. IA uN am ob +In IW 15. 2R On On 1. ]54 clv Tpp ''. m.SH XM ISD +le Ex as. W} ".em a aW ]r' xR 1.% ]Jw .1. al4 1 1m Im ASY an, IN Clw Cmvn. s.v ssei xl(p "'. lW .1 11Y Rs 1]3 tlr -'. em lA 0a i ! 1R aJA. i xn 3R R R cev !. lm 1M .1X ma ..I sm IN 2r ! "i Ifay 1]% BR 1S -H% M UuryNr Ttlamm itl..: pi rNO W Ioi 1R Ha Ha 1011 IOm' I. u AR !R !» xa 23 n ai Md! laas '. Tlp 110 '. tl.n Hm a 1e4 ]N I. � I ai 01. � 01A 651 01. '. % » F.,m . Frbl ]6Y rND MA IIB M HN Hn 94 j )a >N t en Ift 1. 33 ai OA NIRTM"ai "Arai R"as XY.p' 12. mW : 1R Ii ar. I. 63 " xA 1. 50 5» : 1i% 6R 211 xu PCGW Llp Nm9 Fay I. 1W 1]1 ' 11% ima n3� 53 as 115 :R an 6R I R +n I. ai R any. "Ary, fl9] XM' 3a ZY! 1R S23 ]IY ].b 6a oiv 11% ]5% 91» 2 03 i0 p .I- WyMVq 1(O2 a. 10W 1% 111. lid- !W 90 6RilT 2} 11%911 ]yyMy 1JI5 :a. ": 1520 1102 an IF 1. 4N -+m 1p IA 3n 3m 5]Y 71% ": IA11 If IOn 1211 5]Y EA hii YvN 2.1 INR : 6G 51] 1]% W YT I. I. ''. 131 In 'ey 1.» OR Ii). 15. NM -1. bq{pa iY¢n: 6pye. SCd! ]]0 ]0$ -, ISS 1)1. 11% 37%' S0% SBY IN A NNFYTM Svgpe UIrs SGL. OID ]] 10% .OS E B4! ]l1 SIY 6» 1A 1N M N% 1R! 4pn. 21Y 19[N' IA 1. I» 5 d Ir, IIN I I Ir' 2R dR an 02A 59 54 Ia. 31gae 11$i SW 2A0 210 Sw I.I. 21S 1m A. 21. 23 l" x 2R AR ]» 1} 11 10 1R NN aii 1W MYR! IW 1i% EN Iftl'. 1m Ia, IA Ba. 3» 111, 1. 19 1f% E1D NYys 2110 MM 3tl $1] M% 12N 1. 1.3 1 dR 7A ]» 12, -RM, Bl vY Berl:tle Mby 9910 MYq IW 535 N i19. Eh -.121. I2a a. 53 6A 03Y 3R ]]Y 5} Sm . R mrs MYI+m 1951 ''. MYR ]A 4p .114 Ha Etl 1211 a. I. )R 51346 AR !» Rar ai ''. 1% 2A A.-A FA Servo MYtl IU71 III 9 Y6.0 L% 10m IB@ 17. 16 ` 3R B.SY AIY lA " 0% 1Tf i laaal TuYtl IBfl '., TIB! !JA 305 IR 12r 13S! 10 IDem IIN Il3: 93Y 0.)11 60 0.0 IA Atl ''. » 1NA I.A -cmvn Tnrtl 3. 110 as 110 IS @S 10W Sm 54 2A 20 I.. In OR 00% 3m 3An ". e2% rve Gmp 1NYtl ''. am'! ",. TIff . SID SM 1x W W I. Th ''. 2M1 A3% ..110! OR 01 1m I. '..m% NA i An as. 1. Ud 3 a 1% :IDr lid: 1 Sh ''., 2A 2h 2R 40 % ".3» 3P JA 3% R i An lrmtl LWrtiM1 TB ! spar . 21W ! BR %tl R,. I. . 1 1. 1 +B A% faR'! ON In 09 Om a% an A1Me WMa I. IM ]20 21m 11% T1. HV A. )N I Ih 1S OR asy, BR 1 1. 10A 1A NM N9 WN YW.4 IM] '., pM BW !A ]AY : 1 3 aN BOY I Sm slv •R 1. 1336 7 h 31. $IM1 r0G "s A} 1J 1. +Oh Blv 1 am d0 iJ% ! A0 51Y ]3Y 1. PLW PXyeA A)D] p1p i5W 111] 21% : 1 >V'', SS 6.h $tl 23 1. -I S%i 5A 53Y -13A M .Ted[ ,- Ripes Amt '. A 1% 11m 1. a 2h : d:< 1A% 51 50 11% 1x i SRTWa;n 5. W1ev IDZ)0 i mLM ". ]m.V9 ED.OV "', R Sm aL :: 6m 6m IS Im 1. 01% 1]Y ]TI R An., 5 ai o 6aa I. XRW j ID.® � IDXO '., JY � 96 � 2tl 2.: 03 Or ` 9]% If i In Dd 0& Ia. R j lop. 5 Fna 67rY ''. qry/ -: Iv,® 1I1010 mx ixm 11m: 18 1a 0ID Oa ODA 2R JR 1a 1 A AY 0 1 1. .4OW m179 '. AUO '. 2m 31q '. In i 103 1. i 5} 5N "' 1. fA ' 100 50i 17T fin 1. dA IIx': rAGrelmn AWW mTS D am a -1R l 'Ar: 9m IDm 11. v I2M iJ% AR 0sK x3 zs 30% JJM, eu.s- AuYW Ias, p xm ]K m% a; 4n er 1h +r 1» $]x, DAx 0R Bp 9m -2% 1. sya%Ngv2abY • ryszWeY NEO' AW 111 ax ]S eh 1. 2S 2r 5 an 6m1 e11 1m In 3Y XT] !p: 16are 11 1 IA4 AR 93 ]A ]]Y ". ifo 1m 2%; I A, 30K A 311 0 XR O000M0 Jy. ..I JPY affair 2S] ! 17% ": 12N 12i1� 6Y rai '', fA 1h IIY 00 %', 1b R R A00 Jp1 mm0 AY j.1Y 21]5 )3% I. •]v ,� 4Y fA 1} 'A% 01%'. 44% 1P u B% iY }Y 1110 31% At% I]% 1l] $ N 1IY 11^ -: AT&T ps 1A2 VSD 3)W RY pm 0y ly BN 2S~ 23 66Y Vvan 212b1 U50 mO -1% 12m 13S 2h 2n 61% AA% 3L A [mNYR* as .1 ', USO "'. 33% ! 10a 2r A. is 8% % Fm:Y Im% - W 2N 21 RSY 1133 USCAYYr 18ID 5391 .l% Mh 14 IN 10 0» %i NA FonaS BY Us UW! IIW 12M 1R : W ]® 6 e 1 1S SR 0.Rj M M N B% he NLlievn LLW 1 \OW A23 \],Y ", W 3A x9. t. 1N 11 1.A! OR OD% RA sal 1.1 BGE I:e 3]m6 c10 •9.W 51m 15m 15Y 11A 62% ", SsY sR R%n{COmm CaaJ .T. 2)8 GO a9ID lOW M 16m 15s• 5Y ]Sa l4 29% tY III ]Y TELUS Cwp c.�} oe e CPU S1W ASM IOx 1. t5P lA x9. 2A e5% IS%', 1% 4R 2S Ygen Mai tf.M 3U4. ISA {SY -1.. 1N 60 p]% AtY ASY 1. 1.A iii 21% EBmaT/11NR' :. i 1. i BTGOUp Dk UA Ii In JK 10m 10r SS Sn 65% IAA IPS IAS 1% NT $19: {3 19 t9 1S IO% ]I% Il. ly 1 1as 1. 4R RIsi Ontr]YSe hlnp:y EUR a, JAS 30% YY i% ]R In 151 NN » "i aq 65] EUR WW AOY 15y 125: SS s.lv j 1h 50% ♦ SlA PZ T iMnu spa; qRi EUR! 9W 151 mY 17A 11h: SS 6N IN Ir 12» "i 5» 5» IN 15e » R ! I.. NIi 123G I EUR' M 051 M ltr OS 4m A. � xh 15 am li» OR 0R 04 Oh ' 1% IN i h- S.BSYN par -: WO WB ! qfv IS i BN 0.r 27v Z. ]A 13 %! AR 5\x R 1. AbNV Zi i15 CNF: 5150 @> 11% t6♦ IBh' 1S iq 2h 2h - 5]Y S1Y! IR $N -I% R APN WB:su.f 11. off I. 211 24% HS Ila' lY 2T : 2A; 2y ! 111% in 5a, A n 1. Aa 14% R Bep1n 6155 mD Ht » ISh ! 1Y 51 1h 1a ]R In flR BR 01. Olv » R all ffaala EUF" ISW 1019 !]Y 103: A3 !a "'. 1. IS SR ]I%' I 511 51x 2n i W", 111A 1.0 N.5 1a. aZ 3m : ]1%''. d% » T eYrGO+WnY se- 1 9020 SEA .c A,. 5T 16m IA4 04 9 IA I% 30% 5i% 5R $Y 2L 31Y 2Y T.1 S.esn )631 aEF 1As0 1011 J% 18& 1... 80 3 $4 5`% 9I% % 1n WYq... I {6 {ei qh 1{L pL. L.. +le tla eR lJ %; {]% D]% iL.. 21 .Y1j�jI.D�_� 51 1e.Y. 9491 .Ee mW 7.3 39+1 Isn 69 1]. ]e. 35. 5 +% 1- se, 51Y ,D. 93. 1% - .1. siv 6N 19.2% s. Lveo�M.Me sce" ns IISAi ' 4M IA 2% Mpo'.v Mpa -. \Rl ! Z 194 )y A» 53% ': 1N » i AIR Os OAR 111 .7 1R 11. vm 1ar 1.1.: H➢% 2].R 6.]11 0a, % 1 1» MNYTVYSyMn RI® U50 Is 10W BE 31% 9a ]3: 2W IY! 11R IIRi 9R SAK % -)w as- USD AW 38 an 1 W 15.1 1N 13 OIY alA IM ''. NM » SW T�Nan Lp y.BA1Aae L521 SAR 910 a]a R m• 10.a 1 5 s, 14. 13 13 : ]]Y 6» ". ABM DO% 3N 2m R j Y]N pwB b:Blpee ID 510 A I ass Ina AR 13S 13." 1 19 33x 101 %! B.» !R Ih 1]v ''. SR Team £vJ1p.a 1 W as, ayA 11Y ON OS 3R IAa B.TO 5% il{ 1Maen.Narym iS.I UC 12 1W 2» S In 6WI R ii.Na 23n )m 13010 , Ise% .h, EMLmannnlek wE J9a+9 Al. z3B 169 m - ImA11: N AJe ] A. i.] 1% e]% 1.1% 11% ].$A I. "I ]14 '! lAllnMrb ;mep Mw USG M ". 09K tia All 11 iAklov BneA Nb !9 0 S S M M e% fi Y AY R TN�h^n .q w N m M NA 1. 0% R 111enlme An • prua 2 'u . 2m M ._ 9. M A 41% 6x FMe IM x +. 't 5 . ]1 . Is, R l0%: 1K pY 4Y 1. Ybm: cevnlp :J6 SE M1 ._ Q' ]m Sm 57. 2. 2b Sa" 30 % "47% 17% 21. ]Y 3Y. W.I - - - -" III R, Ii li]e {L ..A 14 111 {1 %- _2.R. ].0 Ve er. an IRj Sollm, Company Cmd Morgan Slams, Research e = moo gars 5',anlay Research esnmates Pricing as of November 14, 2016 GLOBAL FOUNDATION Morgan Stanley I —.... 0 Appendix 1 — China • Spectrum In China, the MILT has identffied 3.3.3.661fz and 4.165.OG14z spectrum for SG, of which 3.3- 3.4GHz is only for indoor coverage. The regulator has also conducted consultation on allocating 24.75- 27.5GHz and 37- 42.SGHz for 5G, but with a relatively lower pri- ority. Some spectrum in the 2.6GHz band could also be available for 5G, including 2x20Mhz under - utilized spectrum previouslyaLlocated to CU and CT for TD -LTE, and 85MHz unallocated spectrum. Allocated by the media regulator SAPPRFT, the 700MHz band is cur- rently occupied by China Cable Television Network for TV broad- casting and is unlikely to be available for % in the foreseeable future. Administmtire spectrum assignment at low cost: Consistent with usual practice, we expect the MITT will assign 5G spectrum to opera- Exhibit 52: Spectrum allocation in China } LT CY CJkA7 C?N R CYS Z C%M-t 8'0 A8t 93: Sit [: GLOBAL FOUNDATION tors via an administrative approach without an auction. In ApriL 2018, the NDRC announced the annual usage fee for 5G spectrum with favorable discounts (free for the first three years and at an average 50% discount in the fourth to sixth years) to support SG develop- ment. Assuming each operator receives 100MHz in C -band and 800MHz in mm Wave, we estimate an insignificant annual spectrum cost of Rmb900mn (US$130mn) per operator. We highLight three possible scenarios for SG spectrum allocation in China: e Scenario 1(base case): CU and CT each receive 100MHz in 3.5GHz, whereas CM receives 200MHz in 4.9GHz and some spectrum in 2.6GHz. Underthis scenario, CU and CTbenefit from capex savings via network sharing using adjacent spectrum bands. CM could also save capex using lower spectrum bands, either from the ... 1G1_ `F: ti's 'W Im IM 1= 9 ilia ziX a, C FR xW x s� • 't AMPS b+dm, win 2 cvmeneleu. O i�IV.ffi LTE Scam WIT.C,,r PYdxia M ( l ( M ( TLL l 1G1_ `F: ti's 'W Im IM 1= 9 ilia ziX a, C FR xW x s� • 't AMPS b+dm, win 2 cvmeneleu. O i�IV.ffi LTE Scam WIT.C,,r PYdxia ilia ziX a, C FR xW x s� • 't AMPS b+dm, win 2 cvmeneleu. O i�IV.ffi LTE Scam WIT.C,,r PYdxia Morgan Stanley I Anse • beginning if 100MHz contiguous spectrum in 2.6GHz is immedi- ately available or after refarming existing 4G spectrum in 2.6GHz if <1001il is available. • Scenario 2 (bear case for CM): CU and CT each receive 100MHz in 3.SGHz, whereas CM only receives 2001vlHz in 4.9GHz. This sce- nario implies asymmetric regulation to favor CU and CT at the expense of CM in order to balance the industry s competitive Land- scape. • Scenario 3 (bear case for CT /CT): CM receives 200Mhz in 3.SGHz, whereas CU and CT each receive 1001vlHz 4.9GHz. This scenario implies asymmetric regulation to favor CM at the expense of CU and CT in order to accelerate SG development by allocating the best spectrum to market Leader CM. Exhibit 53: Assumptions for China's 5G capex Capex We think spectrum allocation is the biggest variable for SG capex forecasts. Apart from that, whether and how deeply CU and CT will cooperate on SG network construction is another factor to watch. Other moving parts include the Xi macro cell to be optimized bysmaLl cell, equipment pricing, and transmission capex. We expect China to see the biggest capex increase going into 5G among the four coun- tries, because 1) the Length of the investment cycle will almost double from six years under 4G (2013 -18) to 10 -12 years under 5G, 2) China will not enjoy an equipment cost advantage after switching from being a follower to a Leader, and 3) the government has aggres- sive targets for SG. Dealt active cooperation between CU Cooperation between CU and CT; CM limited scale of sharing behvear CU Specimen and CT; CM capes saving ban relies on 4.921 Z In early stage, and anal CT. CM mainly relies on 4.9GHz for deploying 2.6GHz than saves "pax by adopting 2.6GHz in 5G later cycle Whole cycle "pax (2019 - 2030). Rmb bn, including China Tower • as s. 4G Capes (2019.2025) Rmb M, indu6ng China Tower % vs. 4G Capes to sales (operation; only) No. of 53 m. BTS (20192021) t/o. of 53 macro STS (2030) % vs. 4G current No. BTS Lag term % macro 00 to be optimized by small "11 Transmission "pax Indensity Source: Mogen Stanley Research esh,mahee 2224 2903 3976 70% 122% 204% 1428 1899 2753925 9% 45% 110% Olgavalors 5G "pex to sales ratio will Operators' SG capex to sales ratio will Operators' 5G ca to sales ratio will n"M1Om8%in2019upto16 %inl dsefmm11%m2019upto23 %in rise from 13% in 2019 up 10 43 % in and Men decline greduagy m 7% in the 2022. and Uhen decline gradually to 9% 2= and then dedirha gradually 1011% OIta un In Me long ran In m810 = 75K/250K/500K 1501(1500K71mn 200K/1mN2mn 4mn 5m 7m0 largely the same 25% higher 75% higher 30% 20% 10% 5E %servi "revenue 74 %seMca revenue R,10% service revenue Exhibit 54: Total China 5G capex forecast (2019 -30): Base, bull, bear cases 800 700 600 500 200 100 BA Bee Bear aCapex(LH8) scyex to service revenue (RHS) • S ci, Mogan Stanley Research Exhibit 55: China annual 5G capex forecast: Base case 30% too , 150% Base case'. 5G ca to slam from 2019, 45% pax 25% 80 psak in 2022; Twal SG cepax (201 &2030 to l 40% 70 M122% higher Irish 4G capex(201S20111) 35% 20% W 1 {: 30% 3 50 25% s% ; at 10% & I! 15% 20 0% 5% 10 5% 0% 0% w � 5 p w w w o w w . R iv aaaaaGtiN 5G cipax —as %sehvim revenue Snare Morgan Stanley Research Morgan Stanley i Ee • Exhibit 56: China annual 5G capex forecast: Bull case 00 50% 90 ! Bull Cese'. } 45% 1) Deep edits poopealion beleeen CU and CT: 4 CM apex saving fmm deploying 5G m 2WG 40% 70 band 35% BO I 30% m sB is% 40 .i 30% 30 15% 20 10% 10 { 5% o% m w w w w w w w w w R ,� CAIN 5G spell —ae X vervia revena Sours: Morgan Mwlry Reward, Applications What are Chinese telcos working on? IoT is the key focus for Chinese operators, who are not concerned with fixed- wireLess given comprehensive fiber coverage. They are attaching high importance to 5G applications in their trial programs in 17 cities in 2018. CM is •conducting appLication trials in 12 cities, in addition to network triaLs in five cities. It estabLished three research institutions and 14 open labs to work together with around 220 business partners on the SG application trials. CU established two 5G innovation centers to coop- erate with vertical Leaders and its strategic investors (ALibaba, Tencent, Baidu, and )D.com). CU has identified 10 key applications in the current trial stage, including smart cities and autonomous driving. CT indicated that it has already seen strong SG demand from enterprise customers, particularly related to IoT and industrial internet. What are the other participants in the value chain working on? Internet companies and auto makers are active in autonomous driving tests. The Beijing and Hangzhou governments have already granted approval for real road tests to Baidu, Tencent, Abbaba, Didi, Daimler, and Nio. Beijing currently has 123km of road open for auton- omous driving tests, particularly one in Fangshan with futl 5G cov- erage. State Grid Corporation of China and China Southern Power Grid have partnered with China Telecom and China Mobile, respec- tively, to publish a SG smart grid white paper. They explored the matches between grid demand and tetcos' SG network capabilities, Including network slicing, distribution automation, accurate Load control, usage information collection, and distributed renewable ispower supply. Di, the leading drone manufacturer in China, reLeased its outlook for drones in the SG era in 2017. China Telecom and Huawei conducted an end -to -end SG 360 - degree live transmission in April 2018. GLOBAL FOUNDATION Exhibit 57: China annual 5G capex forecast: Bear case 00 Bern cax. 90 1) Limited sole d eM N betxeen Cu a 00 CT; ]) CM maidy rely m 4 9GHZ Mr 5G 7g M m 50 30 M 10 - w w w w w w w w w w w w . Chine 5G cape. —es %oernm revenue Swace' Morgan 3lanl, Research Infrastructure 50% 45% 40% 35% 30% ZS% m% 5% 1M6 5% 0•b Communications infrastructure companies own assets such towers, smallcells, fiber, and data centers on a passive basis, leasingthem out to multiple tenants/customers including wireless carriers. China Tower has 97% tower market share in China and is viewed as the key beneficiary of a denser network going into the SG era, despite some uncertainty as to China Tower's market share in rolling out smaLL cells. We believe Chinese teLcos will mainly rely on macro cells in the early stages of the SG network roLlout, and weak propagation of the high frequency spectrum requires more towers to provide the same coverage vs. the previous technology cycLe, driving Long -term growth in China Tower's revenue and efficiency. Revenue upside wiLL be strongly Linked to the spectrum allocation results and Chinese telcos' network roLLout pace. The key risk to watch is the potential merger of CU and CT, and the rental formula renewal after 2022. China Communication Services (CCS): CCS is a specialized telecom network constructor for operators and non - operators in China, offering unique exposure to operators' increasing capex going into the SG era. We expect CCS's earnings growth to accelerate from high single digits in 2016 -18 to 11 -15% in 2019 -20. Morgan Stanley I PESt- w w w w w at w at at w w w • W. cell ailas ++ S Mi cell sate. S urm. Cwnpany date, Mwgan Stanley ResewM estimates Data canters: We think massive data processing and storage demands in the 5G era will further accelerate the cellular growth of data center demand. We believe both carriers and carrier - neutral players in the data center market are well positioned to enjoy this fast growth. CT is a leading carrier data center provider in China, fol- • • GLOBAL FOUxoAT1ON Exhlbit 59: CCS is the key beneficiary of increasing capex going into the 513 era 17.0% 5G m..Name revenue ead pm51 150x°' pmwlh fmm hqh eat& dq0 tp dguEb dips 110% 110% TIM i 5.0% m w at w w +CCS Revenue gmMh —CCS Pmrd prxaah sowce Company data Mwgan Stanley Research estimates Lowed by CU and CM. GDS is the leading wholesaleihyperscale carri- er- neutral player, focusing on cloud customers and big internet companies. 21Vianet and Sinnet are leaders in the retail data center market with more diversified customer bases. Morgan Stanley I an.MCH GLOBALFOUNDAiION 0 Appendix 2 — US Spectrum Exhibit 60: 4,95OMHz of mmWave spectrum to come to market in 2018 -19 MHz Offered The FCC kicks off its millimeterwave auctions on November l4, 2018 3,500 with the quiet period already in effect (see Auction 101: Entering the 3,000 Spectrum Frontiers at 28GHz, and Telecom Services & Cable/ 2.500 Satellite: SG Spectrum Prime r ). Although wedo not expect spending 2 .000 in this auction to be very substantial (US$300mn- US$1.3bn), this will 1.500 be the first of a series of auctions in 2018 and 2019 to free up spec- 1.000 trum as we prepare for 5G. In the first round, the FCC is auctioning off 500 850MHz of millimeter wave spectrum in the 27.50- 28.35GHz bands, 0 offering 63bn MHz -Pops Source FCC Nam 142013 `:H 2Oiii 2H2019 Giirvui 2A GHz 1_4_•tr -.. a &" -?SSE L'5- 42 %F!Li (Exhibit 61: AT&T vertz*W Sprint SG NATIONWIDE 56 MOBILE FIXED MASSIVE COVERAGE US spectrum auctions & processes overview MIMO • Budd"' SG netxy w • IautrlHna so nwbik • ta:a�c!1M Canmemial taue:. t.:" SG. 30=.ZDia, "Al. fO in 12 thln, lly rMd wuderr513 1n4 25GHz Auction Value Estimates endof7018 d0zHin Ortober2018 • 9dtier to lre WUn.t:.n; MHz Offered Frequency LOW Mid High Timing 28 GHz Auction 850 2750.28.35 GHz MM $80U__M_. M $13Bn Nov 2018 24 GHz Auction • 700 ............. 24.25 - 24.45 GHz .$_3.._00.. $2Bn Win W $78n _. 1Q 2019e 24.75 - 25.25 GHz 37,39 1,000 37.6 - 38.6 GHz &47 GHz 1,400 38.6 - 40.0 GHz $SBn $38n $5Bn 2H 2O19e Auction 1,000 47.2 - 48.2 Gilt C -Band 180 3.7 -4.2 GHz $1OBn $208n $308n 2019e CBRS 150/70/40- 3.55 - 3.70 GHz $2Bn $4Bn $78n 2020e Total - $328n - • 150 MHz AVaOade, 70 MHz to be auctilloW, 40 MH1 maximum to one party S rm FCC, Morgan Stanley IrenearO estimates Exhibit 62: Initial carrier approaches to 5G differ based on current spectrum portfolios AT&T vertz*W Sprint SG NATIONWIDE 56 MOBILE FIXED MASSIVE COVERAGE HOTSPOTS WIRELESS MIMO • Budd"' SG netxy w • IautrlHna so nwbik • ta:a�c!1M Canmemial taue:. t.:" SG. 30=.ZDia, "Al. fO in 12 thln, lly rMd wuderr513 1n4 25GHz nafionti in. in 2070 endof7018 d0zHin Ortober2018 • 9dtier to lre WUn.t:.n; • 5cq Mhr Ar:dy • Aedlnwu': w,e 5G FodovnM Ior Jditlonal in fin l it M19 f,•p::aye 5r; " ".— t I,, laund+es m 2;19 SC, lowitthq In 2019 Followedbyeddltionel 1,:53 IIt ue nml 1:. v" icduding 50 moWte 5 launches n ;; ",,tra ,wvhr • All operators are moving towards SG Everyone has different starting points No one has clear spectrum across all bands for SG today Source T Mobile Morgan Stanley I - • We estimate that US carriers will spend USS10 -42bn (US$26bn at the mid - point) on spectrum purchases between 2019 and 2021, with Verizon likely to be the biggest purchaser. We esti- mate that the upcoming auctions of millimeter wave spectrum in the 28GHz and 24GHz bands should raise around US$5bn. We are focused on several key variables to determine spectrum spending: l) The outcome of the proposed Sprint/T- Mobile merger. A merged entity would have less need for additional spectrum and could be forced to sell spectrum as a deal condition; 2) Dish's spectrum strategy. Dish has large holdings of unused spectrum which it pro- poses to build out over time with external funding. It is possible that Dish could look to do deals with existing carriers instead; 3) C -band monetization plans. The FCC is considering allowing satellite compa- nies to monetize some portion of the 3.7- 4.2GHz band for wireless services. This has significant potential value, butthe form and timing of any potential monetization remains somewhat fluid. Verizon (US$6.18bn in estimated spend): Verizon is generally seen as the most motivated buyer of spectrum due in part to its Low spectrum per subscriber ranking and significant focus on 5G. The company has built significant holdings in mm Wave and has the ability •to refarm lTE spectrum but appears to have interest in mid -band spectrum. On the other hand Verizon did step back from the AWS -3 auction when prices were elevated, preferring to focus on densifica- tion and deployment of carrier aggregation, L-AA and other tech- niques. Verizon is also actively seeking to delever its balance sheet. AT &T (US$2 -10bn in estimated spectrum spemQ: AT &T seems Likely to be interested in expanding its mmWave holdings in the upcoming auctions, but US$175bn of net debt and the ongoing deployment of 60MHz of fallow spectrum means the company is perhaps Less focused on spectrum acquisition than some of its peers, particularly given the recent Time Warner deal The company seems Less focused on the 56 fixed wireless opportunity than Verizon. T- Mobite (US$2 -12bn estimated spectrum spend): T- Mobile has been active in the CBRS and C -band proceedings and requested a Exhibit 63: Estimating spectrum spending for the Big 4 US telcos Auction Spending Estimates I$bn) Low End Midpoint High End venzon 6.0 12.0 18.0 AT &T 2.0 6.0 10.0 T- Mobile 2.0 7.0 12.0 •Total 0.0 1.0 2.0 Total 10.0 26.0 42.0 Sours Mp,.n S.I, Re., waiverto bid in the upcoming mmWave auctions. On the other hand, the merger with Sprint would put the combined entity in a very strong spectrum position atthe Low- and mid -band levels, with a very dense cell network. This merger would likely dampen their interest in mid -band spectrum auctions, although adding mmWave would likely still be attractive to the merged entity. If the deal does not go through, we would expect to see T- Mobile be more aggressive in pur- suing mid -band spectrum either via FCC processes or with owners of fallow spectrum. Sprint (US$0.2bn expected spectrum spend): Sprint has spent almost nothing on spectrum purchases over the past few years, skip- ping the 600MHz and 700MHz auctions despite Limited low band spectrum holdings. This reflects their elevated financial leverage, which is Likely to continue to Limit future participation in spectrum auctions, although Sprint does appear interested in the upcoming mmWave auctions and also requested and gained an FCC waiver recently. Sprint's 2.SGHz holdings mean it is untikelyto have as much interest in C -band. Capex We estimate that by 2030 there will be 135k SG macro BTS in the US, and around 20% of them will be optimized by small cells. We forecast a cost of US$27Sk per SG macro BTS, and a very slight 1% annual decline rate. We expect US telcos to spend 9.5% of mobile service revenue as capex for SG related transmission by 2030, which will start ramping up gradually from 0.5% in 2019. By operator, we assume AT &T and Verizon will each account for 1/3 of total capex, with T- Mobite & Sprint combined making up the remaining 113. Applications In the US, fixed wireless will be the initial revenue opportunity (via Verizon's SG Home) with mobile offerings following in 2019. If T- Mobile wins approval for its Sprint merger, the company is also Likely to introduce a broad fixed wireless solution, with a focus on rural markets. Over time, we expect IoT, connected cars and smart cities to contribute to revenue growth. From an industry perspective, we expect a broad impact, including in media/entertainment, trans - portation/Logistics/autos, healthcare, agriculture, and energy /utiti- ties. Morgan Stanley I stFa=re • Exhibit 64: T- Mobile believes the merger would significantly increase the com- pany's ability to offer 5G coverage • SC: Thnmghpvr he ( mend Pqn (2021) lMN 994 s. �w,fw _,wu nU... t ,n YT x1) r[^ fa tw Zw oe N IOAC H: A•nyf IY,wglpn t�l4tnr see r f rW V IUIM nF+Y ei p Source: T -M ob�k ca ^Pant data Exhibit 66: Global tower companies Global Independent Tower Companies (Sites 000s) 2,F GLOBAL FOUNDATION Exhibit 65: With the benefit increasing over time €tea tbY Eau .v: T seengb0al tr2 Cmered Pap. 12024) w. rev w. w w w sv .. aw � R.,Wxwp l \vu„lpa M4p1 So,sce T- Mobile company data China American Bhanf Crown SBA Cellnez Sarana Telesltes Tower Intent El Towers Rai Wai Unit Tower Tower Infratel Castle Comm Menara Bersama Source. Compar, data. Magee S nley Research Note Bhani whiner s site case inolixes peer 39,000 of bs awn towers arN the balanre form ns 42 %¢silty interest in hous Towers, the company is expected to have abort IfKIM sites upon cotipletan of its pendirg merger with lrydns Infrastructure plays Towers American Tower, Crown Castle and SBA Communications are the only public tower companies in the US and own a combined 95k macro tower sites, which represent 65% of the total supply. Today, wireless carriers in the US own a very smaLl number of macro sites: white wireLess carriers used to own a significant number of their own towers, they sold the significant majority in Large transactions over the Last 10+ years. In the US, tower counts are expected to grow •modestly in the 1 -2% range annuaLLy over the coming years, in Line with recent trends. While there is some risk of capex spend moving away from towers and towards small cells in a 5G environment, we believe population distribution characteristics provide significant ent. P a protection with about 85% of the US population located in suburban and rural areas that are typicaLLy most efficiently served by macro sites. Fiber & small cell providers Major fiber providers in the US include Crown Castle, Zayo and Uniti, though telecom and cable companies also own a significant amount of assets internally. Fiber owers are the primary deployers of small ceLLs with each node requiring two dark fiber strands. tnaddition to about 40kmacrotowers, Crown Castle owns 60k route miles of fiber primarily in the 25 markets and has over 60k small Celts depLoyed or under contract, which the company has indicated repre- sents roughly half the market, including carrier self performs. Morgan Stanley I aE.-. • Zayo is an international bandwidth infrastructure provider (US, Canada, Europe) and is the largest independent fiber company in the US. Key infrastructure components that Zayo offers to the SG eco- system include fiber (ie. dark fiber, fiber to the tower), colocation data centers. Like the independent tower providers, the national wireless providers are some of Zayo's largest customers. Uniti is a top ten owner of fiber, small cells and macro towers in the United States. Unit! owns 112k fiber route mites and 5.4mn fiber strand miles, in addition to 767 towers and 2.5k small cell nodes. Uniti believes SG networks will require dense fiber networks that driven high incremental margin revenue. For example, Uniti believes it can achieve 709495% incremental gross margin on its small cell nodes and 85 -100% incremental gross margin on its fiber. While fiber owners are generally the most logical providers of small cells given the intrinsic link between the two, we note that there are several companies providing networks, primarily indoor, that do not own underlying fiber assets. This includes Soingo as well as tower owners American Tower and SBA Communications, though we note it remains a very small part of the business for the latter two compa- • nies. Data centers Data centers have been beneficiaries of the increasing data requirements of enterprises (ie. video, storage, security) and 5G will extend and amplify these trends. Scale matters in the data GLOBAL FOUNDATION center industry as enterprises look to deploy multi - geography net- work architectures. All this requires significant capital expenditures. As a result, there has been meaningful consolidation in the industry amongst the carrier neutral providers, as regional providers are acquired by those with national presence and the telecom carriers look to exit the capital intensive industry. Several of the US -based providers have international presence (EQIX, DLR, CONE), which provides an advantage when dealing with enterprises looking for multi - national solutions. QTS is another carrier neutral provider that offers low -cost basis facilities nationally. Network planning & construction Companies such as MasTec and Dycom provide program manage- ment, engineering, construction, maintenance and installation ser- vices for telecommunications providers. Tower companies such as AMT, CCI and SBAC each have network service and site development businesses primarily related to their own assets with offerings related to site acquisition, zoning/permitting, structural analysis and equipment installation. Spectrum plays Wireless network capacity is built on the deployment of spectrum, while spectrum payments either in government auctions or private market transactions represents a significant capital expenditure for telecom companies in addition to my traditional infrastructure com- panies. Exhibit 67: Fiber can offer lease -up revenue that generates significant incremental margin - «bti «ri +t Tnr}ay Taliahassa� EN FlenCa Ilu I l In the US, severatcompanies have rights to significant amounts of spectrum and have expressed potential interest in selling por- tions. These include Dish Network (cur- rently in early stages of deploying an narrowband IoT network), Intelsat & SES (satellite companies with significant 3.7 -4.2 GHz/C -band interests) and Globalstar (a satellite phone company with ownership of 2.4 GHz/S -band spec- trum). - )a roe _... te) ,... "• ° ;' 32 97 960 ''''''``'''_ -$37.000 _11127.030 - RBr 000 Source. wry lnegw pramtatan (9 /6 /16),yieldcwate asannual gross margindiv Ibynet capex In the US, severatcompanies have rights to significant amounts of spectrum and have expressed potential interest in selling por- tions. These include Dish Network (cur- rently in early stages of deploying an narrowband IoT network), Intelsat & SES (satellite companies with significant 3.7 -4.2 GHz/C -band interests) and Globalstar (a satellite phone company with ownership of 2.4 GHz/S -band spec- trum). Morgan Stanley I Araca 0 Appendix 3 — South Korea Spectrum S6 spectrum auctions concluded for the 3.5GHz and 28GHz bands: In Korea, 5G spectrum auctions were concluded without much event, with the three operators spending W3.6tn in aggregate. Operators paid a total W3tn for 280Mhz on the 3.5GHz band to be used overl0 years, and W 600bn on the 28GHz band to be used over the next five years. Given slower - than - expected progress on capex rollout, however, we believe operators will most likely ask for an extension, especially for the 28GHz spectrum which was auctioned under a five -year term. Raaulatoro rariawina options for additional spectrum: The Ministry of Science has started the process of identifying bands for potential future use in preparation for increased demand for SG ser- vices. The 3.7- 4.2GHz band, which is reserved for satellite use, is .being reviewed by regulators, given the specific band has been stan- dardized by the 3GPP. The government is also looking to see if the 1.8/2.%.6GHz bands can be reallocated to SG use. • Exhibit 68: Korea: 513 spectrum allocation by operator 3.5Ghz 3.42 LGU` KT Band (80 Mhz) .,. (100 M Amount paid 810 968 Capex Starting from 2019, we assume 100% of Korean telcos' wireless capex, 40% fixed- One/fiber /transmission capex (not applicable to SKT), and 50% of other capex (mainly IT and infrastructure) will be for 5G. In our base case, the total 12 -year 5G cycle capex (including spectrum costs) is expected to be 42 %higherthan capex spent for4G roU out in 2011 to 2018. If excluding the upfront spectrum cost, we forecast equipment capex for 5G to be 30% higher than 4G. As spec- trum does need to be renewed after 10 years (five years for the 28Ghz band), we have assumed spectrum is renewed at a 30 %pre- mium.ln the bull case, we assume the 5G capex is spread out over14 years, which leads to less intense investment and with spectrum being renewed at the same price. In the bear case, we have assumed 20% higher capex for equipment and spectrum costs increasing by 80% upon renewalas demand drives more spectrum to be auctioned. SKT 0 Mhz) 1,219 28 Ghz 26.50 4 KT LGU SKT I Band (800 Mhz) e,- , " ". (800 Mhz) °,_ �s': (8 00 Mhz) «„ Amount paid 207 208 207 (KRW bn) S rre compaM dada. Morgan S nky Aea rch Morgan Stanley I ... E - 0 Appendix 4 — Japan Spectrum SG spectrum allocation: There are currently three bands that are seen as the main candidates for 5G use in Japan. These are the 3.6- 4.2GHz, 4.4- 4.9GHz, and 27.5- 29.SGHz bandwidths ( Exhibit 8). Recent news has also raised the possibility that the 27- 27.SGHz band- width may be considered as well, but the above three bands are widely known to be the top candidates. 5G spectrum allocation by the Ministry of InternaLAffairs and Communications in Japan is sched- uled for around the end of March 2019. In Japan, there is no auction system and the government is in charge of spectrum allocation. The decision on 5G spectrum allocation is thought to be based on the track record of 5G related use case experiments carried out by the three telecom carriers. Capex • The standard creation process around 5G in Japan is thought to be a combination of two methods. Japan aims for commercialization of 513 in 2020, and in the early stages 5G will run on a wireless network through collaboration between NR and eLTE, thus initially there will be a combination of 513 and 4G areas. eLTE will use the same spec- trum as 4G and is able to use 415 base stations with just a software upgrade. However, as eLTE only allows for same speed of 4G, a new radio system will be used to further increase the speed necessary to meet SG standards. As regards the standardization of new radio in Japan, it will be done in two phases. Phase 1 will focus on increasing speed (eMBB) and will occur in 2020, when Japan plans to start com- mercial SG service. Phase 2 will most Likely be around 2022 and will focus on low Latency and massive connectivity (uRLLC & mMTC). In phase 2, cooperation with eLTE will no longer be necessary and the SG servicewill run soLely on the new radio system, resulting in nation- wide 515 service. • GLOBAL FOUNDATION Applications The Ministry of Internal Affairs and Communications announced the official start of comprehensive 5G demonstration experiments in May 2017. This started a wave of use case experiments by Japan's three telecom carriers.The main experiments centered around trans- mission of high- resolution images and footage, autonomous driving, WAR, data transmission using moving vehicles, construction, and IoT applications (examples include remote operation of construction machinery, remote medical examinations, autonomous driving, remote robot operation, digital signage, 4K video conferencing sys- tems, etc). Use ow experiments by N7T Docomo: NTT Docomo has been actively involved in 5G application experiments with numerous use case tests conducted over the past few years. One notable experi- ment tested the remote operation of construction machinery using real -time imaging transmitted through 5G. This experiment was con- ducted in cooperationwith Komatsu and allowed for the gathering of real -time data from construction sites, along with analysis of that data and remote control of construction machinery in real time. Morgan Stanley I AE,E. • Exhibit 69: NTT Docomo: Real -time remote control and data transmission of construction machinery z y Real -time transmission of data and real -time remote control F5G Base Statior sour.: Nr DO . 5r so r GLOBAL FOUNDATION 5G Transmission Device (mobile station) KDDI has also been active in testing applications for 5G, with one notable experiment involving the transmission of 4K footage to 50 different tablets simultaneously as a new form of stadium entertainment. This technology would allow users to see on- demand 4K footage of the game • they are watching in the stadium from different camera angles. Exhibit 70: KDDI: Transmission of 4K footage to 50 different tablets simultaneously as a new form of stadium entertainment • Source: ODI Morgan Stanley I "v- • Exhibit 71: • u Notable 5G related use-case field trials by NTT Docorie GLOBAL F'XtxSkAT*% Use Case Category iDate Description of Experiment eMBB 05 -2017 NTT Docomo tests transmission of 8K Su erVision images using 5G with NHK. May -17 8K images from moving vehicles. NTT Docomo tests remote control systems for construction and mining equipment using 5G CriC 05 -2017 KDDI, Obayashi, and NEC use 513 and 4K 30 monitoring to remotely control construction with Komatsu. eV2X 11 -2017 NTT Docomo testing 5G to put connected cars into practica use in the Odaiba area. KDDI conduct successful live feed of 4K high - resolution video content in a baseball stadium NTT Docomo and NEC achieve successful 28GHz wireless data transmission between a 5G eMBB, uRLLC 04 -2018 base station and a 5G mobile station in 5G field trials using a car moving at 305kmlh. I Jun -18 KDDI conducts real NTT Docomo, AGC, and Ericsson achieve successful 5G communications using vehicle glass eV2X 07 -2018 mounted antennas for 5G connected cars. 28GHz band compatible antenna achieves BGbps super high-speed communications with high-speed vehicle. Source: Comp W Etta Magan S J, Research Exhibit 72: Notable 513 related use -case field trials by KDDI Use Case Category Date Description of Experiment KDDI and NHK conduct trials of 5G applications, achieve successful real -time transmission of eMBB, eV2X May -17 8K images from moving vehicles. eMBB, eV2X Dec -17 KDDI and JR East successfully test transmission of 8KAK video on moving trains using 5G. KDDI, Obayashi, and NEC use 513 and 4K 30 monitoring to remotely control construction Cric Feb-18 machinery. KDDI conduct successful live feed of 4K high - resolution video content in a baseball stadium MloT Mar -18 usino 50 5G tablets. CdC I Jun -18 KDDI conducts real -time 4K video transmission using 5G drone. Source Company dMA Magan 9aniry Resear& P Morgan Stanley I a r•a =a GLOBALFOUNDAtiON 0 Appendix 5 - Spectrum Exhibit 73: WRC -15's actions WRC•151 ow" on Madlugt•RSelpe Spectrum 3.3-3 4GHz Allocated for mobile and identified for 5G in almost 50 countries EE 3.4 -3. SGHz Gtobaly allocated for mobile and identified for SG 962 3.6 -3. BGHz 4 countries identified 3 6- 3.7GHz for 513 5.15- 5.295GHz Tabled an agenda item for WRC -19 Whether to extend or revise mobile allocation to portions of the band WRC -15 Decision on High4Sand Spectrum be be Studied for Consideration at .WRC -19 144 oband (60 of 34GHz. 04 of 3.63.8GHz) 24.25 -27.5 GHz Existing allocation to mobile an primary basis Sharing and compatibility studies for 5G identification 31.8 -33.4 GHz Consider a mobile allocation & 5G identification 37 -40.5 GHz Existing allocation to mobile an primary basis Sharing and compatibility studies for 5G ideniKication 40.5-42.5 GHz Consider a mobile allocation & 5G identification 42.5 -43.5 GHz Existing allocation to mobile on primary basis Sharing and compatibility studies for 5G identification 45.5-47 GHz Existing allocation to mobile on primary basis Sharing and compatibility studies for SG identification 47 -47.2 GHz Vodefoe, Consider a mobile allocation & 5G identification 47.2 -50.2 GHz Existing allocation to mobile on primary basis Sharing and compatibility studies for 5G Identification 50.4 -52.6 GHz Existing allocation to mobile on primary basis Sharing and compatibility studies for 5G identification 66.76 GHz Existing allocation to mobile on primary basis Sharing and compatibility studies for 5G identification 8186 GHz Existing allocation to mobile on primary basis Sharing and compatibility studies for SG identRcation Solace RM, 513 AMERICAS KT Exhibit 74: • 5G spectrum auction results EE 40(3.54-3.58GHZ) 385 962 40 of oband! Regulator 0811 haw Three 20 (346-3.40GHZ) 192 9.62 144 oband (60 of 34GHz. 04 of 3.63.8GHz) 116MHz in 3,&3 8GHz UK Apr -18 band, and 700MNZ to 02 40 (3 5.3.54GHz) 404 10.11 40 of obard ba auctioned Vodefoe, 50 (341- 3.46GHZ) 481 962 50 of obard SKT 100(3.83 .7GHz) 1,081 10.81 100 ofc -0and. 00D of mmWaw KT 100 (3.53.6Gib) 859 8.58 100 of cbmld, 000 of mmWaw LGU 80 (3.42-3.53141) 719 890 80 of oband. SOD of mmWaw Sand Kim Jun-18 SKT 800(26527.3GHz) 184 023 KT 800 (27.128AGHz) 185 0.23 -GU 500 (28.120.92114z) 184 023 Voole.. 90 of 16GHZ3.8GHz 224 249 90 of oband! Orange 60 of 36GHz- 3.8GHz 150 249 100 of o-band Spain Jul-10 Telef6racz 50 of 3.63Hz- 3.8GHz 121 242 90MHz of oband MasmoW NA NA 1JP. 80MHz of oband Telamm IMga 20 (70DMHz band) 586 29.3 20 of 700MHZ W of oband, 200 of morWave Vadalone 20 (700MHz bend) 589 29.4 20 of 700MH; 80 of ward, 2110 of MW. Iliad 20 (700MHz band) 583 29.1 20 of 700MH; 80 of band, 200 of mmWaw Telecom Itega W (3.7GHz bard) 1,460 183 Vodafone 80 (3.7GHz bard) 1,453 182 10-12. of reserve Iliad 20 (3.7GHz bard) 417 Mg "'Ice Italy 00-18 Wind Tre 20(17GHz bard) 417 20.8 80 ofwrq. 200 of mmWeW Tel. Itals 200 (28GHz band) 28 01 Vodefona 200 MHz band) 20 0.1 Iliad 200 (MHz band) 20 0.1 • Wi nd Tre 200 (28GHZ bend) 28 01 Faatwab 200 (26GHz band) 28 0.1 200 of ..WSW Sou. Companyda,a Morgan Stanley I ^raEPPCe • Exhibit 75: SG spectrum cost CM 3.04.OGHz: US$0.7mn/MHz/year (RmbSmn); Spectrum is allocated by administrative approach; China cu No Auction 4.045.OGHz: US$0.4mn/MHz/year (Rmb3mn): For 5G spectrum, free for first 3 years, and 75 %, 50 %, 25% discount �6.02iHZ USSO.07/MHz/year (RmbO.5mn) respectively for the next 3 years CT Verson US$12bn AT&T US$6bn U.S NA T -Noble US$7bn Spend US$ibn KT US$1.3bn South SKT US$1.Obn NA Korea 204MHz LGU US$0.9bn NTT Docomo C • source Company data Japan KDDI No Auction JPY15 -20bn par operator per year (US$130- 170mn) Soltbank source'. Company data mina NORC, Japan MIC. Morgan Stanley Research eaomalea Exhibk 76: Current spectrum allocation among telcos The total spectrum spending in 2019 -21, including mmWeve and potentially some C band US$1081mn for 3.6- 3.7GHz; US$183 for BOOMHz in 28GHz band US$1043mn for 3.5- 3.6GHz; US$183 for 80OMHz in 28GHz band US$902mn for 3.42- 3.SGHz; US$183 for 80OMHz in 28GHz bard Spectrum is allocated by government. The annual fee is for the usage of all spectrums a00MHZ 10 -2MHz 3G(CDMA2000p4G 900MHZ 204MHz 2G 6.2MHz 3G/2G 1NOMHZ 25.2MHz 213 3Da2MHz 4G/3G 20 -214HZ 4G 1 WOMHz 35MHz Kl(tD) 2100MHz 15MHz 3G rrDSCDMA) 25.2MHz 4G/3G 20.2MHz 4G 2300MHz 50MHz 4G (TD) 20MHz 4G no) 20MHz 4G (TD) 2800MHz 6DMHz 4G (rD) 20MHz 4G (TD) 20MHz as (TD) 800MHZ 5 #MHz 4G 15XQMHz 4G 1ox2MHZ 4G 900MHz I WMHZ 4G 1800MHz 15d MHz(UL) 4G 15z/MHz(UL) 4G lwMHZ CDMA/EVDO 20x1MHz(DL) 2WMHz (DL) 210OMHZ lwMw 20 #MHz 4G 20 #MHz 4G lWHz 4MG 230010Hz 27MHz Warp 27MW WBm 2500MHz 10�MNZ 4G 20x2MHz 4G 3.5GMZ 100MHZ 5G 100MHZ SG 8OMHz SG 28GHz BOOt4HZ 513 800MHZ SG 800MHZ 5G 700MHz IWMHZ 4G 10X2MHz 4G 1WMHZ 4G lWMHz 4G 10 #MHz 4G BOOMHZ 5 #MHz 3G (Comm") 5 #MHz 3G WOMHZ 15X210111Z 4GI3G 1500MHz 10,QMHz 413 1512MHz 413 10,QMHz 4G 1700MHz 2WMHz 4G 2WMHz 4G 15XLMHz 4WG IWMHZ 4G 2000MHz 20Y2MHz 4G3G(CDMA2000) 204MHZ 4G3G 20 #MHz 4G3G 25WMHZ WMHZ 4G (TD) 30MHZ 4G (TD) 35001AHz dOMHZ 4G (To) 40MHZ 4G (To) 4OMHz 4G (TD) a00 MHZ 31.5 MHz 700 MHz 29 MHz 22 MHz 10 MHz 8001850 MHz 22 MHz 24 MHz 14 MHz 17MAWS WMHz 47 MHz 41 MHz 1900 MHz 38 MHz 21 MHz 29 MHz 37 MHz 2300 MHz 20 MHz 2500 MHz 1 W MHz Morgan Stanley I xF - -- 0 Disclosure Section GLOBAL FOUNDATION The Information and opinions in Morgan Stanley Research were prepared or are disseminated by Morgan Stanley & Co. LLC ancVorMorgan Stanley CT.V.M. S.A. and/or Morgan Stanley Mexico, Casa de BoLsa, SA. de C.V. and/or Morgan Stanley Canada Limited and/or Morgan Stanley & Co. International plc and/orRMB Morgan Stanley Proprietary Limited and/or Morgan Stanleylvii Securities Co., Ltd and/or Morgan Stanley Capital Group Japan Co., Ltd. and/or Morgan Stanley Aga Limited and/or Morgan Stanley 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 Legal responsibility for its contents and should be contacted with respect to any matters arising from, or in connection with, Morgan Stanley Research) and/or Morgan Stanley Taiwan Limited and/or Morgan StanLey &Co lmemattonal pLc, Seout Branch, and/or Morgan StanLey Australia Limited (A.&N. 67 003 734 576, holder of Australian hnanciaL services License No.233742, which accepts responsibility forts contents), and/or Morgan Stanley Wealth Management Australia Pry Ltd (A.&N.19 009145555, holder of Australian financial services license No. 240813, which accepts responsibility for its contents), and/or Morgan Stanley India Company Private Limited, regulated by the Securities and Exchange Board of India CSEE10 and holder of licenses as a Research Analyst (SERI Registration No. INHOOOOOTI05), Stock Broker (BSE Registration No INBOTI0S4237 and NSE Registration No. INBANF231054231), Merchant Banker (SERI Registration No. INM000011203), and depository participant with National Securities Depository Limited(SEBI Registration No. IN- DP -NSDL -372 -2014) which accepts the responsibility for its contents and should be contacted with respect to any matters arising from, or in connection with, Morgan Stanley Research, and/or PT. Morgan Stanley Sekumas Indonesia and their affiliates (collectively, Morgan Stanley/). For Important disclosures, stock once charts and equity rating histories regarding companies that are the subject of this report, please see the Morgan Stanley Research Disclosure Websde at .. morganstanleycorn /researchdiscLosums, or contact your investment representative or Morgan Stanley Research at1585 Broadway, ( Attention: Research Management), New York, NY, 10036 USA. Forvaluation methodologyand risks associatedwith any recommendation, Win& orprice target referenced in this research report, please contact i he Cbem SupportTeam as follows: USiCanada +1800 303 -2495; Hong Kong +852 28485999; Latin America +1718 TS4 -5444 (U.S); London +44 (0)20 - 74258169; Singapore +65 6834 -6860; Sydney +61(0)2- 97741505; Tokyo +81 (0)3,6836- 9000 - Alternatively you may contact your investment representative or Morgan Stanley Research at 1585 Broadway, (Attention: Research Management), New York, NY 10036 USA 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 rat receive direct or indirect compensation in exchange for expressing specific recommendations or views in this report Simon Flannery; Yang Up; Seyon Park, YUnchen Tsai; TetsuroTsusaka. CFA; Gary Yu. Unless otherwise stated, the individuals listed on the cover page of this report are research analysts. • Global Research Conflict Management policy _. ... ........... ..... ... .._ _.. Morgan Stanley Re . -: !peen published in accordance with our conflict management policy, which Is available at w . mrganstanLeycorrVinstRaionaVresearcWconfliapobcies. Important US Regulatory Disclosures on Subject Companies As of October -' - .:'i3, Morgan Stanley beneficially owned 196 or more of a class of common equity securities of the following companies covered in Morgan Stanley Research: 21Vianet, Century -ink, Inc grown Castle Corp., KT Corp, Naver Corp, NCSOFT Corp, QTS Realty Trust Inc, Shanghai New Culture Media Group, Uniti Group Inc, Ve ozon Communications, Zayo Group Holdings, Inc. Within the last 12 months, Morgan Stanley managed orm- managed a public offering (or 144A offering) of securities of American Tower Corp., AT &T, Inc, Chmalower Corp Ltd, Crown Castle Corp, CyrusCine Inc, Equinix Inc, Frontier Communications Corp, Intelsat SA, Kakao Corp, Naver Corp, QTS Realty Trust Inc, Softbank Group, T- Mobile US, Inc, Vemon Communications. Within the Lastl2 months, Morgan Stanley has received compensation for investment banking services from American Tower Corp., AT &T, Inc., China Tower Corp Ltd, Cincinnati Bell Inc. Crown Castle Corp, CyrusOne Inc, Equinix Inc., Frontier Communications Corp, Kakao Corp, KDDI, Naver Corp, NTT, NTT DOCOMO, QTS Realty Trust Inc, Softbank Group, Venzon Communications, Zayo Group Holdings, Inc_ In the next 3 months, Morgan Stanley expects to receive or intends to seek compensation for investment banking services from 2IVlaneq Atlbaba Pictures Group Ltd, American Tower Corp., AT &T, Inc, BCE Inc, CenturyLlnk, Inc., China TeLecom, China Tower Corp Ltd, China Unicom, Cincinnati Bell Inc., Crown Castle Corp, CyrusOne Inc, Equinix Inc, Frontier Communications Corp, Globatstar Inc, Gogo Inc, HTHKH, Huayi Brothers Media Corporation, IMAX China, Intelsat S.A., Kakao Corp, KDDI, KT Corp, LG Uplus Corp, Naver Corp, Netmarble Games Corp, Nexon Co Ltd, NTT, NTT DOCOMO, PCCW Ltd, QTS Realty Trust Inc, Rogers Communications, Inc., SBA Communications, SK Telecom Co Ltd, SmarTone, Softbank Group, Sprint Corp, T- Mobile US, Inc, Telephone & Data Systems, Television Broadcasts Umited, TELUS Corp., Unti Group Inc, Venzon Communications, ViaSat Inc, Zayo Group Holdings, Inc. Within the last 12 months, Morgan Stanley has received compensation for products and services other than investment banking services from 21Vianet, American Tower Corp., AT &T, Inc, BCE Inc, CenturyLink, Inc., China Unicorn, China United Network Communications, Chunghwa Telecom, Ciminnat i Bell Inc, Crown Castle Corp., Equinix Inc, Intelsat SA, NTT, Rogers Communications, Inc., SmarTone, Softbank Group, Sprint Corp,T- Mobile US, Inc., Television Broadcasts Limited, TELUS Corp., Venzon Communications, ViaSat Inc, Zayo Group Holdings, Inc.. Within the last 12 months, Morgan Stanley has provided or is providing investment banking services to, or has an investment banking client relationship with, thefollowing company: 2Uaneq Alibaha Pictures Group Ltd, American Tower Corp., AT &T, Inc., BCE Inc, Centurylink, Inc., China Telecom, China Tower Corp Ltd China Untcom, Cincinnati Bell Inc, Crown Castle Corp. CyrusOne Inc, Equinix Inc, Frontier Communications Corp, Globalstar Inc, Gogo Inc, HTHKH, Huayl Brothers Media Corporation, IMAX China. Intelsat S.A., Kakao Corp, KDDI, KT Corp, LG Uplus Corp, Naver Corp, NetmarbleGames Corp,Nexon Co Ltd, NTT, NTT DOCOMO, PCCW Ltd, QTS Realty Trust Inc, Rogers Communications, Inc, SBA Communicat ions, SKTetecom Co Ltd, SmarTOne,Softbank Group, Sprint Corp, T- Mobile US, Inc., Telephone & Data Systems, Television Broadcasts Limited, TELLS Corp., Uniti Group Inc, Venzon Communications, VlaSat Inc, Zayo Group Holdings, Inc. • Within the last 12 months, Morgan Stanley has either provided or is providing nonJnvestment banking , securities - related services to and/or in the past has entered into an agreement to provide services or has a client relationship with the following company: 2lVianet, Abbaba Pictures Group Ltd, American Tower Corp., AT &T, Inc.. BCE Inc, Centurylunk, Inc, China Unicorn, China United Network Communications, Chunghwa Telecom, Guarded Bell Inc, Crown Castle Corp., CymsOne Inc, Equinix Inc., Frontier Communications Corp, Gogo Inc, Intelsat S.A., Naver Corp, NTT, NTT DOCOMO, PCCW Ltd, QTS Realty Trust Inc, Rogers Communications, Inc, SmarTone, Softbank Group, Sprint Corp, L Mobile US, Inc, Television Broadcasts Limited, TELUS Corp., Vedzon Morgan Stanley I .u—cm • Communications, VuSat Inc Zayo Group Holdings, Inc.. Morgan Stanley &Co. LLC makes a market in the semrmes of 21Vianet, Amencan Tower Corp., AT &T, Inc, BCE Inc, CenturyUnk, Inc., China Mobile Umaed, China Unicorn, ChunghwaTelecorn, Cincinnati Bell Inc, Crown Castle Corp., CyrusOne Inc, Frontier Communications Corp, GLobaLstar Inc Gogo Inc, KT Corp, Naver Corp, QTS Realty Trust Inc Rogers Communications, Inc, SBA Communications, SK Telecom Cc Ltd, Softbank Croup, Sprint Corp,T- Mobile US, Inc, Telephone & Data Systems, TELUS Corp, US Cellular Corporation, Verizon Communications, Vii Inc, Windstream Corp, Zayo Group Holdings, Inc. The equity research analysts a strategists principally responsible for the preparation of Morgan Stanley Research 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. Equity Research analysts' or strategists' compensation is not linked to Investment banking or capital markets transactions performed by Morgan Stanley or the profitability or revenues of particular trading desks. Morgan Stanley and its affiliates do business that relates to companies%instmments covered In Morgan Stanley Research, including market making, providing liquidity, fund management, commercial banking extension of credit, Investment services and Investment banking. Morgan Stanley sells to and buys from customers the securitiess,struments of companies covered In Morgan Stanley Research on a principal basis. Morgan Stanley may have a position in the debt of the Company or Instruments discussed in this report. Morgan Stanley trades or may trade as principal in the debt securities (or in related derivatives) that are the subject of the debt research report. Certain disclosures listed above are also for compliance with appUcable regulations in non -US jurisdictions. STOCK RATINGS Morgan Stanley uses a relative rating system using terms such as Overweight, Equat- weight, Not -Rated or Underweight (see definitions below). Morgan Stanley does not assign ratings of Buy, Hold or Sell to the stocks we cover. Overweight, Equal - weight, Not -Rated and Underweight are not the equivalent of buy, hold and sell. Investors should carefully, read the definitions of all ratings used in Morgan Stanley Research.lnaddaion, since Morgan Stanley Research contains more complete information concerning the analyst's views, investors should carefully read Morgan Stanley Research, in as entirety, and not infer the 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 holdngs) and other considerations. Global Stock Ratings Distribution (as of October 31, 2018) The Stock Ratings described below apply to Morgan Stanleys Fundamental Equity Research and do not apply to Debt Research produced by the Firm. • For disclosure purposes only (in accordance with NASD and NYSE requirements), we include the category headings of Buy. Hold, and Sell alongside our ratings of Overweight, Equal - weight, Not -Rated and Underweight. Morgan Stanley does not assign ratings of Buy. Hold or Sell to the stocks we cove[ Overweight, EquaL- weight, Not -Rated and Underweight are notthe equivalent of buy, hold, and sell but represent recommended relative weightings (see definitions below). To satisfy regulatory requirements, we correspond Overweight, our most positive stock rating, with a buy recommendation; we correspond Equal - weight and Not -Rated to hold and Underweight to sell recommendations, respectively. Data include common stock and ADRs currently assigned ratings. Investment Banking Clients are companies from whom Morgan Stanley received investment banking compensation in the last 12 months Doe to rounding off of decimals, the percentages provided in the % of totar column may not add up to exactly 100 percent Analyst Stock Ratings Overweight (O or Over) The stock's tonL return is expected to exceed the total return of the retevant country MSCI Index or the average total return of the analyst's industry (or industry team's) coverage universe, on a risk - adjusted basis over the next 12 -18 months. Equal-weight (E of Equal) -The stock's total return is expected to be in Line with the total return of the relevant country MSCI Index or the average total return of the analyst's industry (or Industry teams) coverage universe, on a riskadjusted basis over the next 12 -18 months. Not -Rated (NR)- Currently the analyst does not have adequate conviction about the stack's total return relative to the relevant country MSCI Index or the average total return ofthe analyst's industry (or Industry team's) coverage universe, on a risk- adjusted basis, over the next 12 -18 months. Underweight (U or Under) - The stock's total return is expected to be below the total return of the relevant country MSCI Index or the average total return of the analyst's industry (or industry • team's) coverage universe, on a risk - adjusted basis, over the next 12 -18 months. Unless otherwise specified, the time frame for price targets Included in Morgan Stanley Research is 12 to 18 months. Other Material Investment Services Clients Coverage Universe Investment Banking Clients (IBC) (MISC) Stook Rating Count %ofTotal Count %of Total I BC %of Rating Category Count %of Total Other MlSC Category Overweight/Buy 1157 37% 305 42% 26% 544 39% Equal- weight/Hold 1380 44% 335 46% 24% 632 45% Not- Rated/Hold 47 1% 7 l% 15% 7 0% Underweight/Sell 553 18% 82 11% 15% 220 16% Total 3,137 729 1403 Data include common stock and ADRs currently assigned ratings. Investment Banking Clients are companies from whom Morgan Stanley received investment banking compensation in the last 12 months Doe to rounding off of decimals, the percentages provided in the % of totar column may not add up to exactly 100 percent Analyst Stock Ratings Overweight (O or Over) The stock's tonL return is expected to exceed the total return of the retevant country MSCI Index or the average total return of the analyst's industry (or industry team's) coverage universe, on a risk - adjusted basis over the next 12 -18 months. Equal-weight (E of Equal) -The stock's total return is expected to be in Line with the total return of the relevant country MSCI Index or the average total return of the analyst's industry (or Industry teams) coverage universe, on a riskadjusted basis over the next 12 -18 months. Not -Rated (NR)- Currently the analyst does not have adequate conviction about the stack's total return relative to the relevant country MSCI Index or the average total return ofthe analyst's industry (or Industry team's) coverage universe, on a risk- adjusted basis, over the next 12 -18 months. Underweight (U or Under) - The stock's total return is expected to be below the total return of the relevant country MSCI Index or the average total return of the analyst's industry (or industry • team's) coverage universe, on a risk - adjusted basis, over the next 12 -18 months. Unless otherwise specified, the time frame for price targets Included in Morgan Stanley Research is 12 to 18 months. Morgan Stanley I xr • Analyst Industry Views Gi..O °.AL FOUNDATION 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 -Une (1): 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 (U. 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 or MSCI Latin America Index Europe -MSCI Europe; Japan - TDPIK; Asia - relevant MSCI country index or MSCI sub-regional index or MSCI AC Asia Pacific ex Japan Index. Important Disclosures for Morpn Stanley Smith Barney LLC Customers - ....._.._._ _...... _. ............. ....... - - - -- Important disclosures regarding the relationship between the companies that are the subject of Morgan Stanley Research and Morgan Stanley Smith Barney LLC or Morgan Stanley or any of their affiliates, are available on the Morgan Stanley Wealth Management disclosure websiteat wwwmorganstanley .con ✓ onUne/rewarchdisdosures. ForMorgan Stanley specific disclosures, you may refer tow . marganstantey .corrVmwarchdisctosures. Each Morgan Stanley Equity Research report Is reviewed and approved on behalf of Morgan Stanley Smith Barney U.C. This review and approval is conducted by the same person who reviews the Equity Research report on behalf of Morgan Stanley. This could create a conact of interest. Other Important Disclosures - - - - - _ ......... _ .. ...__._._ .. ....... -- ..- _- .. - - - - -- ... .................... _ .... - - .._.. - - -- Morgan Stanley & Ca International PLC and its affiliates have a significant financial interest in the debt securities of American Tower Corp., AT &T Inc, BCE Inc, Centuryt-ink, .lno, Cincinnati Beg Inc, Crown Castle Corp,Equmix Inc, Frontier Communications Corp. Intelsat S ,, Kakao Corp, KDDI, KT Corp, Rogers Communications, Inc., SBA Communication, SK Telecom Co Ltd, Sohbank Group, Sprint Corp, T- Mobile US, Inc., Uniti Group Inc, VlaSat Inc Morgan Stanley Research policy Is to update research reports as and when the Research Analyst and Research Management deem appropriate, based on developments with the issuer, the sector, or the market that may have a material impact on the research views or opinions gated therein. In addition, certain Research publications are intended to be updated on a regular periodic basis (weekly /monthly /quarterly /annua0 and will ordinarily be updated with that frequency, unless the Research Analyst and Research Management determine that a different publication schedule Is appropriate based on current conditions. •Morgan Stanley is not acting as a municipal advisor and the opinions or views contained herein are not intended to be, and do not constitute, advice within the meaning of Section 975 of the Dodd -Frank Waft Street Reform and Consumer Protection Act. Morgan Stanley produces an equity research product called a"racticatldeaf Views contained InaTacts"L dea on a particular stock maybe contrary to the recommendations or views expressed in research on the same stock This maybe the result of differing time honuxim, methodologies, market events, or other factors . For all research available on a particular stock, please contact your sales representative or go to Matrix at http //w ..morganstanteycom/mat.. Morgan Stanley Research is provided toourclientsthrough our proprietary research portal on Matrixandatso distributed electronically by Morgan Stanlerytoclients. Certain, but not alLMorgan Stanley Research products are also made available to clients through third -party vendors or redirtnbuted to clients through alternate electronic means as a convenience. For access to all available Morgan Stanley Research, please contact your sates representative or go to Matrix at http✓ /w ,morganstanleycom/matru. Any access and/or use of Morgan Stanley Research is subject to Morgan Stanleys Terms of Use (http //w .morganstanleycon✓tenns.htmO. By accessing and/or using Morgan Stanley Research,youam Indicating that you haveread and agree to be bound byourTerms of Use Chttp //w .marganstanley.comAerms.htmO. In addition you consentto Morgan Stanleyprocessing your personatdata and using cookies in accordance with our Privacy Policy and our Global Cookies PoBq(httplAv .. morgans tanleyconitpnvacyyledge.MmO, including for the purposes of setting your preferences and to collect readership data so that we can deliver better a nd more personalized service and products to you. To find out more information about how Morgan Stanley processes personal data, how we use cookies and how to reject cookies see our Privacy Policy and our Global Cookies Policy ( http// wwwmorganstanley .corr✓privacyyledgehtmO. If you do not agree to our Terms of Use and/or if you do not wish to provide your consent to Morgan Stanley processing your personal data or using cookies please do not access our research. Morgan Stanley Research does not provide individually tailored investment advice. Morgan Stanley Research has been prepared without regard to the circumstances and objectives of those who recelve it 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 an investment or strategy will depend on an Irivestorrs circumstances and objectives. The securities, instruments, or strategies discussed in Morgan Stanley Research may not be suitable for all investors, and certain investors may not be eligible to purchase or participate in some or all of them. Morgan Stanley Research is not an offer to buy or set[ orthe solicitation of an offer to buy or sell any secuntydnstrument or to participate in any particular trading strategy. The value of and income from your investments may vary because of changes in interest rates, foreign exchange rates, default rates, prepayment rates, secunties/Instmments prices, 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 securitiesfnstmments transactions. Past performance is not necessarily a guide to future performance. Estimates of future performance are based on assumptions that may not be realized. If provided, and unless otherwise stated, the closing once on the cover page is that of the primary exchange for the subject company's securities/mstruments. The fixed income research analysts, strategists oreconomists principally responsible for the preparation of Morgan Stanley Research have received compensation based upon various factors , •including quality, accuracyandvalue ofresearch,firm profitability or revenues (which indudefbedincome trading and capital markets profitabibtyor revenues ),client feedback and competitive factors. Fixed Income Research analysts, strategisu' or economists' compensation is not linked to investment banking or capital markets transactions performed by Morgan Stanley or the profitability or revenues of particular trading desks. MorganStantey I wr, =..= GLOBALFOUNBATION • The'Important US Regulatory Disclosures on Subject Companies' section in Morgan Stanley Research lists all companies mentioned where Morgan Stanley owns 1% or more of a class of common equity securities of the companies. For all other companies mentioned in Morgan Stanley Research, Morgan Stanley may have an investment of less than 1% in securties/rnstmmenis or derivatives of sewrties7instmments of companies and may trade them in ways different from those discussed In Morgan Stanley Research. Employees of Morgan Stanley not involved in the preparation of Morgan Stanley Research may have investments in securmes7instruments or derivatives of secuntiestinstruments of companies mentioned and may trade them in ways different from those discussed in Morgan Stanley Research. Derivatives may be issued by Morgan Stanley or associated persons. With the exception of information regarding Morgan Stanley, Morgan Stanley Research is based on public information. 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 Morgan Stanley Research change apart from when we intend to discontinue equity research coverage of a subject company. Facts and views presented in Morgan Stanley Research have not beenreviewedby, and may not reflect information known to, professionals in other Morgan Stanley business areas, including investment banking personnel Morgan Stanley Research personnel may participate in company events such as site visits and are generally prohibited from accepting payment by the company of associated expenses unless pre - approved by authorized members of Research management. Morgan Stanley may make investment decisions that are inconsistent with the recommendations or views in this report. To our readers based in Taiwan or trading in Taiwan secudties/instruments :Information on semmies7instmments that trade in Taiwan B distributed by Morgan Stanley Taiwan Limited (NSTL9. Such information is for your reference only. The reader should independently evaluate the investment risks and is solely responsible for their investment decisions. Morgan Stanley Research may not be distributed to the pubLic made orquoted or used by the public media without the express written consent of Morgan Stanley. Any non - customer reader within the scope of Article 7 -1 of the Taiwan Stock Exchange Recommendation Regulations accessing and/or receiving Morgan Stanley Research is not permitted to provide Morgan Stanley Research to any third party Oncluding but not limited to related parties, affiliated companies and any other third parties) or engage in any activities regarding Morgan Stanley Research which may create or give the appearance of creating a conflict of interest. Information on securibestinstruments that do not trade in Taiwan is for inforationalpurposes only and is not to be construed as a recommendation or a solicitation to trade in such secuntiesrnstmments. MSTL may not execute transactions for clients in these secundWinstruments. Certain information in Morgan Stanley Research was sourced by employees of the Shanghai Representative Office of Morgan Stanley Asia limited for the use of Morgan Stanley Asia Limited. Morgan Stanley is not incorporated under PRC law and the research in relation to this report is conducted outside the PRC. Morgan Stanley Research does not constitute an offerto sell or the scUctation of an offer to buy any securities in the PRC. PRC investors shall have the relevant qualifications to invest in such securities and shaft be responsible for obtaining all relevant approvals, licenses, verifications and/or registrations from the relevant governmental authorities themselves. Neither this report nor any part of it is intended as, or shall constitute, provision of any consultnny or advisory service of securities investment as defined under PRC law. Such information is provided for your reference only. • Mogan Stanley Research is disseminated in Brant by Morgan Stanley C.iM.M. S.A.; in Mexico by Morgan Stanley Mexico, Casa de BoLsa, SA. de CV which is regulated by Comision Nacinnal Banana y de Valoms. Fargo de Los Tamarindos 90, Torte 1, Col. Bosques de Las Lomas Flow29, O520 Mexico City; in Japan by Morgan Stanley MUFG Securities Co., Ltd. and, for Commodities related research reports only, Morgan Stanley Capital Group Japan Co., Ltd; in Hong Kong by Morgan Stanley Asia Limited (which accepts responsibilityfor its contents) and by Morgan Stanley Asia InternationaWmited. Hong Kong Branch; in Singapore by Morgan Stanley Asia (Singapore) Pte. (Registration number 199206290 and/or Morgan Stanley Asia (Singapore) Securities Pte Ltd (Registration number 2000O8434H), regulated by the Monetary Authority of Singapore (which accepts Legal responsibility for is contents and should be contacted with respect to any matters arising from, or in connection with, Morgan Stanley Research) and by Morgan Stanley Asia International Limited, Singapore Branch (Registration numberT11FCO2O7F} in Australia to 'wholesale clients' within the meaning of the Australian Corporations Act by Morgan Stanley Australia Limited A.&N. 67 003 734 576, holder of Australian financial services license No. 233742, which accepts responsibility for is contents; in Australia to Wholesale cbents'and'retaitclients ' wthin the meaning of the Australian Corporations Act by Morgan Stanley Wealth Management Australia Pty Ltd (ABN. 19 009 145 555, holder of Australian financial services license No. 24O813, which accepts responsibility for its contents; In Korea by Morgan Stanley & Co International plc, Seoul Branch; in India by Morgan Stanley India Company Private Limited; in Indonesia by PT Morgan Stanley Sekuntas Indonesia; in Canada by Morgan Stanley Canada limited, which has approved of and takes responsibility for its contents in Canada; in Germany by Morgan Stanley Bank AG, Frankfurt am Main and Morgan Stanley Private Wealth Management Limited, NiederlassungDeutschland, regulated byBundesanstaLtI Finanzdienstleistungsaufsicht (BaFin). in Spain by Morgan Stanley, S.V., SA., a Morgan Stanley group company, which is supervised by the Spanish Securities Markets Commission (CNMV) and states that Morgan Stanley Research has been written and distributed In accordance with the rules of conduct applicable to financial research as established under Spanish regulations; in the US by Morgan Stanley & Co. LLC, which accepts responsibility for is contents. Morgan Stanley &Co. Intemational plc, authorized by the Prudential ReguiaoryAuthorltyand regulated by the Financial Conduct Authority and the PrudentiaLRegulatory Authority. disseminates in the UK researchthat 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 is affiliates. RMB Morgan Stanley Proprietary Limited is a member of the ISE Limited and regulated by the Financial Services Board in South Africa. RMB Morgan Stanley Proprietary Limited is a joint venture owned equally by Morgan Stanley International Holdings Inc and RMB Investment Advisory (Proprietary) Limited, which is wholly owned by FlrstRand Limited The information in Morgan Stanley Research is being disseminated by Morgan Stanley Saudi Arabia, regulated by the Capital Market Authority in the Kingdom of Saudi Arabia , and is directed at Sophisticated investors only. Morgan Stanley Hong Kong Securities Limted isthe liquidity provider /market maker for securities of China Mobile Limited, China Telecom, China Unicorn Listed on the Stock Exchange of Hong Kong Limited. An updated List can be found on HKEx website: http //w ..hka.com.hk. The Information In Morgan Stanley Research is being communicated by Morgan Stanley & Co Intemational plc (DIFC Branch), regulated by the Dubai Financial Services Authority (the DFSA), and is directed at Professional Clients only, as defined by the DFSA The financial products or financial services to which this research relates will only be made available to a customer who we are satisfied meets the regulatory criteria to be a Professional Client. The information in Morgan Stanley Research is being communicated by Morgan Stanley & Co. International plc (QFC Branch), regulated by the Qatar Financial Centre Regulatory Authonty (the QFCRA), and is directed at business customers and market counterparties only and is not Intended for Retail Customers as defined by the QFCRA. •As required by the Capital Markets Board of Turkey, investment information, comments and recommendations stated here, are not within the scope of investment advisory activity Investment advisory service is provided exclusively to persons based cn their risk and income preferences by the authorized firs. Comments and recommendation gated here are general in nature These opinions may not fit to your financial status, risk and return preferences. For this reason, to make an investment decision by relying solely to this information stated here may not bring about outcomes that fit your expectations. Morgan Stanley I w =•r - -51..09 <L �ni NOAP,ON • The trademarks and service marks contained in Morgan Stanley Research are the property of their respective owners. Third -party data providers make no warranties or representations relating to the accuracy, completeness, or timeliness of the data they provide and shall not have liability for any damages relztingto such data The Global Industry Classification Standard (GICS) was developed by and is the exclusive property of MSCI and S&P. Morgan Stanley Research, or any portion thereof may not be reprinted, said or redistributed without the written consent of Morgan Stanley. Indicators and trackers referenced m Morgan Stanley Research may not be used as. or treated as. a benchmark under Regulation FU 2016/1011, or any other similar framework. INDUSTRY COVERAGE: Greater China Telecoms and Media COMPANY (TICKER) RATING (AS OF) PRICE` (11/16/2018) Ansel Lin Yang Uu 21 Wanet (VNETO) Alibaba Pictures Group Ltd (I060.HK) E(05/24/2018) HKS116 0 (03/19/2018) Alpha Group (002292.SZ) U (02/04/2016) Rmb6.68 HK$6.90 Beijing Enlight Media (300251.SZ) +t Rmb8.03 HKBN Ltd (1310.HK) Ciwen Media Co Ltd (002343.SZ) E (08/23/2018) Rmb13.26 0 (01/08/2016) Huayi Brothers Media Corporation (300027.SZ) U (06/17/2016) Rmb5.48 HK$2.92 Shanghai New Culture Media Group (300336.SZ) E (03/09/2018) Rmb5.15 SmarTone (031 5.11K) Television Broadcasts Limited (0511.HK) E (01/06/2016) HK$17.36 U (03/15/2017) Zhejiang Huace Film and TV (300133.SZ) 0 (01/29/2018) Rmb1 0.31 Gary Yu - Historical prices are not split adjusted. ......._.....__.. ----- __.._._. .............. _ _____ _... China Mobile Limited (0941.HK) ..__._._.__ _........ ... ...........__... U (07/05/2017) _.._....... ....._......___.___. - HK$76.85 China Telecom (0728.11K) 0 (10/30/2016) HK$4.06 China Tower Corp Ltd (0788.HK) 0 (11106/2018) HK$1.18 China Unicorn (0762.HK) 0 (03/27/2018) HK$8.80 United Network Communications (600050.SS) U (10/30/2016) Rmb5.59 •China Chunghwa Telecom (24123W) U (08/08/2018) NT$106.00 Far Easton (4904.TW) E(08108/2018) NTS73.70 [MAX China (1970.HK) O (05/1912016) HKS19.00 Taiwan Mobile (3045.TW) E (08/08/2018) NT$109.00 Wanda Film Holding Co Ltd (002739.SZ) 0 (10/1512015) Rmb23.56 Yang Uu 21 Wanet (VNETO) 0 (10/04/2017) US$11.71 Beijing Sinnet Technology (300383.SZ) 0 (03/19/2018) Rmbl4.29 China Communication Service Co Ltd (0552.HK) 0 (03/07/2017) HK$6.90 Dr. Peng Telecom & Media Group (6DOBD4.SS) U (01/07/2015) Rmb8.42 HKBN Ltd (1310.HK) O (05/17/2017) HK$11.34 HKT Trust and HKT Ltd. (6823.HK) 0 (01/08/2016) HK$11.04 HTHKH (0215.HK) 0 (09/17/2018) HK$2.92 PCCW Ltd (OOOB.HK) E (01/1712018) HKS4.35 SmarTone (031 5.11K) O (09/17/2018) HK$9.77 Wangsu Science & Technology (300017.SZ) U (03/15/2017) Rmb8.54 Stock Ratings are subject to change. Please see latest research for each company. - Historical prices are not split adjusted. INDUSTRY COVERAGE: S. Korea Telecoms, Media & Internet COMPANY MCKERR) RATING (AS OF) PRIW (11 /1WN18) Sayan Park ..__._._ . _............ ................. Doublet.) Games CO Ltd (192080 KG) ____. .. . ..... ......... ....................... _.__._ .................. __... 0 (03/23/2018) _ ... .. ... ... ... ........... _.._._._ W61,200 Kakao Corp (035720.KS) U (04/18/2018) WIOZ500 Corp (0302GO.KS) E (08/08/2018) W29,950 isKT LG Uplus Corp (032640.KS) 0 (08/08/2018) W75.350 Naver Corp (035420.KS) E(07/27/2018) W112,500 NCSOFT Corp (036570.KS) E (05111/2018) W466,000 INDUSTRY COVERAGE: Telecommunications .. .. ... .. ... .. .. .. .... ......... . ..... . ......... - - ------ COMPANY MCWQ RATING (AS OF) PRICE- (11/16/2018) Tetsuro Tsusaka, CIA, KDDI (94333) Morgan Stanley I ---- VZ528 GLOBAL FOUNDA'nON • *4.594 NTT DOCOMO (9437J) E (07/23/2015) VZ568 Netmarble Games Corp (251270.KS) E(04/19/2018) W122,5110 Stock Ratings are subject to change. Please see latest Neron Co Lid (36593) F (05/28/2018) V1,114 Pearl Abyss Corp (263750.KQ) 0 (08/29/2018) W188,400 SK Telecom Co Ltd (017670.KS) E (01/11/2018) W279,000 COMPANY (TICKER) Stock Ratings are subject to change Please see latest .... . ...... PFUCL- (11fismis) Simon Flannery research for each company. . ... ... ... ... .. I ... ... ... .. ... ... ... ... ... ... .. ... .... . .... .. .. ... --------- -- ----- Tower Corp. (AMTN) ....... . . . ... -- ------- . .... ..... 0 (08/06/2015) .... . .... ... . ....................................... . .... . .. . . . ...... . . US$162.13 • Historical prices are not split adjusted 0(06/27/2018) US$30.12 INDUSTRY COVERAGE: Telecommunications .. .. ... .. ... .. .. .. .... ......... . ..... . ......... - - ------ COMPANY MCWQ RATING (AS OF) PRICE- (11/16/2018) Tetsuro Tsusaka, CIA, KDDI (94333) E (07/24/2014) VZ528 NTT (94323) 0 (11/17/2012) *4.594 NTT DOCOMO (9437J) E (07/23/2015) VZ568 Softbank Group (9984.T) V8,802 Stock Ratings are subject to change. Please see latest research for each company - Historical prices are not split adjusted INDUSTRY COVERAGE: Telecom Services COMPANY (TICKER) RATING (AS OF) .... . ...... PFUCL- (11fismis) Simon Flannery . ... ... ... ... .. I ... ... ... .. ... ... ... ... ... ... .. ... .... . .... .. .. ... --------- -- ----- Tower Corp. (AMTN) ....... . . . ... -- ------- . .... ..... 0 (08/06/2015) .... . .... ... . ....................................... . .... . .. . . . ...... . . US$162.13 •American AT&T, Inc. (TN) 0(06/27/2018) US$30.12 BCE Inc. (BCE.TO) E(12/17/2015) G$54.80 CenturyUnk. Inc. (CTLN) 0 (03/29/2017) US$18.96 Cincinnati Bell Inc. (CBBN) U (05/25/2018) USSI 7.98 Crown Castle Corp. (CCI.N) 0 (11/11/2009) USS110.59 CyrusOne Inc (CONED) 0 (05/2812013) US$53.52 Equinix Inc. (EQIX.0) E (05/13/2009) USS385.28 Frontier Communications Corp (FTR.0) E (07/18/2016) US$3.75 Globalstar Inc (GSATA) E (04/11/2018) USS0.37 Gogo Inc (GOGO.0) U (11/15/2013) USS6.57 Intelsat S.A. (IN) E(Il/12/2018) USS25.36 OTS Realty Trust Inc: (QTS.N) E(11/04/2014) USS38.67 Rogers Communications, Inc. (RCIb.TO) E(12/17/2015) G970.15 SBA Communications (SBAC.0) 0 (03/2812011) US$170.17 Sprint Corp (SA) L5S6.12 Telephone & Data Systems (TDS.N) E (05/30/2014) US$34.83 TELLS Corp. (T.TO) 0 (04/3012018) C$46.25 T-Mobile US, Inc. (TMUS.0) USS68OB Until Group Inc (UNIT.0) E(06/1412018) US$18.71 US Cellular Corporation (USM.N) E(12/14/2016) US$54.80 Vefizon Communications (VZN) 0 (02/27/2014) US$59.08 \RaSat Inc (VSAT 0) E(12/15/2017) US869.55 Windstrearn Corp. (WIN.0) E (07/0112015) US$4.07 Zeyo Group Holdings. Inc. (ZAYOA) 0 (12/12/2017) US82220 Stock Ratings are subject to change. Please see latest research for each company. - Historical prices are not split adjusted • L • The Americas • 1585 Broadway New York, NY 10036 -8293 United States TeL +1(1) 212 7614000 Morgan Stanley © Morgan Stanley 2018 Europe 20 Bank Street, Canary Wharf London E144AD United Kingdom TeL- +44 (0) 20 7 425 8000 Japan 1-9 -7 Otemachi, Chiyoda -ku Tokyo 100 -8104 Japan TeL +81(0) 3 6836 5000 Afia//Paci is 1 Austin Road West Kowloon Hong Kong TeL +852 2848 5200