摩根士丹利-美股-电信与网络设备行业-如果运营商的资本支出预算持平5G无线资本支出从何而来？-2019.2.6-23页.pdf

James.FMeta.MYuuji.AErik.LCautiousMORGAN STANLEY&CO.LLCJames E FaucetteEQUITY ANALYST+1 212 296-5771Meta A MarshallEQUITY ANALYST+1 212 761-0430Yuuji AndersonEQUITY ANALYST+1 212 296-8284Erik LapinskiRESEARCH ASSOCIATE+1 212 761-5541Telecom&Networking EquipmentNorth AmericaIndustryViewTelecom&Networking EquipmentTelecom&Networking Equipment|North America North AmericaWhere Does 5G Wireless CapexCome from If Carrier CapexBudgets Are Flat?Given carrier capex budgets are essentially flat,expect recentuptick in wireless spending to come at the expense of corespending.Exiting 2019,more cautious CIEN and JNPR,favorCOMM as spend moves to the edge from the core.Generally flat capex budgets causes spending to cycle from core to edge.Wetend to be conservative on core equipment suppliers as 5G investments movesaround spend instead of being incremental(see Exhibit 1).We view capexspending in three buckets core,edge,and device with spend moving from onebucket to another.Device spend is borne less by carriers,but core and edgespend changes largely offset each other given flat capex budgets.As shown inExhibit 3,we classify core investment as fiber,optical,routing,and lab testequipment,with investment lasting from 2016-2020,and the heaviestinvestment being in the 2017-2019 period.We then see spending moving to theedge,classified as RAN,antennas,fronthaul,and tower test equipment.Thiscategory should see investment from 2019-2023,as noted in recentERIC/NOK/COMM/XLNX results,but we would expect the heaviest periods to be2020-2022.Finally comes devices,which add traffic onto the network,whereheaviest investment should be similar to edge network being enabled.Overall,core investment came ahead of edge in the 5G cycle as core investment futureproofs the network,giving a higher ROI as data demand increases.Edgeinvestment follows,with degree of spend largely dependent on business casesthat develops for 5G(potentially prolonging the cycle).5G investments starting with higher ROI categories given past challengesmonetizing 4G investment.While 5G promises new telco business opportunities(as highlighted in recent Morgan Stanley reports 5 Drivers of 5G Value andAutonomous Cars&Telcos:Were gonna need a bigger(5G)Pipe),most 5Grevenue opportunities are years away from being full-fledged business models,slowing the pace of 5G wireless spend.This slower investment was noted byEricsson in its recent analyst day,noting it only expected its networks business togrow 1-3%annually over the next couple of years,and CommScope,despitepositive Q4 NA wireless spend being healthy,remains similarly skittish on ameaningful uptick in wireless spend.In general,we think spend that we haveseen to date(e.g.fiber,optical)is most likely to continue to see investmentthrough 2019 as there will be a way to monetize the investment as more trafficcomes onto the network.While we may start to see wireless spend in 2019,asnoted by ERIC/COMM/XLNX recently,we generally expect the pace to be slowerthan in previous wireless generational shifts,as highlighted previously in ourMorgan Stanley does and seeks to do business withcompanies covered in Morgan Stanley Research.As aresult,investors should be aware that the firm may have aconflict of interest that could affect the objectivity ofMorgan Stanley Research.Investors should considerMorgan Stanley Research as only a single factor in makingtheir investment decision.For analyst certification and other important disclosures,refer to the Disclosure Section,located at the end of thisreport.1February 6,2019 05:01 AM GMT每日免费获取报告1、每日微信群内分享7+最新重磅报告；2、每日分享当日华尔街日报、金融时报；3、每周分享经济学人4、行研报告均为公开版，权利归原作者所有，起点财经仅分发做内部学习。扫一扫二维码关注公号回复：研究报告加入“起点财经”微信群。Global Insight,Learning to Ride a 5G Cycle.Fiber/optical vendors seeing benefit today,wireless vendors likely biggerbeneficiaries heading into 2020.Of covered names,favor GLW given fiber sharethroughout 2019.CIEN is currently seeing a strong benefit from spend and wesee healthy growth from telco through 2019(data center to continue for longer),but we are cautious on valuation and expectations,hence our EW rating.COMM,VIAV should see a stronger benefit from 5G spend as we head further into theyear.While routing is seeing some investment today,ASP pressures andstructural redesigns are creating permanent greater headwinds,keeping uscautious(e.g.UW JNPR).2 AnalysisThe global telecom team recently published a Foundation note,5 Drivers of 5G Value,noting that global 5G capex could be 1.7x the size of 4G capex.This is on top of the autoteams Global Insight from 2018,Autonomous Cars&Telcos:Were gonna need a bigger(5G)Pipe,where it was noted that AV could create opportunity for service providers.Thetelco team note included the caveat that 5G spend would take place over 10-12 years,vs.6-8 years for 4G,and autonomous vehicles are still some distance out.As pointedout in our Global Insight from 2017,Learning to Ride a 5G Cycle,we think an uplift incarrier capex is unlikely.Capital constraints limit investment in networks.As shown inExhibit 1,global carrier capex is expected to stay$300bn into2019,the level it has been at for the past couple of years.Whilecertain regions may experience growth(e.g.China),for the mostpart capex is relatively steady by region(see Exhibit 11 in theAppendix).The limitations in expanding capex are largely capitalconstraints and ROI on investment.Against this backdrop,in order to invest in new priorities like 5Gnetworks,carriers must move spend around.Exhibit 2 attempts toshow a rough guide to how the$300bn of network spending,with it being important to note that only 60%of capex is onnetwork infrastructure.Generally flat capex budgets causes spending to cycle from core to edge.We tend to beExhibit 1:Little Growth Expected in Global Carrier CapexSource:Morgan Stanley Global Capex model.Exhibit 2:Approximate Breakdowns of$300bn Global Carrier Capex BudgetsSource:Morgan Stanley Research estimates,Strategy&,DellOro,Gartner.3conservative on core equipment suppliers as 5G investments moves around spendinstead of being incremental.We view capex spending in three buckets core,edge,anddevice with spend moving from one bucket to another.Device spend is borne less bycarriers,but core and edge spend changes largely offset each other given flat capexbudgets.As shown below,we classify core investment as fiber,optical,routing,and labtest equipment,with investment lasting from 2016-2020,and the heaviest investmentbeing in the 2017-2019 period.We then see spending moving to the edge,classified asRAN,antennas,fronthaul,and tower test equipment.This category should seeinvestment from 2019-2023,as noted in recent ERIC/NOK/COMM/XLNX results,but wewould expect the heaviest periods to be 2020-2022.Finally comes devices,which addtraffic onto the network,where heaviest investment should be similar to edge networkbeing enabled.Overall,core investment came ahead of edge in the 5G cycle as coreinvestment future proofs the network,giving a higher ROI as data demand increases.Edge investment follows,with degree of spend largely dependent on business casesthat develops for 5G(potentially prolonging the cycle).Who has the most exposure?Exhibit 3 highlights which companies have exposure tophases of investment in 5G,with Exhibit 4 giving more detail around level of exposure toservice provider spend.Of covered names,we favor GLW given fiber share throughout2019.CIEN should see a strong benefit from spend as well,though we are cautious onvaluation.COMM,VIAV should see a stronger benefit from 5G spend as we headfurther into the year.While routing is seeing investment today,ASP pressures andstructural redesigns are having a greater headwind,keeping us cautious for now(e.g.UWJNPR).Exhibit 3:View of Investment Cycles in WirelessSource:Morgan Stanley Research estimates.4What categories could see outsized growth?As shown in Exhibit6,5G has certain technical objectives that benefit certaincategories of spend more.Small cells and fiber/optical havebeen the categories where there is outsized interest of spend,something we would expect to continue(particularly whenlooking at US carriers build intentions with 5G).Exhibit 4:Vendors Exposed to 5GSectorWhat are they providing?CompanyTickerRevenueEBITWireless RANEricssonERICB-SK90%$22,946$2,549Wireless RAN/InfrastructureNokiaNOKIA-HE90%$26,153$3,668-HuaweiPrivate60%$100,000$15,000-ZTE000063-SZ70%$15,502$1,042Fiber/Antennas/Small CellsCommScopeCOMM-US80%$4,531$887FiberCorningGLW-US35%$12,217$3,717General InfrastructureCiscoCSCO-US25%$52,289$18,493-JuniperJNPR-US45%$4,495$952Infrastructure for BackhaulCienaCIEN-US70%$3,359$521-LumentumLITE-US40%$1,843$563-AcaciaACIA-US60%$427$74-II-VIIIVI-US30%$1,452$278-InfineraINFN-US70%$1,407$53Technology/Modem ChipsQualcommQCOM-US90%$20,875$6,189Chips for RadiosXilinxXLNX-US15%-20%$3,314$1,122Analog and Millimeter WaveAnalog DevicesADI-US12%$6,277$3,044Network Management SoftwareVMwareVMW-US10%$9,814$3,707-Red HatRHT-US10%$3,791$1,007SemiconductorSoftwareRev%fromCarrier or 5G2019eWireless EquipmentFiberOther Carrier EquipmentOpticalSource:Thomson Reuters,Morgan Stanley Research.Exhibit 5:Recent 5G Piece Assumed Growth in 5G Networks ComingPredominately from Small Cells-2040608010012014016018020020192020202120222023Macro Cell/Small Cell Deployments(000s)Source:Morgan Stanley Research estimates.5What are additional areas of capex risk?As noted throughout,there are significantleverage constraints on service providers,even greater than those during the 4G cycle.With limited 5G use cases to date and regulatory issues(e.g.spectrum,small cell)thatprevent faster rollouts,there could still be limitations in spend.Additionally,M&A andHuawei uncertainty put at risk what capital/pricing is available for capex builds.Wethink there is downward risk to carrier spend from rising interest rates that could limitcapex.Tailwinds could be completion of FirstNet build outs,TMUS 600 Mhz build outs,Spring 2.5 Ghz expansion and AT&T fiber build(all winding down in 2019/2020 period).Exhibit 6:Fiber/Optical/Small Cells to See Most Incremental Spend in 5G5GPromise?Expanded CapacityHigher Data RatesLower LatencyHigher Reliability&AvailabilityLower EnergyConsumptionMore EfficientNetworkWhat doesthat mean?More connectionsHigher speed internetin more areasFaster repsonse timeof the networkFewer drops/deadspotsLonger battery lifeCheaper for carrier toprovide service(perGb)Whatcould it beused for?Wireless broadband,IoTIoTIoT,Autonomous CarsWireless broadband,IoTIoTLower cost ofoperating networkLicensedThose with spectrum assetsExpense,AvailabilityUnlicensed-WiFiCOMM(RKUS),CSCOCBRS yet to be released-CrowdedMicrowaveERIC,NOK,AVNW,CRNTPropagation limited(walls,distance)COMM,GLW,CIEN,NOK,INFNExpenseERIC,NOK,Huawei,ZTE,COMMExpense,High OpexERIC,NOK,Huawei,ZTE,COMMExpense,PermittingCSCO,JNPRExpense,Use CaseExisting InfrastructureCSCO,JNPR,DOX,RHTExisting InfrastructureWho Benefits if Investment?Inhibitors to InvestmentWhat you need to optimize for it?Macro CellsSpectrumFiber/OpticalSmall CellsLocalized Data CentersCloud RANNFVSource:Morgan Stanley Research estimates.6 Why wont 5G cycle track with the 4G cycle?Compelling ROI.Given the poor returns on 4G capex,investment motion in 5G needs tobe opportunities for new revenue,particularly given the densification required in orderto achieve the throughput,reliability,latency available under 5G that make IoT andautonomous vehicle business cases plausible.That level of densification will beexpensive to achieve,meaning that service providers need high confidence that newbusiness opportunities will be realistic and deliver revenue in short order.This willrequire more standardization amongst regulatory bodies involved in automation andtransportation as to what types of devices will be included,what type of data will becontributed to the network,what standards will be created.We are still years awayfrom this being realistic,preventing service providers from making some of thedensification moves necessary.Release of spectrum.The FCC highlighted the steps it is taking to try and speed the roll-out of 5G networks at MWC Amerias in September of 2018(see note).In relation to highband spectrum,it has five auctions planned,with 24GHz,28GHz,and 37GHz all expectedto take place over the next 12 months(28 GHz auction having just taken place).Inrelation to mid-band,it has upwards of 500MHz under review,with the CBRS bandreleased and applications being reviewed for usage(goal to be approved by the end ofthe year).Spectrum at 3.4GHz and 3.7 GHz is under various stages of review,with theFCC trying to move these towards an auction timeline.Regulation.Carriers have been looking to densify their networks with small cells foryears.However,intentions have been far greater than roll-outs,largely as localmunicipalities need to approve new deployments.In September 2018,the FCC approveda rule change to help speed deployment,limiting fees that could be charged andshorten the approval process.However,the heads of the House Energy and CommerceCommittee and the Subcommittee on Communications and Technology have recentlysent a letter to the FCC to challenge these rule changes.As long as the year to getapproval,hour to install framework exists for small cells,densification and 5G build-outs are not likely to proceed to plan.Devices.Meaningful wireless capex spend on 5G does not need to take place until thereis significant traffic being carried over a network.Traffic over a network requires devicesthat have 5G chipsets,of which there are only a select number of phones that will startcarrying these in 2019.A 5G enabled iPhone,something that would require a morewholesome 5G network build-out isnt expected until the 2H20,meaning that arelatively small amount of equipment can go into the network to say there are 5G build-outs before then.Why do carriers need 5G?Competition.5G may open new markets such as industrial IoT applications.Some 5Guse cases require network coverage in a specific location(as opposed to very broadcoverage for traditional consumer telecom services).This may attract new entrants tothe industry.In the US,while the wireless carriers are clearly interested in leveraging 5Gfor manufacturing automation use cases,this is an area where other vendors couldleverage private 5G networks,perhaps using unlicensed spectrum,to provide end-to-endsolutions to industrial customers.Communications equipment vendors could provide7end-to-end private 5G networks with necessary service level agreements connectingvarious elements of the production line,robots in warehouses,distribution,etc.The techcompanies could also play a role here,with Amazon and Google showing a stronginterest in the connected home via Alexa,Nest and other products.It is not hard to seetheir interest expanding to the enterprise market.In Japan,Rakuten received 4Gspectrum in 2018 and plans to spend US$5bn(through end-2028),aiming for 96%nationwide coverage(by end-2025),10mn subscribers(by end-2028),and positive netprofit