评估铜网络退役的影响

Restricted - External TOVERWEIGHTNorth America Cable,Satellite & Telecom ServicesNEUTRALPrice TargetUSD 30.00Price (15-May-25)USD 27.33Potential Upside/Downside+9.8%Source: Bloomberg, Barclays ResearchNorth America Cable, Satellite &Telecom ServicesKannan Venkateshwar+1 212 528 7054kannan.venkateshwar@barclays.comBCI, USViren Uchil+1 (212) 526 7523viren.uchil@barclays.comBCI, USLauren Bonham+1 212 526 3856lauren.bonham@barclays.comBCI, USSiyuan (Cindy) Huang, CFA+1 212 526 7952siyuan.huang@barclays.comBCI, USAmir Saley+1 212 526 7564amir.saley@barclays.comBCI, US 2https://fiberbroadband.org/wp-content/uploads/2023/03/Access-Network-OpEx-Analysis-White-Paper.pdf3https://fiberbroadband.org/wp-content/uploads/2024/06/The_Benefits_of_Retiring_Copper_Today_6.7.24.pdf4Fiber Deep Networks and The Lessons Learned From The Field, ToddLoeffelholz,20175https://fiberbroadband.org/wp-content/uploads/2023/03/Access-Network-OpEx-Analysis-White-Paper.pdf6https://www.fierce-network.com/telecom/verizon-cto-weve-already-converted-45m-copper-circuits-fibercloser to homes in the form of street cabinets and added another layer to the networkarchitecture.The technology framework behind cable and legacy DSL networks is not verydifferentas it usesfrequency division to divide capacity betweendifferentapplications such as voice, data andvideo with DOCSIS being a protocol “hack” similar to ADSL, VDSL, G.Fast etc. The physicalmedium of cable is also copper but the medium is protected by insulation and shielding whichhelps manage interference better andsuffersfrom less signal loss over longer distances, whichgives it a technological advantage over legacy copper in terms of how much capacity it candeliver as well as operating expenses. Legacy cable was deigned for video, which is a highintensity application and transitioned to data and voice while telecom companies transitionedfrom voice, which is a low intensity application to data and video. This is why cable over timeended up with more capacity to deliver high speed broadband because the physical medium,i.e., coax wire, had more capacity to allocate to new applications like data and voice comparedto legacy copper. This in essence gave cable more runway to increase speeds vs DSL but is alsoinstructive to understand the limitations of cable compared to fiber in terms of long-termcapacity upgrade path.Despite these legacy retrofits to increase capacity partly by pushing fiber closer and closer overtime to neighborhoods, last-mile copper still necessitates the use of thousands of centralofficesand hundreds of thousands of amplifiers and repeaters all of which significantly increases opexand capex vs fiber. For instance, one fiber location can potentially serve a 40km diameter2butcopper plants will need hundreds of wire centers and thousands of local nodes to serve thesame area3as copper wires ineffectneed powered amplifiers every 100 meters to 2 km. In thecase of cable, even in the ideal Node+0 coax cable plant architectures, which are still relativelyrare in the industry, homes tend to be at the most ~1000 feet +/- from the last powered node4.More realistically, there are at least 4-5 amplifiers between the last node and the home.In the case of fiber, the distance between the Optical Line Terminal (OLT) or the centraloffice/wirecenter DSL equivalent and the ONT in the home tends to be ~20km with no active elementsin between and fiber can in theory support transport across 100 km with no amplifiers5.These architecturaldifferencesmean that that copper and cable plans have millions of activeelements in the network requiring power to operate and significant maintenance needs, whichincrease operating expenses relative to fiber significantly, as we will see later. AT&T, for instance,expects copper decommissioning to reduce electricity consumption by 1.06mm MWh between2024 and 2028, or ~7% of its total energy use in 2024. For context, Verizon has stated thatconversion of 36 centralofficelocations to all fiber has resulted in operational savings of ~$180million6. AT&T at present has ~4600 centraloffices.More importantly, as these legacy networks age, disruption in copper networks is becomingincreasingly moredifficultto fix given lack of support for legacy equipment and workers withthe technical knowledge base to sustain the plant. This backdrop has in essence resulted intelecom companies being unable to reduce costs proportionate to revenue loss from legacynetworks, which has limited the operating leverage flow through from fiber investments atcompanies like AT&T. 2 7Using copper costs savings overall ranging from $3.7bn (base case on our estimates of ~60% savings by 2030) to $6bn(blue-sky scenario of 100% savings by 2030) and Bloomberg consensus estimates for total company revenues of ~$134bnfor 20308Using copper costs savings in the consumer segment ranging from $2.4bn (base case on our estimates of ~60% savingsby 2030) to $3.8bn (blue-sky scenario of 100% savings by 2030) and our estimate