Investment Highlights

• Driving 5G collaboration. The Malaysian Communications and Multimedia Commission’s (MCMC) final report for its 5G spectrum band allocation for mobile broadband services offers a pragmatic approach to drive collaboration amongst industry players. The MCMC, after conducting a public inquiry beginning in July 2019, identified 700MHz, 3.5GHz, 26GHz and 28GHz as the pioneer spectrum bands in Malaysia.

• Assigning single consortium. Instead of allocating the 700MHz and 3.5GHz bands to individual licensees as for the 900MHz and 1,800MHz spectrums, they will be assigned to a single entity comprising a consortium of multiple licensees. This encourages the sharing of resources to roll out the required 5G infrastructure, which has been estimated to cost up to 10x 4G costs. This single-entity approach is envisioned to minimise costs and prevent duplication of infrastructure against the backdrop of additional spending requirements to enhance 4G networks.

• Only 700MHz and 3.5GHz for now. The MCMC will only release 2x30MHz of the 700MHz band and 100MHz for the 3.5GHz band this year for assignment with the remaining frequencies at a later stage. A total bandwidth of 3200MHz for the two higher 5G bands – 26GHz and 28GHz, will be assigned in two methods i.e. i) a tender process (“beauty contest”) for 4 blocks of 400MHz (1600MHz in total) of the 24.9GHz to 26.5GHz frequency bands on a nationwide basis; and ii) firstcome-first-served basis for four blocks of 400MHz (1600MHz in total) of the remaining 26.5Hz to 28.1GHz frequency bands and open to any parties which include non-licensees for the purpose of deploying localised or private networks.

• Apparatus assignment for better affordability. These assignments will be issued via apparatus assignment (AA) as the appropriate spectrum fee through AA is more economical for network deployment, which should result in cost savings for businesses and translate to better affordability for consumers. We are positive on MCMC’s decision as spectrum fees will add to the burden of existing and prospective operators, given the spectrum assignments in the past.
  
  The “beauty contest” process for the assignment is estimated to commence in 1Q2020. After the assignment processes are completed, the MCMC expects commercial 5G deployment in Malaysia to begin in 3Q20.
   
• No changes to 4G licences. The MCMC affirmed that the existing allocation for 2300MHz for WiMax and 2600MHz bands for 4G will be maintained until 31 December 2021 as all operators are urged to continue deploying mobile broadband services using existing technology in parallel to the necessary preparation towards 5G. The MCMC will undertake the necessary review of the bands in 2021.
   
• Still negative on the rollout of the National Fiberisation and Connectivity Plan (NFCP) over a 5-year period from this year to 2023. The NFCP rollout could cost RM21.6bil, of which 50%–60% may be financed by the MCMC's Universal Service Provider fund that currently holds RM8bil. This is in line with the government's objective to recognise access to the internet as a basic right, ensuring equal access to the internet for both urban and rural residents.
  
  Given TM's role as the national broadband provider, the group may bear up to half of the NFCP cost, which translates to RM2.2bil over the next 5 years. Besides TM's own capex requirements, the NFCP rollout alone translates to 19% of FY20F revenue – already above management's FY19F capex target of 18% and 9% in 9MFY19. Additionally, the thrust of the NFCP towards connecting the rural population could mean that revenue accretion from these investments will be minimal.
  
  While the pragmatic 5G approach by the MCMC partly offers relief for the sector, we highlight that the infrastructure and network costs could be substantively higher and it remains to be seen whether operators can pass on these costs to the consumers with adequate investment returns.

• Maintain NEUTRAL outlook on the sector given the escalating NFCP and 5G capex requirements against the backdrop of government-targeted fiberised ARPU reductions. Our only BUY currently is Axiata, given its low EV/EBITDA valuations and rising prospects for monetisation of its multiple businesses.

Source: AmInvest Research - 2 Jan 2020