Just when I thought TPG was very quiet in its plans to roll out its 4G network, it suddenly announced on 19 Mar that it would provide the elderly with a free mobile plan for 2 years that includes unlimited calls and 3GB of data. That sounds really good on paper, but is TPG's 4G network ready?
According to the terms of the spectrum auction, TPG has to achieve the following network coverage by certain dates following the commencement of its spectrum rights on 1 Jul 2017:
|Coverage ||Quality of Service ||Duration ||Date |
|Street Level ||95% coverage ||18 months ||01 Jan 2019 |
|In-Building ||85% coverage ||30 months ||01 Jan 2020 |
|MRT Underground Stations & Lines ||99% coverage ||54 months ||01 Jan 2022 |
Thus, if TPG's roll-out schedule follows the above milestones, its network coverage is good only at the street level when it launches next year. Within buildings, the coverage will be sporadic. What is the use of having free mobile plans if they cannot be used to make or receive calls within buildings?
Between street level and in-building coverage, TPG only has 1 year to roll out to all buildings. Given that there are so many buildings in Singapore, I do not think TPG can achieve the required Quality of Service (QoS) imposed by IMDA. Thus, if my guess is correct, it will be several years before TPG's 4G network is capable of matching those of the existing telcos.
On the capex to build the nation-wide network, TPG has quoted a figure of $200M to $300M. If you compare to M1's fixed asset costs for "network and related application systems" of $517M as at end Dec 2017, TPG's proposed capex of $200M to $300M is on the low side. No doubt, M1's fixed asset costs include the 3G, 4G and Narrow-Band Internet-of-Things (NB-IoT) networks, but to build a nation-wide 4G network capable of meeting all the QoS standards at $200M to $300M still looks low.
Perhaps, what TPG is thinking of is to leverage on the 700MHz frequencies, which require less base stations than the traditional 900MHz and 1800/1900MHz frequencies. However, the 700MHz frequencies are dependent on the population switching out of analogue TV into digital TV, which is expected to be completed only by end Dec 2018.
In an interview by Zaobao on 27 Mar, M1 also had doubts on whether TPG could complete their network by the end of the year, noting that to set up a network, a telco has to purchase land, build base stations, connect the base stations to the core network and configure them. The largest obstacles are in buying the land and building base stations. M1 has around 2,500 macro base stations, and that does not include smaller base stations which are 2 to 3 times that number. Based on their observations, TPG does not seem to have started building the base stations.
In its results presentation for 1H2018 on 20 Mar, TPG reported actual capex of only AUD29.7M (or approximately SGD30.3M) for the Singapore mobile network. This is only 10% of the estimated total capex of $300M, providing further evidence that TPG has been slow in rolling out its network.
It is not easy to set up a telco network from scratch. Based on the experience of U Mobile, the fourth telco in Malaysia, it had to leverage on Celcom's network (i.e. act as a Mobile Virtual Network Operator, MVNO) while it rolled out its own nation-wide network concurrently. But for the case of Singapore, it is highly unlikely any of the existing telcos would allow TPG to come on board as a MVNO, since once TPG's network is completed, it would pose a threat to the existing telcos' business.
Thus, I do not see TPG as posing a serious threat to the existing telcos any time soon. TPG, I am waiting for you.
P.S. I am vested in M1, Netlink Trust and Singtel.
See related blog posts:
Click here to read the full article