Singapore Telcos – Let Winter Begin
Competition a major feature over the next three years
Question mark over TPG’s Singapore mobile business?
- In FY7/18, TPG spent A$62.3m cash capex on its Singapore mobile network rollout and A$4.4m in FY17, bringing its total capex to A$66.7m thus far. At the current run-rate, we estimate that it is on track to spend a total c.A$93m by year-end. Even after converting into S$, TPG’s network investment to date is much lower than in its original announcement in Dec 2016 (capital investment of S$200m-300m) to establish a mobile network with nationwide coverage by Sep 2018. This has raised questions about its commitment to the Singapore mobile business.
- Moreover, the incumbents have put several measures in place over the past two years to crowd out TPG when it finally launches its service, including:
- competitive offerings from the four Mobile Virtual Network Operators (MVNOs) – Circles.Life (host: M1), ZeroMobile (host: Singtel), Zero1 (host: Singtel) and MyRepublic (host: StarHub) – with the latter three having entered the market this year, and
- the introduction of SIM-only plans and data upsize options. This would probably make it harder for TPG to achieve its target of achieving positive EBITDA when it reaches 5-6% market share in a ‘short period of time’.
- The recent proposed merger between TPG Telecom Ltd and Vodafone Hutchison Australia in Australia raises additional questions on future financial support from TPG to expand the Singapore mobile business (SingaporeCo), as it will be separated from the merged entity via an in-specie distribution of SingaporeCo shares to existing TPG shareholders.
Too early to conclude that TPG will drop out of Singapore
- While its cash capex on the Singapore network rollout to date appears to be much lower than budgeted, during its 4QFY18 conference call (Sep 2018), TPG said that its network now covers more than 90% of outdoor areas in Singapore, and it is on track to meet its licence obligations for 95% outdoor coverage by 31 Dec 2018 (road tunnels and in-building: 31 Dec 2019, MRT underground stations/lines: 31 Dec 2021). TPG also said the capex outlook for Singapore is still in line with its initial forecast.
- On the proposed merger in Australia, TPG says it intends to capitalise SingaporeCo appropriately before its separation from TPG Telecom Ltd. The amount to be injected into SingaporeCo will come from the excess cash that TPG plans to distribute to shareholders prior to the implementation of the merger. TPG said it expects TPG Telecom to have strong cash flow generation and to have at least a 50% dividend payout ratio. This (in addition to debt and vendor financing) may be used by TPG’s shareholders to fund SingaporeCo’s opex and capex going forward.
- TPG has spent S$128.8m (900MHz+2300MHz: S$105m, 2500MHz: S$23.8m) to acquire spectrum in Singapore and c.S$70m on network rollout so far (total: S$198.8m). If it walks away now, the investment in spectrum would not be recoverable as its licence clearly states that the spectrum cannot be used to provide wholesale services to existing mobile operators and spectrum trading is not allowed until TPG fulfills its network rollout obligations completely (i.e. MRT underground stations/lines covered) and receives written approval from the Info-communications Media Development Authority (IMDA).
Investors should stay prudent & assume TPG is still committed
- Given the above, we believe investors should remain prudent at this stage and assume that TPG is still wholehearted in its entry into the Singapore mobile market. As reported by Channel News Asia, TPG’s chairman David Teoh said “As we approach the completion of the outdoor coverage milestone and the commencement of trial services, now is the right time to allow the Singapore business to flourish under its own power. I am tremendously excited about the opportunities for TPG Singapore”.
- As a greenfield operator with an inferior network (at least until end-2021), we believe TPG is likely to offer aggressive promotions to attract subs and drive up scale. We expect TPG to offer SIM-only plans with higher quotas for the same price as current offers on the market or lower, as well as potentially free subscription for the initial 6-12 months. In Mar 2018, TPG revealed plans to offer free mobile data plans (3Gb, local voice) to the elderly for the first 24 months, when it launches.
- We have factored in a 10% p.a. negative impact from TPG’s entry on the Singapore mobile base ARPU (ex-roaming) in FY19-20F, which brings postpaid ARPU to S$28-31 (blend of device-bundled and SIM-only plans) in FY20F. We think this is sufficiently close to the estimated S$25-30 SIM-only ARPU needed by TPG to hit its EBITDA breakeven target in 5-6 years.
Negative Singapore telco earnings trend
- Based on our assumption of a 10% p.a. negative impact from TPG’s entry on Singapore mobile ARPU (ex-roaming), we forecast the industry’s mobile service revenue to decline by 8.0%/6.8% y-o-y in FY19/20F before stabilising in FY21F. Based on our estimates, overall FY19/20F Singapore service revenue would be largely steady at 0.1%/0.7% y-o-y due to growth in enterprise fixed revenue.
- While the growth in enterprise fixed revenue would help to cushion the drop in mobile service revenue, we note that the enterprise fixed business generally earns a lower EBITDA margin (mid-teens vs. mobile services’ 35-40%). As such, we forecast industry EBITDA (Singapore only) to fall by 5.4%/4.7% in FY19/20F. For Singtel, we expect its group EBITDA to grow 5.0%/5.3% y-o-y in FY20/21F, as higher associate earnings and lower group digital life losses would more than offset lower Singapore EBITDA.
- At the core net profit level, we expect Singapore earnings of the incumbents to fall by 10.4%/11.5% in FY19/20F, further chipped off by amortisation of 700MHz and the interest cost on debt to pay for the spectrum (assuming payment in mid- 2019 and licence starts in Jan 2020). Singtel group’s core net profit is expected to rise by 6.5%/7.1% y-o-y, based on our estimates
What if TPG has minimal impact?
- If TPG tries to minimise start-up losses/additional network investment by launching with offers that are not too aggressive or different from those of the incumbents/MVNOs, it is possible that it would not get much market traction. We think the market will closely monitor TPG’s traction and its impact on the incumbents by mid-2019. If TPG’s impact is less-than-feared, the market could re-rate Singapore telco share prices. Based on our scenario analysis, if TPG’s entry has zero impact on Singapore telcos’ mobile postpaid base ARPU (i.e. remaining stable at FY18F levels), our fair values for Singtel, StarHub and M1 would rise to S$3.83, S$2.92 and S$2.79, respectively.
- Investors have started to speculate whether mobile ARPUs can rise back to previous levels if TPG does not have a major impact on the market. However, we believe that as long as TPG is in the market, it would be difficult for the incumbents to roll back existing offers, as doing so will only provide room for TPG to survive in the market.
A StarHub-M1 merger may not get IMDA approval
- On paper, a merger between StarHub and M1 would be the easiest to execute, given that both companies are profitable (i.e. no major earnings dilution for the acquirer). Given the massive overlap in operations, there would also be significant opportunities to reap opex and capex synergies from rationalising network infrastructure, sales and distribution channels and back-end operations. As is, StarHub and M1 are exploring deeper collaboration on network sharing from common antenna solutions to joint upgrading of 4.5G indoor antenna systems at large buildings.
- However, we believe the biggest obstacle to a StarHub–M1 merger is obtaining IMDA approval in the near term. Until TPG can prove that it has a long-term sustainable business in Singapore, the IMDA is unlikely to allow a StarHub–M1 merger as the Singapore market could end up with only two players (Singtel and the merged StarHub–M1) if TPG does not survive in the long run. This would go against IMDA’s objective of having a competitive market to drive lower prices and greater innovation.
Buying TPG could be earnings-dilutive for StarHub or M1
- Once TPG has met all of its network rollout obligations in three years’ time, it may be allowed by the IMDA to sell its network and spectrum to interested parties. If this is the case, we think its asking price will depend on many factors, including the number and quality of its subs. However, based on our projections of its network investment and operational losses, we estimate the minimum asking price would be c.S$600m, assuming no loss for TPG. This would be a sizeable acquisition for StarHub and M1, even if it comes with spectrum, and would be earnings dilutive given that TPG would likely still be loss-making.
TPG may acquire M1/StarHub
- David Teoh, the Chairman of TPG, is known to be a deal-maker and we do not rule out the possibility that he/his team may be able to structure a deal and raise the funding to acquire M1 or StarHub. This would immediately give scale to TPG’s Singapore operations and ensure its long-term financial sustainability, besides extracting synergies.
FY19-20 may face some spectrum auction landmines
2100MHz auction could take place in 2020
- The 2100MHz spectrum licences are due to expire in Dec 2021, 20 years since three blocks (40MHz each) were allocated equally to Singtel, StarHub and M1 for a total of S$100m (or S$33m/block). Typically, the IMDA will re-auction expiring spectrum 1-2 years prior to expiry so that the telcos have sufficient time to reconfigure their networks, should there be any major changes in holdings. This points to 2020 or early-2021 at the latest for the IMDA to hold the 2100MHz auction, with payments by winners to be made by mid-2021.
- Based on the 1800MHz spectrum auction in 2014, we estimate the reserve price for the 2100MHz spectrum would be set in the region of S$16m/10MHz. If the incumbents retain their existing 40MHz holdings (i.e. TPG does not win any), then we estimate the minimum payment would be S$64m, based on reserve price. However, given that TPG may also submit a bid, we believe final prices would be higher than the reserve price. We have factored into our Singapore telco models, 2100MHz final prices at 3x that of the reserve price, i.e. S$192m for 40MHz.
5G spectrum auctions
- The timing of 5G spectrum auctions, i.e. 3.5GHz and mmWave, is less certain. However, if the Singapore government wants to encourage telcos to roll out 5G in 2020, then we expect the relevant spectrum to be auctioned as early as 2H19, with payment due six months before the start of the licence.
- For the 3.5GHz spectrum, we think there may only be 100MHz (3.4-3.5GHz) of bandwidth to be auctioned off in the near-to-medium term, as there are existing satellite operators using the C-band in Singapore (e.g. Singtel-owned ST2) and neighbouring countries. As this is likely to be the main band used for 5G globally, we expect to see strong interest from bidders. Based on the average per unit pricing seen in 3.5GHz auctions globally, we estimate that Singtel / StarHub / M1 may have to pay S$42.7m/32.0m/32.0m, assuming they win 40MHz/30MHz/30MHz.
- For the 28GHz spectrum, the telcos are saying they need at least 1GHz each, with the minimum size for each block at 100MHz in contiguity. Given the short coverage range of 28GHz and abundant bandwidth, we do not expect Singapore telcos to pay high final prices. Based on the final per unit pricing from global spectrum auctions, we forecast that each Singapore telco may have to pay S$101.8m for 1GHz of bandwidth. It is also possible that the IMDA could follow in the Hong Kong Communications Authority (HKCA) footsteps and give out the spectrum for free (if supply is greater than demand), in return for telcos’ commitment to achieve certain 5G population coverage targets within a specified timeframe.
- We believe 5G spectrum auctions have not been taken into account by the market yet (we have not factored them in either). Based on our estimates above, Singtel / StarHub / M1 may have to pay a total S$144.5m/133.8m/133.8m for the 3.5GHz and 28GHz spectrum. This would push net debt/EBITDA ratio at end- FY19/20F to 1.36x/1.24x for Singtel, 1.95x/2.10x for StarHub and 2.13x/2.40x for M1. Assuming no improvements in capex efficiency and additional revenue generation, this would reduce our DCF-based target price by 0.1%/1.5%/3.3%. Our FY20F net profit forecasts (on a full-year basis) would be reduced by 0.1%/3.9%/7.5%.
- Refer to the attached PDF report (extraction from CGSCIMB’s report: Navigating ASEAN Telcos) for tabular figures of scenario analysis.
Sector View & Stock Recommendations
Maintain NEUTRAL on Singapore telcos
- We foresee a negative revenue and earnings outlook for the industry until at least 2020, with market competition expected to intensify in the first two years after TPG’s market entry. Market consolidation is a possibility and if it materialises, this will be positive for the share prices of the incumbents, in our view. However, we believe this is more likely to happen in the medium term (i.e. in 3-5 years), as buyer-seller price expectations are typically more aligned after earnings have come under pressure from intense competition.
- Overall, we maintain our Neutral rating on the Singapore telco sector as we believe the substantial de-rating in share prices since 2015 has largely reflected the risk of more intense competition.
- Upside risks are earlier-than-expected market consolidation, further announcements on sizeable cost savings from restructuring programmes and less-than-feared impact from TPG.
- Downside risks are worse-than-expected competition and higher-than-expected spectrum prices for 2100MHz and 3.5GHz.
- Our preferred Singapore telco is Singtel.
More Info: research.sginvestors.io