Singapore Telco Sector - CIMB Research 2017-09-22: Wait For The Storm To Pass

SGinvestors.io

Singapore Telco Sector 4th Telco TPG Telecom M1 vs SingTel vs StarHub M1 LIMITED B2F.SI SINGTEL Z74.SI STARHUB LTD CC3.SI

Singapore Telco Sector - Wait For The Storm To Pass

Proposed COPIF revision could aid TPG’s site acquisitions. 

Rent-free rooftop MDS to be equally allocated to all MNOs. 

• In the proposed revision of the existing Code of Practice for Info-communication Facilities in Buildings (COPIF 2013), the Info-communications Media Development Authority (IMDA) mandates that building developers/owners allocate the use of rooftop space on a rent-free basis to mobile network operators (MNOs) as the preferred Mobile Deployment Space (MDS) location. This will allow MNOs to use the rooftop MDS to house equipment for the provision of mobile coverage in-building and/or outside the building (External Areas). 
• IMDA intends for existing COPIF principles of equal sharing of MDS among licensees to continue to apply, even with the entry of the fourth MNO (i.e. TPG). The public consultation on the proposed revision of COPIF 2013 closed on 21 Jun, with the final decision possibly by year-end or early next year and implementation soon after.

Existing issues with COPIF 2013 

• Currently, COPIF 2013 only covers the MDS set aside by any building developer/owner primarily for the provision of in-building mobile coverage.
• Therefore, MNOs would have to find space outside the development or carry out commercial negotiations with the building developer/owner to deploy radio equipment for the provisioning of mobile coverage to External Areas, such as public walkways and roads. 
• It is also not mandatory for MDS to be sited on building rooftops, with building developers/owners allowed to decide the location of the MDS. This may not meet the MNOs’ network design and deployment plan, which may cause the MNOs to negotiate commercially for a more suitable location, incurring additional cost and time.

COPIF 2013 revision could facilitate efficient deployment of TPG’s mobile coverage 

• If the revision to COPIF 2013 is passed by IMDA in its proposed form, it could facilitate efficient deployment of TPG’s mobile coverage, by cutting down protracted negotiations with building owners and requiring existing MNOs to create space in an MDS to accommodate it (within a specified time after it has been requested). 
• It could also help to lower cost related to MDS rental, especially for the provision of external area coverage. Moreover, it may help TPG’s rollout of in-building coverage, as a development may be better served by mobile deployments on rooftops of adjacent buildings (i.e. antennas pointed towards it).

NLT’s ICO rate revision may also lower TPG’s rollout cost 

• Besides revision to COPIF 2013, the reduction of NetLink Trust’s (NLT’s) regulated interconnection offer (ICO) rates for non-residential end-user connection and non-building address points (NBAP) effective 1 Jan 2018 could help lower the cost of TPG’s mobile network rollout. We believe TPG is likely to lease NLT’s infrastructure for most of its transmission network, at least in the initial years. 
• Post-revision, the standard installation charges for NBAP have been waived (previous: S$150/connection), while the monthly recurring charge was reduced by 60% to S$73.80 (previous: S$185). 
• For segment fibre lease, the monthly recurring charges for connections between central offices (CO to CO) have been reduced by 28%, while the monthly recurring charges for connections from CO to NBAP distribution points (DP) and further on to termination (TP) points have been cut by 74% and 77%, respectively.

TPG’s speeds to be competitive and coverage to be on par with incumbents 3 years after launch 

TPG has awarded key vendor contracts 

• During its FY7/17 results conference call, TPG revealed that it has awarded key vendor contracts and that work is underway for the implementation of the network, including:
  1. site radio network equipment installation,
  2. primary and diverse data centres, and
  3. core network and backhaul. 
• It said that it is on track to achieve nationwide outdoor service coverage by Dec 2018F, in compliance with its licence obligations. From its online hiring programme, we can see that TPG is still in the process of hiring technical staff. 
• TPG started its recruitment drive in Dec 2016 after it won the fourth MNO licence in Singapore and Mr Tony Moffatt, TPG's general counsel, said then that the local operations are expected to expand to 400 people over time.

Service launch possibly closer to end-2018F 

• We believe that TPG would await IMDA’s final decision and implementation of the revised COPIF 2013 before proceeding full-speed ahead to the site acquisition stage, as that could substantially help the rollout process. However, given that IMDA has to deal with various stakeholders, it may only finalise the revision by early 2018F. In that case, we think TPG may launch its mobile service closer towards end-2018F, which is the deadline set by IMDA for nationwide outdoor coverage (i.e. 18 months from Jul 2017). This would be slightly later vs. our basecase expectations of a mid-2018 service launch.

How is the spectrum situation after NESA and GSA? 

• After the new entrant spectrum auction (NESA) in Dec 2016 and general spectrum auction (GSA) in Apr 2017, Singtel continues to hold the most bandwidth among Singapore telcos at a total 215MHz, including bidding for the maximum allowed 40MHz of 700MHz during the GSA. It has access to all available mobile spectrum, except for 2300MHz (TDD). StarHub has the second-highest amount of bandwidth at 190MHz and similar to Singtel, the only spectrum missing from its holdings is 2300MHz (TDD). M1 has the smallest amount of bandwidth among the incumbents at 150MHz.
• Due to the high bid prices during the GSA and M1’s view that more 5G spectrum will be made available, M1 has the least 700MHz holdings and did not pursue the 2500MHz (TDD) aggressively during the GSA. TPG was allocated the 900MHz and 2300MHz (TDD) at NESA and further added 2500MHz (TDD) at GSA for a total bandwidth of 70MHz, or slightly less than half of M1’s.

Speed should not be a problem for TPG… 

• Based on TPG’s spectrum holdings and using LTE-Advance Pro (4x4 MIMO, 256 QAM), its 4G network should be able to reach theoretical peak download speeds of 300-400Mbps (subject to device availability). 
• While incumbents are able to hit 1Gbps theoretical speeds with the same technology and tri-band carrier aggregation (2x60MHz), we believe TPG’s practical speeds may be comparable or even faster as network utilisation rate should be low in the first few years.
• Simplistically, up to 844k subscribers, TPG will have more bandwidth/subscriber vs. the incumbents.

…but coverage will be inferior initially 

• While speed should not be a major issue for the average subscriber, TPG’s network coverage will likely be inferior to the incumbents’ in the near-term.
• However, complying with IMDA’s stringent rollout obligation, TPG should have comprehensive nationwide coverage within three years of service launch, including in-building (85%, Dec 2019) and underground MRT stations/line services (99%, Dec 2021).

Incumbents’ move while awaiting TPG’s entry… 

Trying to better monetise and lock-in subscribers for 2 years 

• Coinciding with the launch of the latest iPhones, StarHub recently raised the monthly subscription fees across its device-bundled postpaid plans by S$5.10 (SIM-only: +S$2.55) after including unlimited weekend data and increased talk-time quotas.
• Singtel also:
  1. raised the monthly subscription fee of its Combo 3 by S$6 to S$68.90 (now including unlimited talk-time), and
  2. removed Combo 4 (S$82.90), which forces re-contracting subscribers to either downtrade to Combo 3 (+Data X2) or uptrade to Combo 6, price of which has been reduced to S$95.90/month (previous: S$102.90/month). For Combo 4 subscribers that downtrade to Combo 3, the loss of subscription revenue for Singtel is more than offset by lower handset subsidies incurred.
• Singtel also introduced the Data X Infinity, which offers unlimited data, as an addon available to Combos 3, 6 and 12 subscribers for an extra S$39.90/month. We see this as a positive move to lock in subscribers on 2-year contracts at a higher ARPU level, prior to TPG’s entry. 
• In general, both StarHub and Singtel would generate more revenue as subscribers re-contract under the revised plans, assuming the higher price does not trigger subscribers to downtrade to a lower price plan (for Singtel, Combo 3 to Combo 2), shift to SIM-only or migrate entirely to other networks (i.e. M1 or Circles.Life). The additional revenue over a 2-year contract will more than offset the higher iPhone 8 subsidies (vs. iPhone 7 bundles, Sep 2016), although earnings would take a hit first (as subsidies are immediately expensed). 
• M1 has not revised its price plans but this is justified as its subsidies for the iPhone 8 are generally the same as for the iPhone 7. 

• Meanwhile, M1 revised its SIM-only plans in Sep 2017. The mySIM+15 (S$15, 0.5Gb, 100 minutes) and mySIM+30 (S$30, 4Gb, 300 minutes) plans were retired. The entry-level SIM-only plan now is mySIM+20, which has seen a cut in data/talktime quota to 2Gb/100 minutes (previous: 3Gb/150 minutes). M1 has introduced a new mySIM+40 which is cheaper and has a higher 10Gb quota than the previous mySIM+45 (6Gb) plan. However, we believe the take-up of the previous mySIM+45 plan was probably not high (as there were other lower-priced options) and the mySIM+40 could actually generate more revenue for M1 as new subscribers that find mySIM+20’s 2Gb insufficient have no choice but to take up this plan instead (previously there was mySIM+30).
• For its SIM-only plans with 12-month contracts, M1 has also removed the S$15 and S$30 variants, but for its S$20 plan, it has raised the data quota slightly to 5Gb (previous: 4Gb). 
• M1 also introduced an unlimited data (+100 minutes talktime) plan for S$98/month. We are neutral on this as it is priced at a fairly high level (with no handset subsidies) and very few subscribers truly need unlimited data (especially given high fixed broadband penetration in Singapore) 

Block and tackle with MVNOs. 

• There is currently one mobile virtual network operator (MVNO), Circles.Life (CL), in Singapore, which launched its service in May 2016 riding on M1’s network. 
• Circles.Life’s SIM-only (no contract) postpaid service is priced at an attractive S$30 (S$28 + S$2 for free incoming calls) for 6Gb (if subscribers port-in from another network) and 100 minutes of talktime. Subscribers can add on additional data/talktime quotas (S$6-10/1-2Gb, S$4/100 minutes) or pay-per-usage (8 Scts/minute, S$1.20/100Mb). We believe M1’s strategy is to leverage on Circles.Life to eventually compete against the fourth mobile entrant (i.e. better to lose subscribers to CL than TPG as M1 still earns a wholesale fee from CL), while in the meantime, gain market share from StarHub/Singtel for postpaid subscribers who do not want to be locked into contracts and want to buy their own phones.
• Besides CL, MyRepublic has announced to the press that it plans to enter the Singapore market as the second MVNO as early as Oct 2017F. At this stage, it is still unclear whether Singtel or StarHub will host MyRepublic on their network.
• MyRepublic CEO Malcolm Rodrigues said it received calls from two telcos "days after" TPG won the fourth telco licence and "the objective is to make it as difficult as possible for TPG to be successful”.
• We believe that incumbent telcos are looking to use MVNOs, as one of their strategies, to compete against TPG. By doing that, incumbents are hoping to contain the price competition to the lower-end segment and hold prices in the mid-to-high-end customer segments.

What is TPG likely to do and the implications? 

More data, data and data! 

• We believe that TPG is likely to offer SIM-only plans, avoiding the incurrence of expensive upfront handset subsidies. 
• Looking at the SIM-only plans by incumbents, we believe TPG may not substantially undercut the pricing of current entry-level plans, which start at S$20, but would most likely offer a high data/talk-time quota at every price tier. In such a situation, incumbents could raise their value inclusions to either match or at least narrow the gap. The consequence of that is that some of the incumbent’s higher ARPU SIM-only subscribers could downtrade to lower-tier plans, which sufficiently meet their usage requirements after the revision.

SIM-only with phone instalment plans… 

• While we believe TPG will not enter into a handset subsidy war with the incumbents, it could still target the mid- to higher-ARPU customer segments that want a phone bundle by offering 0%-interest instalment plans through tie-ups with selected banks, similar to CL’s plan with phone offers. 
• As Circles.Life’s current plan with phone offers show, for data-centric users, a SIM-only plan with a 24-month 0% instalment plan can be comparable in total cost to an incumbent’s device-bundled plan with subsidies, with the further advantage of:
  1. no upfront device fees, and
  2. monthly bills will be reduced to just SIM-only fees once the phone instalment is paid off (i.e. no need to re-contract into another phone-bundle plan).

…can compete with incumbents’ device-bundled plans 

• If TPG offers substantially more generous data quotas on its SIM-only plans, this could have repercussions for the incumbents’ device-bundled plans. While incumbents may not directly cut the prices of existing plans, they may have to raise the data quotas to ensure the gap with TPG is not too wide. 
• For example, if TPG offers 10Gb for a S$30 SIM-only plan, then a plan with the iPhone 8 will cost a total S$1,868 over a 2-year contract. In that instance, M1 may need to raise the data quota on its i-Lite+ plan, so that a subscriber that takes its Upsized Data Plus option will get a comparable 10Gb with a slightly higher total cost of S$1,956 (we think incumbents should be able to command some premium due to their superior network). While there is no loss of revenue for this subscriber, a subscriber on its i-Reg plan (currently having a 10Gb quota with Data Upsized Plus) may be compelled to downtrade to i-Lite (with Data Upsized Plus) as it now offers the same quota at a lower total cost.

TPG is likely to run subscriber acquisition and retention programmes 

• In our view, as TPG seeks to build up its subscriber base (and keep them), it will run short-term promos and is likely to run subscriber-get-subscriber campaigns and reward subscribers for staying on the network. This is very commonly done by new entrants and is also carried out by Circles.Life, which gives extra data quotas to subscribers that stay beyond a certain period of time or for referring other subscribers.

Telcos with premium subscribers, multi-line/bundling would be less affected 

Low-end segment could get very competitive 

• Especially in the first three years after TPG’s service launch, we believe competition is likely to be concentrated in the lower-end segment, given that higher-end subscribers are less likely to consider TPG’s service given its inferior network (unless TPG is able to secure a domestic roaming agreement with one of the incumbents). 
• In addition, given that there could be two (and potentially more) MVNOs competing against TPG, the lower-end of the market is likely to get very competitive. Among the three incumbents, we believe M1 has a higher mix of lower-to-mid-end customers, as its postpaid ARPU level is the lowest. Meanwhile, we believe Singtel has the highest mix of premium/corporate mobile customers.
• Its postpaid ARPU is higher than M1 but lower than StarHub’s, possibly due to substantial discounts (up to 30%) given to corporate customers and relatively high share of elderly subscribers (long-time customers but with lower ARPU).

Greater subscriber stickiness for telcos with triple/quad-play households 

• We also expect telcos that have subscribers with multiple shared lines and/or tied into triple/quad-play bundles to be less affected by competition from TPG than their peers. These customers are already receiving bundling discounts, which should make TPG’s offers less irresistible, although we do note that pay TV subscribers are declining in Singapore due to pirated set-top boxes and emergence of over-the-top (OTT) video as alternative viewing platforms.

Upgrade sector from Underweight to Neutral 

Stocks have de-rated; few near-term de-rating catalysts 

• M1’s share price, StarHub’s share price and Singtel’s share price have de-rated by 26%, 24% and 7% over the past 12 months (M1/StarHub/Singtel’s share prices: -41%/-25%/+11% over the past three years, see: Share Price Performance - Telecommunication Services Sector) due to the fear of increased competition after the entry of the fourth MNO, as well as the weak earnings performance (roaming decline, market maturity). These stocks have significantly underperformed the STI’s 13% rise over the same period (-3% over 3 years). 
• We expect 4Q17F and 1Q18F earnings of Singapore telcos to be hit by high handset subsidies due to strong demand for the iPhone 8/8Plus/X, as well as increase in per unit subsidies (for Singtel and StarHub). 
• Nevertheless, as we have discussed above, the increased subsidies are more than offset by higher revenues (for Singtel and StarHub) after the monthly subscription fee hikes. As such, beyond 1Q18F, we should start to see the positive earnings effects once the immediate impact from high handset subsidies diminishes. 
• If TPG launches its service at end-2018, the negative impact on Singapore telcos’ earnings would only be felt beginning in FY19F and perhaps to a larger extent in FY20F as 2-year contracts expire (i.e. for those signed in Sep 2017-18).

Upgrade M1 and StarHub to Hold; maintain Add on Singtel 

• The current share prices of M1 and StarHub are now very close to our target prices and we have upgraded them from Reduce to Hold, while we have maintained our Add rating on Singtel. Correspondingly, we raise our sector weighting from Underweight to Neutral. 
• However, we think any uplift to earnings in 2Q-4Q18F would be a short-term reprieve before more competition enters the market in FY19F onwards. 
• Our suggested entry points into M1 and StarHub are still much lower than their current share prices, implying a fairly high margin of safety, which is required to factor in the range of possible earnings impact from TPG’s entry.
• Upside/downside risks to our sector call are better/worse-than-expected impact from TPG’s entry.

Singtel (ADD, TP: S$4.10): 

• We keep our Add call on Singtel, with an unchanged SOP-based target price of S$4.10. 
• We forecast Singtel’s core EPS to fall 2.5% in FY18F due to declining earnings in Optus and largely flat associate contributions, partly buffered by mildly stronger earnings in Singapore. We then see 4.2% yoy growth in FY19F due to earnings growth resuming at associates, led by Telkomsel and Bharti. 
• Singtel’s FY18F EV/OpFCF of 17.3x is in line with the ASEAN telco average of 17.1x, supported by attractive FY18-20F yields of 4.8-5.2%. 
• A potential re-rating catalyst is the declaration of a special dividend in the upcoming 2QFY18 results, following the recent IPO of NetLink Trust. 
• Downside risk is more intense competition in Australia, India and Singapore. Singtel is our preferred Singapore telco pick.
• See report: SingTel - CIMB Research 2017-09-22: Yields And Safety In FY18F; Growth From FY19F

StarHub (HOLD, TP: S$2.50): 

• StarHub’s share price has fallen by 24% over the past 12 months and now better reflects the risks of more competition from TPG, in our view. As such, we upgrade StarHub from Reduce to Hold, with an unchanged DCF-based target price of S$2.50 (WACC: 7.1%). 
• Our earnings forecasts are intact. We expect core EPS to decline by 15.4%/3.7%/15.1% in FY17F/18F/19F, with the big decrease in FY19F due to keener mobile competition upon TPG’s market entry. Our FY17F-19F DPS is at S$0.16 p.a., after StarHub’s DPS cut this year. Its 13.8x FY18F EV/OpFCF is at a 19% discount to the ASEAN telco average, which we think is justified given our forecast of declining earnings.
• Our scenario analysis on the possible earnings impact from TPG suggests that a good entry point would be below S$2.20 (bear case) and exit point above S$2.80 (bull case). 
• A key upside/downside risk is lower/higher-than-expected impact from the entry of TPG.
• See report: StarHub - CIMB Research 2017-09-22: Cloudy Weather But At Least Now In The Forecast

M1 (HOLD, TP: S$1.80): 

• We upgrade M1 from Reduce to Hold, with a 5.9% higher DCF-based target price of S$1.80 (WACC: 7.1%), after rolling forward our base year to FY18F. 
• We believe that M1’s share price is now more reflective of the risk of more competition from TPG, after falling a substantial 26% over the past 12 months.
• Our earnings forecasts are intact. We expect M1’s earnings to decline by 3.2% in FY17F, then 9.7%/21.9% in FY18F/19F upon TPG’s entry.
• M1’s 12.8x FY18F EV/OpFCF is at a 25% discount to ASEAN telcos, which we think is justified given its prospects of declining earnings. Our scenario analysis on the possible earnings impact from TPG suggests a good entry point would be below S$1.47 (bear case) and exit point above S$2.14 (bull case).
• Upside/downside risks are better/worse-than-expected impact of TPG’s entry.
• See report: M1 Limited - CIMB Research 2017-09-22: Competition Risk Is Better Priced In Now

FOONG Choong Chen CFA CIMB Research | http://research.itradecimb.com/ 2017-09-22

CIMB Research SGX Stock Analyst Report HOLD Upgrade REDUCE 1.800 Up 1.700

ADD Maintain ADD 4.100 Same 4.100

HOLD Upgrade REDUCE 2.500 Same 2.500