Telecommunications Infrastructure Scaling a New Normal

Pressure on lease rates… XL Axiata’s (XL) (EXCL IJ, BUY, TP: IDR4,550)and Sarana Menara Nusantara’s (SMN) (TOWR IJ, BUY, TP: IDR4,700)Protelindo subsidiary’srecent sale and leaseback deal for 2,500 towers suggests that industry lease rentals remain under pressure. We think XL’s IDR10m per month/site rental has become the new industry benchmark for future lease rental negotiations upon contract renewals; it was IDR15m-18m/month in the past (typically by the bigger towercos). This wasgiven the greater competition within the industryin recent years as well as the telcos’ strong desire to minimise opex.

….could crimp EBITDA in the long run. The telecommunications tower infrastructure companies’ (towercos)currently attractive EBITDA margins are still – to a large extent – protected by higher lease rates under 10-15 year master leasing agreements with the telcos. Nonetheless, we see downside risk to industry EBITDAmargins going forward,driven by increasing contract renewals,with lower rates also applied to new build-to-suit sites. However, the impact shouldbe mitigated by legacy provisions within contracts that prohibit significant variations to lease charges and the timing of contract renewals.

Microcell poles (MCPs)lookset to be the new growth catalyst
We see strong potential for MCPs (or base transceiver station (BTS) hotels) and fibre infrastructure. This isgiven thesite acquisition challenges in the bigger cities and rising demand for backhaul connectivity. Mobile operators are utilising MCPsas additional layers of capacityin dense locations. SMN is well-positioned to tap on the rising demand via wholly-owned Iforte Solusi Infotek (iForte), which it acquired in 2014. iForte owns 500 MCPs in Jakarta and >750km of fibre laid across dedicated busline network.

Potential tower M&As could drive earnings
Aside from XL, Indosat (ISAT IJ, NEUTRAL, TP: IDR5,650)plans to dispose of more towers to pare down debt. We believe this could be a short- to medium-term M&A catalyst for the sector. Indosat sold 2,500 towers to Tower Bersama Infrastructure (Tower Bersama) (TBIG IJ, NEUTRAL, TP: IDR6,450)in 2012 and, at its recent 1Q16 results briefing, reaffirmed its intent to monetise items listed under non-core assets like towers in subsequent quarters to pare down debt. Of the two leading towercos, we think SMN stands a good chance ofprocuring these towers,given its more manageable debt headroom.

Reinitiate coverageon the sector with a NEUTRAL weighting
At 11x FY17F EV/EBITDA, sectorvaluations are at the lowend of the 15x historical mean. This ispost the sell-down of the last 12-18 months, which we deem as fair. We think the strong earnings growthof the past is behind this sector. Our Top Pick for exposure isSMN as:
i. We think the market has not fully baked-in the potential from its recent acquisition of XL’s towers and its earlier diversification into MCPsvia iForte;
ii. Its 10x FY17 EV/EBITDA is at a discount to Tower Bersama’s14.7x.
The upside risksto our view are inorganic expansion and stronger-than-expected capex of the telcos. Downside risks are stronger-than-expected in lease rental erosions and weaker-than-expected site and tenancy additions.

Re-Setting Expectations On Indo Towercos
We reinitiate coverage on the domestictowerco sector with a NEUTRAL weightingpremised on the following:
i. The consolidation in the telco market that could lead to slower network expansion and co-locations going forward;
ii. The re-farming of the 1800 megahertz (MHz) band by the operators, which reduces the need for additional towers/sites;
iii. Escalating competition within the towerco industry and telcos’ cost management initiatives, which could further exacerbate the downward pressure on lease rates;
iv. Potential regulatory guidelines on active infrastructure sharing (likely by end-2016).
The good years are probably behind the sector. In our view, the strong earnings growth (40% on average between 2010 and 2015) – fuelled by aggressive 3G network expansion by the telcos– is behind the sector. We expect a medium-to longer-term structural decline in the industry’scapex intensity, supported by:
i. There-farmed 1800MHz band;
ii. More selective 4G rollouts by mobile operators utilising existing 3G sites.

In addition, we expect potential guidelines on activeinfrastructure-sharing to also help lowerthe capex of the telcosgoing forward. The domestic towerco sector has witnessed multi-year P/E and EV/EBITDA growth, driven by investors’ penchant for more defensive names that offer earnings certainty.

We re-assume coverage of thesector with a BUY on SMN and NEUTRAL on Tower Bersama.
SMN is our top pick for exposure; we like SMN due to its:
i. Stronger balance sheet, which portends scope for more inorganic expansion;
ii. Better growth prospects from recent towercoacquisitions(including MCPs);
iii. Valuationsmore attractive relative to the industry or peers.

Decent M&A flows and strong balance sheet
SMN has a decent track record for both organic and inorganic expansion over the past decade. This company’sstrategy for inorganic expansion saw notable acquisitionsof Hutchison Indonesia’s (Hutchison) tower assets in 2008 and 2010 (Hutchison remains a major customer today). A more recent and landmark deal with XL saw SMN acquire 2,500 of the former’s towers, boosting its number of tower sites owned to 14,737 (Figure 6 and 7).This is the largest amongst the local towercos. SMN also has one of the strongest balance sheets in the towerco industry, with its1Q16 net debt-to-EBITDA ratio of 0.9xbeing significantly lower than the global peer average of 6.04x andrival Tower Bersama’s 5.1x.This provides good debt headroom for further inorganic expansion going forward.

Attractive growthprospects
We forecast FY15-18revenue and EBITDA CAGRs of 11.8% and 12% respectively for SMN. This is ahead of Tower Bersama’s 6% (revenue) and 7% (EBITDA) CAGRs over the same period. The projections have factored in additional new towers/tenancies from the recently-acquired towers from XL.

We are NEUTRAL on Tower Bersama, as we expect it to record net additional tower tenancies of 1,400 and 1,000 in FY16 and FY17 respectively. Its FY13-14 tower tenancy net adds stood at 2,985/2,772 respectively. We also expect its tenancy ratio to remain flat in FY16 at 1.67, vs 1.65 in FY15. This is because we think the company’s organic growth would be subdued due to the telcos’ capex slowdown. In our view, it would be more difficult for Tower Bersama to grow inorganically due to its high gearing levels.

Longer-term tenancy outlook appears patchy
We gatheredfrom industry sources and telcosthat the sectormay look to scale back its BTS expansion in 2016. Thisfollows the refarming of the 1800MHz bandin late2015 and where existing sites can be utilised to deploy 4G equipment, ie less demand for new build-to-suit sites. Also, we think the telcos have excess capacity on their networks, thanks to M&As and network optimisation/modernisation initiatives, and may not be in a hurry to add capacity.

Unlike 3G, which drove significant capex uplifts in the past (this benefited the towercos), the use of single Radio Access Network (RAN) architecture by the telcossuggests that 4G spendingwould be incremental. Inability to grow due to a tight balance sheet. We noticed that Tower Bersama’s net debt-to-EBITDA ratio reached 5.1x in 1Q16 (a numberclose to its 6.5x covenant) which implies that it has limited room to further leverage its balance sheet to either grow organically (by building more towers ahead) or inorganically.

Limited Positive Catalysts In The Short Term
Pricing pressure from rising competition in the industry
Competition in the tower industry is getting more intense. Indonesian tower companies’ solid EBITDA and tenancy ratios have been attracting some new entrants into the market, in our opinion. These new entrants have created pricing pressure for the big independent tower companies (ITC) like SMN and Tower Bersama. After Telekomunikasi Indonesia (Telkom) (TLKM IJ, BUY, TP: IDR4,000) cancelledthe Mitratel deal with Tower Bersama in FY15 at the behest of its commissioners, the future of Mitratel and its 5,500 towers remain uncertain.

As of FY15, SMN is still the market leader with 12,156 towers, while Tower Bersama is the second largest with 11,389 towers. Solusi Tunas Pratama (STP) (SUPR IJ, NR) owns 6,790 towers, whilePT Inti Bangun Sejahtera (IBS) (IBST IJ, NR) has 2,185 towers. PT Komet Infra Nusantara (KIN), which owns 450 towers, is still flush with capital injections from telecoms and towerco thoroughbred private equity firm Providence Equity. The company plans to roll up selected small local towercos, which Indonesia has dozens.

Bargaining power of ITCs has lessened from competition and consolidation among telco operators. Currently, the lease rate of a tower is IDR15m per month for ITCs. The newcomers in the industry are offering lower lease ratescomparedto those of ITCs such as SMN and Tower Bersama. In 2014, XL sold 3,500 towers to STP and leased it back from the latter for IDR10m per month for each tower without the inflation-escalator clause.

The same ratewas applied in the latest transaction deal between XL and SMN – which indicates that the telcos are pressuring tower companies to lower their lease rates. Based on the latest transaction deal, the monthly lease rate is maintained at IDR10m per tower, which is lower than the average of IDR15m per tower.We think this would become the benchmark for future rental contracts.

In our view, competition among the towercos hasprovided the telco operators flexibility in selling their towers and getting better deals on lease rates. Furthermore, the smaller players are willing to compromise on the terms and conditions of the lease contracts, making them even more appealing to the telco operators. The discount ratesofferedby the smaller players have exerteddownward pressure on SMN’s and Tower Bersama’s new tower lease rates.

Potential pressure on EBITDA margins in the long run
In our view, towercos’ EBITDA margins have not reflected a decline in lease rates,given that the existing lease rates (at a higher rate) are contracted under10-15 year master lease agreements. However, we believe that new lease rentals at the build-to-suit tower sites would be lower than those at the existing sites. In the long run, we expect the average lease rates for existing towers to see only a marginal decline, as:
i. Contracts would expire at different time periods;
ii. It would not be easy for the telco operators to lower the lease rates for existing towers as there are some legacy contracts that prohibit such changes.

We foresee a potential decline in EBITDA marginson the back of:
i. Intensifying competition in the tower industry, which has attracted several new entrants;
ii. Likely lower lease rates from new towers. Based on the telco operators, the lease rates have been declining since last year and they are looking into tower operators that could give them the most favourable lease rate.

To better understand the dilution impact of a decline in lease rates on a new tower over the long run, we conducted an analysis to assess the impact on total return on investment (ROI) per tower site. In our view, every 5% decline in lease rates could lead to a 1.5% drop in IRR for SMN and Tower Bersama. Note that in our scenario analysis (Figures 10 and 11), we assume each new tower hasa 1.7x tenancy ratio from its first year. Realistically, the cash flow per site for the first few years is lower than what we have built into our scenario.

No more high capex on towercos
Indonesian towercos had seen significant revenue growth during the rapid rollout of 3G services in 2012-2013,with Tower Bersama and SMN chalking up an average 61% and 40% revenue growth respectively during the period. We believe that a new technology rollout may not necessarily translate into equivalent growth fortowercos, since not every 4G BTS activated by the telcos requires a new tenancy spot on a tower. Going forward, the Government plans to roll out 4G services on the 900MHz spectrum, with services expanding to the 1,800MHz bandlater.

More than 50% of the telcos’ 3G BTS are running on a single Radio Access Network (RAN), which means thatthey are equipped with the technology to serve 4G LTE devices (only a few adjustments to the BTS are required to switch the stationto 4G from 2G/3G). In our opinion, the telcos will likely invest more in fibre infrastructure rather than adding more new towers to ensure the provision of high-quality 4G LTE services. Thus, we believe the telcos’ capex for BTS in FY16 would be limited as they focus on rolling out fibre optic cables.

Land lease rateslikely to be more expensive
Towercos’ land lease expenses are subject to negotiations with the landlords during each renewal cycle (probably around 10 years). In the long run, we think that the landlords, who are becoming more well-informed,would demand higher lease rates which could result inhigher-than-inflation adjustments on an annualised basis.

Furthermore, we believe that land lease rates in new locations would be higher than those inthe existing ones. In addition, the community consent agreement would become harder and more expensive to obtain. As land gets scarcer, the local community would demand higher compensation from the towercos.