Please note: this comment is my personal opinion and is not investment advice.
The irresistible rise of the Netco
Having tried every other possible strategy first, the telco sector is finally coming round to the idea of creating separate wholesale-only fixed network infrastructure companies (“netcos”), a business model which has scope to allow telco network businesses to grow faster, be run more efficiently and be less tightly regulated. This is now an important theme in the telco sector. But what are the circumstances where such netco separation works best?
Growing momentum
Telecom Italia has recently sold to KKR 37.5% of a company containing its sub-loop fixed network, to help fund a FTTH roll-out in Italy. In November EQT made a friendly takeover approach to KPN according to Bloomberg with an eye on the telco’s network infrastructure asset. EQT funds already own Delta Fibre, a Dutch FTTH network. At the November 2020 Morgan Stanley TMT conference BT’s CEO said for the first time that he would be “open-minded” about selling a minority stake in Openreach once fibre regulation was decided (April 2021 theoretically), after years of BT completely ruling out such an idea. Liberty Global CEO Mike Fries has also recently voiced an interest in the company developing its own infrastructure assets. Further afield Telkom SA is now contemplating the sale of a stake in its Openserve netco operation, alongside tower and data-centre transactions.
The Nordic markets pioneered the netco-serveco approach when funding local FTTH builds starting in the 2000’s, with the local municipality building and operating the FTTH network and providing wholesale access to third party service providers on a neutral basis. New Zealand was the first to break up an existing telco into netco-serveco elements, separating Chorus (fixed network) from Spark (service business and mobile network) in 2011, the government then granting Chorus a subsidised franchise to build and operate fibre (FTTH) across most of the country on a wholesale-only basis. The result nine years later is a country with 80% of homes passed by FTTH and with the combined market capitalisation of Spark and Chorus up 2.5x since separation in 2011. In 2018 a Macquarie-backed consortium bought Danish incumbent telco TDC, has separated it into netco and serveco divisions, and is now upgrading the network to FTTH. In 2019 Altice sold stakes in its newly created Meo FTTH (Portugal) netco and SFR FTTH (France) netco whilst Iliad sold a stake in France FTTH projects.
Underlying economic logic
A full separation of netco from serveco operations has an underlying economic logic, as each have differing economics of scale. Netco’s scale economies are local, focused on maximising the number of customers per home passed by the network (i.e. customer penetration). This network scale resembles an upward sloping S-curve, so a 10% increase in penetration above a break-even point can mean a 25-50%+ increase in profits. In contrast, servecos – the likes of Iliad, 1&1, Freenet, TalkTalk but also the retail arms of the incumbent telcos - tend to maximise value by taking the most customer share they can at the national level, sharing overheads (including brand costs and content rights) across the largest number of customers possible.
These diverging economic forces create an economic conflict of interest between the netco and serveco. The netco prospers by maximising the number of resellers of its network. The serveco prospers by reselling the best network in any given region available at the lowest possible price.
Several factors have come along to shake-up the status quo and improve the prospects for such netco separation deals.
· Telco share price performance has continued to lag the stock-market (the SXKP European telco sector index has underperformed the STOXX 600 by 11% in last 12 months, 23% in the last 3 years and by 43% in the last 5 years, all adjusting for dividends), putting pressure on telco execs to consider new strategies.
· Outside investors are interesting in buying telco networks. Lower interest rates have boosted the value of telco infrastructure assets as institutional investors continually search for yield. Growing end-customer demand for FTTH has transformed the business case for building fibre networks, and their value. Private equity firms are flush with unutilised funds.
· In Europe, the new EC regulatory code is offering a lighter touch regulation of telco networks if they operate on a wholesale-only basis.
Listing subsidiaries for valuation reasons hasn’t worked for telcos in the past
Telco execs have traditionally been very sceptical of full network separation. Spinning out parts of telcos has tended to disappoint in the past, most notably when minority stakes in ISPs and mobile network subsidiaries were listed in the early 2000s (DT-T-Online, FT-Wanadoo, BT – mmO2, FT – Orange, Telefonica –Moviles). These minority listings tended to highlight rather than reduce the sum-of-the-parts discount in the parent company share price and many such minority listings subsequently led to major problems when the telco had to then bundle-in the ISP service and then later the mobile service with the fixed line service to be competitive. More recent minority listings – e.g. DT’s listing of TMUS by merging into MetroPCS and Telecom Italia’s listing of tower-co Inwit – also don’t seem to have done much good to their parent’s stock price so far. In all these cases the potential SOTP upside was also undermined by the telco maintaining control rather than selling outright, a question that will probably recur with Vodafone/Vantage Towers in 2021.
But netco splits have potential industrial benefits
When considering netco separation, we should differentiate between these past ISP/mobile minority IPOs, which were done largely for valuation reasons (and didn’t work), and full netco separation, which can unlock significant industrial benefits, as argued above. Some telco execs are starting to agree, judging from the recent deals and commentary.
However separating the netco and servecos does have costs as well. The netco and serveco lose the informational advantage of being part of a vertically integrated telco, useful for each part in deciding where/when/how to invest. Furthermore the weaker side of the business – the side which has less relative competitive advantage - may lose some market share post separation. Furthermore credit rating agencies do not currently like the loss of financial scale involved in network separation – Moody’s has just cut TI another notch (to Ba2/Negative from Ba1/Negative), partly in response to the KKR-FibreCop deal. As separate businesses netco and serveco also require their own separate managerial functions, creating cost duplication.
So in practice, the overall benefits of a full netco separation depend on the size of the potential benefits vs these potential costs.
When would a netco separation deal work best?
Netco separation will usually entail one side of the telco winning partly at the expense of the other. Both sides will face more competition as well as more market opportunity. Ultimately whether there is an overall net benefit depends on the competitive position of the network vs its customer market share pre split. A telco which has a netco with a strong competitive position but this has not been fully reflected in telco market share due to being shackled to an underperforming serveco would likely be a net overall winner from separation, the gain in market share for the netco being greater than the loss in share by the serveco.
The most obvious case would be a new FTTH network whose serveco arm lacks scale in retail and lacks brand recognition. Many of the recent European netco deals do indeed involve such new (challenger) FTTH networks - i.e. SFR FTTH, Iliad FTTH, IP-only and mas fibre.
Regulated incumbent FTTH networks are also likely net winners from separation, as the serveco has already had to share its market with resellers and the netco may still benefit from separation into a company which is then incentivised to market its wholesale offer more aggressively and which may also benefit from lighter-touch regulation (i.e. more power to raise wholesale price). The incumbent FTTH netco that Altice Europe has created in Portugal (selling 49% to Morgan Stanley Infrastructure) is close to this paradigm.
At the other extreme a telco which has a netco with a weak competitive position but is outperforming on market share could be a net loser from separation. The loss in market share by the netco post separation could outweigh the gain at the serveco. A typical example would be an incumbent unregulated legacy xDSL telco facing overbuild by FTTH, cable and 5G-FWA but whose serveco has so far held onto a high customer market share due to its unregulated legacy incumbent position. There has been one netco-serveco deal not far from this paradigm - Windstream-Uniti entailed the sale & leaseback of Windstream’s DSL network to an SPV (Uniti) in 2015. Windstream entered chapter 11 in 2019, due to financial pressure stemming ultimately from the weak competitive position of the Uniti DSL network it was locked into making fixed payments for.
We must also factor in the relative valuation of the Netco and Serveco. Netco assets tend to trade at a premium to servecos, but the premium very much depends on the quality of the netco. A listed FTTH Netco could trade in the mid-teens EV/EBITDA and sell at 20x to a buyer, whilst listed servecos (TalkTalk etc) trade close to 6-7x EV/EBITDA. On this basis the market share gain by a FTTH netco would be worth 2-3x the same market share loss by the serveco. However, a listed xDSL netco facing competition from faster networks (FTTH, HFC and 5G-FWA) may not trade much above this 6-7x serveco multiple. This issue skews the overall cost/benefit analysis and ultimately means that netco separation is most likely to create value where the network has clear competitive advantage.
Finally, quite how netco and serveco compete post separation is uncertain ahead of a split, but particularly with regards to the serveco where performance is driven more by qualitative factors (brand, NPS, entrepreneurship) which don’t necessarily endure. Relying on upside from the serveco post split is risky. So where the netco has clear competitive advantage this reduces the risk in a netco-serveco split.
So the competitive position of the netco becomes critical to the cost/benefit analysis of a netco-serveco split. In this context it notable that the Macquarie consortium has not yet fully spun-out TDC Net but is instead upgrading it to fibre. Similarly TI has created FibreCop to raise the finance it otherwise lacked to upgrade it to FTTH; likewise KKR’s interest in FibreCop appears to come from the value created in upgrading the TI copper sub-loop netco to fibre.
Euskaltel
So where do cable (HFC) networks sit in this calculation? This brings us to the question of Euskaltel.
Euskaltel is a cable company and MVNO based in northern Spain. Its cable network passes 2.5m homes and has 711k cable customers (28% network penetration). Euskaltel’s cable network is 70-80% overbuilt by Telefonica FTTH and c. 33% overbuilt by Orange FTTH. Euskaltel is expanding into the national market using the Virgin Telco brand, reselling the FTTH networks of Orange, Telefonica and Adamo, and the mobile network of Orange. On its published business plan the company aims to take national market share and grow revenues to eu1.25bn by 2025 (from eu685m in FY19). Euskaltel also has a relatively low cost (<eu100 per home) to upgrade to FTTH, given it owns its local loop ducts which were quite recently built (c. 1995-2010).
Euskaltel’s largest shareholder is Zegona Communications which has a strategy of adding value to telcos it invests in. Euskaltel is reported to have recently hired Lazards to consider the potential for a sale of its network. Lazards advised on Iliad’s sale of a 51% stake in Iliad France FTTH projects to InfraVia in 2019, the sale by Altice Europe of the 49% stake in the Meo FTTH netco to Morgan Stanley Infrastructure in 2019 and the sale by TalkTalk of FibreNation to CityFibre in 2020.
Should Euskaltel go ahead and sell the HFC network or upgrade it to FTTH first? We consider our two factors identified above – the relative competitive position of the netco compared to the customer market share of the telco pre-split. Euskaltel’s HFC network is in a relatively weak competitive position, extensively overbuilt by FTTH. Whilst Docsis3 HFC networks compete theoretically with FTTH on downstream speeds, they lag on upstream, and can also suffer on downstream during busier periods. Furthermore, customers show a general preference for future-proofed fibre vs cable, whether this is rational or not. As an HFC network the Euskaltel netco would have to offer a wholesale price at a discount to the rival wholesale FTTH offers of Telefonica and Orange to compete, and be a price-taker in the market. Orange FTTH could be particularly disruptive as it has no price-discrimination obligation.
Euskaltel’s 28% penetration (connections relative to homes passed) corresponds to 35% market share (assuming Spain’s average 80% fixed line penetration of homes). Such a market share already reflects an uncompetitive netco position, having ceded market share to FTTH. Serveco might therefore increase market share by being able to resell Telefonica and Orange as well (which it is already doing outside the Euskaltel network areas). However part of this benefit to serveco would likely come from switching customers away from netco. Netco might have to discount wholesale price more to keep serveco customers. This would lower the value of netco considerably, both in terms of its cashflow and the valuation multiple applied. Euskaltel is therefore an example where the lack of clear competitive advantage of the netco significantly skews and undermines the cost/benefit of a split.
Euskaltel’s netco would be in a better position if it had been upgraded to FTTH prior to separation, being able then to command a higher wholesale price, more market share and a higher (potentially much higher) valuation, along with greater leverage capacity.
Conclusion
Network separation is a not a panacea, and should in any case be considered for its potential industrial benefits. Apparent sum-of-the-parts upside should only be considered as a secondary factor, if at all. The competitive position of the netco is absolutely critical to the cost/benefit of a split. Cable (HFC) netco deals may struggle to create value if the netco is overbuilt by FTTH. So Euskaltel is fortunate in having a cheap path to FTTH, which it should probably pursue before considering any network split/disposal. Liberty Global will also have to weigh its options carefully. And we should see this netco question as one factor driving a general trend towards cable companies upgrading to full FTTH.
Please note: this comment is my personal opinion and is not investment advice.