Telecom Carriers: Why Aren’t These Cyclicals Cycling?
Paul Sagawa / Artur Pylak
sagawa@ / firstname.lastname@example.org
April 2, 2012
Telecom Carriers: Why Aren’t These Cyclicals Cycling?
- Telecom ought to be cyclical, but lately, it hasn’t been. Other asset intensive, low marginal cost industries have seen capex turn upward over the past few years, driven by recovery, low interest rates and tax incentives, while capex by telecom leaders VZ and T continues to set new historical lows despite their complaints about overcrowded networks and pleas for new spectrum. The best explanation is a lack of competition, spurred by the business shift to wireless where asymmetrical spectrum holdings exacerbate a tepid rivalry. The result is wireless prices that are second only to Canada, and strong and stable cash flows for the market leaders. Recent actions, such as the blocked T/TMobile merger and the proposed VZ/Cable tie up suggest both an intention by industry leaders to thwart rivalry, and the intent of the current FCC to promote it. We are concerned that the strong recent returns for the top carriers are vulnerable to government policy and renewed rivalry, and prefer secondary carriers, tower companies and equipment suppliers.
- Telecom carrier capital spending and cash flows no longer seem to follow the cyclical pattern typical of asset intensive, low marginal cost businesses. High fixed cost businesses – e.g. airlines, railroads, heavy manufacturing, and chemicals, et al. – ordinarily follow a well-worn path of rising cash flows to rising investment to plummeting returns to sharp capex cuts. Telecom, despite one of the most asset heavy profiles in the economy, no longer seems to follow this path. While the rest of the economy sees rising capex, spurred by recovery, interest rates and tax incentives, VZ and T are taking network spending lower and raising prices – pleasing investors and vexing regulators.
- The reduction in capital intensity has coincided with the ascension of wireless, despite the apparent wireless data crisis. Wireless has risen from 24% of telco revenues to 66% over the past decade, while capex/sales has fallen sharply. On face value, this is curious, as US carriers have been vocal about looming capacity shortages in the face of explosive demand for mobile data. Wireless capacity can be added in two ways – spectrum or capex – and while both AT&T and Verizon have chased the former via aggressive deals, neither has been willing to invest in new cell sites.
- Verizon and AT&T’s superior spectrum holdings already give them enormous competitive advantage, evident in market share, profitability, and pricing. T and VZ operate primarily in the 850MHz and 700MHz bands, allowing cells with nearly 50% better range and more than twice the coverage area than in the 1.9GHz band used by Sprint, T-Mobile and others. This means much better coverage with many fewer cells and thus, less investment. Verizon and AT&T’s coverage advantage has allowed them to build leading share while blunting price competition from their less fortunate rivals. Over many years, share advantage has translated to significant scale advantage, but has not translated into price cuts for customers.
- International markets show the same pattern of spectrum inequality, market concentration, and muted investment. Typically, the first carriers in a market operate on more attractive frequencies, giving them cost and coverage advantages over later entrants in higher bands. The extent of the benefit is tied to the population density of each country. In countries where spectrum allocations are skewed, market share is concentrated and population is sparse, prices tend to be high. This is clearly the case in the Canada and the U.S., where monthly wireless bills are the highest in the developed world.
- We are concerned that the currently strong telecom industry cash flows are unstable and could be severely affected by regulation or changing rivalry. Historically strong cash flows, from high prices and low capex, have allowed T and VZ to sustain nearly 6% dividend yields and strong share price appreciation. This Pax Telco–mana is vulnerable to regulatory and judicial actions by the government to promote competition, threats by non-traditional competitors and/or an aggressive competitive move by either of the leaders.
- U.S. regulators may act to promote competition, which could yield increased spending and renewed price rivalry. The FCC and DOJ blocked T’s T-Mobile acquisition, and may do the same to VZ’s spectrum deal with cable MSOs. Congress has authorized the FCC to auction as much as 120MHz of highly desirable spectrum, with leeway to set rules to promote competitive balance. A Republican FCC could yield rules more favorable to the market leaders, so we expect the current commission to push to set the auction conditions as quickly as possible.
- Cash rich Internet companies have strong incentive to promote a competitive wireless data market. Web leaders, like Google, Apple, Microsoft, and Amazon, are focusing on their mobile products for much of their anticipated growth, and to various degrees, have been aggressive investors in infrastructure. These companies could inject capital to support second tier carriers, or even bid on spectrum directly, looking to promote lower prices and higher caps for mobile data.
- We see significant risk to the top telecom carriers. We prefer secondary carriers, equipment vendors and tower companies. T and VZ have delivered extraordinary returns over the past two years, and still support nearly 6% dividend yields. We are concerned that the cash flows underlying these earnings and pay-outs are at risk to future competition. We prefer secondary carriers (which could benefit from new spectrum), telecommunications equipment vendors (which have suffered a multiyear slump,) and to a lesser extent, tower companies (which are already richly valued).
It Quacks Like a Duck
There are few industries more asset intensive than telecommunications. There are essentially no marginal costs to carrying one more telephone call, one more text message or one more web site download. Comparing this to other notoriously asset heavy businesses like electric utilities, semiconductor manufacturing, railroads or airlines on depreciation to COGS shows that telecom carriers need take a back seat to no one (Exhibit 1). The business is almost all fixed costs, with some semi-variable costs that can be cut with some advance notice, and the merest whisper of truly variable costs that rise and fall with usage, moment to moment (Exhibit 2).
Exh 1: Depreciation to Cost of Goods Sold across Capital Intensive Industries, 1980-2011
Industries like this are almost always cyclical in nature. As demand rises and capacity fills, prices rise to reflect scarcity. With higher prices, one competitor breaks ranks and adds capacity in a bold play for market share and the others are forced to follow or cede permanent scale advantage to the others. Over capacity then yields falling prices and plummeting returns, at least until the low prices draw enough demand to start the cycle all over again. For many decades, even with the AT&T monopoly and tight government return on investment regulation, telecom spending was cyclical. Post Bell System break-up, the cyclicality was even more evident, culminating with the bursting of the Internet bubble in 2000. Since then? Not so much.
For the two top telecom carriers, Verizon and AT&T, capex as a percentage of sales has slid steadily to historic lows, while operating cash flows have remained historically strong, yielding consistent and excellent free cash flow returns (Exhibit 3). Meanwhile, smaller carriers have not fared quite so well. The next level of carriers, led by Sprint and including Deutsche Telekom’s T-Mobile USA subsidiary, regional carriers US Cellular, Metro PCS and Leap Wireless, CLECs Century Telecom and Cincinnati Bell, and alternative long distance players Level3, Global Crossing (acquired by Level 3), and Qwest (Acquired by CenturyTel), have all experienced a cash flow squeeze (Exhibit 4).
Exh 2: Sales to Fixed Assets across Capital Intensive Industries, 1980-2011
Exh 3: AT&T and Verizon CAPEX and Cash from Operations as a Percent of Sales, 1984-2011
Exh 4: Capex as a percent of Sales, Second Tier US Carriers*, 2002-2011
Can You Hear Me Now?
In the past decade, carrier wireless revenues have grown rapidly while their wired network businesses have contracted, raising mobile services to 45% and 63% of total revenues for industry leaders AT&T and Verizon. With mobile sales now pre-eminent, perhaps wireless is just different? A cursory examination would suggest this is not so. Wireless-only carriers in the US and abroad have shown cyclicality in their capital spending in previous periods, and the ratio of depreciation to COGS reveals substantial capital intensity in the business. Moreover, most US markets feature at least 4 and as many as 6 networks offering service (Exhibit 5).
Exh 5: US Wireless Carrier Metrics, Q4 2011
Lately, US carriers have been raising hue and cry over spectrum. The wild popularity of smartphones has saturated the data carrying capacity of 3G wireless networks, according to the carriers, making it imperative that the FCC free up additional radio frequencies for their use. Indeed, AT&T’s stated objective in its recent attempt to acquire Deutsche Telekom’s T-Mobile USA subsidiary was to add its 49MHz of spectrum to its own 88MHz. By AT&T’s reasoning, reducing the number of nationwide competitors by 33% played no role in the deal, but the deal was struck down by the FCC as anti-competitive anyway. The commission is also taking a close look at Verizon’s proposed $3.6B deal for 40MHz of unused spectrum currently licensed to a consortium of cable operators, a transaction that carries with it a commitment by both sides to market each other’s products (and by implication, stay out of their businesses, thank you very much). Verizon and AT&T have also been active in lobbying Congress to block the FCC from imposing restrictions on how much spectrum they might each purchase in future auctions. Clearly, America’s top wireless providers are not shy about spending money on spectrum.
Of course, adding spectrum is not the only way to add capacity to a wireless network. The classic method is called cell splitting, whereby a second base station can be added to a cell experiencing congestion, with each base station taking half the traffic. This is not as expensive as it sounds – today, a new base station costs less than $40,000, with additional costs for renting tower space and establishing means to backhaul traffic (Exhibit 6). Splitting a cell may not even require finding a new physical location, as modern antennae can be finely tuned to focus on a defined slice of area served by a tower, allowing a second base station to share space and backhaul with the first. AT&T turned to this technique in spades in the early part of the millennium, when it needed to squeeze its TDMA-based customers into a narrower band of spectrum to free channels in which to deploy GSM and WCDMA. Indeed this effort is evident as a modest bulge in the company’s capex between 2002 and 2005.
Exh 6: LTE Deployment Economics compared to DOCSIS and FTTH
Exh 7: Base Station Scale
Verizon and AT&T COULD do this again. They could expand 3G capacity in their most congested geographies by cell splitting, and in doing so, free up further frequencies for their 4G LTE networks. They could also augment their cell networks with tiny “femto” cells tied to wired networks or even with WiFi (Exhibit 7). Yet, the two market leaders would rather drive capex lower, even as a modest economic recovery, record low interest rates and unusual tax incentives to invest have driven capital spending in the rest of the economy higher. Over the past three years, the direction of capital spending for most asset intensive industries has been up (Exhibit 8).
Exh 8: Capex as a Percent of Sales, Capital Intensive Industries*, 2006-2011
Instead, the top telecos would rather risk shareholder money on M&A transactions that face regulatory scrutiny. Indeed, AT&T’s star-crossed T-Mobile deal cost its shareholders a $4B penalty along with the transfer of some of AT&T’s supposedly scarce spectrum back to DT. Adding in legal costs, other transaction costs and the time value of money, it would seem that AT&T’s resources would have been much better spent buying equipment and splitting cells, IF the objective was simply to add network capacity.
Exh 9: Relative Coverage Advantage of Lower Frequency Spectrum
All Networks are not created equal
Like in most countries around the world, the U.S. wireless carrier industry grew in stages. The first two licenses were granted in 1982 to the soon-to-be-broken up AT&T and to a series of new companies, led by industry pioneer Craig McCaw and LIN Broadcasting. These first licenses were in the 850MHz band, very attractive spectrum just above the broadcast TV band that enabled very long range with high quality reception in and out of doors. Eventually the licenses in this band were aggregated and acquired by the present day AT&T and Verizon. The second set of licenses were auctioned in 1994, and offered frequencies in the 1.9GHz band. This spectrum is vastly inferior to the AT&T/Verizon spectrum, limiting the maximum effective size of a cell site to less than half that of the incumbent networks with commensurately worse reception (Exhibit 9). The spectrum disadvantage is most acute in more sparsely populated areas with flat topology, where traffic rarely requires cell splitting and the lower banded operator can make do with dramatically fewer cell sites. Conversely, congested and/or mountainous geographies blunt the spectrum advantage, as capacity considerations can dictate cell sizes well below maximum at either frequency.
Over time, the natural superiority of the two incumbent’s spectrum holdings has translated to substantial market advantage, despite the periodic release of further, generally higher, bands to the mobile communications market. Together, Verizon and AT&T hold 64% of the market, and importantly, generate 86% of the industry’s operating cash flows, despite a roughly 25% price premium over the next largest competitors, Sprint and T-Mobile.
It’s a World-Wide Phenomenon!
Globally, the picture is similar with highly concentrated wireless markets, as measured by the Herfindahl index, widely used as a screening tool by antitrust regulators in determining whether a proposed merger is likely to raise antitrust concerns. When applied to wireless networks in various countries, a pattern of uncompetitive environments emerges: politically enabled monopolies, natural monopolies, duopoly/oligopolies, and planned competition (Exhibit 10).
In China and Mexico, the market is dominated by a single player with over two thirds market share. China Mobile controls 67.8% of the Chinese market which translates to over 650M subscribers, over 6 times the size of AT&T or Verizon. While it is also the 8th largest company in the world in terms of market cap, it is also a publicly traded state-owned enterprise that enjoys protectionist benefits. China Mobile’s main competitors China Unicom and China Telecom , with 20% and 10% market shares respectively, are also publicly straded state-owned enterprises, suggesting that the state sanctioned dominance of China Mobile is extremely unlikely to see vigorous rivalry any time in the future.
Exh 10: Wireless Market Competitiveness and Population Density by Region/Country
In Mexico, Carlos Slim’s Telcel subsidiary of America Movil has 70.8% of the market. The company has licenses for key spectrum in the 850MHz and 1.9GHz ranges in all nine geographic regions of Mexico. Monopolies are commonplace in Mexico and the result of powerful interests successfully controlling weak government institutions and influencing political parties. Despite this control, Telcel’s network only covers 63% of the country geographically and 90% of the population, as competitive pressures are not sufficient to motivate Telcel to invest in broadening its coverage. Not surprisingly, the company has numerous complaints on Mexican consumer websites, while Mr. Slim recently ascended to the title of “world’s richest man”.
Japan presents an example of a politically enabled monopoly to a lesser extent. NTT Docomo is majority owned by Nippon Telephone and Telegraph (NTT), which was broken up in the 1999 much like Ma Bell in 1984, and still about one-third government owned. Despite privatization and the introduction of competition, NTT Docomo still has nearly 50% market share in Japan. Japan’s population density and mountainous topology negate much of the advantage of lower spectrum bands, but scale advantages in amortizing fixed costs and branding with Japanese consumers remain potent advantages for the original incumbent. Competitors KDDI and Softbank have recently fought back, gaining some subscribers at the expense of NTT due to number portability and deals to carry Apple’s iPhone.
The advantages of spectrum assignment are most evident in sparsely populated markets, such as Canada, Australia, and Sweden. Each country has a market share leader with about 40% share. In Canada, Rogers has a greater spectrum allocation than its competitors and offers coverage to 95% of the Canadian population with 2G GSM and EDGE. The company has been able to cover more of Canada’s population because of its allocation of lower frequency spectrum (850MHz), which is more powerful and requires fewer cell sites for a given area than 1.9GHz. A Canadian government study found that for a small Canadian city the 1.9GHz spectrum requires 30-50% more cell sites while for a larger area with more peripheral coverage requirements twice as many cell sites may be necessary. In Australia, the story is similar with Telstra’s spectrum allocation in lower bands vs. competitors with much higher frequency assignments. At 850MHz, Telstra claims to cover 99% of the Australian population with its 3G network. In Sweden, TeliaSonera has 44% share and licenses for lower frequency spectrum. New entrants are unlikely since it is hard to justify entering a small yet capital-intensive market.
In larger European countries, competition resembles the US market with an oligopoly with 2-4 dominant operators. In the UK, O2 and Vodafone each have about a quarter of the market while Orange and T-Mobile have combined their networks into a joint venture called Everything Everywhere with about one third share. Vodafone and O2 have the most desirable spectrum at 850/900MHz while T-Mobile and Orange operate in the 1800/2100MHz range. Given the UK’s status as one of the most densely populated countries in Europe, operating at a higher frequency is not as big a competitive disadvantage (Exhibit 11).
Exh 11: Spectrum Holdings in France and the UK
France has a similar structure to the US with a duopoly of France Telecom’s Orange and Vivendi’s SFR dominating the market followed by Bouygues Telecom with fewer spectrum holdings in densely populated areas (Exhibit 12). The French market is notable with the recent launch of a disruptive fourth wireless player called Free Mobile. The company is the fourth entity to be awarded spectrum in France. While it currently covers only 27% of the population with its network, Free Mobile was able to strike a roaming agreement with Orange. ARCEP, the French telecom regulator made it a requirement for incumbent carriers to offer 2G roaming on their networks to new entrants. Free Mobile is unique in that it uses a combination of WiFi, HSPA, and 2G to offer its service. Free is also a provider of ADSL broadband service in France and uses 5M WiFi access points located in customer homes and businesses serve as the backbone of its network. When not near a WiFi access point, devices switch the HSPA networks built by either Free or roaming partner Orange. Free Mobile launched its service January 2012 and has since taken about 201,000 of Orange’s subscribers in about one month. Free Mobile offers an unlimited voice, data, and text messaging plan for 20 Euros, or $26.50 at current exchange rates. Incumbent carriers have introduced low priced plans to compete with Free. ARCEP has since expressed concern about the state of the wireless market as incumbent carriers have faced sales and margin pressures responding to the entry of Free Mobile.
Exh 12: Average Carrier Spectrum Holdings by Country – Europe
India on the other hand, seems to present the case of a highly competitive market. It is the second largest mobile market in the world with nearly 900M mobile subscribers, competition is planned and the market is fragmented with not one player having more than 20% market share nationwide. Aside from being the second most populous country in the world, India is also the second most densely populated country in our sample and it makes sense to deploy higher frequency spectrum. India is noted for having a sophisticated regulatory regime with complex licensing systems based on regions. Each player is noted to have a set of licenses that cover part of the country and no single mobile operator has licenses that cover the entire country. The result is oligopolistic competition regionally with the appearance of perfect competition nationally. The complexity of India’s spectrum licensing scheme is not without controversy. A scandal involving the allocation of 2G spectrum in 2008 led the Indian Supreme Court to nullify 122 2G spectrum licenses in February 2012. Various officials and executives were accused of corruption for undervaluing the spectrum. These licenses will be reallocated in a competitive bidding process in the coming months. It remains to be seen if Indian network carriers will be allowed to consolidate and truly national carriers will emerge.
We Can Fix It! … My Brother’s Got an Awesome Set of Tools!
Back in the 1950’s, the Tennessee Valley Authority noticed something interesting. On every bid for heavy equipment for its power system – e.g. turbines and the like – the losing bids would come in exactly the same. Given the highly complex nature of the products in question and the supposedly secret bidding process, this was curious indeed. In fact, a cartel of heavy equipment suppliers were meeting ahead of every bid, deciding who would be next to submit a winning bid according to the phases of the moon. Strangely enough, the cartel would also agree to the value of the losing bids, a degree of unnecessary cooperation that proved to be their undoing. In 1961, 50 executives were fined, with 9 GE and Westinghouse managers receiving jail time.
More recently, U.S. courts levied fines of $1.7B against 21 international airlines, including British Air, Singapore Airlines, Air France, Korean Air and other, for conspiring to set artificially high fuel surcharges on flights. Fifteen airline executives faced criminal charges. This is merely the latest in a long string of price fixing schemes that have been uncovered in the airline industry over many decades.
In high tech, 7 manufacturers of large format LCD displays, led by Samsung and Sharp, were recently fined more than $500M for having conspired to fix prices between 1996 and 2003. In 2002, the US DOJ launched a probe into price fixing allegations against the DRAM industry, resulting in guilty pleas and fines against Micron, Infineon, Samsung, Hynix and Elpida.
All of these industries are high fixed cost, low marginal cost businesses. When marginal cost is near zero and there is ample capacity, almost any price can make a contribution against fixed cost, and something is better than nothing. Without variable costs to set a floor, these industries have often experienced bouts of crushing price competition, conditions that typically last until enough demand is stimulated to fill capacity and quell the competitive rivalry. Under these circumstances, it is difficult to deliver sustained returns – most of these industries show dramatic fluctuations in ROIC over time (Exhibit 13-14). This, in turn, is an enormous temptation for should-be rivals to co-operate on pricing.
Exh 13: Return on Invested Capital by Industry, 1980-2011
Exh 14: AT&T/VZ versus Industry, Return on Invested Capital, 1980-2011
Ma Bell Dies Hard
For most of its history, telecommunications was spared the fate of many of these asset-heavy, cash flow poor businesses through a combination of monopoly and regulation. Economic theory and practice show that businesses that manage to drive off or buy out competition and establish insurmountable barriers to entry can constrict investment and raise prices – essentially the early story of the fledgling AT&T that led to its regulation as a “common carrier” beginning in 1949. Thereafter, AT&T’s “rate of return” was managed by the FCC through strictly controlled prices, giving AT&T incentive to grow its revenues by increasing its investment back to its network, cyclically compressing free cash flow until rate relief kicked in. Of course, this mechanism was famously discarded for many of AT&T’s business with the 1982 consent decree that broke the giant into 9 separate companies. Over the subsequent three decades, the pieces of Ma Bell groped their way through massive industry change, driven not by long-distance competition as originally expected, but rather by challenges from the rapidly expanding cable industry and wireless. Eventually, the seven original RBOCs and the long distance operation recombined to form today’s AT&T and Verizon, with the former Mountain Bell, now Qwest, the lone outlier.
The US telecom industry has been roughly stable since the original AT&T was finally acquired back into SBC to form the current company. That stability appears to have been very good for the big two, as capex has drifted lower and operating cash flows higher in the years since. While competition between AT&T and Verizon has certainly erupted, it has been largely contained to a brand fight at retail, with AT&T’s iPhone coup overshadowing general service pricing that has remained remarkably similar and consistently higher than rates offered by their smaller competitors. This lack of aggressive pricing has resulted in average monthly wireless bills that are second only to Canada amongst developed economies, with the two North American markets both 30% higher than the number three Finns (Exhibit 15).
Exh 15: Minimum Cost of Complete Mobile Service Packages by Country
What Are You Gonna Do About It?
The FCC and the DOJ are concerned. AT&T’s planned acquisition of T-Mobile USA and its 10.3% market share and 49MHz of spectrum was nixed for its competitive impact. Verizon’s announced plan to acquire 40MHz of spectrum currently held by a consortium of cable MSOs and to enter into a cooperative cross-selling agreement is currently under scrutiny. In both cases, regulators have focused on the competitive impact of concentrating spectrum licenses into the hands of the top two wireless players. AT&T’s argument that the 30% of nationwide market share left to Sprint and a handful of regional competitors after its deal constituted real competition for consumers was poorly received, despite a lobbyist fueled din in support of the deal from corners of Congress distrustful of government intervention. Verizon’s proposal will likely get its day in court before the possibility of a reprieve from a more industry friendly administration. Early signs are negative, with the explicit cross marketing agreement between Verizon and its putative competitors a further anti-competitive wrinkle for the government to consider.
Even if Verizon’s deal is approved, there are other government threats to the cozy VZ-T cash flow club. Most importantly, new spectrum auctions may better equip challengers to compete with the big two. In February, Congress finally authorized the FCC to conduct incentive auctions of extremely attractive 750MHz band spectrum currently controlled by television broadcast stations. These channel allocations are based on 50 year old broadcast technology – the stations could deliver their programming over much smaller slice. Moreover, these stations were deliberately spaced across the band to avoid the interference commonplace during the analog era, leaving irregular blocks arrayed both frequency and geography. The incentive auctions will offer these station owners a portion of auction proceeds to agree to move to smaller slots in contiguous channels, thus freeing large blocks conducive to modern wireless communications.
Exh 16: National Broadband Plan Proposed Mobile/Broadband Spectrum
The auctions themselves are a bit of a compromise between political forces that favor using the auctions to spur competition and those that favor a light regulatory hand on the market leaders. The auction of up to 120MHz will be open to all comers – including Verizon and AT&T – so the market leaders will be able to add to their holdings (Exhibit 16). However, the FCC will be free to set limitations on how much carriers in the market will be able to have and/or buy. This provision assures that a fair size slug will be won by someone other than Verizon or AT&T. It also reiterates the FCCs mandate to manage spectrum holdings for competitive impact, supporting its authority to block Verizon’s cable deal if it deems it necessary and to review the impact of any spectrum deals by either of the two. This reduces the chance of a deal for Dish Network’s 40MHz spectrum holdings, but increases the likelihood that those holdings can be combined with new frequency blocks to create a viable market entrant. We note that such a market entrant could then sell wholesale capacity to the big two, but that the spectrum license asset would remain independent.
How Ya Gonna Pay For That?
Supporters of the incentive auction proposal seduced Congress with whispers of up to $35B paid into the US Treasury. The big question: Who, other than Verizon and AT&T, would have the financial resources to pay those sort of prices? Sprint is a financial mess and T-Mobile owner Deutsche Telekom seems more interested in cutting and running than doubling down. Cable operators are trying to sell their current spectrum holdings to Verizon, so why would they be interested in buying even more?
The answer is likely outside the line-up of usual suspects. The crew of cash rich internet leaders, with their strong vested interest in seeing fast cheap and unfettered wireless data services, are particularly intriguintg. While Apple, Google and Microsoft might have reservations about competing directly with Verizon and AT&T, given their dependence on them to distribute their platforms, they could be interested in providing or facilitating wholesale capacity for the industry, or potentially, using the spectrum to offer fixed residential wireless broadband service in direct competition with cable operators. Admittedly, this scenario seems to be more Google-like than either Microsoft or Apple, but all of these companies surely see the benefits of promoting wireless competition. Extending the circle a bit further, Facebook and Amazon also fit the profile of cash rich and clear beneficiaries of a more competitive and open wireless industry.
While the concept may seem farfetched, remember that Intel, Samsung and Motorola invested billions of dollars into Clearwire to fund that company’s network based on the now dead-end WiMAX technology. Clearwire still holds 150MHz of spectrum licenses in the 2.6GHz frequency range and is looking for funding to build over its orphaned network with the industry standard LTE technology. We note that Clearwire’s spectrum allows a maximum cell area less than half that of either Sprint or T-Mobile, and a small fraction of that available to the market leaders. That the Intel, Samsung and Motorola were willing to take that bet suggests that there may be other outside participants from the technology world willing to invest in wireless buildouts in more attractive frequency ranges.
Purely financial players may also be intrigued by the impending auction. Phil Falcone’s Harbinger Capital raised billions of dollars to fund LightSquared’s proposed LTE network using spectrum licensed for satellite operations, although the project ended badly when it was determined that the network would interfere with the GPS system on an adjacent frequency. A network in pre-approved, prime spectrum could be a far more attractive investment. Other peripherally associated companies could see wireless as a business extension. Dish Network is already planning its own wholesale LTE network on non-interfering satellite spectrum in the 2GHz band, with the possibility that the FCC may execute a swap that could put Dish in more attractive spectrum in the 1.7GHz range. We note that Dish also holds a 6MHz piece of spectrum in the prime 700MHz block that is too small to be useful for LTE on its own, but could be valuable in conjunction with further purchases in the licensing auction to come.
Scenarios, I’ll Give You Scenarios …
One plausible scenario is that nothing really changes. Spectrum auctions are delayed, and political change brings an FCC not inclined to manage competition in wireless markets. Verizon and AT&T continue to maintain a significant price umbrella, with high prices and usage caps to keep mobile data traffic to a manageable growth curve. In such a scenario, the two top carriers could enjoy strong returns on capital with dampened industry growth. While good for VZ and T shareholders, this scenario would be difficult for the mobile device industry, as the utility of wireless data would be limited to customers able to afford the most expensive pricing plans. It would also be bad for telecommunications equipment vendors and tower companies, as dampened mobile data demand would be accompanied by slowing network expansion and further reductions in capital spending. It would also be bad for secondary carriers in the market, which would see Verizon and AT&T extend their cost advantages leaving them at the mercy of the pricing umbrella set by the market leaders.
Telecom industry investors seem to be betting on some variant of the status quo, but we see the scenario as inherently unstable. Not only does the threat of government intervention and the potential for third party market entry remain, but the relationship between Verizon and AT&T could change as well. After AT&T’s failed attempt at merger, Verizon has a significant upper hand between the two, with a more broadly deployed 4G network and better spectrum holdings. A weakened AT&T is dangerous. Five years ago, with AT&T falling behind after its costly transition to GSM, AT&T broke ranks to offer the iPhone as an exclusive, taking significant share but simultaneously jail-breaking tens of millions of customers from the carrier restricted data access that had previously been the norm. If AT&T decides to go for market share, things could change in a hurry.
Another plausible scenario is that meaningful competition emerges via government encouragement and new investment. A vitalized Sprint or T-Mobile, or a rogue network funded by internet money could push an aggressive unlimited data agenda that forces prices lower and broad capacity expansion. Such a scenario would be very bad indeed for AT&T and Verizon, but unambiguously good for telecommunications equipment, towers, mobile devices. It might be good for secondary carriers, depending on the circumstances of new investment and new competition. It might be bad for fixed networks, cable and other wise, as well priced unlimited wireless data plans could easily compete for residential broadband as well (Exhibit 17).
Exh 17: The Winners and Losers: Most Likely Scenario
We are inclined to believe that reality will look more like the second scenario than the first. History suggests that cooperation between competitors rarely lasts for very long. Eventually, someone breaks ranks. We also believe that the incentive auction process will be well underway this year, and that limits on overall ownership will be in place. This could well be a catalyst that breaks the status quo, particularly if it draws internet players and their financial strength into the discussion. Recognizing risks that won’t be resolved until the auctions are complete, we still favor investment in mobile devices and towers as bets on market change. We are also intrigued by telecommunications equipment stocks, which have been notoriously poor performers. Any movement toward competition in the US market would also be favorable for these stocks, particularly if it is echoed by similar movement in global markets. On the other hand, we are skeptical that the leading telcos can sustain their returns on capital. Investments in smaller carriers may pay off handsomely, but remain quite risky under current conditions.
Exh 18: Annual Shareholder Returns by Telecom Sector, 2000-2011
Exh 19: Average P/E by Telecom Subsector
Exh 20: Average P/E by Telecom Subsector, 1990-2011
Exh 21: Average P/E by Telecom Subsector, 2008-2012