Wireless Carriers: Will the iPhone5 or New Spectrum Revive Carrier Competition?
SEE LAST PAGE OF THIS REPORT Paul Sagawa / Artur Pylak
FOR IMPORTANT DISCLOSURES 203.901.1633 /.1634
sagawa@ / email@example.com
October 3, 2012
Wireless Carriers: Will the iPhone5 or New Spectrum Revive Carrier Competition?
- The 2007 introduction of the iPhone changed wireless competition, stymieing carrier ambitions of AOL-like walled gardens and setting the subsidy bar much, much higher. For AT&T, the exclusive deal with Apple was worth it for the big market share gains and rising ARPU, but for #1 Verizon, the impact was unequivocally negative. Over the next five years, the industry balance of power stabilized, with the dominant players holding the line on data caps, service pricing, subsidies and capital investment, leading to notable outperformance by both stocks. However, historically, co-operation amongst wireless carriers has been unstable. The iPhone5, with high subsidies and new support for LTE service could spur more vigorous rivalry in the near term, with AT&T under pressure to counter Verizon’s superior 4G coverage. New spectrum, transitioned to commercial use by existing licensees or FCC auction, could also shift the competitive balance, particularly if new players – e.g. Google, Apple, Microsoft, etc. – step in to help smaller carriers co-ordinate network sharing and finance build-outs. These dynamics call into question the sustainability of oligopolistic returns for the leaders, and thus, the extraordinary dividend yields that they have supported.
- The iPhone changed everything for the wireless web. Prior to the iPhone, carriers sought to control the wireless internet through tightly curated, premium priced, packages of content exclusive to their brands. The iPhone took advantage of the growing reach and speed of 3G networks to offer a true browser, augmented by “app” shortcuts to popular web services, and an intuitive and powerful touch screen interface. AT&T, which had spent 5 years changing its network technology, had fallen far behind Verizon and agreed to unlimited open internet access, full Apple control of branding and apps, and extraordinary subsidies to get an iPhone exclusive. The deal reversed AT&T’s market share slide, and forced competitors to counter with unlimited plans and hefty equipment subsidies of their own.
- Over the past 5 years, carrier rivalry has stabilized. Now, T and VZ have the iPhone, and while carrier aspirations of controlling content are long dead, data usage caps, high prices and falling network capex are driving record cash flow for both leading operators and attractive dividend returns for investors. The question is whether the peace can be sustained – history shows that periods of strong carrier returns have been short lived.
- The iPhone5 could change everything again. Verizon has a huge advantage in 4G LTE, with a big head start on coverage and more and better spectrum than AT&T. Meanwhile, the iPhone5 finally adds LTE, 18 months after the first 4G Android devices hit Verizon’s network. To date, the expiration of AT&T’s iPhone exclusive has not driven the churn that many expected, but the stark difference in coverage and network capacity vs. Verizon, along with unlimited data offers from iPhone newcomer Sprint, could spur market share movements. Détente thus disrupted, usage caps could be lifted, prices could be dropped, and capex could be increased.
- The spectrum map STILL favors Verizon. With approval for its spectrum purchase from the cable industry, Verizon has two nationwide blocks to devote to LTE – its initial 4G deployments are in the 22MHz C-block VZ picked up in the 2008 700MHz license auctions, and the company now has 40MHz in the well supported AWS band. With these two widely supported bands, equipment and devices for Verizon’s network straightforward and easily available. In contrast, AT&T is deploying its LTE in 12 to 24 MHz of 700MHz spectrum, but lacks a national footprint and plans to augment its initial build-out with a variety of unusual frequency blocks, including some which do not have natural 2-way pairing and cannot be supported until the next version of LTE. The spectrum hodge-podge makes it more difficult to provide coverage and to source equipment or devices.
- New spectrum could also intensify competitive rivalry. DISH is awaiting an FCC ruling on its plan to use 40MHz of its spectrum for terrestrial services. Clearwire plans to build out LTE in its 150MHz of spectrum, but is cost and coverage disadvantaged by its very high frequency, requires specific equipment and devices, and has been challenged in raising financing. The FCC plans to issue rules for a new auction of up to 120MHz of the 700MHz band, with the proceeds to be shared with the TV station owners currently using it. Given the large amount of spectrum already controlled by T and VZ, it is likely that their participation in any auctions would be limited, and any attempt by them to acquire further spectrum through acquisition would be closely scrutinized.
- Internet players are a spectrum wildcard. Google, Apple, Amazon, Microsoft and others have substantial wallets and vested interest in seeing cheap and unlimited wireless data services. While we would not expect these companies to enter the carrier market directly, they could finance and support network building by others, perhaps coordinating wholesale network builds and facilitating network sharing agreements that would give smaller operators the capacity and cost structure to compete with the leading carriers on price and service.
- Increasing rivalry is a risk to AT&T and Verizon cash flows. Carrier cash flows are highly sensitive to changes in ARPU, customer acquisition costs and capital spending, all of which could move against the leading operators should either the iPhone5 or new spectrum catalyze competition on price, subsidies or data limits. Both telecom leaders currently generate operating cash flows after capex of greater than 11%. AT&T pays out roughly 70% of that cash as dividends, with Verizon paying out only half as much. Should rivalry push cash generation into single digits, where it has been during other stretches of more vigorous rivalry, even Verizon’s lower payout could be put at risk.
- Device platform players and small carriers would benefit from more vigorous competition. Google, Apple, Microsoft and Amazon are obvious beneficiaries should wireless prices fall and/or data caps become less restrictive. If new spectrum were to enable a viable wholesale LTE network, smaller carriers like Metro PCS, Leap Wireless and US Cellular would benefit.
First, the conclusions…
Telecom carrier dividends are a siren song for many investors. In a world of negative real interest rates, 4.5-5% yields with strong share appreciation over the past three years to boot from seemingly stable companies like AT&T and Verizon is hard to ignore. Moreover, the recent history of pricing discipline, new data usage caps, and falling capex have driven strong cash flows and the light of hope that the once rivalrous market leaders can sustain a peace to the benefit of their shareholders.
History suggests otherwise. In wireless markets all over the world, carriers have periodically stumbled into relative peace, but as service prices rise and subsidies and capex fall, someone eventually breaks ranks in a bid for share. The classic “prisoner’s dilemma” dynamic is familiar to anyone who has invested in other capital intensive businesses – there is a reason that heavy manufacturing and airline profits tend to be cyclical. Less than 6 years ago, AT&T, having lost significant share to Verizon during an expensive and disruptive transition to a new network technology, broke ranks to cut its exclusive deal for the original iPhone, bringing huge handset subsidies, unlimited data plans and the death of carrier controlled content portals in the process. In this light, it is wise to consider current conditions to be an unstable equilibrium.
The catalyst that disrupts the tacit cooperation could come from many directions, but the iPhone5 launch might be the nudge. Verizon enjoys a substantial advantage in 4G LTE, serving 5 times as many markets as AT&T within a 22MHz national band of prime spectrum, with a second 40MHz national block ready to go after its recently completed deal with the cable industry. AT&T’s roll out was held up during its unsuccessful bid for T-Mobile and its spectrum is piecemeal, with one or two 12 MHz prime spectrum blocks in most markets, and a less attractive 17MHz block stitched together through recent deals. The new iPhone finally supports LTE, and if Verizon starts taking market share, AT&T could be pushed to answer with lower prices, removing data limitations or boosting subsidies.
Competition could also come from below. Verizon and AT&T enjoy scale and spectrum advantages that give them lower costs and superior service vs. all other wireless competitors. New spectrum could close the gap, particularly if deep pocketed internet leaders like Google or Apple, with obvious vested interest in seeing cheaper wireless data, get involved in financing wholesale networks to be shared by the smaller carriers. Both Dish and Clearwire have plans to build such networks in their own substantial spectrum holdings, while the FCC expects to auction as much as 120MHz of prime spectrum in 2-3 years, sharing the proceeds with TV stations owners. Upgrades to LTE, already on their way to market, will increase both user speeds and system capacity, eventually allowing wireless to compete directly for residential broadband service as well as mobile data, thus, making an investment all the more attractive to an internet industry also stymied by the market power of cable operators.
Given the asset intensive, fixed cost heavy nature of their business, falling prices, increasing subsidies and/or rising capex would have an unusually painful effect on carrier cash flows, and thus, on their ability to fund dividends. AT&T’s dividends already eat up 70% of cash flow after capex, with Verizon paying out 35% of the cash generated by its business. Maintaining these levels will require an unlikely status quo in the competitive balance.
Life has been pretty good for AT&T, Verizon and their investors over the past few years. YoY, both stocks are up more than 25% to levels last seen in 2008, while delivering dividend yields of better than 4.5% (Exhibit 1). Revenues have been trending upward for both companies, while operating cash flows have held steady against capital spending that has been falling for Verizon and ticking upward slightly for AT&T. Wireless pricing has been benign, with both leading carriers able to institute firm data usage caps to conserve network capacity and forestall network capital investment. Churn for both carriers is now at all time lows, with less than 1% of subscribers cancelling service during the most recent quarter, and reducing pressure on customer retention and acquisition spending.
Exh 1: Dividend Yields of AT&T and Verizon versus 10 Year US T-Bill
It wasn’t always this way. Just six years ago, AT&T was on the ropes after having spent billions of dollars changing from its original 2G network built on the American IS-136 TDMA digital standard to a new network based on the more broadly adopted European GSM standard. In the process, service quality suffered as the spectrum used in the old network was shifted to the new one, squeezing capacity for existing users. Verizon had taken full advantage of AT&T’s plight to emphasize its own network capacity and reliability and enjoyed a substantial market share gain as a result.
Indeed, Verizon had its way on most things. Carriers held the upper hand with cell phone suppliers. Devices would sport carrier logos, sometimes to the exclusion of the handset maker’s own logo. Manufacturers were forced to replace their own software with user interface programs developed by the carrier. Data users were shunted through walled gardens of carrier curated content intended to reinstitute the AOL model of the Internet. High data prices and poor mobile browser performance further strengthened carrier control. Handset subsidies, long a bane to carrier economics worldwide but a proven tool for poaching market share, were at least stable and under $200.
Enter the iPhone
A desperate competitor is a dangerous competitor. In 2007, AT&T announced its exclusive deal with Apple for the original iPhone. There would be no carrier logo on this phone, nor would it sport AT&T developed bloatware or push users to AT&T’s content service. Data usage would be unlimited, with a fully functional browser and app shortcuts sending data over the carrier’s new 3G service. Within a year, AT&T was subsidizing the iPhone to the tune of $400 a unit, immediately turning industry customer acquisition costs on their ear (Exhibit 2). In one fell swoop, US carrier domination of device suppliers was over, as AT&T’s competitors all dropped their own proprietary user interfaces and content gardens, scurrying for hot Android smartphones to fill their iPhone voids. Along the way, unlimited data usage clogged airwaves prompting Verizon to lead a charge into faster and far more efficient 4G LTE. Carrier economics suffered accordingly.
Exh 2: AT&T vs. Verizon – Postpaid Churn and ARPU post-iPhone debut
In 2011, AT&T’s exclusive lapsed and Verizon finally got the iPhone. However, the new 4GS model did not support the new LTE standard and the carriers were resigned to another year supporting iPhone users on their already burdened 3G networks. Consequently, aggressive share gains would only add to the network congestion, and rather than invest in more 3G capacity with 4G already in play, Verizon led a charge toward higher prices and usage caps that was echoed by AT&T. This dynamic was the backdrop to the perfect world snapshot that is the US wireless market today.
Second Class Citizens
So where were Sprint and T-Mobile in this picture? Struggling for their lives despite the substantial price umbrella set by their larger rivals. Ordinarily, market leaders with congested networks and high prices might expect a challenge from smaller, hungrier rivals willing to trade margins for market share, but not so much in the US market. It boils down to spectrum (Exhibit 3).
Exh 3: US Wireless Carrier Metrics, Q2 2011
AT&T and Verizon are the heirs to the original two spectrum licenses granted in each market by the FCC in the 850MHz band during the early 1980’s. This band is extremely attractive for terrestrial wireless communications, as lower frequencies travel farther and penetrate solid objects better than higher frequencies, thus enabling each cell to cover more territory and to deliver better signals inside of buildings or behind hills. In contrast, Sprint and T-Mobile’s networks were first established in new spectrum blocks at much higher frequencies sold by the FCC at auction in 1993. At 1.9GHz, these blocks require roughly 80% more cells to offer the same geographic coverage and will not support indoor service nearly as well as a network running at 850MHz (Exhibit 4).
Exh 4: Relative Coverage Advantages of Lower Frequency Spectrum versus Higher Frequencies
This spectrum disadvantage has been an albatross for the second tier carriers from the beginning, leaving them with higher network costs, less coverage and poorer service quality than the market leaders. Thus, Sprint and T-Mobile have been betwixt and between, with costs too high to undercut the leaders by more than a little and with service too poor to compete at price parity. Over the years, the leaders have gained further leverage through economies of scale, and have been aggressive in adding to their spectrum advantages as new blocks have come available. Now, the two top carriers behave as though the smaller carriers don’t matter, largely because they don’t.
Q: What has very high fixed costs, very low marginal costs and is extremely asset intensive?
The answer is – a cyclical business. Telecom, like airlines, semiconductor fabrication, and paper mills, has a long history of cyclical behavior. Demand builds, capacity fills, prices rise, competitors build more capacity, and prices fall. Businesses like this often have a difficult time generating consistent return on invested capital over time, but also often tend toward anti-competitive behavior (Exhibit 5). A quick search on famous price fixing cases brings a list decidedly tilted toward cyclical industries, and several, including telecom and electricity, have operated as tightly regulated monopolies around the world for much of their history (Exhibit 6). With this context, high fixed cost, low marginal cost, asset intensive businesses are generally either consistently profitable or competitive but never both. Even when peace and profitability holds for longer than expected, eventually someone breaks ranks, as the lure of market share gains or the market leader’s advantage grows too great.
Exh 5: Return on Invested Capital Across Cyclical Businesses
Exh 6: Famous Price Fixing Scandals
Exh 7: US LTE Coverage by Carrier, Q3 2012
The iPhone5 has the potential to upset the peace. Apple took its time to support 4G LTE, only just adding the standard more than a year after its Android-driven competitors. Having invested in a spanking new coast-to-coast LTE network covering more than 5 times as many markets as any of its competitors and with capacity to burn, Verizon is in prime position to pick off its rival’s customers, most of whom will still be competing for airtime on the jam packed 3G networks (Exhibit 7).
Moreover, Verizon’s spectrum holdings position it to maintain a considerable network advantage for the foreseeable future. Verizon has 62MHz of clean spectrum – with essentially national coverage and the full support of wireless semiconductor suppliers – devoted to its LTE network. The initial deployment, in 22MHz of extremely attractive 700MHz band spectrum bought at the 2008 FCC auction, offers outstanding coverage from big cells with excellent in building penetration. Future expansion will come in the 40MHz AWS block Verizon recently purchased from a consortium of cable operators. In contrast, AT&T has deployed LTE into 12 to 24MHz blocks of spectrum in the 700MHz band purchased at the same auction (Exhibit 8). AT&T’s licenses do not have the same comprehensive national coverage that Verizon’s do. AT&T plans to augment its primary deployment with collection of additional blocks, including some which do not have natural 2-way pairing and cannot be supported until the next version of LTE. The spectrum hodge-podge makes it more difficult to provide coverage and to source equipment or devices.
Exh 8: Distribution of 700MHz Spectrum Licenses
Under these circumstances, an aggressive competitive play could be irresistible to Verizon, while AT&T may look to pre-empt Verizon with moves that could upset the high-priced, data limited, low capex nirvana in which they currently live.
Barbarians at the Gate
There are other players who have the potential to spoil the party. While the secondary competitors, like Sprint, T-Mobile, MetroPCs, US Cellular and Leap Wireless, lack the spectrum assets and financial resources to compete with Verizon for leadership in LTE, wholesale networks built by independent parties and shared amongst the smaller players could help to narrow the gap. Dish Network holds licenses to a 40MHz block of spectrum adjacent to the AWS band already in use for cellular services and is expected to gain FCC approval before the end of the year to build a wholesale LTE network. Clearwire, after its ill-fated roll out of a network based on the now orphaned WiMax standard once championed by Intel, Motorola and Samsung, is using its huge 150MHz swath of high frequency 2.5GHz spectrum to build its own wholesale LTE network. While Clearwire’s spectrum is substantially disadvantaged relative to the 700MHz blocks used by Verizon and AT&T for coverage and in-building penetration, the sheer size of the block would allow extremely fast speeds, particularly matched with an updated version of the LTE standard that will be available for deployment next year. Done correctly, this network could even compete for residential broadband service against cable modems.
More 700MHz spectrum could also come to auction within the next 2-3 years as the FCC proceeds with its plans to reclaim bands currently used for TV broadcasts through incentive auctions designed to hasten the process by sharing proceeds with the station owners currently holding the licenses. The FCC projects that up to 120MHz in total could be made available through this process (Exhibit 9). The FCC is also expected to issue new rules that will limit the amount of spectrum that could be controlled by a single company, weighted to the attractiveness of the various spectrum bands. Verizon and AT&T, already using attractive spectrum in the 700MHz and 850MHz bands, could find themselves restricted in the incentive auctions as a result.
Exhibit 9: National Broadband Plan Proposed Mobile/Broadband Spectrum
The biggest wildcard could be the participation of the cash rich internet community. Google, Apple, Microsoft, and Amazon have obvious interest in thwarting a regime of high wireless prices and data usage caps. Google is a particularly intriguing potential player, given that it dropped out in the late stages of bidding during the last 700MHz auction and that it recently invested $400M to build an experimental residential fiber network in Kansas City. A Google funded LTE network in 700MHz spectrum to be shared by the secondary carriers in the market could be a wicked foil to the Verizon-AT&T duopoly.
Back to those Dividends…
AT&T is currently paying out 70% of the cash it generates from operations after capital spending toward shareholder dividends. Verizon is less generous, paying out 35% of cash after capex (Exhibit 10). Cash generation is highly sensitive to customer ARPUs, acquisition and retention costs, and capital spending needs, all of which could be unfavorably affected by renewed rivalry. While wireless dominance has allowed Verizon to sustain double digit cash flow after capex as a percentage of its sales for the decade since the 2001 recession, AT&T dipped into single digits in the two years prior to the iPhone launch. Looking further back, neither company was able to sustain double digits during the highly competitive 1990’s (Exhibit 11). Moreover, a more competitive wireless market moving forward would undoubtedly have negative connotations for sales and earnings growth as well.
Exh 10: AT&T and Verizon – Dividends as % of Free Cash Flow
Exh 10: AT&T and Verizon – Cash Generation as % of Sales
On the flip side, any breakdown in the co-operation between the two leading wireless carriers that resulted in lower prices, faster speeds and less restrictive data usage plans would be unambiguously good for the Internet platform players – Google, Apple, Microsoft and Amazon. It would also be good news for smaller players like Metro PCS, Leap Wireless and US Cellular who could gain national coverage and improved cost competitiveness if a wholesale network becomes a reality.
Exh 11: Winners and Losers