Summit Ridge Group - Valuation & Financial Consutling for Media Telecom and Wireless
Telecom, Media and Satellite
Valuation & Financial Consulting Practice

Downstream Implications of AWS-3 Auction

Posted on: December 24th, 2014 by Armand Musey

While there has been extensive discussion about the AWS-3 auctions, there has been less discussion about the long-term industry impacts of the auctions:

1)    Upcoming 600 MHz Auction Prices – Higher transaction prices from the AWS-3 auction could imply higher prices for the 600 MHz auction. However:

  • With the telecom industry spending ~$45B on the AWS-3 auction, the industry may need to replenish its financial reserves before incurring similar amounts or more on the 600 MHz auction. For comparison, the CTIA reports that 2013 CAPEX totaled only $33B and has averaged $26B annually over the past 10 years.
  • The two carriers with the deepest pockets, AT&T and Verizon, already possess deep spectrum holdings below 1 GHz. Therefore, substantial bidding by them would likely be based on a desire to foreclose future entrants as opposed to based on their auction use needs.

2)    Industry Returns –The US wireless industry is unlikely to earn an acceptable return on its AWS-3 investment.

  • The auction alone represents an investment of approximately $140 for every US subscriber connection! Furthermore, this expenditure excludes any build out necessary to actually use the spectrum.
  • The industry is unlikely to increase prices by the few dollars per month it would take to justify the spectrum purchases and associated buildout. Even before the auction, industry pricing and returns have been falling as the market reaches saturation, so the necessary pricing increase is unlikely

3)    DISH Dynamics – A widely touted reason for the high AWS-3 auction prices is involvement by DISH Network-backed entities in order to increase market comparables for their AWS-4 spectrum. However:

  • DISH may actually want to buy AWS-3 spectrum, perhaps the adjacent J-block spectrum, to increase its capacity and to force T-Mobile or Verizon to the table for a future alliance.
  • DISH is facing a 40% build out requirement (by population) by 2017 and 70% by 2021. If it misses the first requirement deadline, the second requirement deadline will also advance a year 2020.  Given the enormous engineering challenges involved, time is running out for DISH. DISH is not likely to enter the wireless market by itself as a new entrant nor has it has not been successful in finding a partner. Buying spectrum that T-Mobile and Verizon need, even if at inflated prices, may force them to the table and help DISH monetize its AWS-4 spectrum.

Ultimately, the reasons for the inflated AWS-3 auction prices while near-concurrent non-auction spectrum transactions at much lower prices, remain unclear. We expect release of auction results next month to shed additional light.

New Draft Paper on Spectrum Demand Forecast Error

Posted on: August 18th, 2014 by Armand Musey

Aalok Mehta (PhD candidate at USC ) and I just finished a draft of a paper we are scheduled to present at the Telecom Policy Research Conference (TPRC) next month. The paper, “Overestimating Wireless Demand: Policy and Investment Implications of Upward Bias in Mobile Data Forecasts,” is about forecast error in demand estimates for spectrum.

The abstract is below, and a full copy is available at –


In this paper, we present evidence of persistent errors in projections of wireless demand and examine the implications for wireless policy and investment. Mobile demand projections are relied upon in academic and government research and used for critically important telecommunications policy decisions, both domestically and internationally. The Federal Communications Commission, for example, used such projections to estimate a 275 MHz spectrum shortage by 2014 and featured such estimates in the U.S. National Broadband Plan as evidence for allocating additional spectrum for cellular services. The International Telecommunications Union Radiocommunication Sector endorsed in 2006 an estimate of a 1,280- to 1,720-MHz spectrum shortage by 2020. In hindsight, both estimates clearly overstated demand; however, they remain the basis for policy direction, and their underlying accuracy has not been evaluated in a systematic manner. The potential for persistent bias in these projections may allow policy errors based on these data to compound over time as opposed to self-correct. Biased industry projections also cause problems for private investment. Investors rely on industry projections as a significant input in their investment decision-making process. An upward bias in wireless projections can result in overallocation of capital in licensed wireless systems versus other industries where, on a marginal basis, capital might be more productive. It may also cause investors to overpay for wireless assets. As this overallocation becomes apparent, prices typically drop and the industry must downsize or correct. This creates unnecessary volatility and may drive away future investments.

To investigate upward biases in wireless projections, we examine major historical projections of wireless demand—including those from Cisco, Coda Research, Yankee Group, and the ITU—and compare them to updated projections over time and actual results. Assuming such projections are neutral, the number of projections that overestimate demand should be roughly equal to those that underestimate demand. Likewise, we would expect that, as projections are updated over time, errors on the upward side should roughly equal those on the downward side. However, the evidence reveals a persistent tendency to overestimate in both number and value. Of the past seven Cisco mobile traffic forecasts for North America, for example, overestimates were nearly twice as frequent as underestimates (19 vs. 10). Overestimates are also on average of greater magnitude than underestimates (103 vs. 81 PB/month). Moreover, the longer the timeframe, the more likely the estimates are likely to be over-projections due to exponential growth issues. Many of these errors result from failures to anticipate or correct accurately for new business practices, such as mobile offloading or tiered data plans.

Wireless policy decisions have long timeframes; more than half of recent U.S. spectrum reallocations have taken a decade or longer, and incumbent users often possess significant technical and political power to stall or limit spectrum reorganizations. Persistent upward biases in mobile demand projections thus have significant, long-term policy implications, including possibly overweighting spectrum policy towards expensive clearings for new licensed bands over more flexible shared or unlicensed management systems. Such biases may also lead to businesses overvaluing spectrum licenses and to governments overestimating the contributions of exclusive spectrum allocations to deficit reduction and economic growth.

Our findings suggest the mobile industry contains much higher levels of inherent demand uncertainty than is commonly estimated and that business and governments may not be fully factoring it into their policy decisions. To reduce dependence on uncertain estimates, government officials should consider assigning spectrum allocations with greater flexibility of use.  Additionally, we recommend that policymakers use demand projections only when they are transparent, have authorial accountability, and comply with processes to reduce conflicts of interest.

U-verse is the Regulatory Key to the Proposed AT&T/DIRECTV Merger

Posted on: June 19th, 2014 by Armand Musey

AT&T ‘s proposed $48 billion acquisition of DIRECTV is based on two simple business facts – 1) AT&T has broadband but needs a stronger video product to compete with cable; and 2) DIRERCTV needs a voice and broadband product to compete with cable. By combining services the can save money on customer service, back office and an estimated $1.6 billion in annual programming costs. On the surface, it seems like a merger that will create a stronger competitor to cable systems everywhere. So what’s the regulatory concern?

The first issue is that AT&T has a video offering with its U-Verse service. U-verse is a state of the art high-end fiber to the node (FTTN) service that also offers video, much like Verizon’s FiOS. Thus in markets where U-verse operates, video competition will be reduced. But this is not an insurmountable hurdle.

AT&T/DIRECTV will argue that consumers will still have two other options in those markets – DISH and the local cable company, not to mention OTT services such as Netflix, Hulu and Amazon Prime. Moreover, mobile broadband speeds are approaching the point where they will be able to offer a competitive video service in the foreseeable future.  In necessary, AT&T could also a third party to market U-verse to ensure it competes with DIRECTV – much in the way Time Warner allowed Road Runner to market its internet service after the AOL merger.

The second issue is that content providers will likely object on the grounds the combined entity will have disproportionate buying power. But we don’t believe this will be a significant regulatory issue. The buying power of the combine entities is roughly 26 million homes. This is well below the 30% informal market share screen used for identifying disproportionate media buying power. And it’s certainly well below the buying power of a combined Comcast/Time Warner merger.

The third and largest hurdle to FCC approval is the FCC’s analysis of merger’s medium to longer-term impact on broadband deployment in the US.  And this issue is far more complicated.  This is because building out U-verse is very expensive. The current U-verse business model is dependent on a combination of revenue from voice, video and data subscriptions. In many areas AT&T cannot justify it even with these three revenue streams that total an average approximately $106 per month. If AT&T acquires DIRECTV, they would be able to get video subscribers from DIRECTV, a much more economical system for video broadcasting.  U-verse would lose the video revenue stream and it would be harder to justify additional U-verse build-out. Moreover, any additional build-out might be done cheaper and with less capacity as it would not need to offer speeds sufficient to support a robust video offering. AT&T’s ability to save on these U-verse capital expenditures is undoubtedly one of the primary benefits of the merger. But it’s also one of the potential problems from the FCC’s perspective.

The FCC’s mission, to a large part, is to facilitate the deployment and adoption of advanced communications services. But if AT&T acquires DIRECTV and is incentivized to slow U-verse build-out, that goal is undermined.  And it is significant. U-verse uses the most advanced communication technology and is run by one of the country’s largest telecom behemoths. This is at a time when the FCC is already on the defensive due to a perception that landline Internet speeds in the US are slower than other countries and putting the country and an economic disadvantage. Moreover, if the FCC approves an AT&T/DIRECTV merger, they will be hard pressed to deny a merger between DISH and Verizon, potentially creating an opportunity for them to reduce investment in their FiOS system. A slowdown of U-verse deployment, not to mention a slowdown of FiOS deployment, would put a massive dent in the country’s broadband growth. Is this what the FCC Commissioners want to see happening on their watch?

For the AT&T/DIRECTV merger to succeed, it likely will need to offer significant evidence that it will not slow its investment in new broadband deployment, particularly U-verse.  AT&T is emphasizing its potential savings on programming savings when, the regulators’ eyes are going to be on the potential CapX savings that could slow the country’s broadband deployment.


FCC Auctions May Disappoint – Supply of Money Less than Supply of Spectrum

Posted on: April 22nd, 2014 by Armand Musey

There’s a legitimate question as to whether the wireless industry has enough money to pay historical prices for spectrum the FCC intends to auction over the next 18 months. The FCC intends to auction approximately 150 MHz of spectrum between the AWS-3 and the Broadcast Incentive auction (assuming 65 MHz at AWS-3 and 85 MHz in the Broadcast Incentive auction) by the end of 2015.  This approximately 47 billion MHz/PoPs based a US population of 310 million. At historical prices of approximately $1.00 per MHz/PoP for AWS-3 spectrum and $1.50 per MHz/PoP for 700 MHz spectrum ($1.28 from 2008 auction adjusted for inflation), the AWS-3 auction would raise $20.15 billion and the broadcast incentive auction would raise $39.525 billion, for a total of $59.657 billion, which we round to $60 billion.  Unfortunately, its unclear the wireless industry has, in aggregate, this much money to spend.

Concentrated Industry

Total wireless industry cash flow is approximately $35-$40 billion. Moreover, it is highly concentrated. AT&T and Verizon have 64% of the industry subscribers and over 80% of the profits. Sprint and T-Mobile have approximately 18% and 12% marketshare. After AT&T and Verizon, industry cash flows fall rapidly. Both Sprint and T-Mobile are cash-flow negative after accounting for debt service. There are approximately 100 smaller carriers. And since most are privately owned, their financials are unavailable. But, it is unlikely the total cash flows of the smaller carriers total more than $1 billion annually.

Currently, AT&T and Verizon control 77% of the wireless broadband spectrum below 1 GHz and approximately 50% of the spectrum between 1 GHz and 2.5 GHz.  For AT&T and Verizon to get 60% of the spectrum auctioned and T-Mobile and Sprint to get 20%, the Big-4 will need to spend $48 billion assuming historical pricing discussed above.  AT&T and Verizon would need to spend roughly $18 billion each (assuming both win 30%) while Sprint and T-Mobile would need to spend $6 billion (assuming both win 10%).  This would leave the smaller carriers to spend $12 billion for the 20% remaining spectrum, which would be virtually impossible given their size. This leads to the potential that smaller carriers get even less, and that industry concentration increases.

Rural Focus Doesn’t Help Smaller Carriers Enough

However the reality is a bit more complex.  The Big-4 will largely be bidding for spectrum in the largest markets where they need it the most (and where in the Incentive Auction, supply is likely to be less).  Conversely, the smaller carriers are likely to bid in smaller markets, where pricing is likely to be lower.  We’ve noticed a trend towards greater disparity in spectrum pricing between large and small markets. We believe it is reasonable to assume they will only need to pay 50% of average national prices, or $6 billion to win 20% of the auction at historical prices. This leaves the Big-4 needing to pay $54 billion for 80% of the spectrum to maintain historical prices. We believe both of these are very tall orders.

Investors are Likely to Lower Participation

Historically, investors (aka, spectrum speculators) participated in spectrum auctions by bidding for underpriced spectrum with the hopes of selling to a carrier at a higher price when they were ready to use it. But given the increased concentration of the wireless industry since the last major auction in 2008, investor participation is likely to be lower. Simply put, there are fewer carrier for them to sell to, making their holding spectrum a more risky proposition. Of course, some investors are likely to participate in both the AWS-3 and the Broadcast Incentive auctions, but probably at lower levels than in the 2008 700 MHz auction (Auction 73).

AT&T Threat to Limit Participation in FCC Auctions Must be Taken Seriously

AT&T recently suggested it might reevaluate its participation in FCC spectrum auctions if certain conditions were not met. While many industry observers dismissed this as merely a negotiation tactic – and it may very well be – it must be seriously considered. AT&T is not only talking about not participating aggressively in the spectrum auction, but it is also acting that way. In spite of the run-up to a very expensive auction, AT&T has continued to freely spend cash on its generous dividend (over 50% payout ratio) and its stock buy-back program (despite a cash balance of only $3.44 billion), which seems incongruous with the notion that it is getting ready to spend $15-20 billion on an FCC auction.  Without aggressive participation by both AT&T and Verizon, the auction prices will almost certainly fall well short of expectations.

Spectrum Price Increases have Tapered

Additionally, the rapid rise of Wi-Fi offloading, now accounting for nearly 50% of wireless traffic and the proliferation of small cell sites has put a damper on spectrum prices. Recent sales have shown only mid-single digit increases.  Verizon’s sale of 700 MHz A-block spectrum to T-Mobile for $1.85 per MHz/PoP is particularly striking. This represents a 16.3% increase (2.5% annual rate) from the $1.59 per MHz/PoP paid for the same spectrum in Auction 73 in 2008. However, when it was auctioned, the spectrum was facing significant impairments. There was no prospect of alleviating the channel 51 interference issues and handset interoperability appeared to be a long-term challenge. Since then, the handset interoperability issue has been resolved via industry agreement and the broadcast incentive auction is likely to resolve the channel 51 interference issues.  Even with these “upgrades” to the spectrum, it still sold for single digit returns.

Credit Markets are Open – for Now

The credit markets are open for risks loans and interest rates are near historical lows.  This may improve the ability of carriers to borrow to finance their FCC auction participation. However, credit markets can close quickly and there is no guarantee they will remain open to support bidders for the FCC spectrum auctions. 

Declining Profitability of Wireless Industry Pressures Prices

There are two main reasons a wireless company bids for spectrum. The first is that they need the spectrum to serve their customers. In this case, they will pay as much as needed to avoid the cost of cell-splitting, up to the amount their customers will pay for the marginal capacity. The second reason carrier bid for spectrum is to keep it out of the hands of their competitors so they can maintain higher pricing. The largest carriers seem to have focused on the latter. Verizon is currently running its entire LTE network, a network carrying approximately 20% of the nations wireless traffic, on a single 22 MHz block on the 700 MHz band.  Additionally, Verizon’s customers are disproportionately focused in high-density urban areas. This would imply the entire country could operate on about 100 MHz of the 618 MHz currently allocated for mobile broadband. Yet Verizon appears eager to bid for more spectrum. Of course future services will require additional spectrum. But the US “spectrum crunch” appears to be more an issue of allocation than absolute spectrum availability.

But hoarding spectrum from competitors makes less sense when reliable Wi-Fi is available and small cells are increasingly affordable. We’re already seeing carrier pricing fall in recent months. Should this trend continue, there will be less motivation to pay higher prices for incremental spectrum.

New Entrant Could Change Dynamics

The biggest factor that could help maintain pricing would be the arrival of a new entrant. Given the maturity of the U.S. wireless market, it is unlikely a new entrant will arrive unless they bring significant synergies. The most obvious of these is DISH Network/EchoStar. They already have 40 MHz of AWS-4 spectrum and 5 MHz of 700 MHz spectrum. Additionally, they have 14 million customers they are already billing and serving, relationships with consumer electronic distribution channels and a strategic need for a data service to stem their subscriber losses.  But they have appeared hesitant to enter and are likely waiting for the right partner.  This could be a company such as Google, but it could also be an existing wireless carrier such as T-mobile. If the latter, that carrier’s participation in the FCC auctions might be tempered as DISH/EchoStar would significantly augment their spectrum holdings. But if it is a company such as Google, teamed with DISH/EchoStar they could bid aggressively to shore-up spectrum holdings and support historical spectrum prices.

FCC and Industry Options

If the AWS-3 auctions don’t go well, the FCC will likely delay the Broadcast Incentive auction and wait until the industry has additional demand for the incremental spectrum.  The FCC has until 2022 to complete the Broadcast Incentive auction. The alternative is to complete the auction at lower prices. While it is good to put spectrum in the market, if pricing is too low, carriers may be forced to write down the value of their current spectrum holdings on their financial statements. This could have the effect of causing them to blow bank covenants and/or making additional financing harder. The prospect of sharply reduced auction pricing is likely to cause the carriers to pressure the FCC to delay auctions or reduce their size. The other option would be for the FCC to allow the largest carriers to control an even larger amount of spectrum.  If this occurs in the Broadcast Incentive auction, it would cement AT&T and Verizon’s positions dominating critical Sub-1GHz spectrum. The FCC clearly has some tough decisions ahead.

The Road to Approval for a Sprint/T-Mobile Merger May Run Through DISH Network

Posted on: March 17th, 2014 by Armand Musey

It’s no secret that Sprint (S) and T-Mobile (TMUS) have been exploring a potential merger and that there are significant concerns over whether such a merger would be approved. Sprint and T-Mobile argue that the US wireless market is a near-duopoly dominated by ATT (T) and Verizon (VZ) with a combined market share of approximately 61%. Sprint and T-Mobile have approximately 18% and 12% market share respectively and the remaining 9% of the market is split between approximately 100 smaller carriers. The argument is that by combining, Sprint/T-Mobile would have economies of scale needed to more effectively compete with ATT and Verizon. They argue the significant synergies ($1.5 to $2.0 billion annually) would allow them to offer higher speed data and other advanced services and/or reduce prices in ways they otherwise couldn’t afford.

Many market legal observers are understandably skeptical the FCC will accept this argument (Note: communications industry mergers require both DoJ and FCC approval, but FCC approval is general considered harder. This because the FCC has a more nuanced “public interest” standard for approving a merger). In particular, a merger that involved a reduction from four major participants to three is often hard for regulators to accept. This is particularly under Democratic administrations and for merger involving consumer services.

DISH to the Rescue?

DISH may offer the solution assuming FCC would like to see more competition to ATT and Verizon, but is concerned about a reduction to three major competitors in the wireless industry. DISH, or someone working with them, is probably the only viable new entrant in the mature US wireless industry. DISH has 14 million existing customers they are billing. They already have relationships with critical retail distribution channels and hardware manufacturers. And DISH has approximately 55 MHz of spectrum nationally. Moreover, the DBS industry is seeing its subscriber base shrink because it lacks a broadband option that a wireless service might offer.

It’s also clear the FCC would like to see DISH as a new entrant to the wireless industry. The FCC has bent over backwards to give DISH terrestrial use of the AWS-4 spectrum at no additional cost, and then an option to convert all of it to downlink if they bid for the H-Block spectrum (an auction that more resembles a negotiated bid than an actual auction).

Despite these advantages, DISH appears hesitant to enter the highly competitive wireless market. But what if Sprint/T-Mobile threw DISH a lifeline to enter the market in order to get approval for their own merger? Sprint/T-Mobile could offer DISH preferred roaming rates on their networks until DISH completed its network rollout, sharing of other facilities for a similar period, ensure that phones built for Sprint/T-Mobile also worked on DISH’s frequencies and a host of other benefits to facilitate DISH as a rapid new entrant. The prospect of DISH as a strong rapid new entrant to maintain four major players just might be enough to convince the FCC to allow the Sprint/T-Mobile merger.

The obvious question is whether supporting a new entrant is too much for Sprint/ T-Mobile to offer to get a merger approved? With synergies of roughly $1.5 to $2.0 billion annually, they would be incented to make some significant concessions. Also, if DISH decides to enter later without these advantages, perhaps motivated by a desire to offer broadband to stem subscriber losses, they may need to enter aggressively with very low prices. Thus, from a Sprint/T-Mobile perspective, it might be better to support DISH’s entry in an orderly fashion sooner, rather than risking a highly aggressive “scorch the earth” entry a year or two later.

Intelsat’s Yearend Results Highlight Dependence on Next Generation Epic Satellites

Posted on: February 24th, 2014 by Armand Musey No Comments

DISCLAIMER: The analysis below is for information purposes only. We believe, but cannot guarantee that the information contained in it is accurate. Moreover, it should not be considered a recommendation to buy or sell any security of any kind. Please consult a qualified advisor before making significant investment or business decisions. See our Terms of Service for additional information.

Intelsat went public in April 2013, partially based on the expectation on page 2 of their F-1 IPO fling that after several years of slow growth Intelsat was “well positioned to experience growth in free cash flow in the near future.” One of the four factors that would drive this growth was “Our [Intelsat’s] $3.7 billion fleet investment program that began in 2008 was substantially complete by the end of 2012, enhancing our future revenue potential.” Since it’s IPO, management of this heavily leveraged company has deftly exploited the strong bond markets to lower debt service and further reduce operating expenses. It would appear Intelsat is delivering on its promise to pay-down debt and increase the value of its equity. Intelsat’s Q4 results, while in-line with guidance, put the final nail in the coffin of any hope that its pre IPO expectations of near-term free cash flow growth will materialize.

Investors have been skeptical of Intelsat – its stock is up only 7.27% from its April 2013 IPO, compared to 34.65% for the NASDAQ and 19.11% for the S&P500. Intelsat’s Q4 results not only dashed hope for near-term growth, but they unveiled new concerns about revenue and backlog decline. This makes the company long-term success increasingly dependent on the success of the upcoming Epic satellites to offset the declines and allow them to service their debt over the long run.

Intelsat’s revenue was slightly down year over year- $2,604 million vs. $2,610 million for 2012. Granted that US sequestration process has taken its toll, as has pricing pressure in Africa. Guidance for 2014 was even lower – $2,450 to $2,500 million. At the midpoint, that is an additional 5.7% decline in revenue. According to its IPO prospectus, Intelsat typically has 82% of its revenue in backlog at the start of the year, so it only needs to sell 18% to make its year. To have a nearly 6% decline in revenue when 82% of revenue is in backlog, suggests a much larger projected decline in sales activity during 2014 – perhaps as much as 30%. Obviously, if sales productivity falls 30% revenue will eventually fall 30% as the company works through its backlog. Moreover, management made it clear that it lacked growth potential with it current satellite fleet – the same fleet that less than a year ago was touted as an engine of growth! Instead, investors will need to wait until 2016 for the new Epic satellites for any hope of growth. The question is whether the company can wait that long and what happens if the promised growth from Epic does not materialize.

A 30% revenue decline would be problematic for Intelsat given its debt load. At $2,604 million in revenue Intelsat generated $2,032 million in EBITDA.  Assuming a similar 78% EBITDA margin, at the midpoint of revenue guidance, that will be $1,915 in EBITDA in 2014. According to management, the company has approximately $950 million in debt service, $525 million in capx and about $75 to $100 million in other expenses. This leaves approximately $450 million in cash flow for debt service and other expenses for 2013 and $339 for 2014. This is consistent with management guidance that the company intends to pay down about $400 million of its $15.3 billion in debt during 2014, but some of that will come through reductions in its current $250 million cash balance.

Intelsat’s cash flow seems comfortable on the surface, but we note that it has 43 satellites. Assuming a 15-year average life, they need to launch approximately three satellites a year to replace each of these revenue-generating assets. This costs far more than $525 million, most likely around $900 million (Note: depreciation was $736 million in 2013, but satellite costs have increased, moreover capx guidance over the next three years is considerably higher, ranging form $575M to $850M per year), putting Intelsat dangerously close to cash flow break-even. Moreover, we understand that Intelsat does not carry in-orbit insurance for it satellites. Industry history suggests geostationary satellites fail at a rate of approximately 1% a year. Intelsat generates, on average, about $60 million in revenue a year from these satellites that require approximately three to four years to build and launch. Thus one or two satellite failures could cause problems for Intelsat’s ability to service its debt. Additionally, as the Epic satellites launch, Intelsat will begin recognizing its prepaid deferred revenue, which currently stands at $888 million, but based on management guidance is likely to total between $1.0 and $1.1 billion by the end of 2015. This revenue, which management suggests is amortized over the life of the satellite, or about $70 million per year, will not generate any additional cash, lowering its cash EBITDA by that amount absent an increase in revenue. Moreover, those customers who prepaid year in advance, likely received large volumes of capacity at deep discounts. This could, “flood the market” to an extent and make it difficult for Intelsat to sell at full price. As a result of these risks, Intelsat is heavily dependent on its bet that the Epic system will allow them to lower their long-term capx and increase their revenue.

Intelsat’s problems are compounded by declining backlog. Yearend backlog is down from $10.7 billion at the end of 2012 to $10.1 billion at the end of 2013. Backlog has been trending down quarter over quarter during the past four quarters. The curious issue is that the company was able to come very close to 2012 revenue despite the backlog decline. This suggests the company was able to get to its 2013 revenue via sales contracts of shorter duration. In fact, the decline in backlog suggest the total value of sales made in 2013 was $600 million lower than their 2013 revenue, a decline of about 25%. However, management was clear that it was the short-term government business that took the greatest hit in 2013. So why did backlog decline? We can only surmise that the company signed a disproportionate amount of shorter-term network services business as opposed to longer-term media contracts with large media organizations. Network business is generally less attractive due to both its shorter duration and the fact that network service customers are typically smaller and are higher credit risks. This may also explain the spike in bad credit, from $8 million $30 million. The additional problem with declining backlog is that it lowers the revenue visibility of the company. It’s this visibility, along with strong credit markets, that have allowed Intelsat to borrow at attractive rates. A continued backing decline threatens its ability to borrow at low rates as debt becomes due over the next several years.

Competition is also increasing. O3B recently launched service, and other high throughput satellite (“HTS”) will be following closely in 2014 and 2015. These HTS satellites have several times the capacity of current satellite. They will likely drive pricing down for point-to-point network applications that constitute approximately 60% of Intelsat’s revenue. But Intelsat will likely benefit from increased revenue on its next generation Epic satellites that should offset this loss and may even increase Intelsat’s revenue. However, new entrants, including ABS and NewSat are also launching HTS satellites. They are new operators, and unlike Intelsat, can reduce pricing without risking revenue on their legacy lower through-put satellites. At the same time, the march of fiber deployment is continuing in Africa. Demand and margin erosion is likely to continue and even accelerate. The net impact of the above issues is that changes in market dynamics have largely offset Intelsat’s progress on cost-cutting and debt refinancing. It’s still not out of the woods yet.

Five Reasons the Comcast/Time Warner Cable Anti Trust Concerns are Overbown

Posted on: February 17th, 2014 by Armand Musey No Comments

There has been a lot of market speculation that the Comcast’s proposed acquisition of Time Warner Cable might be blocked on anti-trust concerns. We believe these concerns are overblown.

1) The Cable Operations Don’t Compete in the Same Markets, so Industry Concentration Doesn’t Change.

The combined company would serve over 30 million homes, which initially seems to significantly consolidate the industry. But the DoJ must analyze the market from the perspective of where they are sold – the customer. There are only a few areas where customer currently has a choice between Comcast and Time Warner – and in those few areas they may have to divest operations or negotiate an arrangement that provides a similar result. The company has already indicated it intends to sell territories accounting for approximately 3 million homes and these are undoubtedly on the list of territories to be sold. So, generally speaking, cable industry concentration will be the same since almost no one will have fewer choices for video as a result of a merger. This is completely the opposite situation that would occur if DIRECTV and DISH Network were to merge. In this case, most consumers would be reduced from three choices (cable and two satellite companies) to two (Cable and one satellite company). Of course, the growth of Verizon’s FiOS and ATT’s U-verse and other networks is slowly eroding even that argument.

2) Vertical Integrations Issues Manageable

The other anti-trust issue relates to vertical integration. Comcast owns NBC and some of its broadcasting stations. This puts the combine company in a stronger position to hurt DIRECTV or DISH Network by withholding NBC content from them. Of course, Comcast could have done this before – and this was a concern about the NCB/Comcast transaction. But it seems to have been resolved and there have not been serious complains that Comcast/NBC was using their content leverage unfairly against DIRECTV and DISH Network. Also, regulators generally are less worried about abuse of market power due to vertical integration than they are about horizontal integration as the negative impact on end-users is less clear. It’s hard to see why the DoJ would suddenly assume the vertical integration issues would now be larger simply because Comcast is larger. Any vertical integration concerns could be solved by clearer requirements that the combined company sell any of their proprietary content to DIRECTV and DISH Network (and other competitors) on terms similar to those they supply the other cable companies they don’t compete against.

3) Leverage Over Content Providers Helps Consumers

Another issue raised is that the combined company will pressure content providers to supply better terms. But this would be considered a consumer positive and a reason for regulators to support the deal. It would give the combined company the ability to lower prices or offer additional services for the same price to compete more aggressively against satellite and Over the Top (OTT) Internet video service such as Netflix and Hulu. Of course there is no guarantee they will do this, but there is no reason to believe that a having a cost advantage would allow them to raise prices or reduce services. Far from being a reason to deny the merger, this is a strong supporting argument.

4) Access to New Content and Programming Diversity Concerns not a Major Barrior

There is a potential public policy concern that such a large company handing media distribution would have inordinate influence in determining which media content can survive. It’s possible that some content would be unviable without carriage by a distributor that controls 30% of the nation’s eyeballs. This type of market power due to vertical integration is experienced in other industries. And it’s precisely why the company would seek to sell about 3 million viewers – to keep under the 30% threshold regulators often us when looking at abuse market power with respect to suppliers in vertical integration situations. Unless the DoJ or the FCC decides that the media industry is, for some reason, dramatically different than other industries, this should not be a problem. Meeting the 30% guideline should serve to handle the regulator’s concerns. Finally, Comcast and Time Warner don’t control all of their content. They are required to carry the content of local television stations via “must carry” rules.  So content providers seeking carriage on Comcast or Time Warner cable systems can also bypass Comcast and Time Warner and get that carriage via carriage on television broadcast stations.

5) Net Neutrality and Other Irrelevant Issues

The FCC is clearly disappointed in the recent DC Court of Appeals decisions striking down Net Neutrality (Verizon v. FCC). Among other issues, is the fear the cable companies will be able to reduce video competition by restricting OTT video provides including Hulu and Netflix. This merger does not change the impact of that decision. However, the proposed Comcast/TWC merger gives the FCC the ability to condition the merger on them agreeing to net neutrality rules. The FCC can thus guarantee 30% of the country will have the net neutrality that they are not allowed to impose via a rulemaking. This opportunity may be the strongest reason the FCC has for supporting the merger.

Other opposition to the merger, about having so many people in the country dependent on one provider, security concerns and the unpopularity of the industry among consumers are political issues and not rooted in traditional anti-trust law. Of course the FCC uses a more subjective “public interest” test than the DoJ, so these political issues could still come into play. But based on an basic traditional anti trust analysis, the deal looks in good shape.

Wireless Demand Projections – Elasticity of the Top-1% Percent Users

Posted on: January 26th, 2014 by Armand Musey No Comments

Wireless network traffic growth projections have been significantly cut since the FCC’s 2010 National Broadband Plan. The National Broadband Plan projected wireless network demand in 2010 to be 468 billion MB, vs the 398 billion MB actually carried in 2010 according to the CTIA. In 2011, the spread got wider, 1090 billion MB projected vs 890 billion MB actually carried. And in 2012, the FCC projected 2219 billion MB was even further from the actual traffic of 1468 billion MB. Extrapolating these actual numbers to 2015 suggest the current allocation of 608 MHz of spectrum for mobile broadband will be close to actual demand levels in major metro areas. Of course traffic will continue to grown past 2015, but the “crisis” does not appear to be nearly of the proportions projected a few years ago.

Current network growth projections, combined with increases in Wi-Fi offloading and small cell deployment call into question the National Broadband Plan’s central conclusion that 300 additional MHz of mobile broadband spectrum would be needed to meet demand by 2015. As we approach 2015, it’s become obvious the spectrum crunch is not as severe as once projected. Large swaths of mobile spectrum remain undeveloped – including prime sub-1 GHz spectrum in the 700 MHz band auctioned over six years ago.

The difference between the lower wireless network traffic today and earlier higher projections largely comes from two primary: 1) Increased Wi-Fi offloading from 15% in 2009 to over 40% today; and 2) the phasing out of unlimited data plans in the US. While the impact of Wi-Fi offloading is straightforward, the impact of the phasing out of unlimited data plans is less obvious.

Cisco’s  “Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, 2012–2017” finds that from 2010 to 2012 unlimited data plans fell from 81% of data plans to 45%. During this time, the usage of the top 1% of users fell from 52% in January 2010 to 24% in September 2011 to 16% in September 2012. This drop alone can account for the most of the shortfall in wireless traffic growth versus earlier projections.

There are two likely possibilities in terms of wireless traffic growth: The first is that wireless traffic demand of high volume users is high elastic. As users are required to pay for incremental usage, they quickly limit their use. When incremental usage is essentially free, they find ways to use it with traffic that is not important. The second is that carriers had a small number of high usage customer they were able to cull and this reduced their traffic growth rate over the past few years. If the latter is true it suggests the slower growth over the past few years was a one-time event and that growth will accelerate again soon.

But wireless traffic may grow in any case. With threats of an imminent spectrum crunch behind them, carriers appear to be returning to unlimited data plans. Sprint recently announced a new service, Spark, that advertising download speeds of 50-60 Mbs and potentially increasing to 150 Mbs in the future. Presumably this will come from developing its EBS/BRS spectrum, spectrum that has been largely fallow. At the same time, Carriers have been finding enough spectral efficiencies with LTE technology to resume offering unlimited data plans. Presumably if demand starts to grow too rapidly, they will begin to scale back some of these unlimited data packages.

Meanwhile, spectrum prices in the secondary market appear to have stabilized. And no major wireless operator is even participating the H-block auctions underway or bidding for the LightSquared assets in the current bankruptcy auction.  This further suggests the need for additional spectrum is not approaching the crisis levels forecast in 2010.

Lower than projected wireless demand calls into question the need for the planned FCC spectrum auctions over the next few years and pricing the FCC may get if the auctions go ahead as scheduled.

Three Downstream Implications of a Sprint/T-Mobile Combination

Posted on: December 17th, 2013 by Armand Musey No Comments

While there has been significant industry discussion about a possible Sprint/T-Mobile combination, there has been little discussion about potential implications for other areas of the industry. Here are three:

1) Broadcast incentive auction (and other upcoming FCC spectrum auctions) could suffer. T-Mobile, having essentially no spectrum below 1 GHz was expected to bid aggressively in the broadcast incentive auction. Sprint has approximately 13 MHz of SMR spectrum, so it is not quite as constrained with respect to lower frequency spectrum. If the Sprint and T-Mobile are tied-up trying to get approval for a merger or the early stages of integrating a merger, its quite possible that they won’t expend significant resources on the incentive auction or other auctions. We’re already seeing anemic interest in the H-block auction with both Sprint and T-Mobile declining to bid.

2) Tower operators could suffer. As the number of major bidders for tower space declines, the bidding leverage moves to the wireless companies and away from the tower operators. Sprint and T-Mobile both have significant amounts of PCS spectrum (est. 41 and 30 MHz respectively) – many of their antennas may be consolidated over time. We’ve previously discussed that the tower operators have not been participating proportionately in the increase in wireless traffic since so much of that growth is going to Wi-Fi and small cell sites where tower operators usually don’t participate. This trend shows little sign of slowing.

3) DISH may need to act decisively and bid for T-Mobile themselves. It’s becoming increasingly clear DISH needs to enter the wireless business to offset a maturing (declining) core television business. But it’s unlikely DISH wants to enter without a major industry partner. With fewer major partners to negotiate with, the bidding dynamics turn against DISH and the increased consolidation will likely make it even harder for a new entrant to enter alone. This may force DISH to make a serious bid for T-mobile to essentially “buy a partner.”

While the US wireless market is relatively mature, it certainly remains dynamic.

“The Spectrum Handbook” 2013 Released

Posted on: November 26th, 2013 by Armand Musey No Comments

Summit Ridge Group,LLC is please to announce the publication of The Spectrum Handbook 2013. This 182 page book is available in hardcopy and e-book version through major on-line resellers.

Executive Summary
This Handbook has three objectives: 1) to serve as a primer for explaining the complex issues around the use of electromagnetic spectrum; 2) to analyze, from both an economic and a legal perspective, the regulatory processes being considered or underway to reallocate or change the use of spectrum bands; and 3) to be a reference source for industry professionals. Part I of the Handbook provides an overview of the spectrum and the regulatory process. Part II of the Handbook explains the various available spectrum bands, discussing their range, location, and physical properties and how these impact their ability to be used. An analysis of the current allocation of these spectrum bands in the United States follows. Part III contains detailed explanations of the various spectrum band plans. Throughout the Handbook, we provide links in the footnotes to sources for additional information.
From a macro-perspective, regulators worldwide are currently reallocating spectrum from underutilized applications to the burgeoning mobile wireless broadband applications. Given the needs and importance of wireless broadband, from an economic and social perspective, this trend is likely unstoppable. The FCC is allocating both licensed spectrum (including the broadcast incentive auction) and unlicensed spectrum (including the 3.5 GHz and 5 GHz processes). Unlicensed (shared) spectrum is one approach to minimize disruption from these reallocation efforts and expand utilization is the small but significant spectrum sharing movement. Spectrum sharing is simultaneously proposing to improve spectral efficiency and calling into question the need for licenses altogether.
As a result of the existing processes underway to improve the efficiency of spectrum allocation along with new technologies that further improve efficiency, the extent of “spectrum crunch” (i.e. the apparent lack of available spectrum) is poorly understood and hotly debated. Summit Ridge Group does not believe it is likely to bring information access to a grinding halt in the United States. Rather, we may see temporary congestion while regulators approve new reallocations of spectrum and spectrum-sharing plans, and service providers build out services on new spectrum. These processes, combined with wireless carriers’ improved ability to regulate customer data usage—primarily by charging higher fees and/or capping usage, and offloading traffic via Wi-Fi and other technologies—should allow operators to continue to provide reliable service in the face of increasing demand. These trends are also likely to temper the increase in spectrum prices in the future.


New White Paper: Right-sizing Spectrum Auction Licenses: The Case for Smaller Geographic License Areas in the TV Broadcast Incentive Auction

Posted on: November 21st, 2013 by Armand Musey No Comments

Summit Ridge Group, LCC  just completed a white paper covering a significant issue related to the upcoming FCC Broadcast Incentive Auction. The paper, Right-sizing Spectrum Auction Licenses: The Case for Smaller Geographic License Areas in the TV Broadcast Incentive Auction was sponsored by the Competitive Carriers Association.  The Executive Summary is below.

Executive Summary

The wireless sector is a key contributor to economic activity and growth. Over the next several
years, wireless service providers are expected to invest $25 to $53 billion upgrading and expanding
their networks to deploy 4G mobile broadband across the nation. All told, wireless broadband
investment and the services and innovation supported by such investment are expected to add
between $259 and $355 billion to US GDP each year through 2017.
The Federal Communications Commission (“Commission” or “FCC”) is currently designing the
largest ever auction of terrestrial wireless spectrum, currently planned for late 2014 (the “Incentive

The purpose is to free up to 120 MHz of prime spectrum in the 600 MHz band,
currently licensed to over-the-air TV broadcasting, to be repurposed for licensing for mobile
broadband and other higher value wireless services. To accomplish this goal, the FCC proposes to
use a two-part auction process in which broadcast television license holders submit bids for
relinquishing their licenses (“Reverse Auction”); and commercial broadband providers bid to
acquire licenses to the spectrum freed up (“Forward Auction”). The FCC is currently evaluating
various auction design elements to promote competition in the auction. To best ensure this
important goal, the FCC is considering a number of auction design features, including spectrum
aggregation limits, constraints on the types of bidding allowed, and the appropriate framework to
use for the license territories to be used in the Forward Auction. This paper focuses solely on this
last issue. We explain here how adopting appropriately small-sized geographic territories is
necessary to promote competition and other important economic and social goals, while noting that
right-sizing the license territories may not by itself be sufficient to ensure adequate competition
and participation in the Forward Auction. For example, the Commission could adopt smaller
license sizes and still end with an auction where the two largest wireless carriers aggregate all of
the offered spectrum. Such an outcome would be inconsistent with the goal of promoting
competition in wireless services.

The territory size used for spectrum licenses is as important for valuing spectrum as the parcel size
is to real estate value. If all plots were 50 acres, parcels in Manhattan would be too expensive and
too large for most; this might compel buyers interested in a small parcel in Manhattan or a parcel
in New Jersey adjacent to Manhattan to bid for land they don’t want. Alternatively, otherwise
qualified buyers might be prevented from buying land altogether. Analogously, wrong-sizing
spectrum license territories to be used in future spectrum auctions, and in particular the Incentive
Auction, is likely to result in significant and unnecessary inefficiencies in the allocation of scarce
radio frequency spectrum resources. For carriers who are compelled to bid for wrong-sized
spectrum license packages, the added cost may be sufficient to discourage their participation; or if
they do participate, they are less likely to offer successful bids; or if they are successful, they will
have fewer resources available to deploy services using the spectrum. In each case, the efficiency
of the auction and the larger goals of the process suffer.

This paper explains why sufficiently small geographic areas, such as Cellular Market Areas
(“CMAs”), are the correct license territory framework to use to ensure that licenses are right-sized
in the Forward Auction. Industry participants and the FCC have successfully used smaller
geographic license sizes to auction spectrum in the past, and doing so in the Forward Auction
offers important advantages. Using smaller territories is better than using the larger Economic
Areas (“EAs”) because smaller areas efficiently match the needs of bidders to the spectrum they
seek. Their use ensures that the planned auction will reallocate spectrum resources efficiently
while promoting competition, economic growth, and universal broadband service.

Smaller license areas are better than EAs because smaller areas will help to maximize the amount
of spectrum that is repurposed for the Forward Auction. Specifically, smaller areas should increase
the ability to allow for market variation in areas where limited amounts of spectrum are procured
through the Reverse Auction, while reducing the amount of spectrum lost due to international
border coordination with Canada and Mexico or other encumbrances. Smaller geographic license
sizes should also maximize opportunities for efficient participation by both large and small
wireless service providers, and promote efficient build out of spectrum acquired through the
Forward Auction. Looking at past auctions, evidence suggests that auction proceeds would be
optimized through the use of smaller areas as opposed to EAs. Moreover, using smaller territories
is more consistent with the long-term direction of efficient spectrum management reform and
future wireless markets, including access to spectrum through secondary market transactions.
Finally, this paper rebuts some of the arguments made to date against the use of smaller geographic
license areas.

Is “Terrestrial Bias” a Political Problem for the Satellite Industry?

Posted on: October 15th, 2013 by Armand Musey No Comments

Is “Terrestrial Bias” a Political Problem for the Satellite Industry?

I recently caught-up with a European wireless executive and we compared notes on the communications industry. While discussing different technologies, I mentioned the progress satellite broadband was making in various countries. He responded with an interesting insight. Simply put, local politicians are eager to support terrestrial wireless and wireline projects because they create jobs in the areas they serve. However, satellite does not. Most satellite jobs are created in far aware satellite manufacturing and operations locations. As a result, local politicians often have a “terrestrial bias” when crafting policies.

Such terrestrial bias may be more prevalent in Europe than it is in the US. In the US, the FCC stepped in at the national level, and has limited the ability of towns and landlords to prohibit satellite dishes. But politicians in Europe, where higher unemployment levels are common, may be more sensitive to job creation than U.S. politicians. At the same time, most of Europe is, for the most part, sufficiently wealthy and densely populated to afford a variety of terrestrial technologies. In developing countries, satellite is often the only viable option, so job creation potential of alternative technologies is not an option.

Does this difference explain, in part, why satellite broadband has been relatively slower to develop in Europe than in the US or in developing countries? I’m not sure, but it’s a deceptively simple idea that seems to explain the data.

Takeaways from CCA Convention in Las Vegas

Posted on: September 19th, 2013 by Armand Musey No Comments

The Competitive Carriers Association held their annual convention earlier this week in Las Vegas.  A few takeaways:

1) Concern over LTE upgrade cycle. Small to mid-sized carriers appear anxious about the costs and benefits of LTE. They seem to realize they will need to do it eventually in order to remain competitive. But they are not certain there will be an appropriate return on the investment. Verizon’s LTE in Rural America, and a variant offered by Alcatel Lucent, are designed to provide shared infrastructure for smaller carriers to make this transition. But it comes at the price of reduced flexibility.

2) VoLTE, VOIP and other technologies are eliminating the divide between voice and data traffic. It threatens to convert their traditional high price per bit voice revenue into low price per bit data revenue. Carriers were flocking to presentations by various OTT service providers and to their booths in the exhibition hall.

It’s widely known that the cable industry’s Internet pipe is threatening to overtake and replace its core video services business. As a result it is scrambling to offer a host of over the top (OTT) services to supplement it dwindling video revenue. Now the same appear to be happening to the wireless carriers.

3) Shared infrastructure, of all types, generated significant interest. Lots of companies offering shared network management software, subscriber management software, customer applications and even shared LTE networks. The shared infrastructure was squarely aimed at smaller carriers who did not have the scale to develop these products in-house but wanted to be able to be competitive with the larger carriers.

4) Tower operators starting to embrace small cells. Crown Castle (Ticker: CCI) had a detailed set of slides in their booth showing the efforts to build small cell networks. This dovetails with our earlier analysis regarding how the increased use of small cells was limiting the tower operators benefit from the growth in the industry. They are responding and joining the effort. It remains to be seen how they will compete with smaller systems integration firms.

5) There was the ever present discussion of the broadcast incentive auction and what it might mean for smaller carriers – would they get additional needed spectrum or would it go to larger carriers leaving them increasingly marginalized?

“The Spectrum Handbook 2013″ Released

Posted on: September 15th, 2013 by Armand Musey No Comments

Summit Ridge Group, LLC is pleased to announce the release of its publication, The Spectrum Handbook 2013. The 182-page publication provides an overview of spectrum-related issues globally from technical, regulatory and business perspectives while acknowledging the many public policy issues. The Handbook is intended to be a useful resource for new attorneys, bankers and government staffers who need to quickly get up to speed on the industry as well as a reference tool for experienced practitioners. J. Armand Musey, CFA JD/MBA said, “As far as we can tell, there is no other publication which assembles this much information on the subject in one publication. With that said, we intend to update and improve each year and welcome feedback.” The Handbook has received acclaim from leaders in law, finance and public policy organizations.


Preview The Spectrum Handbook 2013 >>


Buy on Amazon (Paperback) >>

Buy on Amazon (Kindle) >>

More Evidence on Widespread Telecom Pricing Pressure

Posted on: August 18th, 2013 by Armand Musey

Last week I attended the 16th Annual Oppenheimer Technology, Internet and Communications Conference in Boston. Over 100 companies also attended, with 5 to 7 presenting simultaneously in different rooms, so unfortunately I could not attend all of them and concentrated my time on the communications presentations including: American Tower [AMT], AT&T [T], Ceragon Networks [CRNT], Cogent Communications [CCOI], Frontier Communications [FTR], KVH Industries [KVHI], RigNet [RNET], Verizon [VZ], and Vonage [VG] among others.

The major takeaway was pricing continues to fall for most players in the telecom industry. This is particularly true with bandwidth pricing at all levels – fiber, wireless and satellite. New technologies involving improved spectrum utilization are proliferating – new wavelength division in fiber, Wi-Fi, small cell offloading and LTE for wireless and new high throughput satellites for increased satellite capacity.

Frontier indicated their advantage that is allowing them to succeed while many other rural telcos are struggling is their investment in their infrastructure. Cogent indicated their advantage was buying distressed fiber systems pennies on the dollar that is now far more efficient than competitors’ copper-based solutions. AT&T and Verizon are benefiting from massive infrastructure that allows them to serve complex corporate applications. Vonage, which lacks such advantages, discussed a new $9.95/month service that seemed poised to barely break even and a move to the Brazil market which they hope will be less competitive. Meanwhile, the satellite players are struggling to come to terms with the potential disruption of new high throughput satellites. These new satellite also provide a full network solution but may bypass traditional systems operators giving the satellite operators a greater share of the end-customer value. Meanwhile, Qualcomm is seeking to use satellite spectrum in a ground-based system to provide in-flight broadband at much great rates that currently possible.

Given the rapid bandwidth deflation, companies hampered by prior investments (and associated debt service) in older less efficient systems are at risk for cannibalization by competitors with more cost effective systems. This trend is nothing new in the TMT sector, but the scope of it appears to be larger than previously seen. It may lead to industry consolidation based on a company’s (in)ability to upgrade to the most efficient system versus it’s competitors.  It’s also consistent with our earlier thesis of hardware/infrastructure (network equipment, satellites, small cells, wave division multiplexers etc), becoming a larger component of the communications value chain.

Why the Wireless Tower Operators Haven’t Participated in the 2013 Market Rally

Posted on: August 5th, 2013 by Armand Musey

Why the Wireless Tower Operators Haven’t Participated in the 2013 Market Rally

The S&P and the NASDAQ are up approximately 20% year to date. Yet the three largest tower operators (American Tower [AMT], Crown Castle [CCI] and SBA Communications [SBAC]) have not participated in this rally (YTD returns between 1% and -11%) despite the rapid growth in wireless usages, their high leverage and a strong bond market that has allowed them to refinance at attractive rates.

I was at an industry dinner recently that included investors and tower industry experts where we discussed these issues. It was a wide-ranging discussion, but a few things stood out for me:

1)   The major wireless carriers publically indicate are uninterested in unlicensed shared spectrum. But at the same time they have rapidly moved to off-loading traffic in the most valuable urban areas to Wi-Fi. Depending on which estimates one considers, carriers currently offload approximately 50% of their traffic to Wi-Fi.

2)   Wi-Fi siting is becoming easier in many areas due to some local zoning changes including allowing for light pole mounting and similar improved siting policies that are becoming more common. This is likely to further expand Wi-Fi offloading.

3)   The rapid move to Wi-Fi offloading is likely a significant part of the reason prices for U.S. licensed spectrum has largely stopped increasing over the past few years. One prominent study concludes international spectrum prices have fallen rapidly. Tower operators generally do not host Wi-Fi sites. Thus, towers operators now only participated in about half of the US wireless traffic (and falling).

4)   As spectrum prices have flattened, it is easier for wireless companies to increase capacity on their licensed spectrum (which carries their non Wi-Fi traffic) by buying additional spectrum. Carriers are now somewhat less inclined to continually make capital expenditures to divide cell sites compared to when spectrum prices seemed to be headed endlessly upward. This has allowed the wireless industry to expand capacity while adding fewer new antennas on towers.

5)   Major carriers have obtained much of their additional spectrum from industry consolidation and then increasing utilization of the acquired carrier’s spectrum (T-Mobile/MetroPCS, Sprint/Clearwire. ATT/Leap and many other smaller transactions). This has enabled them to take over antennas and existing leases on towers as opposed to building out the new spectrum from scratch.

6)   The consolidation of the carriers has also likely begun to change the negotiating dynamics of between the tower operators and the wireless carriers – there are now fewer wireless companies to compete for space on any given tower.

The tower industry is not without a bright side. Federal rules, including the “shock clock” that requires municipalities to quickly respond to tower siting applications are likely to improve tower siting in suburban areas. Upcoming FCC spectrum allocations, including the incentive auction are likely to increase demand for tower capacity as those frequencies are built out nationally as is the FirstNet, the proposed federal safety communications network. However more wireless capital expenditures are being made on spectrum and on Wi-Fi networks (often paid for by end customers) as opposed to cell splitting of licensed spectrum on towers. As a result, the tower demand are poised to grow at a slower rate than overall wireless capital expenditures.

Wall Street Finally Starting to UnderstandHigh Throughput Satellites

Posted on: July 17th, 2013 by Armand Musey

A recent Morgan Stanley research report  suggest they are cooling on their view of SES, in part, due to the expanded capacity upcoming High Throughput Satellites (HTS) may bring the industry. Each of these satellites has 40x to 100x the capacity of a traditional single-beam satellite. The Morgan Stanley report estimates that industry capacity may increase by 75% by 2016. We suspect the actual capacity increase may much higher by 2017-2018. While this is not surprising, it is surprising that it is taking the investment community so long to realize it. We’ve been discussing this risk for quite some time.

The cynical side of me suggests the large banking fees from debt refinancing and a recent industry IPO may have subconsciously encouraged Wall Street analysts to be a bit less aggressive at evaluating this risk than they might otherwise have been. Now that those deals are largely behind us, it’s safer for them to dig into these uncomfortable issues. To be clear, I am not accusing them of anything inappropriate – but human nature is such that its often hard to criticize the hand that feeds you. However, it’s more troubling to me that the credit rating agencies, the supposedly most independent group of analysts, do not seem to be seriously evaluating this risk.

The irony is that SES is probably one of the least exposes FSS operators to a potential capacity increase due to HTS. Most of SES’s business in Europe and in the US is broadcast traffic where HTS don’t have an advantage. Intelsat is probably one of the more exposed FSS operators to HTS price competition due to its larger amount of point-to-point traffic it carries.  And Intelsat is much more heavily levered, thus amplifying any potential impact on the equity. Of course, Intelsat will be launching its own HTS satellites (its Epic system). But the industry capacity increase may create significant downward pressure on point-to-point pricing that it may dwarf any increase in traffic they may see on the Epic satellites.

We don’t pretend to have all of the answers to this issue, but it’s an issue the investment community needs to do a better job of evaluating. And the industry should also do a better job of explaining how it will manage the complex scenario of a massive capacity increase over the next few years.

Is there an Attractive Business Model for Shared Spectrum?

Posted on: July 14th, 2013 by Armand Musey

The telecom industry has seen a tremendous increase in interest in unlicensed shared spectrum over the past year or so. Wi-Fi has been an amazing success, now carrying nearly 50% of wireless traffic. Moreover, the U.S. government has over 400 MHz of additional spectrum in the NPRM stage that it is planning to make available on a shared based. At industry conferences around the country, equipment providers, including Alcatel, Google/Motorola Mobility, Qualcomm and others are eagerly explaining their forthcoming products to take advantage of this new spectrum. However, there has been less clarity about what the business models for shared spectrum service providers will look like.  There is even less clarity as to when and how investors will commit money to shared spectrum. [note: in the context of this article “shared spectrum” refers to spectrum, usually unlicensed (including white spaces and wi-fi), where numerous new users are possible at any time. It is not referring to coordinated sharing of licensed spectrum].

Service providers have trouble justifying building out infrastructure on shared spectrum. Others users sharing that same spectrum can potentially limit their access it. This makes is difficult to guarantee any level of quality of service or even know how many customers their network might support. At the same time, unlicensed spectrum eliminates the license as a barrier to entry. Therefore, if a service provider is successful in a specific market, another entrant can move into that same market and on that same spectrum. The new entrant can not only undercut the incumbent, but also reduce their share of available spectrum. These factors have made it difficult for investors to get excited about investing in service providers using shared spectrum. So far commercial service on shared spectrum has largely been limited to small operators in rural areas. These entrepreneurial service providers are often nimble enough to survive in these small markets where larger operators cannot. But these are generally niche markets.

One solution to jump-staring a viable shared spectrum business model would be for a third party to build out infrastructure on shared spectrum and lease it to multiple service providers.  This would be similar role to that currently provided by tower operators to mobile operators. But instead of building the just the towers, they would also build and run the networks that multiple service providers could share. Service providers would largely handle marketing, billing etc. much like wireless MVNOs. Well capitalized the tower operators such as American Tower and Crown Castle are probably in the best positioned to fill this infrastructure role. However, the tower operators have shown little interest. This may be because they are typically highly leveraged and desire to keep their business low risk and their client base filled with large credit worthy mobile operators.

It’s appears shared spectrum use will, at least for now, continue to be largely limited to short-range private networks – as wi-fi is primarily used today. Shared spectrum is attractive for organizations to provide wireless capacity within and around small areas they control – offices, restaurants, auditoriums etc. However, these networks are generally not open for public access – only for those they want to serve – customers, employees etc. Moreover, they serves small areas where it’s often feasible to offer wireline service. In this way it’s effectively serving as a “private wireline substitute” as opposed to the traditional goal of wireless – offering connectivity where wireline is not viable. The social benefits of these networks are fall short of networks that cover large areas and are open for public use.

Another possibility is that shared spectrum will simply be used by traditional wireless operators – AT&T, Verizon, Sprint, T-Mobile etc – for supplemental capacity during peak periods. They already have customers, lower cost of capital and economies of scale. But even these operators also have the same challenge of not knowing how much capacity they will have on the spectrum when they build out which will likely cause them to be more cautions in building out than they would be with licensed spectrum. And use of shared spectrum by major operators will hardly unleash the flood of new entrepreneurs with new technologies that the government is hoping for.

No matter how the share spectrum industry evolves, if it is successful, the real winners are likely to be the equipment providers. Any service provider will need equipment and given the challenges of the using the spectrum they will be eager to use the most cost and spectrally efficient technologies.

I’ll be speaking at the Super WiFi Summit in Las Vegas on August 28th on a panel titled: “VCs and Investors Buying into White Spaces” where, along with a distinguished panel, we will be discussing these issues.


Conference Summary – More Weak Arguments from Big Companies

Posted on: June 16th, 2013 by Armand Musey No Comments

I attended the Optimal Coevolution of Mobile Broadband Technology and Spectrum Policy workshop in Washington D.C. on Friday.  A number of leading spectrum policy and technology figures were present.

Key takeaways:

1)   Blair Levin (Aspen Institute Fellow and an author of the National Broadband Report) indicated that earlier drafts of the National Broadband Plan included a larger reduction of broadcast spectrum than the proposed 120 Mhz. Seemed to suggest that even with broadband speeds significantly lower than the National Broadband Plan’s goals of 100 megabits per second, there would be no need for broadcast television. Presumably, all of that spectrum could be redeployed for mobile broadband or other high value uses.

2)   Thomas C. Power (presidential advisor for Technology Policy) indicated the federal government was moving on numerous fronts to encourage agencies that were not efficiently using spectrum to give it up or share it. He discussed the presidential memorandum released earlier that day. This includes greater reporting requirements, enhanced monitoring of usage levels, and aggressive inventorying of government usage. Mr. Power seemed confident these measures would shake loose spectrum for sharing.  I am not as confident as he is that the executive office will be able to get many agencies to return or share spectrum. But any measures that increase the cost for agencies to keep it may be part of a solution to encourage them to give eventually share or return it.

3)   A representative from a major equipment provider gave a presentation that included advocating Wi-Fi spectrum expansion. A leading industry economist asked him if there was any indication Wi-Fi bands were congested or might become so in the future. Incredibly, neither this representative nor anyone else in the audience of industry experts could think of any real evidence of the need for more Wi-Fi spectrum! Most reported Wi-Fi congestion in office is the result of poorly set-up systems that can be easily solved. I am not arguing that there is no need for additional Wi-Fi capacity, but rather that the case for it has not been made. The best argument for additional Wi-Fi spectrum might that the panned FCC approval of Wi-Fi “signal boosters” will extend the Wi-Fi range and reduce frequency reuse. But if there is no clearly documented demand for additional Wi-Fi spectrum, they why is there such a push for more of it? Who benefits when additional Wi-Fi bands are added and everyone needs upgrade their equipment to make sure it is compatible with all of the new devices being made for the new bands? (sic)

4)   Another representative from a large telecom company had little response to a question about why the industry is largely focused on continuing bandplans based on equally sized pairs of uplink and downlink spectrum when most research shows there is between 8 and 30 times more downlink traffic than uplink traffic. This wastes a large portion of the uplink spectrum. The representative had little response other than an acknowledgement that the asymmetry is driving femtocells to use a TDD format. Obviously, a long term plan to increase the amount of downlink spectrum relative to uplink or a move to a TDD based system (TDD sends and receives on the same spectrum) would be politically difficult. But such a plan could increase the effective amount of available spectrum by perhaps 30% to 45%. However, there is little industry discussion except occasional ideas of “supplemental downlink spectrum” for the new 600 MHz band plan. Others from the equipment side seem to always bring up obscure scenarios as a reason for keeping equal-size spectrum pairing. These include parties where young people send video clips to their friends and thus there is more uplink traffic. Two weeks ago I participated in a panel where an industry representative indicated that the asymmetry between upload and download was relatively new – starting with the advent of Internet capable phones – and might not last (as if the Internet was just a fad). I am really not sure who benefits from wasting spectrum with the current equally sized band plans. Perhaps it’s the companies without TDD type technology?

While I am not surprised industry representatives make self-serving arguments in an attempt to influence their regulatory interests. That is their job and it happens in all industries and is not necessarily inappropriate. However, I am shocked at the hollowness of the arguments currently being made by senior officials in the telecom industry. This observation dovetails with my earlier post of how the large industry players are driving the industry policy discourse by the sheer size of their bullhorns and crowding out other views – sometimes with positions for which they have almost no supporting information. Last month I was at another industry workshop and wrote a blog post in response to a similarly absurd suggestion from a senior trade association official that the U.S. was behind other countries, including Europe in mobile broadband. There are a lot of smart people in the telecom industry, isn’t it about time we try to raise the level of the discourse?

Kudos to Peter Rysavy and John Mayo for organizing a high quality event.

Wireless Industry Lobbying Mutes Alternative Views

Posted on: June 11th, 2013 by Armand Musey No Comments

The wireless industry invests roughly $30 billions of dollars annually in capital expenditures building out their networks and acquiring spectrum. The major carriers average $5 billion to $10 billion in capital expenditures each. Given that the industry is heavily regulated, it would be irresponsible for them to not protect their investment with appropriate political lobbying efforts. It has been reported that the Big-4 and the CTIA spent over $53 million in lobbying efforts in 2012 ( and occupy three spots of the list of the 30 largest spenders on lobbying. This is disproportionate to the industry’s size in the economy, but perhaps appropriate due to the regulatory intensiveness of the industry. Their lobbying budget allows the wireless industry to maintain constant vigilance on pending legislation and FCC policy changes.

The large wireless service providers complete comments on almost every docket (AT&T filed 11 comments on the broadcast incentive auction process alone), and conduct numerous meeting with members of Congress and their staff. This gives the wireless industry significant access to and likely clout over spectrum policy. For example, when the FCC attempted to prevent AT&T and Verizon’s participation in the broadcast incentive auctions, Congress quickly responded with legislation that instructed the FCC not to prevent any qualified bidder from participating.

The risk of these large lobbying expenditures is that the power of a few strong industry players often sets the narrative of how an issue is explained. It is often difficult for an alternative viewpoint, particularly that of a potential new entrant, to get heard. The major wireless industry players have stressed the “spectrum crunch,” without much supporting information, as the basis for their need for significant amounts of additional spectrum. It is only after a few years, that significant numbers of people are now questioning seriously the data upon which this is based. The opponents of the “spectrum crunch” hypothesis don’t have conclusive information on their side either. But their relative lack of funding put them a few years behind in the public/political debate.

Over the next few years the wireless industry will be facing major decision on many issues such as, spectrum auctions, shared spectrum, spectrum caps, industry consolidation and other issues. It would be a shame if alternative views are drowned out.

Putting the State of the U.S. Wireless Industry in Perspective

Posted on: May 13th, 2013 by Armand Musey No Comments

I’ve neglected this blog for a while as I am working on my forthcoming “Spectrum Handbook 2013” which I expect to release in a few weeks. Please contact me if you’d like a copy. But I did have a chance to attend the Winnik International Telecoms and Internet Forum in Washington, D.C. on Thursday and Friday May 9th and 10th sponsored by Hogan Lovels. The Thursday afternoon program focused predominately on international trade issues, while Friday’s full day session focused on telecom issues with a significant level of international comparison.

It’s easy to complain about the state of the US Wireless industry. I recently took a transatlantic flight and the Internet access on Turkish Airways over the Atlantic Ocean was better than the Internet access in the International Terminal at JFK Airport! Indeed, much has been said and written about the U.S. spectrum crunch and the inadequate regulatory responses as well as the obscene international roaming charges U.S. subscribers often pay and the hassle of locked phones. But as I sat through the presentations at the Winnik Forum, and better understood the challenges facing other countries, and cross referenced facts in other publications, it occurred to me that the U.S. situation is really not that bad. Here are some facts:

  1. The U.S. has 608 MHz of spectrum allocated for licensed wireless broadband – up from 170 MHz in 2006. This is more than almost any other country except Germany with 615 MHz. Most European countries have hundreds of MHz less.
  2. Highly touted Asian competitors including Japan and China also behind on spectrum – 500 MHz and 227 MHz respectively.
  3. The U.S. is also ahead of the curve on unlicensed spectrum with approximately 500 MHz of new unlicensed spectrum in the NPRM stage. It’s the only country with rules for white space devices.
  4. The United States has seen wireless CapX increase by about 75% since 2007 according to Goldman Sachs. Multiple sources show that Europe has seen a significant decline in CapX over the past several years.
  5. The U.S. currently has 70% of the world’s LTE subscribers. While this number is sure to decline as countries with large populations add service, this shows the U.S. market is massively ahead of what other countries have achieved. I recently tested my broadband speed my 4G iPhone in Manhattan on a Saturday night – the most crowded market in the country – during a crowded time. I had 14 Mbps download and over 1 Mbps upload. On a Sunday afternoon, I had 24.4 Mbps download and 5.3Mbps upload!
  6. In addition to terrestrial wireless, the U.S. is also a leader in advanced satellite broadband with ViaSat’s Exede and Hughes Network Systems’ Jupiter offerings. These satellite offerings, although lagging Cable and DSL, are becoming increasingly competitive. Additionally, the U.S. has the world’s only mobile satellite radio system – SiriusXM Satellite Radio.
  7. Despite enjoying these technical advances, according to the CTIA, U.S. consumers also benefit from the lowest pricing and highest wireless usage rates in the developed world.

The U.S., as with other countries, also has plans to add several hundred MHz of licensed spectrum over the next several years. While these plans are behind schedule, so are those of most other countries. Currently, wireless carriers have build out perhaps 250 MHz of the 414 MHz available below 2.5 GHz and perhaps 1/3 of the 194 MHz of spectrum in the 2.5 GHz band. They have still significant amounts of spectrum to build out before they need more – not to mention that other countries, such as Japan, have much denser cell sites.

I am don’t mean to be an apologist for FCC delays or to suggest “we’re fine since everyone else is even worse-off.” I am all for having a healthy does of dissatisfaction to avoid complacency. And given the economic and social benefits of increased broadband, the U.S. wireless industry is right to aggressively goad regulators to deliver as much as possible. But let’s not lose sight of the big picture:  the U.S. wireless industry is working reasonably well.

Quick Analysis of Intelsat’s Q1 Financial Results

Posted on: May 12th, 2013 by Armand Musey No Comments

I’ve neglected the blog for a while as I am working on my forthcoming “Spectrum Handbook 2013.” I expect to release it in a few weeks. Please contact me if you’d like a copy. But I did have a chance to review Intelsat’s Q1 results. Results were somewhat better than I expected – modest growth, stable margins, etc. A few thoughts – none of which should be viewed as a recommendation to buy or sell securities or any kind:

A) IPO WAS PROPERLY PRICED — Intelsat’s IPO looks like a success. Priced at $18 below its filing range and with a sharply reduced size, the stock has held its own and even exhibited a respectable IPO “pop.” While the IPO valuation represents little upside for Intelsat’s private equity investors and I was skeptical there would demand based on the company’s risk, it seems the valuation was right for the market.

But we are still a bit confused about who owns the stock. We surveyed the leading industry consultants and none of them received more than a single investor inquiry during the IPO process! Moreover, not a single investor asked a question during the recent quarterly conference call! We’ve not sure if this is good or bad, but is certainly seems strange. Please let us know if you have any insight into the mysteriously silent Intelsat shareholders.

B) CORPORATE STRUCTURE SIMPLIFIED — Intelsat continues to flatten its corporate structure. The entity formerly known as Intelsat S.A. (the highest reporting entity) will be renamed Intelsat Investments SA and will cease to report as the only debt at that level has been retired. The top level, formerly know as Intelsat Global Holdings, will be renamed Intelsat, S.A. and will be the reporting entity of most relevance to equity owners. This will make life easier for those following the company but may put some org chart designers out of work.

C) Q1 REVENUE WAS OK — Management emphasized the 2% growth year over year compared to Q1 2012. We are not sure this is quite apt. A more relevant comparison is the drop-off after Q4’12 compared to Q4’11. Intelsat’s Q1’13 revenue of $655.1M was $21.3M lower than Q4’12 revenue of $672.4M. By contrast, its $644.2M of revenue Q1’12 was only $9M lower than its Q4’11 revenue of $653M. Thus, Intelsat had a larger revenue drop from Q4 to Q1 than in the prior year – over twice as large. Moreover, in 2012 Intelsat saw five satellites come on line to help increase its growth through 2012. This will not be the case in 2013 when it does not plan to launch any satellites. Some possible reasons for the large Q4 to Q1 drop-off include:

  1. Possibly due to an effort to “shore-up” Q4’12 prior to a planned IPO left fewer low-hanging sales in Q1’13 as well as the impact of management’s expected distraction during the IPO process. To be clear, we are not suggesting management engaged in anything improper.
  2. Sequestration is hitting the industry, reducing government related sales.
  3. We have heard from multiple sources in the industry (but not enough to have a solid conclusion) that Intelsat sales efforts have been lackluster compared to other players in the industry. Intelsat made sharp cuts to its G&A costs over the past few years as previously discussed. These cuts may be starting, at the margin, to hamper sales efforts. This could be particularly true with government sales where contract lengths are getting shorter, increasing the volume of contracts and associated administrative work. Intelsat may have prioritized its staff to work on longer-term, higher margin media business.
  4. Media revenue was up 6% – this suggests Intelsat is competing well in that market – slightly higher than the overall media market. This contributed to an increase of 5% for transponder services and 4% for network revenue as a whole.
  5. Channel services and Internet trunking is declining at a rate of $30M annually. We think this trend is likely to continue as the fiber build out continues, particularly in Africa.
  6. Off –network services declined $13M or 17% – largely due to lower hardware sales and 3rd party sales. These sales are lower margin than Intelsat’s overall business, causing a decline in direct cost of revenue by $7

D) SHRINKING BACKLOG? -– Intelsat’s backlog fell by approximately $300M during Q1’13 t0 $10.4B. This represents a dismal book to bill ration of approximately 0.5:1. However, major contracts are often lumpy by nature and one quarter is not sufficient to indicate a trend. However, CEO Dave McGlade indicates that the decline was due to lack of launches that normally increase backlog. As there are no launches scheduled until the second half of 2014 (and launch dates are often delayed), this seemed to suggest Intelsat’s backlog would continue to decline. At a $300M loss per quarter, backlog would still by $8.6B by the end of Q3’14. We are not suggesting backlog will fall this far.

E) FLAT 2013 REVENUE GUIDANCE — Intelsat is projecting 2013 Revenue of $2,615M to $2,640M. At the midpoint, this is an annual growth rate of 0.665% versus 2012’s 0.840% growth. Given the lack of new satellite comes on line in 2012 and the impact of sequestration, this may be somewhat aggressive, but not unreasonable.

F) SUCCESSFUL DEBT REFINANCING — Intelsat has done a great job of leveraging the strong bond markets to lower its interest rate and spread-out its debt maturities to eliminate the bulge of debt maturities it previously had in 2017. The effect of these transactions is to lower debt payments by $170M. Pro-forma for the 2013 transactions, Intelsat lowered its debt to adjusted EBITDA multiple from 7.85x to 7.55x. Of course, these calculations do not include the $854M in deferred revenue that, in our view, is effectively debt.

G) INSURANCE PAYMENTS RECEIVED — Intelsat received at total of $491M in insurance proceeds from Intelsat 19 and Intelsat 27. Most of this was used for debt repayments. As we mentioned, earlier, the Intelsat 27 failure seems like a blessing in disguise.

H) AMBITIOUS SATELLITE CONSTRUCTION — Intelsat announced four new Epic satellites in addition to the one under construction. One was generally expected, the other three were not. This represents a bold bet that Intelsat will be able to replace multiple satellites with a single Epic satellite. The first will be go into service until 2016. Intelsat also has for “regular” satellite under construction, bring a total of nine satellites currently under construction.

Dave McGlade indicated that approximately 20% of the first Epic satellite, IS29, has been presold to Panasonic Avionics, Harris Caprock and MPN. We suspect this represents significant portion of Intelsat’s $854M of customer prepayments. CFO, Michael McDonnell indicated IS29 has the equivalent of 270 transponders. If 54 transponders equivalents (20%) have been pre-sold and pre-paid, that represents a lot of capacity for which Intelsat will not receive additional cash after the launch. McGlade indicated Intelsat would seek similar customer pre-commitments for the other Epic satellites.

I) CAPX GUIDANCE — Capx Guidance is $600M to $675M for 2012, $575M to $650M for 2014 and $775M to $850M for 2015 when their launch cycle ramps-up. Although management did not provide guidance, we’d expect high Capx in 2016 and 2017 as the rest of the Epic satellites are launched. So its not clear Intelsat’s Capx holiday will last.Note that Intelsat’s Capx guidance includes capitalized interest. After satellites launch, this interest is then expensed. This is proper accounting, but it makes Intelsat’s true total interest payments somewhat confusing to track. For example despite all of Intelsat’s refinancing over the past 15 months that lowered Intelsat interest costs by hundreds of millions annually, Q1’13 interest expense was $318.4M vs. $312.9M in Q1’12. This was due to the five satellite launches that resulted in expensing interest related to those satellites as opposed to capitalizing them.

BOTTOM LINE: Intelsat ‘s quarter was decent give market condition. But its debt situation, in our view, remains precarious. Intelsat has approximately $2B in annual EBITDA with little to no growth, over $1B in interest expense and $500M to $1B in annual Capx. This leaves little room for handling setbacks such as satellite failures or business slowdowns. In 2004, for example Intelsat lost two satellites in orbit. We understand that Intelsat only insures its launches and does not generally carry in-orbit insurance. At $300M per satellite, each with average annual revenue of $40M to $70M and a 3-4 year replacement cycle, an in-orbit satellite failure or two could be quite challenging. Intelsat is also exposed to other risks including refinancing its debt as it comes due when the markets may not be as cooperative. The IPO, however, gives Intelsat additional options to handle financial setbacks. These include rights offerings and debt to equity conversions. But these are all potentially dilutive to stock holders.

Unbelievable Lack of Consensus on Spectrum Supply & Demand

Posted on: April 9th, 2013 by Armand Musey No Comments

I’ve recently spoken at several wireless industry conferences where I’ve had a chance to informally discuss various topics with a number of well-regarded U.S. industry experts. In the course of these conversations, the discussion often turned to the issue of supply and demand for wireless spectrum. While it’s normal for opinions about the future to vary, I was struck by the extent of the disagreement on this fundamental topic. Here’s a quick summary of the debate.

The Supply Debate

The supply side of the equation is more complex than the demand side. Much of the uncertainty is due to the difficulty in predicting the outcome and timing of various government efforts to increase spectrum availability. But the following is know:

  1.  The 2010 National Broadband Plan seeks to add 300 MHz of spectrum for mobile broadband by 2015 and 500 MHz by 2020. The largest part of that is the 120 MHz the FCC seeks to pull back from television broadcasters via the reverse broadcaster incentive auction process.
  2. The PCAST report, suggests the government share 1,000 MHz of federal spectrum below 3.7 GHz with commercial users.
  3. There is also a movement afoot to force spectrum license users to allow others to use their spectrum when they are not using it (see:

Cumulatively, these plans, if realized, could massively increase spectrum supply for wireless services.

But these processes are somewhat behind schedule – some say hopelessly. Some say few government agencies will ultimately agree to share spectrum. They also suggest wireless companies will be reluctant to be “second class citizens” on government spectrum. They will resist use smart radios needed to determine when and where they can use it.

The television broadcasting incentive auction is currently getting a great deal of attention. Some say it won’t actually happen due to political issues while others say it’s on the road to happen in the next 18 months and will release 120 MHz of spectrum nationwide into the market. They note that all of the other spectrum reallocation efforts are all years behind schedule. Others say it will fail because the spectrum is too valuable for the broadcasters to sell and any price the FCC will likely offer. At the same time, another leading expert told me that the broadcast incentive auctions may not happen because so much of the government spectrum is being opened for sharing per the PCAST recommendations. Nearly 500MHz of shared spectrum allocation is currently in the NPRM process – 3.5GHz [100-150MHz], 4.9GHz [50MHz], 5Ghz [195MHz] and there is also some effort going on in the 1750-1850MHz [100MHz] band. This camp argues that the incentive auction may be called off because the broadcasters’ spectrum simply won’t be needed.

The Aereo, Inc situation creates ad additional wrinkle. Some broadcasters (including Fox) are now saying that if the Second Circuit decision affirming Aereo’s right to sell their over the air broadcasts through the internet, they may stop broadcasting over the air and only sell to satellite and cable systems. If this happens, television broadcasters might not have much incentive to keep their spectrum and would eagerly participate in the auction process. Ironically, in this case it would not be the weakest broadcasters who would participate, but rather the strongest. The strongest broadcasters, especially those with network affiliates, have leverage with the cable companies. The smaller ones are more likely to be dependent on “must carry” rules that are tied to broadcasting over the air. Of course the Aereo’s legal situation is not completely clear. A lower level (district court) in the Ninth Circuit (covering nine western states plus Guam and the Northern Mariana Islands) ordered a similar service to shut down. The judge indicated the broadcasters were likely to win their claim of copyright infringement. Should the case be appealed and the Ninth Circuit rules opposite the Second Circuit, it would set the stage for the Supreme Court to decide. The broadcasters may also seek regulatory relief. Broadcasters have a great deal of influence in Washington. However, the government is encouraging television broadcasters to leave their spectrum. Passing a law strengthening their hold on spectrum would seem inconsistent.

The Demand Debate

Like the argument with respect to supply, there is little consensus about the rate of future demand growth. The issue is somewhat more fundamental – predicting consumer usage patterns and improvements in spectrum efficiency. Some experts point to high historical demand growth and the increased rollout of smart phones, which tend to use multiple of the amount of spectrum of other phones and the rollout of 4G which only encourages increased internet usage including media downloading.  Cisco, for example, projects a 10-fold growth in mobile demand from 2012 and 2017. Currently approximately, the FCC has allocated approximately 608 MHz to wireless broadband. The “spectrum crunch” side argues that such demand growth will outpace any likely increase in supply.

Others say the demand growth is exaggerated (See They point to data caps that most operators have put on their new plans. Some evidence shows that the growth in data traffic has slowed. Newer 3G and 4G protocols are far more spectrally efficient that earlier versions. Additional rapid Wi-Fi off-loading – estimated at close to 33% of wireless traffic in 2011 and growing – perhaps to as much as 60% today — is absorbing much of the demand growth. Skeptics of the “spectrum crunch” point to the a study cited in the PCAST report indicating that less than 20% of the capacity in prime spectrum bands under 3.7 GHz is utilized even in the most congested areas (See PCAST Report, p.99) . The PCAST Report argues that more efficient spectrum use could increase effective capacity by a factor of 1,000 (p. vi).


It’s rare for an industry as mature as the U.S. wireless industry to have such little consensus about medium-term supply and demand growth. But a lack of consensus attracts investors because it creates an opportunity to make money for those who can get it right. Currently investors are bidding-up smaller broadcasters in major markets with the hope of selling back to the FCC at an even higher price. Broadcasters expecting the auction to be less attractive are eagerly selling to them. Wireless companies are making major spectrum transactions with little industry certainty about future supply that could significant impact the values of the spectrum they are trading.

The Intelsat IPO – A Setup for Dilution?

Posted on: April 5th, 2013 by Armand Musey No Comments

DISCLAIMER: The analysis below is for information purposes only. We believe, but cannot guarantee that the information contained in it is accurate. Moreover, it should not be considered a recommendation to buy or sell any security of any kind. Please consult a qualified advisor before making significant investment or business decisions. See our Terms of Service for additional information

Intelsat announced it will be going forward with an IPO scaled-down twice – from $1.75-2.0B last summer to $0.75-1.0B two months ago and now somewhere in the $0.5B range. Market demand was cited as a reason. The fact that market demand has been shrinking during a period of increasingly strong stock markets, currently near all-time highs, is not encouraging.

It’s somewhat counter intuitive that the deal would be lowered to address market demand. Intelsat’s biggest challenge is managing its debt load of over $16B. The IPO proceeds will largely be used to pay-down debt. But the smaller $0.5B IPO would pay down only a small fraction of that debt. It would leave the IPO investors in a more risky position than a larger IPO that would pay down more debt and bring leverage ratios to more normal levels. Therefore, one would think that a smaller IPO would be less attractive to the market as it would leave purchasers with stakes in a riskier company.

On the other hand, there is the age-old issue of supply and demand. Selling shares in a more risky company is harder to justify to savvy investors. But with a smaller deal, they need to sell it to fewer people. It just might work, particularly if it is sold to retail investors.

But what will the IPO investors get? They will get shares in a company with more debt than it can comfortably manage. What will the company get? They will get a publicly traded stock. With a publically traded stock they will be able to do things like offer bondholders incentives to covert debt to equity, rights offerings to raise more money and other tools to reduce the debt load. This is great for everyone, except the stockholders who may see endless dilutive deals as a result.

The IPO would also create an exit path for Serafina (the investment vehicle for BC Partners and the other PE owners) who are at the end of a PE firm’s typical 5-year holding period. When their lock-up period expires, Serafina and other holders of the approximately 79% of the shares not sold in the IPO, will be able to dribble their into the open market. But that creates overhang risk for the new public shareholders. Such selling, or even the risk of it, may keep share prices depressed.

Intelsat’s Financials Explained in Three Easy Tables

Posted on: March 20th, 2013 by Armand Musey No Comments

DISCLAIMER: The analysis below is for information purposes only. We believe, but cannot guarantee that the information contained in it is accurate. Moreover, it should not be considered a recommendation to buy or sell any security of any kind. Please consult a qualified advisor before making significant investment or business decisions. See our Terms of Service for additional information.

There has been a lot of confusion about Intelsat’s debt and operating performance.  We believe an analysis of a few publicly available numbers will clear-up much of this confusion and thereby advance industry discussion.


Revenue 2,5132,5452,5882,610
Income for Ops (ex. asset impairments)1,0451,0231,1691,186
Amortization of Intangible Assets(146)(130)(106)(92)
General and Administrative Expenses(260)(220)(208)(204)
Intelsat S.A. EBITDA (adjusted) 1,9731,9892,0172,016



As the chart above indicates, Intelsat’s revenue has been nearly flat over the past four years, increasing $97M, from 2,513M to 2,610M, an annual rate of 0.95%.

Income from Operations (excluding a $499M and $110M asset impairment charge in 2009 and 2010 respectively) increased $141M between 2009 and 2012, from $1,045 to 1,186M, or 3.22% annually. However, over $54M of this increase (more than 1/3) was driven by the tapering-off of amortization of backlog and customer relations from the 2008 acquisition. This non-cash item declined from $146M in 2009 to $92M in 2012 and will continue to decline independent of the company’s actual performance.  Additionally, decreases in General and Administrative expenses, which declined $56M annually from $260M to $204M between 2009 and 2012, provided an additional $56M of the increase in Income from Operations.  These two items together accounted for all but $31M of Intelsat’s increased Income from Operations over the past four years. Without these items, Income from Operations would have increased at an annual rate of 0.74%, closely in line with the revenue increase.

Intelsat’s annual EBITDA increased $41M, from $1,973M to 2,016M, or an annual rate of 0.54% over the past four years. But without the reductions in General and Administrative expenses mentioned above, Intelsat’s 2012 EBITDA would have been about $7M LOWER in 2012 than in 2009. Interestingly, Intelsat’s EBITDA was $700K lower in 2012 than 2011 Typically a company planning an IPO works very hard to ensure that its most recent financials show some improvement, even if it is small.

Our Take-Aways: Intelsat’s revenue is growing slower than inflation. As a result, inflationary growth in some expense items is starting to put downward pressure on operating results. Cost cutting to maintain margins, in our opinion, is likely to be unsustainable in the long-run.


Cash from Operating Activities 8741,018916833
Δ Customer Pre-payments 64172296124
Cash From Ops Excluding Customer Prepayments810846620708
Sub Interest Paid in-Kind (298)(245)(27)(5)
Cash from Ops (Exc. Pre Payments and PIK)512601592703
Depreciation (Exc. Intangible Amort)(658)(669)(664)(673)



Intelsat’s reported Cash Flow from Operations (includes interest payments) have been declining for the last three years and are now lower than in 2008. This is in large part because Intelsat has restructured a significant portion of its non-cash pay debt (discount or paid in-kind) so that it is now cash-pay. As a result, although its interest expense has declined almost $100M (due to lower rates), its cash payments have risen significantly. Since all the interest needs to be paid eventually, theoretically it does not make an economic difference in the long run. However, larger required cash payments do reduce the company’s flexibility in the short-run. 

Rising customer prepayments have materially supported Intelsat’s Cash Flow from Operations. Essentially, Intelsat has been collecting money from its customers years in advance of providing service (presumably at a discount to normal rates) and then recording it as revenue as it provides the service over time. In our view, it is effectively a form of borrowing. The customer prepayments increases Cash Flow from Operations in the year the money is collected, but decreases customer payments in later years when the service is provided.

As all interest will need to be paid eventually and since increasing customer borrowing is not sustainable, we have attempted to adjust the Cash Flow from Operations to reflect an elimination of non-cash pay interest and customer prepayments. The result of these adjustments is there is a clear trend of increasing underlying Cash Flow from Operations. Cash Flow from Operations has increased approximately $200M since 2009. Approximately half of this increase ($96M) is due to the lower interest expenses and another $56M is due to lower General and Administrative Expenses. We do not believe significant additional reductions in either of these expenses are likely in the future. This limits future increases in Cash Flow from Operations absent increased revenue.

Finally, we note that Cash Flows tend to be more “lumpy” than accrual results. So we would tend to read less into the lower 2011 and the higher 2012 Cash Flow numbers. Since the EBITDA numbers are similar in both years, we suspect much of this swing is due to timing issues. For example bills paid Dec 31th vs Jan 1st..

Our Take-Aways: Intelsat seems able to generate approximately $600M to $700M in sustainable operating cash flow per year, about equal its normalized depreciation. This means that if everything goes just right, Intelsat has enough cash flow to replace the fleet as it currently is, but little left over for debt repayment or other contingencies.


20082009201020112012Δ ’08-’12
Headline Debt14,87315,32115,91716,00215,9031,030
Net Debt14,40314,84315,22415,61315,7161,313
LT Deferred Revenue161255407724834673
Net Debt + Deferred Revenue14,56415,09815,63116,33816,5501,986



Intelsat’s debt is increasing more than is initially obvious. In addition to the increase of the face value of its debt by $1.03B, from $14.833B at the end of 2008 to $15.903B at the end of 2012, the company’s cash has also declined $283M during this period, from $470M to $187M. As a result, its net debt at the end of 2012 is actually $1.313B higher than it was at the end of 2008.

The customer prepayments discussed earlier effectively increase Intelsat’s debt, in our view. But they are not as obvious as traditional borrowing. If one does not take deferred revenue into consideration, one will, in our opinion, not understand the company’s true debt levels. Intelsat’s long-term deferred revenue balance (excludes balances to be settled within a year) grew from $161M at the end of 2008 to $834M at the end 2012. At the end of 2012 Intelsat had $84.1M of deferred revenue for which it expected to provide services in 2013. It will not receive additional cash for this. Unless Intelsat is able to continue to borrow from customers at similar levels, this practice will lower cash flows in future years.

When both the decline in cash and the increases in long-term deferred revenue are considered, Intelsat’s “true debt” (our term) has increased $1.986B since the end of 2008, significantly more than the $1.030B “headline” number.  We should note that after the end of 2012, Intelsat 27 suffered a launch failure. The $400M+ in expected insurance proceeds should increase Intelsat’s cash balance and somewhat reduce Intelsat’s net debt.

So what did Intelsat get for this additional $1.986B in debt? Primarily, it received an extra $1B in net satellite fleet assets net of depreciation. If Intelsat limited its Capx to its operating cash flow, it would have been stretched to keep-up with existing fleet depreciation. Intelsat also took almost $500M in losses on early exchange of debt to lower its long-term interest costs. Unfortunately, it did not get much more revenue or EBITDA or Cash Flow from customers.

Our Take-Aways: Intelsat has added almost $2 billion in debt over the past four years, but has little to show for it in terms of improved operating performance. Intelsat has barely enough internal cash flow to maintain its fleet at current investment levels, leaving almost none to pay-down its debt or deal with contingencies, setbacks that may arise, or an industry downturn. Absent a method for converting its debt to equity (perhaps through an IPO or a restructuring) its onerous debt load appears to be effectively permanent.

Our Conclusions

Intelsat has barely enough money to replenish its fleet (Note – see April 7 update below on this issue). If the company is completely stable or grows, that may not be problematic. But an industry downturn, even a modest one, or a company setback, could upset this delicate debt balance of operating cash flow vs. CapX requirements. As Intelsat does not have cash flow to materially retire debt, it may be exposed to the risks of perpetually needing to refinance the debt as it comes due.

The customer pre-payments are an area to watch because they can eat into future cash flows as the company is obligated to provide service in future years but can’t collect additional revenue. If this balance increases, and customers at some point stop effectively extending credit to Intelsat, it may have to provide significant services without revenue. This could lead to a cash crunch. So far, we have seen no evidence that this is happening.

Update 3/21/13 @ 12:10pm — Please send any comments or questions to We’d like to make sure we are communicating as clearly and accurately as possible when presenting this somewhat complicated financial issue.

Update 4/7/2013 @ 6:07pm — It’s been pointed out to me that I was being conservative on the replacement capx assumptions. Typically replacement capx is higher than depreciation because new Capx is replacing old PP&E.  The new PP&E has generally increased in cost since it was purchased several years earlier. Thus to account for inflation and revenue growth (assuming no real revenue growth in this case), the replacement capx should be: depreciation * (1 + nominal growth)^(average life of assets * .5).

Intelsat’s 2012 depreciation of $673M on $9,421M of total assets (gross of depreciation), suggests an average of a 14 year depreciation life, roughly consistant with the average life of satellites, their largest asset. According to the St. Lewis Federal Reserve, the producer price index for equipment has increased 1.6% annually or about 12% over the past 7 years. Thus the replacement CapX should be (1.12 *673M) or $753M. If Intelsat has to replace its assets at higher rates based on inflation, it becomes much harder for it to juggle its cash flow vs. capX requirements.

Does Intelsat’s Recent Management Shake-up Indicate Lack of Confidence in the IPO?

Posted on: March 13th, 2013 by Armand Musey No Comments

Intelsat’s recent management shake-up may suggest management has given-up on the IPO. If not, the company sent negative signals to investors that may setback the IPO prospects even further.

Intelsat management shake-up included several senior-level promotions and the departure from daily operations of their highly credible General Counsel, Phil Spector. Mr. Spector also ran the company’s business development efforts. Ordinarily, this would not raise a lot of eyebrows. People are promoted and leave companies all the time. But this seems different.

Intelsat is heavily levered with a credit rating several steps below investment grade. Many analysts are concerned about it’s ability to manage the debt, which is dangerously close to (and may, in my view, even exceed) the company’s total value. Most analysts agree that its only likely way for Intelsat to repay its debt outside of a restructuring is through a successful IPO. But the IPO process is itself hampered by the company’s debt load. Additionally, Intelsat has shown very little growth over the past few years. And it just reported disappointing Q4 results that suggest growth, might be slowing even further. The impact of softening government demand and challenges in Africa are likely, in my view, to get worse over time – not to mention the industry price-risk associated with new HTS satellites coming on line. In the best case, the needed IPO is going to be challenging.

Its unlikely Phil Spector would step-down voluntarily if he felt the company was on target for a successful IPO. According to my reading of Intelsat’s 2012 10K, he has an incentive package including vested shares and options it valued at over $22m. A successful IPO would likely increase this value significantly. But Intelsat can repurchase those shares at fair market value as of his departure. This, as I understand it, would deprive him of any upside from the IPO unless it happens within six months. If an IPO were likely, he would probably have wanted to stay on to ensure recognizing that upside.

But its also unlikely Intelsat would push out a credible long-term General Counsel just prior to launching a challenging IPO, absent egregious behavior on his part. It reduces the perception of depth and stability of the management team. And if he had been involved in egregious behavior, its hard to believe they would have offered him a conciliation seat on the board.

I’ve never met the new General Counsel and Chief Administrative Officer, Michelle Bryan. But I hear she is quite competent and hope to meet her someday. However, if I were going to hire a General Council prior to a tricky IPO, I would do a search for one with recent IPO experience, not someone who has been running HR for six years and whose prior GC experience is notable for work with distressed companies in the bankruptcy/restructuring process. A 2003 US Airways press announcing her departure, indicates she spent her last year, “overseeing the legal and government affairs responsibilities associated with the airline’s successful application for a federal loan guarantee and its restructuring under the U.S. bankruptcy code” and was lauded for “keeping the [bankruptcy] process on track and successful.” According to Wikipedia, US Airways was one of the first major airlines to terminate its pilot’s pension program to cut costs. After leaving US Airways, Ms. Bryan spent nine months at Laidlaw, where, according to an Intelsat press release upon her hiring, she addressed corporate governance and compensation issues – this was during Laidlaw’s recovery from bankruptcy (late 2003/early 2004) when such issues are common. No mention was made of any involvement in Laidlaw’s NYSE listing in 2004. If Intelsat is headed for a restructuring, her skills seem perfectly aligned for the job. Simply put, when highly levered company swaps a well-regarded member of senior management, for no apparent reason, in favor of one with seemingly notable restructuring experience, it is not a comforting signal to potential equity investors.

There is also the curious issue of the management promotions shortly after reporting disappointing year-end results and publically downsizing the IPO target from $1.75B to between $750M to $1B (the company needs to maintain at least a public façade of a planned IPO or its credit ratings will be in jeopardy). Dave McGlade was promoted to Chairman, Michael McDonnell gets the added responsibility for business development, Michelle Bryan as previously mentioned, was promoted to General Counsel and Chief Administrative Officer. Additionally, Thierry Guillemin was promoted to Executive Vice President and Chief Technical Officer. Mr. Guillemin may be a great executive (I don’t think I’ve met him) and he may be totally blameless for last month’s Intelsat 27 launch failure (which may have even unexpectedly helped the company) and deserving of a promotion. But, from a PR perspective, a huge technical failure is curious backdrop against which to promote the top technical person. If this is the way the board handles disappointing results – management promotions – it is, in my view, going to make investors additionally skeptical about participating in the IPO.

But I doubt the savvy private equity firms controlling Intelsat are this seemingly daft and/or tone deaf to investor concerns. A more plausible scenario, in my view, is that after the Q4 numbers and the Intelsat 27 failure, it became increasingly apparent to senior management that the IPO wasn’t going to happen. A restructuring was increasingly likely and the management team’s equity incentive packages were now largely worthless. To keep the team at Intelsat through the process, however, the board may have seen fit to give them promotions to burnish their resumes so they can get better jobs when its over. This may have been more palatable than giving them additional cash compensation. But at age 62 and with a stellar resume, this would likely be less appealing to Phil Spector. By splitting his titles (GC and head of Bus Dev) they could offer two other promotions. In the process, they get a GC with the HR and Legal skills needed in the event of a restructuring.

This may or may not be what’s actually happening – I don’t know. But the recent actions, in my view, are highly consistent with this scenario. Some industry observers attribute recent management changes to internal politics. It’s certainly possible that management priorities put internal politics above giving the critically needed IPO the best possible chance, and that the board was willing to go along with them. To be clear, I have no knowledge that this was the case. But if it was, that’s another reason investors might be cautious about investing in the IPO. Either way you look at it, the recent Intelsat management moves do not seem to bode well for an IPO in the near future. I hope there is another, more innocuous alternative that I am missing.

UPDATE: 3/14/13 @9:50AM —  A few people have pointed out that Intelsat picked probably the worst possible time to make these awkward announcements — a week before the Satellite 2013 Conference. This almost assures they will receive the maximum possible industry and press scrutiny during the conference. I honestly can’t think of a logical reason for this timing.

Could Incumbents Wireless Cos Dominate Unlicensed Spectrum?

Posted on: February 25th, 2013 by Armand Musey No Comments

Significant amounts of government spectrum are headed towards being made available for shared unlicensed use. However, industry commentators have barely analyzed the potential impact on the industry structure from the availability of this new spectrum. The common initial reaction, and one suggested in the PCAST Report is that new entrepreneurial unlicensed providers will emerge to create a vibrant new group companies offering new services and better prices. This view presumably implies new unlicensed service providers may put significant competitive pressure existing licensed service providers as they lose their monopoly/oligopoly status their spectrum licenses afford them. While, this may happen, its also possible that the impact of the new spectrum may not be so radical. The existing licensed providers may simple leverage their competitive advantages to dominate unlicensed spectrum as well.

A major obstacle for new unlicensed service providers seeking to create significant industry positions is obtaining the financing needed for the infrastructure buildouts their services require. This is particularly true if they can’t find credible barriers to entry in order to prevent rapid commoditization of their services. But even in an optimal scenario, it unlikely individual small service providers will be able to finance major regional or national buildouts on multiple spectrum bands.

As mentioned in a prior post on unlicensed spectrum, hardware providers with leading technology, whose equipment all services providers will need, are probably in the strongest position within the current unlicensed spectrum ecosystem to get financing and maintain strong margins. Emerging service providers will undoubtedly push hardware providers for vendor financing and many are likely to provide valuable assistance. But it’s unlikely hardware providers will be willing or able to provide enough financing to jump-start massive buildouts by small unlicensed service providers and also fund the service providers’ negative operating cash-flows until breakeven.

Even potential new service providers, such as DISH, with its significant financing ability and strategic advantages including a 14 million-subscriber base, are hesitant to undertake the cost and challenges enter the already competitive US wireless market. It’s unlikely that lowering (or eliminating ) the price DISH paid for its AWS-4 spectrum would motivate them to change their mind and aggressively enter the market. This calculation for small startup service provider without DISH’s advantages is likely to be even less favorable.

Existing wireless major wireless service companies and/or tower operators, however, are well positioned to undertake these buildout projects. Both of these groups have significant financing ability and relatively low costs of capital that emerging unlicensed providers lack. But they also have other advantages and motivation that may make undertaking a significant wireless buildout a more compelling and viable option than for any new entrant.

Existing wireless companies have long complained of a spectrum shortage and they already have built out most of the country on multiple bands. Expanding their current coverage to encompass additional unlicensed spectrum bands is likely to be significantly less expensive for them than for any new entrant. It may even be cheaper than their current cell splitting plans on licensed spectrum.  Moreover, unlicensed buildouts could allow them to offer (and charge) their existing subscribers significant additional data capacity. Existing service providers would also be able to leverage existing relationships and purchasing power with critical equipment manufacturers and other service providers.

Tower operators have one of the most expensive pieces of the wireless ecosystem – tower space. Many of their towers have additional capacity that they could potentially use to accommodate additional antennas and other equipment for unlicensed use. If a network needs additional towers, they are in the best position to build them and to find other the tenants to offset the cost. If the tower companies provide the service on a wholesale basis to multiple service providers, perhaps for peak demand use to supplement their licensed spectrum, it could allow them to side-step conflicts of interest that might undermine its existing tower leasing business and allow them stay closer to their core competencies.

Wireless service providers and tower operators may work together in building out infrastructure on unlicensed spectrum. In fact, they face challenges that may require them to partner on some projects. The tower operators are not experts in service provision, nor will they be eager to offend their current licensed tenants. The wireless companies will likely need lease concessions from tower operators to justify the project. In many case the tower operators may have RF engineering data on various sites that a wireless service provider does not have outside their current bands.

The potential opening of significant amounts of unlicensed spectrum is a watershed moment in telecom industry. The history of its unfolding has yet to be written. However the current providers superior financing, economies of scale and customer relationships may yet leave them with sufficient barriers to largely neutralize unlicensed competition in this area, even without their spectrum oligopolies. This may leave new unlicensed spectrum service providers at the fringes of the industry or in narrow verticals in which the major operators don’t participate.

Why the Intelsat 27 Failure May Be Good News for Intelsat

Posted on: February 5th, 2013 by Armand Musey No Comments

The launch failure of Intelsat 27 (“IS 27”) as it tries to show growth before a planned IPO was widely viewed as terrible news for Intelsat. However, it may be the opposite.  Of course Intelsat will lose some expected revenue.  But the launch failure also gives Intelsat insurance money that may be more valuable than the incremental revenue of an in-orbit IS 27. Like many satellite ventures, IS 27 may have seemed like a good investment when it was planned, only to look less attractive years later as service commencement approached with the company in an increasingly precarious financial situation. The insurance policy gives Intelsat a chance to unwind that investment and put it to more pressing needs.

Intelsat needed the revenue from IS 27 to enhance a growth story to support a long planned IPO. However, the reality is that Intelsat is a mature company that does not earn enough return on assets to cover even the cost of its debt.  A successfully launched IS 27 may have increased growth modestly from its 51-satellite fleet. But these would likely be offset, in the near term, from declines in government revenue as the wars in Afghanistan and Iraq wind down. Additional short-term pressure comes from increased competition in Africa from both space and terrestrial players. In the longer-term, increased capacity from HTS satellites flooding the market with massive amounts of capacity would likely limit Intelsat’s growth. Given that the company currently has a debt/EBITDA ratio of 8.3x, possibly exceeding its value, it’s hard to see how there is much if, any equity, to sell in an IPO.

The Intelsat IPO story was likely based on a highly optimistic scenario. It also gives debt investors a somewhat brighter glimmer of hope for repayment of its subordinated debt outside of a restructuring. This hope may help the company, at the margin, issue debt on slightly better terms. In our view, the real focus of Intelsat management is to manage its difficult debt situation. The launch failure may have provided a bit of an opening.

Intelsat’s February 1st press release indicates IS 27 was fully insured. Various trade press publications including and and pegged the insurance value at around $400 million. This seems high to us, compared to the typical $250 million cost of a standard C/Ku-band satellite and an estimated additional $50 million for the UHF load. Perhaps there are other factors we don’t know that added another $100 million. And the full insurance amount might not be available to the company. Intelsat has acknowledged taking customer prepayments and may owe certain customers a fraction of the insurance to refund pre-paid capacity on IS 27. The company also has a number of partially damaged satellites that it can use to accommodate some customers. However, the extent of this capacity is, admittedly not clear to us from outside the company.

Intelsat itself is likely worth approximately 7x-9x EBITDA. So unless Intelsat will lose more than about $35 to $45 million in EBITDA annually (after juggling some customers to other satellites and losing others it cannot) from not having IS 27, it is probably better off pocketing the ~400m insurance money and using it to pay down debt. That would be more valuable than having a functional IS 27 in orbit. Based on the increased competition in Latin America and the dismal outlook IS 27’s UHF payload, this might very well be the case.

One downside from not replacing IS 27 is that Intelsat’s auditors might require them to write down the value of the 55.5 deg west orbital slot. Currently, Intelsat lists $2.388 billion in orbital slots on its balance sheet (perhaps a topic for a future post). While writing down the value of one or more orbital slots would have no direct financial impact, it would put yet another, dent in the company’s IPO story. Even if replacing the satellite is a close call, the chance to pay down debt and reduce the company’s risk is likely compelling. Don’t hold your breath waiting for an IS 27R announcement.  Even if there is one in name, it probably won’t look much like the original.


Super Wi-Fi Presentation – January 2013

Posted on: February 3rd, 2013 by Armand Musey No Comments

I participated in a panel at the Super Wi-Fi Summit in Miami on January 31st titled, “The Financial Community and Shared Spectrum: Looking for Common Ground.” Other panelists included: Bill Evenson of CIT; Chuck Walters of Walters Associates; and Thomas Carroccio of the Law Offices of A. Thomas Carroccio PLLC (Moderator). My slides from this investment panel are attached.

One of the issues that resonated was how spectrum valuation might change if an when shared spectrum use becomes more prevalent. One of the key aspects to spectrum licenses is that they provides the licensee with a monopoly over the spectrum (or an oligopoly over spectrum within a range) and the monopoly profits that go with it. As unlicensed entrants enter the may potentially erode these monopoly profits. This may partially explain the reluctance of licensees to embrace any type of shared spectrum use – even if it is secondary to theirs on a non-interference basis. While this is not a significant valuation issue for licensed spectrum today, it may become significant if the spectrum sharing movement grows.

Unlicensed Spectrum and the Challenge of Finding Investment Capital

Posted on: January 27th, 2013 by Armand Musey No Comments

White spaces use is becoming increasingly visible way to use spectrum more efficiently and jumpstart the use of new services. However, the industry will need significant investment in order to build a substantial industry. Traditional telecom service providers have FCC spectrum licenses that can serve as collateral for investors. White space service providers don’t have these licenses, making the investment case somewhat different and, in some respects, a bit more challenging.

The issue of not having licenses has two impacts on investors. First, the white space service providers do not have the valuable license to help collateralize an investment. Without the licenses, loans to white space service providers would, in general, be more risky. Without other assets or proven cash flows, loans may be nearly impossible to get. On the other hand, users of unlicensed spectrum don’t need to bid for, buy, lease or a license either, potentially reducing their capital needs. The most significant challenge in finding equity investment is that a lack of license reduces barriers to entry for competitors in the industry. If a white space operator is successful, it’s easier for competitors to enter their market and lower prices.

Lots of successful businesses operate with low barriers to entry: restaurants, systems integrators, many websites etc. But most of those businesses can be started on a shoestring with minimal investment from friends and family, and then find growth investment once profitable. White space services companies, however, will often need considerable capital at an early stage to build out their networks, but customer equipment etc.

Normally such early stage money would come from venture capital firms. But venture capital firms are typically looking for companies that will give them a potential home run, not merely a reasonable return. They generally expect many, if not most of their portfolio companies to fail. They hope for a few investments that will provide gigantic returns to make-up for the losses in the others. This is why they often invest in companies with legal protection from competition such as companies with biotech patents, technology patents or licenses. With low barriers to entry, it will be hard for white space operators to make a case they can deliver potential for home runs. The other major sources of equity financing, private equity firms, are not as focused on such home runs. But they typically invest in later stage companies that have strong cash flows they can borrow against. They can get a strong return by leveraging their investment. Unfortunately, white space startups are rarely in this position either.

Given the lack of licenses or cash flow, white space startups will likely have to focus on other creative ways to develop barriers to entry. For example, they could try to get long-term customer contracts, exclusive relationships with key distribution partners, or even use proprietary equipment standards. The players in the industry who are best positioned to raise institutional funds are the equipment provides who have valuable proprietary technology. Service providers would be well advised to seek alliances with trusted equipment providers to assist with their financing. Of course, in the white spaces industry, many of the equipment providers are seeking funding themselves. It won’t be easy, but creative solutions are possible.

The Shifting Balance Between Infrastructure vs. Service Providers Now Favoring Infrastructure

Posted on: January 17th, 2013 by Armand Musey No Comments

In the wireless industry hardware providers seem to be gaining the upper hand over the wireless service providers. Examining the wireless value chain, its clear that hardware and spectrum are the only areas with increased demand. The wireless service infrastructure consists of 1) spectrum; 2) network hardware; 3) network integration and management; 4) marketing, billing and customer care; 5) distribution; and 6) handset manufacturing.

Wireless operators perform network integration and management (#3) and marketing, billing and customer care (#4) while they share distribution (#5) with consumer electronic outlets. Network integration and management has been commoditized for a long time. Customer growth has stagnated, putting marketing, billing, customer care (#4) and distribution (#5) in commodity positions as well. While the iPhone briefly pulled handset manufacturing out of a commodity position, declining demand and the rise of Android systems, suggest it is slipping back to a commodity position again.

The net result is that the benefits of increased wireless minutes are accruing to spectrum (#1) and network hardware (#2). The two are obviously complementary –as demand grows, wireless operators have a choice of spending more on increasing spectral efficiency by purchasing hardware to divide cell sites and/or invest in technologies such as wi-fi offloading, or buying additional spectrum.

Attempting to improve their negotiating position with network hardware manufacturers is one reason for the wireless consolidation – particularly cross-border consolidation that yields few other operating benefits besides increased purchasing power with network hardware manufacturers. For example, Sprint’s potential acquisition of Clearwire is likely, in part motivated to help its parent Softbank get equipment manufacturers to support the TD-LTE standard Softbank uses.

The improving position of hardware manufacturers is probably not limited to the wireless industry. Its likely one reason Google is investing in fiber networks, a reason behind the recent discussion of the European single telecom network. Even in the satellite industry, the most disruptive force is coming from the new Ka-band spot-beam satellite that are allowing new entrants such as O3B, NewSat, ViaSat and other to enter and threaten existing operators with older legacy equipment. As spectrum demands continues to grow faster than supply, network equipment companies capable of offering the most efficient systems are likely to take an increasing portion of the value chain.

Technology Won’t Decide the Ka-band vs. Ku-band HTS Battle

Posted on: January 6th, 2013 by Armand Musey No Comments

Technology Won’t Decide the Ka-band vs. Ku-band Battle for HTS Satellites

The current debate about the advantages of Ka-band vs. Ku-band for next generation High Throughput Satellites (HTS) misses the point. Endless pages have been written about Ka-band advantages of higher theoretical throughput vs. Ku-bands superiority in sub-optimal weather vs. the costs of end user equipment. Focusing on these technical issues misses the larger point of the strategic issues that will likely drive marketplace acceptance.

The old adage “satellite wars are won on the ground” is likely to be true with Ka vs. Ku-band HTS. While it’s possible that either Ka-band or Ku-band will be a technical disaster making it commercially unviable, the odds that a large portion of engineers missed this is highly unlikely. While they will have differences, both are likely to ultimately work reasonably similarly from a customer’s perspective. Rather, the Ka-band vs. Ku-band is likely to follow the path of previous battles over technical approaches, ranging from the battle over railroad track widths (gauge) in the mid 1800s to VHS vs. Betamax of the 1980s and the more recent GSM vs. CDMA battles. In each of these, the battle for the market had nothing to do with the technical advantages of either approach. Rather the winners were based on decision by those who controlled enough of the market to force it on the rest.

In the case of the railroads, there were over a dozen different standards for railroad widths before 1860. It wasn’t until the Eastern railroads using 4’ 8 ½” gauge rails convinced the new railroads in the west to adopt the same standard that the others were put in such a minority position, they finally had to undergo the expensive conversion process or retrofitting their tracks.

With Betamax vs. VHS, most people agreed that Betamax offered superior picture quality. But JVC (who controlled VHS), quickly licensed its technology to a multitude of manufacturers, driving prices lower and creating competitive innovation. The death knell for Betamax was the greater availability of Hollywood movies on VHS.

The European standardization on GSM has quickly built demand for hardware using that technology. Hardware manufactures spent tens of billions of dollars developing and improving it. Despite the success of CDMA in the US, the use of GSM as the defacto European standard has given it such a head start, that most other countries have simply copied it.

With HTS, there is no new market to connect to as with the western expansion of the U.S. railroads, no content issues as with VHS vs. Betamax and no regulatory commitment as with GSM vs. CDMA. However, support from the hardware manufacturers and sales channels is critical. In these areas, Ka-band has the advantage. Ku-band HTS is largely an Intelsat approach (along with a few much smaller players). While the Intelsat’s EPIC satellites will have C-band and Ka-and coverage, they seem to be emphasizing the Ku-band capacity, leaving them the odd man out in the HTS world. Intelsat is the world’s largest satellite operator, giving it significant influence in the industry. But its financial situation undercuts this influence. It will be a challenge to convince manufacturers and sales channel partners to make significant long-term commitments to a technology from a company that is in the financial distress Intelsat faces.

A full analysis of Intelsat’s financial position is beyond the scope of this posting. However, Intelsat’s riskiest bonds are rated CCC+ by S&P (Caa2 by Moody’s). This rating places the debt well within the “high yield” (junk) category, four levels below investment grade. S&P defines the CCC rating as: “Currently vulnerable and dependent on favorable business, financial and economic conditions to meet financial commitments.” Even Intelsat’s safer bonds are rated B by S&P (B3 by Moody’s), three levels below investment grade. S&P defines the B rating as: “More vulnerable to adverse business, financial and economic conditions but currently has the capacity to meet financial commitments.” Should Intelsat be forced to restructure, they will likely find it hard to continue to spend large amounts of money developing a new HTS market.

Ka-band HTS solutions include the SES-backed O3B, Viasat’s Exede and Hughes Network Systems/DISH’s Jupiter platform and Inmarsat’s Global Express, to name a few. This is a powerful collection of players. While the hardware for each will not be completely interchangeable, many of the components and engineering analysis will be complementary. If a hardware manufacturer needs decide between Ka-band development with several possible major customers or Ku-band development with one potential major customer that has financial challenges and a few smaller customers such as Thaicom, the choice seems simple.

Intelsat, however, has made a clever move to attempt bridge this potential problem with influencing distribution channels. It has secured prepayments from resellers, largely to fund construction of its new EPIC satellites. This has the resellers significantly invested in the outcome and less willing to quickly switch to other approaches should Ku-band get off to a slow start. Harris CapRock, KVH, MTN and Panasonic Avionics  – four important resellers and manufacturers – seem to support Intelsat’s Ku-band approach. Of course, Intelsat also benefits from its large Ku-band customer base. Its Ku-band HTS approach may help them convert them to HTS more economically. But Intelsat will need to be careful about aggressively converting existing customers to a new system that will undoubtedly have much lower pricing on a per MHz basis.

It remains to be seen if Intelsat’s moves will be sufficient to overcome the collective power of the other market participants. Another possibility is that two standards thrive side-by-side as the PC and Mac have done. However, history suggests this rarely happens.

ITAR Restrictions on US Satellite Exports to be Loosened

Posted on: December 28th, 2012 by Armand Musey No Comments

Under the 2013 Defense Authorization bill approved by both houses and awaiting President Obama’s expected signature, the highly controversial restrictions on US satellite exports under the International Traffic and Arms Regulation (known as ITAR) are set to be loosened. Specifically, the bill allows the president to remove commercial satellites and satellite components from the U.S. munitions list while allowing the president to keep certain technologies on the list. Exports and technology transfer to China, Cuba, Iran, North Koran and Syria would still be prohibited.

The ITAR restrictions on satellite technology, created significant hindrances for U.S. Satellite manufacturers (as well as others, particularly in the aerospace industry). Marketing meetings with foreign customers required, among other things, government chaperones and background checks on all attendees. Manufactures were also highly limited in their ability to describe their products to these customers. However satellite sales involves selling a high-priced, high technology product, often exceeding $100 million each and each satellite requires a great deal of customization for the end-customer. The bureaucracy behind ITAR was enormous. An executive of a satellite launch company once said, in response to my question on a panel at a conference, that they had 15 full-time people dealing with ITAR issues! That does not include the outside attorneys for both sides, the state department staff and others. Customers found the inability to freely share information frustrating and, in many cases took their business to manufacturers in countries without these restrictions.

The ironic of the situation is that its hard to see how the ITAR restrictions on commercial communications satellites do anything to prevent information from getting to other countries. Satellite customers in foreign countries simple started doing business with other (mostly European) satellite manufacture who are able to freely share that very same information! In the seven years after the ITAR restrictions, the Center for Strategic International Studies (CSIS), estimates the US share of the satellite manufacturing industry fell from 73% to 25%. Based on my industry experience during this time, this number seems about right. ITAR has also hurt satellite component manufacturers. Some foreign manufactures have gone out of their way to build “ITAR free” satellites – ones that don’t use any US components on the U.S. munitions list and this avoid any ITAR issues. The net effect was that the US dominance in this strategically important industry was destroyed. It created an opportunity for other countries to catch-up in this field – and they did.

The challenge for policy makers is not only to figure out how such a senseless idea came about, but also to understand why it took so long to fix and how to prevent this from happening again in the satellite industry or elsewhere. Why has the Department of State continuously stressed the security importance of these rules even though they seem logically absurd? Why did Congress, for so many years, accept these explanations despite having a large number of well-paid and articulate satellite industry executives and lobbyists repeatedly explaining this absurdity to them and their staffs? If it takes 14 years for the satellite industry to correct this problem, it’s scary to think about the problems this kind of political leadership could have on an group or an industry with fewer resources.

Takeaways from Reverse Incentive Auction Webinar

Posted on: December 20th, 2012 by Armand Musey No Comments

Takeaways from Reverse Incentive Auction Webinar

The reverse incentive auction webinar took place at Pillsbury Winthrup’s Washington DC offices yesterday December 19th.  A replay is available at  (note: paid registration required, CLE credits may be available). In addition to the opening remarks, there were three panels – Legal, Valuation and Policy. The key takeaways from each section are:

I. Opening Remarks –

(Harold Furchtgott-Roth, former FCC Commissioner, Introduced by Barlow Keener, Keener Law Group)

1)    Noted that increasing spectrum available for its highest use, mobile broadband was an important component to increasing economic growth.

2)    Noted that the wireless industry was one of the fastest growing segments of the economy.

3)    Indicated that the current spectrum inflexibility was costing consumers hundreds of billions of dollars annually. This was particularly important given the current budgetary concerns and the lack of growth in the economy.

II. Legal Panel –

(moderated by Marty Stern of K & L Gates. Rebecca Hansen from the FCC provided an overview of the auction. John Hane, Pillsbury Winthrup and Brian Madden, Lerman Senter offered legal comments.)

1)    The FCC is officially deciding between a sealed bid process whereby participants indicate the price at which they will sell their spectrum and a descending bid auction, whereby the FCC offers a price to broadcasters in a given market and then lowers it until only the number of acceptances they would like remain. However, most observes believe the FCC is headed towards the descending bid auction.

2)    FCC will score bids based on a unit price per market and on how it impacts repacking and then decide which ones to accept.

3)    Many details about the number of options broadcasters will have are not settled.

4)    Brian Madden stated that broadcasters need to think differently about the value of their businesses, because in the reverse auction process, the value will not be based on traditional broadcaster valuation metrics, but rather the value to a different user.

5)    John Hane raised concerns about many details including how the repacking, which may change the frequencies for many broadcasters will impact their coverage areas, how interference issues result and how they might be compensated for losses. He suggested that the FCC take time between rounds to evaluate the potential impact of repacking.

III. Valuation Panel –

(moderated by J. Armand Musey, Summit Ridge Group, LLC. Panelists included three distinguished economists: Prof. Thomas Hazlett, George Mason Law School; Coleman Bazelon, The Brattle Group; and Mark Fratrick, BIA/Kelsey.)

1)    Prof. Hazlett discussed the history and process that lead to the auction.

2)    Coleman Bazelon indicated that FCC will only need to conduct the auction in the top 30 markets plus some adjacent markets to reach their goes of freeing 120 MHz nationwide. FCC bidding likely to result in prices in well in excess of their enterprise value, perhaps as much as 3x-4x.

3)    Mark Fratick said that in recent months investors have purchased smaller broadcasters at premiums of 40%-50% above historical prices.

4)    Panel agreed that some broadcasters, by virtue of their strategic position, such as overlapping major markets may be particularly important for the FCC to acquire and may achieve particularly high prices. However, it is difficult for a broadcaster to understand if they are in this position or a much weaker one. This requires significant technical, economic and legal analysis.

5)    Panelists agreed that recent increased allocations and proposed allocations of wireless broadband spectrum would not reduce spectrum prices in the forward auction relative to prior auctions. Increased demand for wireless broadband has more than offset those spectrums increases, even after accounting for new technology that is increasing spectrum utilization efficiency.

IV. Policy Panel –

(moderated by Barlow Keener of Keener Law. Panelists included Prof. Preston Marshall, Univ. of S. California; Fred Campbell, CLIP; Trey Hanbury, Hogan Lovells.)

1)    Panelists indicated that there might be additional auctions in the future and that this may not be their only exist opportunity.

2)    Panelists agreed that it is in the broadcasters’ interest to invest in understanding this auction process.

3)    Broadcasters may be surprised with the value they could achieve and may want to take advantage of it. If not, such information could better prepare them in the event of a future auction.

4)    Prof. Marshall noted that future developments in cognitive radio may allow broadcasters to co-use spectrum with television broadcasters and avoiding the need for future auctions. Other panelists noted that such technology is not currently commercially viable.

5)    Panelists agreed that the auction rules should be kept simple to maximize the number of participants on both sides. If the rules are too complex, it creates uncertainly and people instinctively avoid uncertainty.

Discount Code for Reverse Incentive Auction Webinar

Posted on: December 16th, 2012 by Armand Musey No Comments

A previous post (see: describes an upcoming 3-hour webinar on the reverse incentive auction. If you enter the code: AMWEB, you will get a 36% discount on the $150 price, bringing it to $95.

The webinar is about the FCC TV Incentive Auction (Spectrum Act 2012). We have Coleman Bazelon, The Brattle Group; Barlow Keener, Keener Law; Mark Fratrik, BIA / Kelsey; and Rebecca Hanson, FCC among others.  Former FCC Commissioner Harold Furchtgott-Gott is introducing the webinar. Use the following link for a full description:

Reverse Incentive Auction Webinar – December 19th

Posted on: December 11th, 2012 by Armand Musey No Comments

I’ll be moderating a valuation panel discussion as part of a three hour webinar on the FCC’s Reverse Incentive Auction process. Here is a link to the site describing the event: :

Incentive Auctions: What Broadcasters Need to Know

  • A 3 Hr. a Video Virtual Event (Webinar).
  • December 19th starting at 8:45 AM EST

The discussion will focus on challenges facing the television broadcasters as they decide whether to participate in the auction and how to calibrate their bids for optimal results. Participants will discuss the requirements in the Spectrum Act, the FCC’s recent Notice of Rulemaking Proposal (NPRM) as well as unresolved legal issues. Other panels will include a discussion of valuation and policy issues facing broadcasters.

As a virtual event there is a charge of $150 to attend, however government employees (this means you Congressional staff, FCC, NTIA etc.) should email me for free or per diem rates as needed.  3 hours of CLE credit are also available.


8:45 AM – 9:00 AM Introduction – Opening Remarks:

  • Harold Furchtgott-Roth, Former FCC Commissioner (Opening Remarks)
  • Barlow Keener, Keener Law Group (Introduction)

9:00 AM – 10:00 AM        Legal Perspective:

The Rules of the Auction Process

The opening panel will provide an overview of critical elements of the broadcast incentive auction as well as unsettled issues in the FCC’s recent NPRM. It will compare the upcoming reverse incentive auction with prior FCC auction processes. The panel will also discuss legal appeal process for the reverse incentive auction and potential pitfalls auction participants should consider.

  • Rebecca Hanson                 FCC
  • John Hane                           Pillsbury Law
  • Brian Madden                     Lerman Senter
  • Marty Stern                          K&L Gates

10:00 AM -11:00 AM        Valuation:
Valuing The Reverse Auction Assets and Related Business Issues

The second panel will build on the legal explanation of the first panel. It will consider how the auction process and other factors will impact auction valuations and broadcasters decisions to participate. Panel topics will include general auction valuation expectations, variations in spectrum value by type of broadcaster and geography, and valuation considerations for broadcasters to consider when participating in the auction.

  • Coleman Bazelon                Brattle Group
  • Mark Fratrik                         BIA / Kelsey
  • Major Investment Bank      (Invited)
  • J. Armand Musey               Summit Ridge Group


11:00 AM – 12:00 PM      Policy Perspective:

Important Social, Political and Economic Policy Implications

Our final panel will step beyond strict legal and economic considerations and evaluate the FCC’s goals versus those of the television broadcast industry. The panel will discuss issues including implications for news access in rural areas, implications for other FCC license holders, and other policy concerns.

  • Preston Marshall                Univ. of S. California
  • Fred Campbell                     CLIP
  • Trey Hanbury                      Hogan Lovells
  • Barlow Keener                    Keener Law Group

As you can see, we have gathered leading experts in the legal, financial and policy files to discuss the implications of the FCC’s upcoming television broadcast incentive auction on the television broadcasting industry.

I am looking forward to your participation.

J. Armand Musey, CFA

A Layman’s Guide to the Possible Intelsat IPO

Posted on: November 29th, 2012 by Armand Musey No Comments

Given all of the industry questions about a possible Intelsat IPO, I decided to spend some time an analyze the situation. This resulted in a 26 page analysis (e-mail me at for a copy. Here is the executive summary:

Executive Summary

In light of its May 18th $1.75 billion IPO filing, satellite industry observers are actively discussing the possibility of an Intelsat IPO. However the rumors and facts have often been difficult to disentangle. This report will attempt to shed light on the facts for forward the discussion.

Serafina, an investment vehicle controlled by several investment firms including BC partners, completed their acquisition of Intelsat in February 2008. Typically, a private equity fund has a five to seven year life and thus times its investments exits at the lower end of that range. February 2013 will be the five-year mark for Serafina’s investment; it’s natural to wonder about their exit strategy. It’s not uncommon for PE firms to use an IPO as an exit strategy and many intuitively see Intelsat’s IPO filing as the likely exit strategy. But there is much more to the story.

Intelsat lacks the growth characteristics of a typical IPO candidate. And its current debt load makes any equity value very risky. Simply put, high risk and low growth is not a compelling investment opportunity. While Intelsat could potentially attempt to go public based on a high dividend, much like a utility company, its high debt would make the dividend highly risky. Other FSS companies with lower debt are offering dividends near 4% annually, meaning Intelsat would have to offer significantly higher dividends of offset the higher risk. But higher dividends would not counter the company’s need to lower its fixed payments on debt and other capital.

Rather, the purpose of an Intelsat IPO would likely be to reduce its debt burden and create a vehicle to allow it to eventually refinance or convert other debt securities into equity-related securities. Intelsat would still be a low growth company, but if it could make a compelling case for continued moderate growth, and an ability to manage its post-IPO debt, the leverage could increase the return to the stock, making it a very reasonable IPO candidate. But a number of things would have to go right for the company. An IPO would create a public vehicle that could ultimately facilitate an exit for Serafina, but is unlikely to create an exit opportunity for in the near-term.

Update: April 10, 2013: We’ve decided to release our Summit-Ridge-Complete-Intelsat

Implications of Expected FCC Limits on DISH’s AWS-4 Spectrum

Posted on: November 21st, 2012 by Armand Musey No Comments


The FCC recently indicated it was leaning towards limiting DISH’s use of its 40 Mhz of AWS-4 spectrum it acquired from Terrestar and DBSD North America to prevent interference with the H-block which runs from 1915 MHz to 1920 Mhz and 1995 MHz to 2000 MHz. DISH’s spectrum currently runs between 2000 Mhz and 2020 MHz on the uplink.

The proposed order would shift the establishment of the AWS-4 spectrum up 5 MHz to 2005 to 2025 MHz to avoid interference with the adjacent H-block . An adverse FCC decision would further complicate DISH’s entry, not only by impairing 25% of its uplink spectrum, but also by forcing it to restart the slow approval process of its LTE network and possibly complicating its efforts to obtain hardware for the system. DISH chairman, Charlie Ergen indicated to the Wall Street Journal that this was a potential “Game Changer” that puts all options on the table including bringing in a partner or selling the spectrum entirely.


Its important to keep in mind that Charlie Ergen was a professional poker player before starting DISH. He knows how to take calculated risks, and it appears this is the game his is playing with the FCC.

The DISH Perspective

DISH is in a bind. It’s having difficulty competing against high-speed two-way cable systems and new entrants such as Verizon’s FiOS. Consequently, it has begun to slowly lose customers. At this point there are two primary options for EchoStar to stabilize and grow – 1) Partner with a telco company that could give it a two-way service and operating synergies (and to whom it could provide a video service); 2) start its own telecom service.

Both DISH and DIRECT have tried various alliances with wireline telcos in the past but none of them seem to have worked for a multitude of reasons. Moreover, the wireless companies do not seem to have expressed a strong interest in a partnership with DISH. This leaves DISH with the possibility to starting a new wireless company. DISH has acquired significant spectrum holdings and technology that position itself to do just that, but without making long-term commitments that are hard to unwind. This makes it hard to decipher if DISH actually intends to enter the wireless market or if its just threatening to enter so as to engage the major wireless providers to consider an alliance.

However the more the wireless industry consolidates, the harder it will be to acquire smaller players and other assets needed to build a major new entrant. Consolidation also allows the large players to achieve greater economies of scale and gain greater leverage over distribution channels and suppliers. This cements their dominant position and compounds the challenges of a new entrant such as DISH to enter the market. But its not clear DISH has another viable alternative.

The FCC Perspective

While, from the Wireless industry’s perspective, keeping DISH out of the industry would be a welcome relief against a significant competitive threat, the FCC undoubtedly sees it differently. The FCC probably views the risk that DISH does not enter the market is a bad omen for additional competition and new services. At the same time it wants to maximize the value of all spectrum, including the H-block.

The wireless industry is rapidly consolidating – As we’ve noted before, in the past 12 months the following major wireless industry transactions have reshaped the industry: Verizon’s purchase of Spectrum Co, signaling the probably end of the cable industry’s interest in entering the industry; T-mobile’s acquisition of MetroPCS and Sprint taking a majority stake in Clearwire. And there is significant market speculation about Leap wireless combining, possibly with AT&T. T-Mobile USA COO recently indicated that the “four major players could go down to three.” Additionally, many industry insiders are speculating about significant additional consolidation within the tier two players over the next year or so.

The FCC is very aware of the need for competition to keep prices down. In the wireless industry globally, one of the best predictors of low prices is the number players in the market. One of the best predictors of innovation is the presence of emerging players who often take risks and attempt do things differently. Moreover, as discussed in a previous posting wireless industry acquisitions potentially put pressure on the spectrum auction process and spectrum valuations, possibly jeopardizing the FCC’s ability to get appropriate bids from wireless operators in certain markets.

Given its resources, its spectrum assets and potential synergies, no entity is positioned to enter the industry and shake things up than DISH. Charlie Ergen started DISH on a shoestring, launching its first satellite in 1995 and competing against GM backed DIRECTV and the cable industry. This forced greater competition and an acceleration of the digital cable rollout. The FCC need DISH once again, but they also know his options are limited. The poker game continues.

Spectrum Property and Spectrum Valuation Presentations

Posted on: November 12th, 2012 by Armand Musey No Comments

We’ve just added two presentation to our website that might be of interest:

1) An Introductory Framework for Wireless Spectrum Valuation

2) Broadcasting Licenses: Ownership Rights and the Spectrum Rationalization Challenge: Why the Government is Buying Back its Own Spectrum.

The first presentation is what the title implies — It outlines a basic framework for analyzing wireless spectrum valuation. Its reviews basic methods of valuation as well as the advantages and disadvantages of each. The presentation also reviews special considerations in comparing spectrum that an appraiser needs to consider. Finally it analyzes issues specifically related to television broadcast and mobile wireless spectrum as they relate to the upcoming incentive auctions.

The second presentation is an analysis of the quasi-property rights held by television broadcasters and its impact on the FCC spectrum rationalization process including the planned incentive auctions. Analysis of the text of relevant legislation, legislative and judicial history as well as public policy implications. This presentation was prepared for a talk given at Columbia Law School in November 2012 and is based on an paper published in the Columbia Science and Technology Law Review (Spring 2012) which examines the showdown between television broadcasters and the government in light of the FCC’s plan to reallocate currently licensed broadcast spectrum to significantly higher value mobile broadband use. The government seeks to do so in an economically, socially and legally efficient manner, and has indicated that it seeks a reallocation via a voluntary reverse auction process. Nonetheless, any spectrum reallocation proceeding raises the question of whether, and to what extent, television broadcasters ultimately possess rights to licensed spectrum, and what type of compensation, if any, they would be owed if the FCC takes their spectrum licenses involuntarily.

Wireless Consolidation’s Impact on the Incentive Auctions

Posted on: November 5th, 2012 by Armand Musey No Comments

Since the discussion of the television broadcast incentive auctions began a few years ago, the US wireless industry has significantly consolidated. In the past 12 months the following major transactions have reshaped the industry: Verizon’s buys Spectrum Co, signaling the probably end of the cable industry’s interest in entering the industry; T-mobile buys MetroPCS and Sprint takes a majority stake in Clearwire. And there is significant market speculation about Leap wireless combining, possibly with AT&T. Each of these deals likely eliminates a major bidder in the incentive auctions and, to varying degrees, reduces the need for incremental spectrum for the combined entity. The impact on bidding in the upcoming incentive auction is unclear – its only one datapoint – but it points to potentially lower demand and/or auction prices than initially anticipated. There will undoubtedly be large demand for spectrum in the largest markets that greatly exceeds the price television broadcasters value the spectrum. However, in some of the smaller markets within the auction process (the auction will largely be confined to the major markets, although many more markets would be subject to repacking), this might not be the case. Its a complicated business and engineering issue. But potentially, it could complicate the FCC desire to reallocate 120 Mhz nationwide to mobile broadband.

Incentive Auction NPRM & Plight of Low Powered TV Stations

Posted on: October 25th, 2012 by Armand Musey No Comments

While there has been a lot of discussion about the FCC’s October 2nd Notice of Proposed Rulemaking (NPRM) for the incentive auctions, the plight of low powered television stations has been somewhat neglected. These stations hold secondary licenses and are allowed to operate on their frequency allocation subject to them not interfering with the primary licensee of the spectrum.

The NPRM indicates the FCC does not plan to allow the low powered stations to participate in the incentive auction. This was not surprising as their participate was not clearly authorized by the Spectrum Act that authorized the auctions. But the FCC indicated it does not intend to allow them to participating in the repacking of the remaining television spectrum.

This would seem to suggest that any low power station on spectrum where the primary licensee returns their license to the FCC in the auction, or if their spectrum is repacked so their current frequency is no longer used for television, they will no longer be able to broadcast. The FCC seems to intend to allow some “displaced” low power stations relocate, but notes, “Only a limited number of available channels may exist following the repacking process, limiting the relocation options available to displaced low power television and translator stations” (p.119).

The difference which which the Full power and Class A stations are being treated versus the low power stations highlights the political nature of the decision to compensate the broadcasters. In the case of the low powered stations the FCC is suggesting it may simply exercise its power to reallocate the spectrum without compensation. They had the same legal ability to do that with the full power and class A station, but given the political and pragmatic issues involved with pulling the broadcasters’ licenses, Congress stepped in, via the Spectrum Act, to ensure the broadcasters were compensated.

It remains to be seen if the FCC will stand by the policy. If it does, it will be interesting to see if it decreases investors’ willingness to invest in businesses using shared spectrum or secondary licenses.

Foreign Wireless Companies bBuying their US Counterparts….

Posted on: October 17th, 2012 by Armand Musey No Comments

The past few weeks we have seen two major wireless transactions announced.  First Deutsche Telecom backed  T-mobile announced an acquisition MetroPCS and late last week the Japan’s Softbank announced it was paying $20 billion for a 70% interest in Sprint.  Why does foreign firms find US wireless companies so attractive?  On first blush, it does not make a lot of sense as there are no obvious operational synergies between wireless companies in different countries with difference languages.

Part of the issue appears to be that its not the US companies that are so attractive, but rather that their home markets are less attractive.  The wireless markets in both Japan (and most of Asia) and Germany (as well as most of Europe) are more mature.  Moreover their economies appear to be stagnating. These companies appear to be desperately looking for growth and the US is one of the few wealthy countries with significant growth potential, on relative basis.

But the bigger question is why other US media and communications companies, who have potentially much greater synergies, are not stepping up to the plate in bidding for the wireless companies? The synergies between a major cable or DBS operator and a wireless company could be huge — a single bill, customer support, bundled packages etc.   Maybe such alliances are the next step for the foreign owners?

The only explanation I can think of right now is that the foreign buyer have lower cost of capital in their home countries and it will be easier for them to finance the LTE rollout and that this gave them an edge in the bidding. Comments welcome.

Super WiFi Summit Takeaways

Posted on: October 7th, 2012 by Armand Musey No Comments

I just returned from the “Super WiFi Summit” in Austin, TX (Oct 3-5).  The range of speakers – from major companies including Microsoft and Texas Instrument to start-up entrepreneurs and various policy experts (including Blaire Levin and Michael Calabrese as well as some international policymakers)  indicated a robust WiFi industry.  A general theme was ne of disappointment in the limited regulatory support for increased whitespace spectrum in the FCC’s NPRM for the Broadcast incentive auctions.  Several ideas were floated to ease the spectrum crunch by allowing increased spectrum flexibility, particularly via spectrum sharing.  There was a general belief that if the industry can show significant technical progress and market acceptance, the spectrum will eventually be made available.  But the fear was that without additional spectrum guarantees, it will be harder to garner additional investment to make those advances possible.

My general thoughts were that as spectrum sharing technology improves, the once unquestionable belief that service providers need exclusive use of spectrum to justify the investment in building out services on that spectrum will erode.  At the same time, it’s clear the FCC needs a more flexible framework to allow faster spectrum reallocation in an increasingly dynamic communications sector.  Spectrum sharing fits the bill — except that it does not generate revenue for the government at a time when budget deficits high and the U.S. Treasury is looking for revenue from all possible sources including spectrum auctions.  A few creative ideas were floated about paying for spectrum on an annual or even a per-use basis.

Second Paper on TV Broadcasters’ Property Rights in their FCC License Published

Posted on: September 26th, 2012 by Armand Musey No Comments

My second paper on television broadcasters’ potential property rights in their spectrum was recently published in the Columbia Science and Technology Law Review. This paper is a detailed legal analysis of the subject.  My prior paper on the subject, was a policy-oriented paper about potential equitable solutions to the dilemma of reallocating television spectrum.

Broadcasting Licenses: Ownership Rights and the Spectrum Rationalization Challenge

J. Armand Musey

Columbia Science & Technology Law Review, Vol. 13, 2012

This Article examines the showdown between television broadcasters and the government in light of the FCC’s plan to reallocate currently licensed broadcast spectrum to significantly higher value mobile broadband use. The government seeks to do so in an economically, socially and legally efficient manner, and has indicated that it seeks a reallocation process that is voluntary for broadcasters. Nonetheless, any spectrum reallocation proceeding raises the question of whether, and to what extent, television broadcasters ultimately possess rights to licensed spectrum, and what type of compensation, if any, they would be owed if the FCC takes their spectrum licenses involuntarily.

This Article finds that broadcasters have a very weak property rights claim over their spectrum licenses. However, broadcasters may be entitled to due process before their licenses can be taken involuntarily; they are almost certainly entitled to seek judicial review of any adverse FCC decisions. Such review would extend the already lengthy FCC rulemaking and adjudication process, and further delay spectrum reassignment. For practical political reasons, including maximizing revenue from future spectrum auctions, the most expedient way to reallocate spectrum is to incentivize the broadcasters to voluntarily participate in a reallocation plan by providing compensation beyond the legally required minimum.

Introductory Presentation on Summit Ridge Group, LLC

Posted on: August 1st, 2012 by Armand Musey No Comments

Summit Ridge Group, LLC’s new introductory presentation regarding the valuation and strategic advisory services we offer.

Summit Ridge Group, LLC
from Summit Ridge Group, LLC

Feedback and questions are always welcome.

Telecom Exchange 2012

Posted on: June 29th, 2012 by Armand Musey No Comments

I spent an interesting day at Telecom Exchange 2012 on Wednesday June 27th at the fabled Cipriani on Wall Street. While there were also plenty of major company announcements, I’ll leave commenting on those to the many qualified industry observers and instead focus on macro connections between industry segments. The conference was a step away from my traditional satellite/wireless/media background, but it was fascinating to see how much common ground exists between them.

The satellite industry often thinks of itself as a unique small speck of the communications industry under constant threat from fiber. As it turns out, revenue from the fiber network carriers is growing, globally, at about 5% annually. Traffic is growing faster, but declining pricing limits the revenue growth – about the exact same story as the satellite operators! A quick Google search shows that Insight Research recently predicted global telecommunication industry growth revenue to increase 5.3% annually over the next five years. Even in the booming wireless carrier market IBISWorld estimates only 7.8% annual growth over the past five years. Essentially, the communications industry appears to be remarkably evenly mature across segments.

Viewing maps of fiber being laid in new markets, by companies such as China Unicom, highlighted the competition between fiber and satellite as fiber reaches more and more remote outpost. However, most companies at the conference indicated that most carrier industry growth was coming from 1) Cloud computing applications (seemed to be on every company’s agenda) with various strategies and seeming no consensus on the right one; 2) Low latency networks for high frequency trading (companies such as Hudson Fiber Network) and 3) 4G backhaul (also on everyone’s agenda), all applications having very little impact on the satellite industry. This seems to be somewhat at odds with industry research suggesting rapid carrier growth in Asia, the Pacific Rim and Latin America driven by their growing middle classes. This discrepancy may be explained by the mix of companies to whom I spoke, most of whom operated primarily in the US. But this may also be somewhat analogous to the satellite industry’s counterintuitive experience of seeing most of its growth from new applications in the United States and Europe as opposed to developing regions with less existing infrastructure.

The structure of the conference was also remarkably similar to that of the satellite industry. The rows of booths of companies selling networking hardware is really no different than any satellite conference. Of course, there were no satellite teleports exhibiting, but plenty of telecom hub operators. There were no satellite component manufactures exhibiting traveling wave tubes and the like, but there were plenty of companies selling optical repeaters and similarly analogous equipment. The leading industry news sources including Datacenter Dynamics, Globalvision Magazine, and Telepresence Options were also there. I spoke to a representative from 451 Research and was impressed with the depth of the independent research available in this, much larger sector.

Television Broadcast Spectrum Reclamation Through Eminent Domain Paper Published

Posted on: June 6th, 2012 by Armand Musey No Comments

My paper “How the Traditional Property Rights Model Informs the Television Broadcasting Spectrum Rationalization Challenge” has been published in the Hastings Communications and Entertainment Law Journal – 34 Hastings Comm. & Ent. L.J. 145 (2012). While some of the issues have been overshadowed by recent incentive auction legislation, the conceptual framework is likely to be relevant for future spectrum reclamation processes.


This paper examines the prospective role of zoning rights and eminent domain in the FCC’s challenge of reallocating underutilized television broadcast spectrum for use in significantly higher value mobile broadband applications. The government must reallocate the spectrum in an economically and legally efficient manner, balancing the interests of the politically powerful broadcasters and those of society as a whole. Recently, the government has decided to explore ways to incentivize the broadcasters to voluntarily return their spectrum licenses. From a strictly legal perspective, the broadcasters have a relatively weak claim to property rights. However, the government, has indicated it seeks an incentivized voluntary return of spectrum by the broadcasters. The government’s plan to take the broadcasters’ spectrum voluntarily, rather than involuntarily, suggests that the government intends to effectively treat the broadcasters as though they have certain property rights to their spectrum licenses – specifically the right to control possession. Given the government’s de facto recognition of elements of property rights of the broadcasters, the twin theories of zoning and eminent domain can inform a mutually acceptable solution between the government and the broadcasters. These traditional property rights theories suggest that a payment to the broadcasters, some continued retransmission guarantees and cable or satellite subscription subsidies to affected viewers, totaling of $0.10 to $0.19 per Mhz/POP would likely to be fair to all parties. In addition, the threat of “down zoning” the television broadcasters’ rights incentivizes the broadcasters to accept an equitable offer put forth by the government. The government would be able to auction the spectrum for $1.02 to $1.22 per Mhz/POP a net profit of approximately $28.5 to $38.3 billion.