After patiently waiting for over three years to make Iridium a significant holding, we recently did so with gusto, turning it into one of our largest longs.  Iridium has been written up on VIC as a long in November 2009, a short in December 2009, a long in November 2010, and a short in February 2011.  We have owned a small position for the entire time and can say from experience that the shorts have been right on the stock price, so far: the price is slightly down, while the S&P 500 has risen 60%.  But the shorts have been wrong on the business.  Several feared risks have come and gone: increased competition in the voice services market, end-of-life for the existing satellites, execution risk in developing new satellites and their launch plans, and financing for the new satellites.  Meanwhile the business has shown steady growth.  In the three years from 2009 to 2012, subscriber count rose 59%, service revenues rose 28%, adjusted EBITDA margins rose from 40% to 52% (due to very high incremental margins on new service revenue), and adjusted EBITDA rose 56%. 

With that kind of profit growth and a flat stock price, Iridium has reached a valuation of 4.9x 2013 EV/EBITDA and 6.4x 2013 adjusted EPS.  (The adjustment merely adds back the non-cash tax expense; Iridium will pay “negligible” cash taxes through 2020).  The stock price has lagged the market lately because of one factor that is about to end – a revenue growth slowdown caused by the U.S. government’s Afghanistan drawdown – and one fear factor that is almost certain to be resolved in Iridium’s favor – the upcoming renewal of its Department of Defense contract.  The company also has two new positive catalysts that will play out over the next few years:  Iridium just signed Caterpillar as a major new customer who will integrate Iridium tracking devices into its equipment, and it just signed NAV CANADA as its foundation customer for its new Aireon air traffic tracking business.  We believe the end of the Afghanistan drawdowns starting with next quarter’s earnings and the DoD renewal will be significant catalysts that could lead to a doubling of the stock price in short order.  We calculate today’s DCF-based fair value at $33.40, with 375% upside from today’s price.  No investor is going to pay that much today given that the big cash flows don’t start for five years (even though investors regularly do pay that much for enterprise software companies with similar cash flow profiles).  But the point remains: the margin of safety here looks very large.

The investment case has one (and only one) catch:  Iridium will spend capex of $2 billion over the next five years to build and launch its next-generation satellite network.  The debt funding for this spending will drive its net debt / EBITDA ratio from 2.2x as of year-end 2012 to a peak of about 5x in 2015 before falling rapidly thereafter.  With $200m of EBITDA that is stable, growing fast, and can largely be used to pay down debt (because there are no cash taxes), we are not concerned with that peak leverage.  We show the capex and leverage progression below.  After Iridium NEXT is completed in 2017, capex will fall to $20m per year, and free cash flow per share will balloon to $4 per share – on a $7 stock – and then keep growing.

BUSINESS OVERVIEW

Iridium owns and operates a fleet of 66 satellites, along with several orbiting and ground spares in case of individual failure, which provide voice and data communications anywhere on the planet.  It is important to know that these satellites are in low-earth-orbit, 485 miles above the planet’s surface; most “communications satellites” are in geostationary orbit at 22,236 miles (46 times as high), and some others are in medium-earth orbit at 12,552 miles (26 times as high).  Here is a correctly-scaled depiction of the different orbits:

Source: http://en.wikipedia.org/wiki/File:Comparison_satellite_navigation_orbits.svg

This distance advantage creates far less latency (delay) in communications and helps explain why feared competition in, say, handheld voice communications from Inmarsat’s geostationary network always seemed overblown to us.  Those fears have in fact come to nothing.

The current fleet is approaching its end of life.  The company is well along the process of deploying a new fleet, dubbed Iridium NEXT, which will again consist of 66 operating satellites along with six orbiting spares and nine ground spares.  The new fleet will be backwards-compatible with existing ground equipment but will allow significantly increased data speeds and throughput.  Launches are scheduled for 2015-2017.  The $2 billion in remaining expenses are fully funded by a syndicated bank loan that is guaranteed by the French government export lending authority.  (Iridium’s main contractor is France’s Thales Alenia Space.)  Thanks to the guarantee, the loan is low-cost (under 6% including commitment fees and insurance premiums) and has a repayment term from 2017-2024.  

Iridium splits its revenues six ways: government voice services, government data services, commercial voice services, commercial data services, equipment sales, and engineering & support.  Services revenue is the core.  Steady subscriber growth is driven by the development of new applications and increased penetration of those applications, which is in turn driven by continuously falling unit costs, continuous improvements in equipment performance and costs, and the continued proliferation of data communications generally.

Commercial voice services is the largest category at 45% of 2012 revenues.  This business line is the casual observer’s classic view of Iridium: someone standing in the middle of a desert talking on his brick-sized handset.  Voice services enjoyed consistent 10-14% annual subscriber growth for many quarters until the last three.  The change was the U.S. military’s drawdown in Afghanistan, which prompted a large drawdown in civilian contractors and other in-country personnel who used Iridium.  Recent subscriber growth slowed to 6% in 1Q13.  Management believes that the Afghanistan effect is almost done and that more normal growth should resume in a quarter or two.  Along with steadily rising subscribers comes steadily falling ARPU, currently $43 per month and dropping at 8-12% per year.  (This ARPU drop is made possible as Iridium adds volume to its fixed-cost network.)  The interplay of subscriber growth and ARPU falls yields a consistent “normal” revenue growth of 6-9%, which fell to -2% to +1% over the last three quarters.

Commercial machine-to-machine (M2M) data services is the business’s gem, with 11% of 2012 revenues and growing rapidly.  This service line helps track and provide data communications for, e.g., truck fleets, ships, construction and agricultural heavy equipment, and pipelines and drilling rigs.  Iridium’s subscriber growth has been 30-50% per year, ARPU declines -2% to -12%, and revenue growth 20-45%.  High growth in this segment should continue for years, thanks to ever-increasing data connectivity and rapidly falling M2M equipment costs.  Iridium has a pure-play peer in this area, Orbcomm (ORBC), which was recently written up by danconia17; if you look into Iridium, I suggest you read the Orbcomm write-up or their investor deck for color on this market segment.  In the past, the two businesses appear not to have directly competed much.  Iridium offers lower latency, greater throughput, full polar coverage, and a better network architecture, but it charges a $16 monthly ARPU, while Orbcomm is without those advantages and charges only $5.50 ARPU.  Any customer who could be adequately served by Orbcomm used Orbcomm, and those who needed more functionality paid up for Iridium.  That has changed, though; Iridium just made a huge encroachment onto Orbcomm’s turf with a new contract to be Caterpillar’s “primary provider.”  More on that development below.

Government voice services represent 15% of 2012 revenues.  These are largely the same handset-based services as for commercial users, only with Iridium’s “foundational” customer, the U.S. Department of Defense.  Other U.S. agencies also buy under the DoD contract, including the State Department, Homeland Security, Customs, and FEMA.  This revenue piece took the biggest hit from the Afghanistan drawdown, with subscriber growth falling from +20% in 2010 to -3% to -8% over the last few quarters.  Government voice ARPU is much more stable than commercial voice ARPU at -2% per year, but that has still led to recent 6-8% revenue declines.  Again, these declines should end in the next quarter or two.

Government M2M data services represent only 1% of 2012 revenues but are also growing rapidly.  Subscribers have been growing at 20-45%, ARPU falling 0-6%, and revenues growing 18-40%.

Equipment sales are 24% of 2012 revenues but are a trailing indicator rather than a driver of the business; new subscribers yield a one-time equipment sale.  Equipment revenues have stayed roughly flat even as subscriber additions keep growing, thanks to falling equipment costs and prices.

Engineering and support are 4% of 2012 revenues and are, like equipment, a function of new subscribers and new types of applications for data service.

Putting all this together, the only real business driver is the recurring service revenue and, behind that, subscriber growth (because ARPU trends are steady).  Total service revenue growth was 11% in 2010, 11% in 2011, and dropped to 4% in 2012 with the Afghanistan draw-down.  The company has guided for total 2013 subscriber growth of 15-20%, up from 1Q’s 14%, and total 2013 service revenue growth of 8-10%, up from 1Q’s 3%.  Underneath that level, they say government service revenues will fall in 2Q and 3Q but grow substantially in 4Q.  For the last three quarters, equipment revenue and engineering and support revenue have been strongly negative, which is unimportant to the business health or value but which dragged total revenue growth modestly negative (-2% to -5%).  That headline revenue decline presumably has scared off the quant-based traders.

TWO LARGE NEW BUSINESS OPPORTUNITIES – AIREON AND CATERPILLAR

Iridium has created a new joint venture called Aireon with NAV CANADA, Canada’s traffic control agency, to create a new business that tracks aircraft via satellite instead of radar.  Because tracking over oceans can be more precise with satellites than with radar, traffic control agencies can assign much more efficient flight routes without fear of collision, which will save fuel, reduce emissions, and reduce congestion and delays.  Iridium cites a study estimating $125m in annual fuel savings just for North American flights and $6-8 billion over 12 years globally.  The necessary transmission equipment will be added to the Iridium NEXT satellites as a “hosted payload.”  Aireon intends to charge traffic control entities (and, indirectly, the airlines), a portion of the estimated fuel savings.  Iridium will benefit from Aireon three ways:

(1)    Aireon will pay Iridium $200m in “hosted payload” fees to help defer the cost of the Iridium NEXT launches.

(2)    Aireon will pay Iridium annual data service revenues, just like any other customer; assuming Aireon signs a few more customers, the contract caps these revenues at $20m per year, which is a small but nice boost to the current service revenue base of ~$280m.

(3)    Iridium will keep half of Aireon’s profits.

NAV CANADA was Iridium’s natural JV partner because it is the world’s largest handler of oceanic air traffic, thanks to the polar flight routes to North America from both Europe and Asia.  In late April, NAV CANADA and Aireon announced that they had signed a contract for Aireon’s provision of services to NAV CANADA (which is distinct from Iridium’s $20m/year contract with Aireon for providing data to Aireon).  Management expects Aireon to sign several other air traffic control agencies as customers.  The last to sign will likely be the world’s second-largest, the U.S. Federal Aviation Administration, because it is a pure public agency that must ask Congress for more money while the other countries’ controllers are privatized for-profit entities.  Management expects an FAA deal, but not until late 2014 or 2015.

Also, as part of its first-quarter earnings release and call in April, Iridium announced a “large, multi-year agreement” with Caterpillar to be its “primary provider” for M2M services.  The contract has a non-disclosure provision, and management has not yet put out a promised press release on it, but this deal is clearly a game-changer.  Up until this point, Orbcomm has dominated the heavy-equipment niche; it lists as its customers not only Caterpillar, but also Komatsu, Hitachi, Volvo, Hyundai, and Sumitomo.  Two factors appear to have changed.  First, Iridium is now much more cost-competitive.  Iridium’s ground equipment cost has fallen from over $300 per unit to $100 (versus Orbcomm’s $50-100), and its service ARPU has fallen more rapidly than Orbcomm’s.  Just as importantly, customers appear to be moving away from simple telematics – a “ping” a few times a day to track location – and towards more complex communications, which favors Iridium. 

The extent and speed at which Iridium is displacing Orbcomm at Caterpillar is unclear; Orbcomm has already responded that it still has a “strong” Caterpillar relationship with an order backlog of over 10,000 units and 2 years.  But when we spoke to Iridium management, they told us: “we are using the words ‘primary provider’ because that is what the contract says.”  During Iridium’s earnings call, CEO Matt Desch spoke about the Caterpillar deal at length, and unless he is spinning well beyond what is acceptable (which is not his style), it is highly significant, both because Caterpillar is a large customer and for what the deal says about Iridium’s newfound ability to serve a wider range of M2M customers.  Here are his comments, with emphasis added:

[W]e recently closed on a multi-year agreement with Caterpillar to become its primary provider for satellite M2M data services. We haven't had an opportunity to issue a press release yet as we just got the deal signed, but I wanted to emphasize this agreement with you today as it's central to our growth strategy in the M2M business. This watershed agreement with the world's leading heavy equipment manufacturer underpins our strategy to provide real-time data communications for telematics asset and fleet management and location-based services for global OEMs and aftermarket solution providers.

Caterpillar appears to be adopting a comprehensive telematics strategy for a big part of its extensive product line and we think our new contract clearly demonstrates that the market demanding the highest quality and best value is picking Iridium for their needs. We expect this will be the first of a number of significant contracts in the coming quarters and years and believe that this deal will be meaningfully contributing to our commercial M2M subscriber and revenue growth beginning in 2014. . . .

[W]e've been working on this a long time and it's important to note kind of Caterpillar has embarked on at least appears to us a corporate strategy to connect via telematics everything that they manufacture. Satellite M2M we think will be a big part of that, but I think that sort of -- I think they were working at a bit more opportunistically, but now it’s taken sort of a corporate wide approach and that led them to rethink who they were going to use long term for satellite M2M and I think they really looked at the industry extensively and we reiterated many times why we think we would be selected by these kind of suppliers, it's latency, it's speed, it's a two-way attributes, it's sort of the uncompromising nature of if you're making all those assets and they are showing up every part of the planet not worrying about where they are and how fast that you can respond to them. And sort of a more of a portfolio of product because this isn't just a SPD, but includes two way, higher speed services to connect larger assets and I don't say just track, but actually connect to provide information about the health of the vehicle where is the, how it's performing and other attributes that they want to find out about it. . . .

I would say what's been different a little bit in the last two or three years that's made us more capable now than we were, say, five years ago when perhaps say they were getting more started in this area or even before you know, is the cost of our devices has really, really come down over the last couple of years. The size and scope of those. So, we have traditionally play more at the higher level of the M2M market but now I think this contract shows that we can compete very effectively at what I would call the lower, lower ARPU but you know, wider scale larger volumes kind of area of the industry. . . .

I think Caterpillar, potentially long-term could become one of our largest M2M customers. . . .

One of the reasons why Caterpillar was very important to us is because as Caterpillar goes, so goes the rest of the industry. There are bellwether kind of player, who are highly noticed by all their peers and I think when they make a selection, it's respected by the rest of the industry. . . .

After the earnings call, management told us that Caterpillar was very clear about its reasons for selecting Iridium: its network’s much better latency and throughput, and its more robust R&D program and coming product pipeline.  No competitor can match or is likely to match either one of these.  The future looks bright for Iridium’s M2M business.

STEADILY GROWING MARGINS AND PROFITS

Because Iridium’s network is almost entirely fixed-cost, steadily increasing subscribers and revenue generates steadily increasing margins, even with the ARPU declines.  For years we have calculated the effective incremental margin on year-over-year revenue growth, and although the number varies greatly in any given quarter, the multi-quarter average is consistently around 66%.  Service gross margins have steadily risen from 60% in 2007 to 78% in 2012 and 80% for the last 3 quarters.  “Operational EBITDA” margins have risen from 32% in 2007 to 54% in 2012 and 56% for the last three quarters.  We expect margin growth to continue for many years.  Management has guided to a 60% “operational EBITDA” margin in 2015.  Although we do not count on it in our model, even higher margins should be achievable later on; Iridium’s peers Inmarsat, Intelsat, and SES each have EBITDA margins in the mid-70’s.

Note that there are three relevant “EBITDA” numbers:  A standard EBITDA definition takes net income and adds back interest, taxes, D&A, and a small amount for non-cash purchase adjustments caused by Iridium’s 2007 bankruptcy.  For its “operational EBITDA” number, the company adds back two more items.  The first is operating expenses associated with the un-launched Iridium NEXT (mostly R&D); we view this adjustment as appropriate to show the true earnings power of the business post-launch.  The second management adjustment is stock compensation; we refuse give a free pass to a company that still pretends like it’s 1999 and that dilution from stock comp is costless.  Thus the 2012 financials look like this:

64m net income

- 1m interest income                                                                                   

+ 30m taxes – note that the tax accrual is all non-cash

+ 81m D&A

- 0.5m purchase accounting adjustments

= 174m “standard” EBITDA

+ 24m Iridium NEXT expenses

= 198m our adjusted EBITDA (52% margin)

+ 7m stock comp

= 206m management’s “operational EBITDA” (54% margin)

THE DEPARTMENT OF DEFENSE RENEWAL OFFERS UPSIDE, NOT DOWNSIDE

Iridium’s government business, which is 23% of service revenues, has been operating under a 2008 Department of Defense (DoD) contract structured as a one-year commitment with four one-year options to extend.  The contract expires on September 30, 2013, and the RFP process for a new contract is underway.  The need to renew this contract creates uncertainty for Iridium, but very little risk and substantial upside.  The three theoretical risks are (1) DoD switches providers, (2) DoD ceases or materially cuts back on the services it buys, or (3) DoD materially squeezes Iridium on price.  None of these seem at all likely.  Iridium’s services are deeply embedded in DoD’s operations, can’t be performed using non-satellite equipment (else they already would be), and can’t be switched to another provider.  Back when Iridium went bankrupt in 1999, its parent Motorola originally intended to de-orbit the satellites.  DoD told Motorola that the Iridium network was already an irreplaceable asset for it and insisted that Motorola find a way to keep Iridium operating.  (The rumored threat: “If you ever want another defense contract again, you had better keep those
birds in the sky.”)  By now DoD has spent over $200m to build its proprietary Iridium gateway and other ground equipment, all of which would be squandered if it stopped using Iridium.  Indeed, although its service contract needs renewing, DoD has already signed (last October) a new $47m five-year contract to upgrade its proprietary gateway in anticipation of Iridium NEXT.  (See http://investor.iridium.com/releasedetail.cfm?ReleaseID=717407.)

Put simply, DoD has no alternative to Iridium; even apart from the sunk equipment, system costs, and upheaval from switching, no other network is capable of meeting DoD’s requirements.  And as for a major price/margin squeeze: that is just not done in defense contracting.  We are not aware of it occurring in any aerospace defense contract renewal involving a highly heterogeneous product. 

Far from presenting real risk, the renewal’s most likely outcome is a meaningful increase in government revenues and profits.  Management says their approach to DoD’s sole-source RFP is to offer additional functionality and value in exchange for higher revenues.  Even apart from whatever new functionality might be non-public, we know of two significant growth opportunities involving “Netted Iridium.”  Prior to 2010, Iridium allowed push-to-talk functionality – effectively creating a traditional multi-user tactical radio network via satellite, as opposed to merely two-person point-to-point phone calls – only within the footprint of a single satellite.  “Netted Iridium” allows push-to-talk across the entire global network, which is a highly valued function for DoD.  Up until now, the required handsets were sole-sourced and cost $7,000 each, which limited Netted Iridium’s usage to “the guys in the mountains” (e.g., special forces) and very senior officers.  Iridium and DoD are now working to dual-source those handsets and drive down their cost significantly, which could expand Netted Iridium’s user base many-fold.  They are also working to incorporate Netted Iridium functionality into all DoD radio handsets.  DoD’s existing installed handset base is roughly 500,000, compared to only 8,000 Netted Iridium handsets deployed today; that change would increase usage exponentially.  (Iridium’s investor deck mentions this opportunity almost in passing; see page 23, “Potential for Iridium waveform to be part of high volume, standard (Sincgars, JTRS) software-defined radios.”)  We are not counting on such a huge win, but it is nice to know that the possibility exists.

For all of these reasons, we believe management is more than justified in providing the guidance it has: government revenue declines for the next two quarters but then “substantial” growth in 4Q and beyond, after the contract renewal.

THE COMING CAPEX AND DEBT BULGE

To get Iridium NEXT launched and operational, Iridium will spend $3 billion of total capex, with $1 billion spent so far and $2 billion to go, plus another ~$25m/year of expensed SG&A.  (Other maintenance capex over this period is trivial, roughly $10m per year.)  That spending is significant, especially compared to the current market cap of roughly $650m and enterprise value of $1.2 billion.  However, $255m of that cost will be recouped through hosted-payload fees -- $200m from Aireon and another $55m from two other customers.  Iridium’s relatively large, steady, and growing earnings are more than enough to support the resulting debt load.  Management has guided to a peak leverage of 5.0x EBITDA in 2015 with pay-down of 0.5 leverage turns per year thereafter.  Our own bottom-up numbers yield a slightly lower peak leverage.  Our model assumptions are not certain enough to conclude that management’s guidance is conservative, but they are enough to conclude we should not be concerned about the capex and debt load:

VALUATION

This stock is very cheap on any metric.  On the assumptions detailed below, which include zero profit contribution from Iridium’s equity stake in Aireon, our 10-year DCF model yields a value of $33.40, 375% above today’s price.  The real value from an owner’s perspective starts accruing in 2018, after the Iridium NEXT capex is done.  Free cash flow per share spikes to almost $4, keeps rising, dips in 2021 as cash income taxes begin, then rises back above $4 and keeps going:

On traditional PE and EV/EBITDA multiples, the stock is cheap even on current earnings and over the next few years:

Key assumptions:

• Total revenue growth of 7% in 2013, 10% in 2014-2015, and 9%/8% in out-years, driven by double-digit commercial service revenue growth, flat/negative 2013 revenue growth for the combined other revenue lines, and 4-5% growth thereafter for other revenue lines
• “Operational EBITDA” margin rises from 53.6% in 2012 to 59.6% in 2015 and 61.0% in 2020, then steady.  (I have specific assumptions for each cost line, but let’s not get lost in the weeds.)  Stopping the margin growth at 61% may be conservative.
• Additional Iridium NEXT expenses as shown above, continuing at $10m/year forever
• Zero profits from the Aireon equity stake; I am being conservative because I cannot estimate those profits.  The separate service revenue paid from Aireon to Iridium is embedded in the double-digit service revenue growth.
• 0% taxes through 2020, caused by R&D tax credits and by accelerated depreciation.  35% thereafter.
• Share count grows at 2%
• Capex falls from the specified numbers in 2013-2017 to $20m thereafter
• 8.8% WACC, composed of a 6% cost of debt and 12% cost of equity, using today’s ratios of debt and equity.  (The actual WACC will dip well below that for the next several years as debt grows.)
• 3.5% terminal growth rate
• Iridium will likely need a 3^rd-generation fleet 15 years after Iridium NEXT, so new capex should be centered around 2030.  At 3.5% inflation on NEXT’s costs, the 2030 nominal capex would be $5 billion.  Discounting that cost back to today at 8.8% yields a present value cost of $1.2 billion, which I subtract from the terminal value.

RISKS

We do not view any of these risks as both material in their potential impact and at all likely to occur:

• Department of Defense contract renewal: As discussed above, we see very low odds for a materially bad change to the contract or a failure to renew.  We do see some short-term headline risk from a potential delay in signing the renewal, but that would be merely noise.
• Launch failure:  A satellite blow-up on launch or burn-up after a misplaced orbit is a fairly common occurrence.  (Orbcomm had one recently.)  The stock of whichever company owns the satellite usually dips in reaction.  This, too, is short-term noise.  When you are launching 72 satellites over three years and each is slotting into an existing network, if one blows up, you collect your insurance payment and move on to the next one.
• New competition:  In terms of competition from other satellites, Iridium’s niche is close to being a natural monopoly.  No one today competes much with them on their capability set, and no one is likely to spend the $4-5 billion necessary to create new competition.  The real threat would come from ground-based substitutes.  Most tangibly, there is speculation that someone could create a network of terrestrial flying drones omnipresent enough to provide global coverage.  Dreamers cite the distance advantages versus “communications satellites,” but those advantages would exist only versus far-away geostationary or medium-earth-orbit satellites, not versus Iridium’s low-earth-orbit network.  Drones would have an obvious up-front capex cost advantage, but we are not sure they would have a total cost advantage given their much higher operating expenses and lower lifespans.  (How many drones would it take to provide global coverage just over land – 10,000?  50,000?  Each of them being fueled, piloted, serviced, and repaired?)  Drones are also unlikely to be able to provide the service reliability of a satellite network.  In any event, such a network would take many years to get into the air, and to our knowledge none are even in planning.  Iridium equity is worth its current market price even if the business’s terminal value beyond 10 years is near zero.

• Revenue growth turns positive in 3Q or 4Q
• DOD contract renewal is announced
• Further details provided on the Caterpillar contract; Caterpillar revenues begin
• New headline-worthy M2M contract wins
• Further Aireon customer contracts announced
• Much higher revenue and profit growth in 2014, flagged by management’s full-year guidance in the 4Q earnings release