Description

Summary

Iridium operates a constellation of 66 satellites in low earth orbit (“LEO”) with a useful life is 15 years. It cost $3B to build and launched in 2018 after several years of delays replacing its dying gen1 constellation. The company’s shareholders have been reveling in that accomplishment since and for good reason - it was a challenging process with funding concerns. However, getting through that capex burden does not magically turn a crappy, low-growth business into a good one. The market disagrees. The stock is up from $12 in 2018 to $40 now. The majority of this return has come via multiple appreciation, and valuation now sits at a rich 19x fwd EBITDA. This outperformance comes against a backdrop of most satellite stocks struggling. Its best comp (Inmarsat) was taken private at 8x in 2019. This begs the question - what did Iridium do so right? The answer cannot be found looking at forward estimates, which have been steadily declining. Over the last year, the stock is up 60% despite 2021 consensus EBITDA falling from $390M to $369M. The stock is getting a free pass because of COVID, but this misses the point. Because of the constellation’s limited useful life, slower monetization is a permanent loss of value. Also, COVID represents no thematic benefit moving forward. I realize that IRDM is just one of many aggressively priced equities in this market, but within that crowd, it stands out because of 1) a deeply flawed free cash flow narrative fed by management, 2) heavily disrupted end markets, and 3) a lack of exciting growth avenues. 

The stock has outperformed because management has been able to convince shareholders to 1) pretend future capex requirements disappear and 2) ignore adverse competitive changes in the industry. The bull case revolves around the capital return window between constellation builds with good outcomes tied to growth. The problem is the new constellation failed to unlock meaningful new market opportunities and the legacy products (largely sat phones) can barely support any growth. Since launching Iridium NEXT, core commercial service revenue (excluding hosted payloads) grew 10% in 2018, 8% in 2019, and less than 1% last year with 2021 guided +3%. Pre-COVID, 2020 service revenue was guided +6%-8%. The deceleration is clear enough, and the business is starting to look like the gen1 constellation. After sat phone demand peaked in 2012 (entering secular decline), Iridium’s 6-yr EBITDA CAGR from 2012-2018 was below 5%. That calculation again excludes hosted payloads, which are fixed payments that temporarily juiced growth optics but are flat moving forward and not representative of the core business. Moreover, a significant portion of hosted payload revenue recognition is non-cash.

The constellation rebuild effectively amounted to maintenance capex, which makes Iridium’s business model look pretty horrible in my opinion. It can be summarized as “invest a ton of money launching a constellation and then hope that the industry does not change much over the next 15 years.” Maybe the sleepy satellite industry could have supported Iridium’s business model a decade ago (when they initially planned the investment), but things have changed. The pace of innovation has accelerated dramatically featuring huge gains in capex productivity. Easy access to capital coupled with technological improvements has led to an outpouring of investment in space. There are many new LEO projects underway (SpaceX, OneWeb, Telesat, Amazon) and massive amounts of GEO and MEO capacity coming (VSAT-3, mPower, Konnect VHTS, Jupiter-3, Inmarsat I-6). This is all well documented- https://advanced-television.com/2021/02/12/report-satellite-broadband-capacity-to-grow-tenfold-by-year-end/  

Iridium is not a technology company and outsourced its build to Thales, the French gov backed aerospace company, in 2010. The timing was such that Iridium missed out on most of this innovation cycle. The gen2 constellation remains only capable of serving niche applications and unable to tap into high growth markets. So while shareholders keep celebrating the capex holiday, does it continue to be a fun holiday if the growth plan fails? Large cracks are already showing through multiple guidance cuts for the key “broadband” product (Certus). These have thus far been brushed aside but have severe implications for growth and deleveraging over the medium and long term.

Zooming out, what business can realistically hope to be successful with a mostly static product over a 15-year period? The LEO business model is famously difficult. There is a reason why every LEO has gone bankrupt thus far, with Iridium’s gen1 constellation one of the prime examples. Having a reorg’d equity was key for Iridium making it this far. Who knows if SpaceX can ultimately achieve its goals, but they already have 1,000 satellites in orbit with a plan to replace them every 5 years. The disruption is happening either way. How could it possibly make sense that Iridium is insulated from these broader developments? The company’s pitch used to be “sure we have slow speeds but at least we offer polar coverage and low latency.” Those are no longer differentiators, nor were they sufficient to generate much growth previously.

Management claims they are protected from new supply, but that is BS. They employ circular reasoning. On the one hand, they describe capacity growth by competitors as “commodity broadband” outside their focus. On the other, they talk about maritime broadband as their main growth market where they are competing within the same “commodity” umbrella. You cannot have it both ways. Also, at this point we know management’s spin to be false empirically. Broadband deflation driven by capacity growth were the underlying factors driving multiple guidance cuts for Certus. This trend is only going to get worse. Management sold investors on attacking Inmarsat’s base of business while failing to acknowledge/realize that Inmarsat’s L-band business was already being aggressively disrupted by better technology. It is not hard to imagine Iridium’s LSD growth flattening out and then turning negative over the next 5 years as L-band services keep getting displaced while battling price pressure. Net leverage is 4x and the model is very sensitive to the downside. 

For more context, Katana wrote up IRDM as a short in 2017 and that post provides good background that I won’t cover here.

Key points

1) Flawed FCF narrative spoon-fed by management

The company likes to highlight its FCF profile. The following chart was provided in the Q4 call –

Using $1.67 of FCF/share, IRDM is trading mid-20s on that number. Ok, maybe that is fair. However, this calculation fails to incorporate replacement capex. Without replacement capex, the business, with certainty, will literally fall out of the sky. The company was just battling this prior to 2018. Here is what capex looks like over the last decade - 

Any reasonable valuation methodology should incorporate the capex required to sustain the business. $3B divided by 15 years is $200M. IRDM’s D&A was $300M last year. Adding replacement capex to the table above eats up the entire cash flow. Just because Iridium’s capex is extremely lumpy does not change the fact that it is mandatory to preserve terminal value. Capitalizing trough capex would be laughably inappropriate, and yet, this is exactly what the CFO continually pushes. Here is a snapshot from the 4Q18 call:

With the completion of Iridium NEXT Iridium has become a rare company. One that is characterized by robust EBITDA growth and material levered free cash flow. In a few minutes I'll introduce some measures of Iridium's levered free cash flow we think our investors should consider moving forward…

With our newly issued EBITDA guidance, investors can calculate the financial metrics that should be additional consideration for our valuation going forward. Today Iridium is a completely different company than the one you have known for the last 9 years. Today we have a very different risk profile with material free cash flow generation capability. I'd expect that investors will increasingly consider metrics like levered-free cash flow, levered-free cash flow growth, levered-free cash flow yield, levered-free cash flow conversion and CapEx intensity in assessing the fair market valuation.

He goes as far as suggesting the comps he thinks investors should use. From 4Q19 call:

We continue to believe that Iridium's free cash flow statistics compare quite favorably to companies in the communications infrastructure space that trade at much higher multiples than Iridium. For example, certain fiber companies trade in the high teens and certain data center companies enjoy a multiple of 20x, while the tower sector trades around 25x EBITDA. We think this is an important investment consideration in evaluating Iridium.

You can find this messaging all over their calls. It is a red flag whenever management tries to push valuation metrics on investors. Moreover, tower and data center companies have far less technology risk, and I have no idea how he justifies EBITDA growth and positive levered FCF as “rare.” Those qualities define large swathes of the public markets. Analyst coverage is sparse and low quality. Sellside was quick to eat this up. 

Maybe one could argue that the next constellation will cost less. Ok, but that view ushers in a bigger problem. Satellite cost curves have indeed declined dramatically, which opened up a huge amount of competition. Lower cost curves are also not mutually exclusive with lower total dollars invested. Gross investment industry-wide has been rising. SpaceX is planning to spend over $10B on its LEO constellation. This is an arms race and IRDM’s L-band network was made for another era. It would be ridiculous to assume Iridium can grow into broadband markets over the long-term while investing less dollars while its competitors pour money into new capacity.

2) Ongoing industry disruption creates big overhang on key growth story

Broadband is where the demand is for obvious reasons. Customers across all end markets want more data and faster speeds for less money. Adding a broadband product (Certus) was a key factor highlighted by management when justifying the investment to replace its constellation. For years, Iridium was guiding to $100M of run-rate Certus revenue exiting 2021. Here is the CFO talking about the product at a 2017 conference –

Similarly, Iridium Certus, so you can imagine, our existing telecom network is in late '90s, vintage telecom network, you can imagine the data speeds you had in the late '90s. We're similarly situated in our existing network. Our broadband offering is 128 K. Under NEXT, that goes to 1.4 MB. And then all of a sudden now, we are front and center in providing broadband services to maritime, aviation and land applications that we couldn't service today. And so we estimate, that's $100 million of incremental revenue as we exit 2020. So clearly, a bright future for this company.

Management ended up sneakily changing this guidance to $100M factoring in legacy OpenPort service, which was $25M. So $100M dropped to $75M incremental for Certus exiting 2021. Ok fine. They were still jazzed. The product launched in early 2019 and here is the CEO talking about it at the March 2019 investor day:

So the broadband opportunity we've mentioned, we think is today about $700 million primarily driven by Inmarsat, and we will over time split in that market somehow with them. I'd like to take it all away, but we'll have some large percentage of that. But we put our stake in the ground and our growth rates to say $100 million but really this narrowband I don't believe investors are appreciating nearly enough because I believe this is a long-term multibillion-dollar market, and we're extremely well positioned.

After just one full quarter of operation, management cut its target by -33% to only $50M incremental. Pre-COVID, the company clearly demonstrated it had little handle on the broader satcom market that it sought to enter. Now with COVID headwinds, there is effectively zero chance they hit this lowered bar. Certus is currently run-rating at $11M of contribution calculated as the annualized growth in broadband since they started reporting it in 1Q19. Management’s error was obvious. By benchmarking Certus to Inmarsat’s Fleetbroadband, they missed that L-band service was already hemorrhaging share to Ku and Ka providers. Inmarsat used to be public, and you can see their vessel counts and ARPU declining rapidly well before Certus launched – 

There was no overlap among analysts covering both companies so nobody bothered to push back on Iridium’s assumptions. The problems at Fleetbroaband and Certus are structural. Iridium has a narrowband network that utilizes a small slice of L-band spectrum – an important difference versus the larger satcom market that has been predominantly investing in Ku and Ka. Spectrum can be a complicated topic but just think of it as the raw material that determines how much data you can transmit over a connection. To compare to terrestrial examples, L-band is like 3G and Ku/Ka 4G/5G. Subscribers want streaming and live video so naturally players like Intelsat, Eutelsat, SES, Viasat, Hughes, Inmarsat, SpaceX, Amazon, and Telesat are all investing in Ku and Ka capacity to make that happen. Iridium holds a license for 8 MHz of contiguous L-band. By comparison, SpaceX is using 2 GHz (or 2,000 MHz) for just downlink. The concept of “capacity” in Iridium’s network is not even discussed because it is so limited. So while we have no metrics to use, suffice it to say that a handful of SpaceX sats have more capacity than the entire IRDM network. 

To provide a real-world comparison, Certus launched in maritime with the following plan – 350 Kbps download and 50 MB data caps for $380/month plus you need to buy a terminal, which cost in the thousands. As an alternative, KVH, a maritime reseller of Intelsat and others’ capacity, was offering 200 MB and 5 Mbps speeds for $99/month with no upfront terminal cost. So if you are a vessel owner, KVH offers faster download speeds and over 90% cheaper data on a $/MB basis. KVH is one of many VSAT providers including Marlink and Speedcast that all offer far superior and cheaper service. No wonder Certus got off to a rocky start. KVH points to the existing pool of L-band subscribers as their low hanging fruit to convert to VSAT Ku/Ka connections in the prevailing “upgrade cycle” (this slide from their investor day) – 

Within VSAT connections, IRDM only plays a backup/safety role, which is low ARPU, low relevance. For example, Inmarsat offers its backup L-band service for free within VSAT plans. Backup connections is where Iridium has had modest success to date. It will likely keep growing here, but being a backup cannot support accelerating growth. Its OpenPort product on the gen1 constellation was already filling this role. Over the long-term, the backup TAM should be shrinking as Ku and Ka capacity fills in.

Iridium has since upgraded speeds to 700 Kbps and eventually plans to max out at 1.4 Mbps. This hardly moves the needle relative to the competition, and the bigger problem is the data caps. Iridium’s thin capacity means that service tiers need to be priced high to limit usage or else overall network performance will deteriorate. It is analogous to the issues facing Gogo’s ATG network. As Certus tries to add aviation and gov customers, it will face the same problems. And yet the CEO keeps pushing this idea that Ku and Ka is “commodity broadband,” implying Iridium’s is somehow better. This is a ridiculous spin on an obvious problem. We are to believe low speed and low throughput is better? Broadband is a commodity. Iridium happens to be high cost.

3) Limited growth potential

It is hard to see how Iridium can generate service revenue growth above LSD even in a good scenario when you break down each category – 

• Commercial satellite phones (36% of service revenue) – Iridium is best known for its sat phones, which remain industry leading but operate within a flat-at-best demand backdrop. This is a long-term loser through VOIP and terrestrial expansion. Sat phones still account for almost 50% of commercial service revenue and close to 60% of total service revenue when you group in the government contract (predominantly phones). Its 6-yr CAGR in this line is 0.3%. They used to report volume trends in the 10-K and airtime usage was down every year between 2011 and 2017. Occasional price increases have been propping it up. 

• IoT (22%) – This is a sexy name for connecting Garmin devices (notably the inReach handheld) and heavy equipment location trackers. This had been growing low teens but decelerated to MSD pre-COVID due to growing ARPU pressure. COVID disruption then dropped IoT performance to flat in 2020. MSD growth here is a reasonably good projection for them moving forward.

• Gov contract (22%) – This is an all-you-can-eat contract for sat phones and IoT and is contracted through 2026. It grew 4% last year and will CAGR at just 1.5% through the end of the contract. At expiry, the 10-yr rev CAGR from Gov will be just 2%. Hard to believe this can accelerate.

• Hosted payloads (13%) – This consists of fixed payments from third-party payloads Iridium added to its satellites. A large portion is non-cash deferred revenue recognition from upfront lease payments. In the FCF chart above, you can see $22M stripped out last year. AIreon, the larger of the hosted payloads, has been unable to afford long delayed payments. This line has been the primary driver of service rev growth since 2017, but will decline -6% in 2021 and then flatten until contracts with Harris and Aireon expire in 2027 and 2030, respectively.

• Broadband (7%) – The Certus product discussed in the previous section plus legacy OpenPort. This line probably has a couple years of modest growth in niche applications (back-up connection in maritime, small vessel connections, cockpit safety services, and some small gov utility). It will be battling steady ARPU pressure and could flatten out and begin declining over the next 5 years. As noted, it has already been massively underperforming vs plan.

• The company also sells some low margin equipment and engineering support (20% of total rev).

• Combining the above, the company’s only growth drivers are IoT and broadband, which make up 30% of service rev. Giving them credit for steady MSD growth in IoT and mid-teens growth in Certus yields +LSD overall service revenue. 

Valuation 

An easy and generous way to value IRDM would be to look for comps in IoT, which is IRDM’s best segment. 

• Orbcomm (ORBC) is an IoT pureplay that trades 12x fwd EBITDA. It uses terrestrial and satellite connections, which reduces capital intensity and lowers technology risk relative to IRDM. Therefore, I would argue ORBC’s multiple represents a generous valuation for IRDM and would yield a price target of $25. 

• Garmin (GRMN) has been Iridium’s fastest growth vertical. Iridium is far more capital intensive than Garmin, and using an EPS basis appropriately includes depreciation expense. Iridium’s EPS will be negative this year and consensus is just under 10 cents for FY22. GRMN trades 23x FY22 EPS. Applying that multiple to IRDM’s 10 cents yields a stock price of $2.30.

A more intense way to value Iridium is to assess the potential for capital returns to the equity during this capex holiday, which management guided to last until 2028. At that point in time, the current constellation will be 10 years old with 5 years remaining before it starts to fail. The last constellation took 8 years to build and launch. This plan assumes the new one will take 5 years. Once the holiday ends, capex will ramp and the business will turn heavily FCF negative again. The balance sheet at the end of the holiday needs to be in a position to support the rebuild. To illustrate this point, this is what IRDM’s cash flow looks like over the last decade (it consumed $1.7B of cash flow b/t 2010-2018):

Net leverage today is 3.9x and projecting the full holiday and rebuild cycle shows how limited capital return potential will be unless IRDM can accelerate growth meaningfully. Here is the exercise I did:

• Project out the business by line. My assumptions include: 

• Sat phones flat

• IoT reaccelerating to +MSD

• Gov and hosted payloads we know from contracts

• For Certus, pickup in net dollar growth for the next two years approaching peak then declining in the out years

• Opex should start to inflate LSD after cutting costs last year

• Incorporating interest expense, tax rates, and capex guidance – all knowable

• Giving them credit for remaining Aireon hosting payments ($146M)

• Let’s assume they pay a $2.15 dividend per year through 2027 (the end of the holiday). That amounts to $300M gross outflow per year on the diluted share count. You’d get back $15/share of your basis, but then as you start layering in build costs for the next generation of satellites, net leverage balloons to 8x. The company will again be facing bankruptcy risk and the equity terminal value would start approaching zero. In this scenario, the stock is worth what? The PV of dividends plus some small remaining terminal value in a tough industry. Call it $10.

Here is my excel output for those curious (not sure how formatting will turn out with this upload) –  

I acknowledge that the scenario I am running above includes lots of unknowable assumptions. Maybe it is too pessimistic but I could easily argue that I’m being too optimistic about sat phone and broadband trends. We can see the broader trends and both products could turn negative over the next 5-10 years. The point is that it is hard to see how this asset and this balance sheet can justify capital return anywhere near what the stock price suggests. These limitations are already emerging. The company just announced a $300M buyback authorization through 2022. That does not amount to much on a $5.6B market cap over two years. Capex was guided to increase from $35M to $45M for the next 2 years. Core opex was also guided to pick up after cutting costs aggressively last year. The company historically has spent very little in sales and marketing or R&D. This is catching up with them.

Again maybe the pushback on this analysis is that the next constellation rebuild will cost less. However, that does not fix the bull thesis. We know Iridium does not have any meaningful competitive advantages by looking at their track record over a long period of time. Returns on invested capital are dictated by forces outside their control, and therefore investing less in the future constellation will yield lower earnings power. Said another way, if replacement capex for Iridium’s next constellation is much lower, then its earnings power on the existing constellation will be declining that much faster.

Why now? 

The company is finally entering a period where they need to show hockey-stick like growth from Certus. Assuming equipment and engineering revenue are flat, consensus estimates imply service revenue grows 4% in 2021 doubling to 8% in 2022. If Iridium cannot deliver, patience should be running thin. Declining estimates should catalyze a much-needed re-rating on a levered and highly capital-intensive business facing a shrinking TAM. The biggest risk to my thesis is that nobody will care and IRDM bulls keep on eating up that levered FCF story capitalizing trough capex. Note the stock recently ripped up to $50 after ARK disclosed it bought 5% of the company. It then fell back to $40 as rates moved higher and insider selling picked up. 

Other risks and considerations

• I did not discuss Aireon much here. At times, Aireon has received huge hype and valuation credit. Iridium owns 22% of Aireon’s equity in addition to its hosting relationship. Aireon has been unable to land a big contract with the FAA nor pay Iridium its original $200M hosting contract (only $54M to date). Both of those deliverables are ~3 years late. For anyone taking a sober look here, assuming much value from Aireon is speculative at best. Its track record is objectively poor.
• Inmarsat is launching more L-band capacity this year. This will be a notable headwind in 2022 on top of the broader Ku and Ka-led deflation
• Iridium has been good at refinancing. They currently pay just L+2.75% with a 1% floor. Hard to see this going lower but I’ve been wrong so far. 
• The company is launching Certus for IoT this year. I do not really know what that means. Had the growth story all along been about IoT then maybe I would find this more actionable. Instead, management focused on broadband and has thus far whiffed massively. So while I find it hard to give them much credit, I have not done the work here beyond the existing Garmin and heavy equipment demand centers. Maybe IoT has bigger growth potential than the MSD I am giving them credit for.
• Maybe IRDM can push price harder in sat phones. I have no reason to think so but maybe they could corner a dying market
• There are two businesses trying to disintermediate Iridium in the sat phone market – AST & Science and Lynk. These companies aim to make smartphones compatible with satellite coverage. The former just went public via a SPAC at a $1.8B valuation. This would be a killer for Iridium. However, I have not done the work to assess how realistic this is. I add it here as a potential wild card that would significantly improve the short thesis.

Catalyst

Certus growth stalls

Estimate reductions

More clarity on competitive threats

Buyback and divi amounts dissapoint