We're live from Orlando. Welcome, everybody. 45th Annual Raymond James Institutional Investor Conference. I'm Ric Prentiss, head of telecom services, but now more broadly TMT. We did pick up Disney, Warner Bros. Discovery as well as Paramount this past fall and really excited to move into our satellite space. If you saw the signs out there with Raymond James 45th logo, you'll see a picture of a satellite out there now and the little like all the little icon memes out there of what this conference is all about. So we're happy that Tom Fitzpatrick can join us, CFO of Iridium, to show us a few slides, update us on the company, and then Tom, I'll do a fireside chat format and then we'll save some time at the end for Q&A. So Tom, welcome. My 28th year, but a lot of them for you too.
Thanks, Ric. Thanks very much. Good afternoon, everybody. Nice to be with you all. Just a couple introductory slides here and I'll be talking about some non-GAAP measures. I would just direct you to our website for a reconciliation to the GAAP measure. I'm going to show you some statistics that we're pretty proud of in terms of our long-term growth in service revenues and EBITDA. You know we were able to deliver those results because of the uniqueness of our network. It's one of a kind. It's a low Earth orbit mesh network operating in the L-band. As I said, it's one of a kind. It does things for our customers that nobody else can. That's why we've won to date and that's why we're going to continue to win, we think, through the rest of the decade.
I'm going to refresh your memories for those of you who didn't see our presentation in September at Investor Day and what we think the financial prospects are over that time period. Let's just define terms here. In terms of 2023, our total revenues were $791 million. Of that, 74% is recurring service revenue. That's what drives the profitability of the business. It's recurring in nature and very profitable revenue. The other revenue is engineering and equipment revenues. Unlike our terrestrial brethren, we actually make money when we sell equipment and so it's a fine line of business for us, but it pales by comparison in terms of its ability to drive growth in EBITDA. Our service revenue breaks down. $219 million is our commercial voice and data business. Fastest growing segment is IoT.
That's got a 20% CAGR in subs for 10 years, 1.7 million subscribers. And then our oldest and largest single customer is the U.S. Department of Defense, $738 million in 2019. So long-term track record of growth. You see a 7% CAGR dating all the way back to 2014 and 9% growth in EBITDA. And that just demonstrates the inherent operating leverage of the business. Variable costs are very low to produce incremental minute of use on our network. And so as service revenue grows, EBITDA grows faster. And we've proven that year in and year out for a very long period of time. And I'm here to tell you today we think that continues through until 2030. Again, our ability to deliver these kind of numbers is based on our network competitive advantage. It is fully intact.
There are some new entrants to the marketplace that we compete very well against or we stand up very well against. We're actually complementary to their offerings. Their offerings have certain deficiencies that we're able to complement. We think that our network competitive advantage is very much intact. That advantage will continue to yield growth in service revenue. I'm going to take you through that in detail in terms of where we think our prospects are through 2030. Again, just like we saw over the last 10 years when service revenue grows, EBITDA grows faster. Just by the same token, our free cash flow grows faster still because our CapEx is flat and will be flat through the rest of the decade as we continue to enjoy our CapEx holiday. Free cash flow will grow even faster than a quite fast-growing EBITDA.
Really excited today. We announced our first acquisition in Iridium's history of a company called Satelles. Satelles, as it says, uses the Iridium constellation to provide PNT solutions for alternative PNT service and service to data centers, et c. This company is one that we've had a very longstanding relationship with. As I said, they use the Iridium constellation. They use a channel on our network to deliver their service. They were founded in 2013, offering satellite time and location. Their founders are thought leaders in the whole area of PNT. They're PhDs out of Stanford. They have something like 40 patents of their technology that they built, again, that rise on the Iridium network. They're regarded as the leading provider of positioning, navigation, and timing. Iridium's a 20% shareholder.
We, I think, acquired our first stake about seven years ago and we actually made three investments in this company. We've had a board seat for some time and so we have watched very closely the developments with their business and the progress of their business as the funnel filled up, opportunities got capitalized on. So we're really bullish about the business in terms of government opportunities. It's called alternative PNT service. I'll give you an example. So GPS, which is everywhere, basically. It's in all kinds of infrastructure. They cite a statistic that there's more GPS receivers on the Earth than there are toothbrushes. So it's everywhere. It's in critical infrastructure, but GPS is inherently vulnerable. It's vulnerable to bad guys spoofing the network and sending false signals or blocking the network altogether. You've heard accounts of that, I think, in the Ukraine conflict.
I'll give you an example of a drone, military drone. I think we believe it was a military drone. It was using GPS to get its coordinates and landed where it thought was safe territory. It found out it was bad guys had spoofed the location and the drone landed in enemy territory and the drone was taken over by the bad guys. With Satelles, what Satelles does is its GPS authentication. Using completely different satellites, the Iridium satellites, it determines with high degree of accuracy what the location is. So it would be used as an alternative to GPS or GPS authentication. So that has lots of applications in critical infrastructure and military applications and it's really resonating. I believe the market sizing for PNT is something like $3 billion by 2032.
So we think we're getting a head start in what is going to be a very fast-growing market. Similarly, in commercial applications, cybersecurity, so geo-authentication. So it's another layer of authentication. So you have your password, your secret code, and then your user ID, and then geo-authentication. So there's a chip in your computer that says you're in China even though you're supposed to be in Los Angeles and you don't get authenticated onto the network. So really excited about that. Now, Iridium PNT signal, the Certus signal, is 1,000 x stronger than GPS. And that's owing to the fact that, one, the GPS satellites are something like 15,000 miles from Earth. Ours are 500 miles from Earth. So the signal's stronger and it uses the paging channel on our networks, which is inherently strong. So all of those attributes make the Certus offering really compelling.
It adds to our differentiated position. It augments our portfolio with the government as well as commercial services and broadens our relationship with key technology partners. As I said, we've had a pretty significant peek under the covers here by our relationship with the company and we're quite confident that this service revenue that we got about in the area of $5 million bucks out of our sales to Satelles as a wholesaler to them for their ability to use our network, that that goes to $100 million in 2030 and there'll be retail revenues that Iridium will produce. It'll be accretive to us, our EBITDA on 2025, and their vision and culture aligns with Iridium's. Like I said, we know these folks really well. The purchase price is $115 million for the 80% of Satelles that we don't already own.
We'll finance it with a tack-on to our Term Loan B. Let there be no mistake, we will not take our foot off the gas in terms of our pacing of share buybacks. We're happy to tack on to our term loan, take the leverage up a tad because it'll de-lever back and all of our long-term targets for leverage will remain unaffected by this slight detour in taking the leverage up. But look for us to continue the pacing of share buybacks. So what I want to review with you is just the exact slides that we put up at our Investor Day in September of 2023. Let me say, we see no competitive development since our Investor Day. So what I'm going to largely do here today is to reiterate those long-term financial targets that we issued not too many months ago.
But the headline here is we don't see any material competitive development. And it's kind of the tenets of what we're building here, which is we're going to show you what our capacity for returns to shareholder are. Just let's review what the tenets were. Uphold the CapEx holiday. Some investors have known that we promised the CapEx holiday for as far back as 2010 when we began the construction and we delivered on it. Our CapEx holiday is very much here now and you're seeing the results of it and that will continue through 2030. EBITDA is going to scale based on operating leverage just like it's always done. As a result, we'll grow levered free cash flow.
We said in the Investor Day, look, when we think about our prospects for throwing off free cash flow, we're not banking on our equipment revenues, which were sky high in 2022, fully 30% higher than the prior five years. We said we've modeled them down and adjusted our cash flow expectations accordingly through 2030. And what you're seeing in 2024 with our short-term guide is we're saying, hey, equipment's going to be down as we get back to more normal levels. There's a point of view that says that moderation in equipment revenues is a harbinger of softness in the business. Let me say unequivocally that is not what the company thinks. We saw what spiked in 2022. What spiked was principally in IoT. The channel was worried about supply chain issues and they overordered. They basically stocked the channel so that they wouldn't stock out.
What we saw was that kind of that stockpiling start to come down in the second half of 2023 and continue down in 2024. So in terms of our EBITDA guidance, it's got some pressure in 2024, but that's simply just the air being let out of the balloon because of the supply chain issues. I'm going to show you some numbers in a minute that show that there is no sign of weakness in our IoT business. That's the central plank of our growth story and that remains very much intact. So the notion that equipment revenue's going down portends any kind of weakness we reject. We'll maintain our debt profile through 2030, maintain and grow the dividend. We grew the dividend, just announced it's going to grow by 6%, and we'll opportunistically repurchase shares. There's an authorization. Total authorized is $1 billion.
I think we have $700 million left to go. I just said, well, we will not slow down the pace, notwithstanding the Satelles acquisition. So here's the exact revenue projection that we made at the Analyst Day that service revenues, again, that which drives profitability, drives growth in EBITDA, will grow from $588 million in 2023 to $1 billion in 2030. Again, I'm reiterating that today. We identified and we'll reiterate that these are the five planks that we see for that growth. First and foremost, IoT. Our growth in IoT dwarfs the contributions from any other organic growth that we see. They're all other fine sources, but IoT is the principal growth driver. And finally, tuck-in acquisitions. So IoT, D2D, telephony, midband, and tuck-in acquisitions. So how are we doing against that six months later since when we made this projection?
Well, let's just look at IoT. What do you see there? Again, it's our biggest single growth plank in our 2030 projections. It grew by 13% in 2023. That's the same rate of growth that it grew in 2022. So no falloff in growth in 2023. And we expect 2024 to be at or better than 13%. D2D, we were way down the road with Qualcomm on a direct-to-device solution. Technically, it worked. It was a technical success, but Qualcomm couldn't get the sale closed, if you will, with the handset manufacturers. And so they kind of put that initiative to the side. We pivoted and announced Project Stardust, which will be a standards-based solution. We fully expect to be a meaningful player in the direct-to-device in 2030. We modeled Qualcomm very conservatively, so we think that Project Stardust will take up where Qualcomm left off.
Our telephony and midband businesses are performing as we expected. We just announced the Satelles acquisition, a central plank of our growth story, which we think is $100 million in revenues in 2030. As I said, very confident in that number. So you put all that together and what do you see? You see an 8% growth CAGR from 2023 to 2030. So in line to actually a little bit better than what we did over the past 10 years. So the growth story is intact. What we know about growth and service revenue is EBITDA grows faster and free cash flow grows faster still. So all systems are go from our perspective based on our results since the Investor Day. One final point on this chart. What you don't see on this chart as one of the central planks is our broadband business.
It's a fine business, but it's less than 10% of our revenues. Where we see our space is as a backup to VSAT and maritime. We said for a long time that we have no interest in playing in the commodity broadband space. Well, a small portion of our broadband revenues, where we were used as a primary, basically touches the broadband or the VSAT space. What we've seen is a substitution of our product for lower-priced VSAT. We've said that that'll work its way through in the next couple of quarters. But the point is we saw that coming. So when we configured our $1 billion, we were aware of what the pricing levels were there and factored that in and view this as a non-issue. But much has been made about it by certain analysts and we just say, A, we knew about it.
It doesn't affect our $1 billion, and isn't that what you care about anyway? And view it as pretty consistent with what we've said: commodity broadband is something we have no interest in. And so when you roll all that together, that generates $3 billion in capacity for shareholder returns. And so at the time, it was 50% of our market cap. Today, it's 85% of our market cap. So that one thing has changed. And so call that to investors' attention. But the meat and potatoes of our story has remained unchanged from when we told it in September, and we're as confident as we were then in telling it today. And so the investment highlights strong cash flow generation. We see leverage at below 2 x as we exit 2030. We've had $1 billion in authorizations for share repurchases. That's a capacity for $3 billion in shareholder returns. We increased our dividend to 2% yield with stocks at current levels and the free cash flow yield is fully 9%. So with that, I'll turn it over to you, Rick.
Right. Come on over. Appreciate those words. Yep. Obviously, the stock has come under pressure. What I want to start with is how do you get the comfort of the competitive power of your moat, your ability to operate against what I'll call the boogeyman? SpaceX, Starlink's got everybody spooked out there that they're going to come in and just run roughshod over the satellite space and maybe terrestrial space. What makes you different and what makes your moat powerful?
So the L-band, first and foremost. So they're in the K-band. We are routinely put on ships as a backup to Starlink. And in fact, it's a really compelling offering versus Viasat/Inmarsat. The combination of Iridium with Starlink is way faster, both as a backup and as a primary. Starlink's faster than Viasat. We're faster than Inmarsat and cheaper. And so that's an offering that's going to resonate, but it also highlights the limitation of the Starlink offering. It's in the K-band. And so that's what we take comfort in, first and foremost. And the notion of competition, if you just think about go through our segments. In commercial voice and data, we've had competition there for a long time. And it's cheaper. It's cheaper than us. Inmarsat's cheaper than us. Globalstar's cheaper than us. But 80% of our users are enterprises and first responders.
Cheap isn't what matters. Coverage is what matters. Functionality is what matters. So there's been cheaper alternatives for 15 years to our solution in commercial voice and data. So we don't think a cheap alternative is going to resonate any more than it has historically. In IoT, I just showed you we're going to be 13% or better in 2024 in a much bigger business. How are you able to keep that up in the face of Starlink? They're nowhere near that space, okay? You should consider what's fastest growing there is in personal communications, so devices from Garmin. There's competition to Garmin. There's Bivy and Somewear and other players who compete with Garmin. Every one of them uses the Iridium network. So why is that? It's because of the unique functionality of the Iridium network in the L-band. It's small antennas, low power requirements.
Take it anywhere and it works. And so that's a tall order to fulfill. Our DoD contract, okay? DoD contract was the last time it was let in 2019. 2019, do I have that right? Yeah, 2019 was a sole source, 7-year sole source. Why is that? Because they want traffic to be able to be encrypted and land in their proprietary gateway in Hawaii. Well, nobody else can do that, right? It doesn't matter how fast it is. It's these unique requirements that just because you have fast data speeds don't get the job done. So we just go through each element of our business and say, geez, we don't see any major impact from them based on our situation.
You go back to Bell Atlantic band. I go back to BellSouth. A lot of people in this room probably couldn't spell L-band if we gave them the L. What does L-band mean? Why is it so different? Can't someone else get some?
L-band is, it's allocated by the ITU. And so we have a global allocation of L-band that Motorola got for us, geez, in the early 1990s. And so there's no more to be had. And so the players that have it are Inmarsat, Globalstar. Globalstar, basically, Apple contracted for 85% of their capacity for their direct-to-device solution. And so there's not L-band to be had.
Okay. You lay out a compelling story about the free cash flow, revenue growth, EBITDA growth, free cash flow growth, where you're focusing on the free cash flow. We chatted last night at dinner a little bit about given the stock used to be the buyback was 50%. Now it's 85% of your value out there. Why not take leverage up? Why not go after this stock opportunity if you think it's as attractive as you are laying out for us?
Well, we're taking leverage up. I just told you we're going to take leverage up. We wouldn't consider taking our foot off the gas when the stock buyback in the face of the Certus acquisition. We'll take leverage up because leverage is going to come right back down. We'll exit 2030 below 2x . The business naturally delivers. So we will continue. I mean, we bought $700 million worth of our stock back already. There's another $300 to go get. We'll exhaust the current authorization and we'll put up another one.
Good. Okay. What do you think the catalyst is to get the stock going forward then? What is it the market doesn't get that you get that you'd like them to get, but what's the catalyst we can look to and say, hey, this will get people comfortable?
We'll put up the numbers. Let me say we will. But there was a point of view that there was weakness in our 2024 numbers. Let's just take that. We had a longstanding guide, have a longstanding guide in what our service revenues would be that we put out in 2021. And we said our service revenue growth will average high single digits for the period between 2023 and 2025, right? And we reiterated that all the way through. We're going to beat that because all you have to believe is that in 2025, we do 5.5 or 5.2. So that guide is completely retired, if you will. There was a point of view that it was soft. Well, then 2023 had the benefit of a price increase. We said it was non-recurring. It'll happen kind of every few years, 5 years or so.
So if you do the math, there's nothing soft about the 2024 guide. And I just kind of laid on you $100 million worth of revenue in Certus that if you plunk that into your equation, you can see that it's not that what we've got to do is not that far off of where we're running currently. And we said tuck in acquisitions. So something else like a Certus, we'd be all over that. And just for the avoidance of doubt, Certus is not only $100 million worth of revenue that we're very confident in. It's accretive to the $3 billion in capacity for shareholder returns because it pays for itself over that time more than pays for itself over that time period. So we'll pay off the $115 and have extra cash left over to do more shareholder returns. So there's other opportunities out there like that that we're looking at. So all I can do, Ric, is say we're very confident in our growth profile and we're a cash flow machine and we'll continue to be so through 2030.
Okay. The accounting life change confused some people. I viewed it as good news that you're extending the accounting life of your satellite constellation from 12.5 to 17.5 years. Off the record, I think it's longer than that even, probably. Walk us through what that change did to kind of the view on CapEx. But was it due to book revenue and EBITDA as well?
Yeah. So when we launched the constellation, we had to pick a depreciable life. And so you have to have anecdotally, we said, look, our last constellation lasted 20 years and was working just fine when we decommissioned it. We think this is going to last at least 20 years. That means there's a 10-year CapEx holiday. But the design life that the manufacturer published was 12.5 years. So a principle of accounting is conservatism, right? And so you use the conservative number until you have a basis, not an anecdotal basis like what the last one lasted. So what we have now is we launched our first satellites in 2017. We see the performance. We haven't had a failure of a satellite. We just launched 5 new spares.
You put all that into the equation and then you can say, the evidence shows that 12.5 years is too short. It needs to go to 17.5 years, which actually lines up really close with the CapEx holiday. So now that would say the existing constellation is fully depreciated in 2035. And that feels like that's about when you'd be launching some new satellites that you start spending money on in 2031. So it's kind of it all lines up. It has the effect of reducing our hosted payload revenue recognition because the revenue gets recognized over the useful life of the constellation. So it's a finite amount of revenue. Take Aireon, for example. Their hosting is $200 million bucks. We were amortizing that over the 12.5-year life b ut now it says, whoa, we're going to have to bring that in over a longer period of time. So it brings the revenue recognition down by $9 million bucks. Has no effect on cash flow, etc. It's just an anomaly of the accounting rules.
Okay. Last one I want to close with. We'll take most of the questions down in the breakout because obviously, a good deck, good presentation for people that know the story or don't know the story. But clearly, the direct-to-device, the satellite communication into smartphones, disappointing with Qualcomm. How far back do you think that knocked you as far as timing? And remind us how much you have to spend to kind of get back to where we want to be. And competitively, what does that mean then?
So yeah, Qualcomm was a disappointment, particularly because the technical solution worked just fine. I mean, one thing, it was like, yeah, we missed our guest and this doesn't work. It works just fine. But it was an issue of a meeting of the minds between the seller and the buyer. So the seller was Qualcomm and the buyer was the Android community. And the Android community and we weren't at the table, but one can imagine that it had to do with an increase in costs to the BOM, right? And wedding yourself to a proprietary solution and one vendor. Well, the standards-based solution is not. There's no wedding to one vendor. There's going to be lots of chip makers that make the standards-based chip. And so yes, it's a delay. We now need to work on our network.
You see us spending $5 million bucks incrementally this year on R&D. We took our CapEx up a bit. So if Qualcomm had occurred, that wouldn't have happened. So look, you have to roll with the punches here and we're rolling with the punches. Will the standards-based solution be less than Qualcomm? I don't know about that. My bet would be it's going to be a bigger opportunity because you don't have the purchase price objection of a handset manufacturer. They buy the chip from any one of dozens of chip manufacturers. Costs them nothing incrementally. And so whereas the Qualcomm chip likely would have found its way, there would have been a line drawn by Samsung, hey, we're only going to put it into this number of high-end phones. It's not going to go in everyone.
Whereas now, it can proliferate the whole product portfolio because there's no incremental cost. And so now the Verizon rep will be able to sell satellite connectivity to a lower-end user and a higher-end user. So we'll see how that all settles out. But it's an opportunity any way you look at it, from not just through 2030, but this is a decades-long trend. I mean, there's going to come a day when every smartphone has satellite connectivity in it, it seems to me.
Great. Well, let's break it there. We'll take the rest of the questions down in breakout. Thanks, everybody.