Amadeus：旅行行业的IT底座 - [Business Breakdowns，第237期]

Today, we are breaking down the travel IT giant Amadeus. If you have booked travel with a travel agent or done business travel and used their system, there’s a very good chance that was done through Amadeus’s plumbing. They are IT for travel, and it goes well beyond what you see or interact with as a consumer, doing this within the airlines and hotels, managing inventory, and handling all of the behind-the-scenes work that makes these logistical engines actually flow.

My guest is Ben Needham, portfolio manager at Ninety One Asset Management. He joined to cover why Amadeus might have some misconceptions in the market, the business and how it has grown and captured such dominant market share over the years, and what that looks like in terms of a value proposition on all sides of the equation. We also get into the AI debate and what that might mean for Amadeus, and what a strong balance sheet could represent for Amadeus relative to some of its peers.

So, a very interesting niche business with a lot to discuss here, and Ben was a great guest to cover it all. Please enjoy this breakdown of Amadeus. All right, Ben, I’m excited to have you here to talk about Amadeus. This was admittedly one that took some time for me to wrap my head around as I was doing some initial research on it.

An easy place to start was just to get your overview description: how you would paint the picture of this business to a listener who is vaguely familiar or not familiar at all with this name.

1. Amadeus Runs Travel IT

Thanks for having me. I think the best way of summarizing the business is as the gorilla of travel IT, but a friendly gorilla, which is able to grow well ahead of a structurally growing market—i.e., travel. Importantly, this is in a low-risk way, given its diversification across geography and travel provider customers, and the inflation-linked nature of how the revenue model works.

So, what do they actually do? There are 3 main parts of the business to understand. Firstly, they have a distribution business, which links up travel sellers with travel providers in the indirect travel channel. It essentially aggregates content from the travel providers—i.e., airlines, hotels, and rail companies—and supplies that to a pretty fragmented global travel-provider ecosystem.

This includes corporate travel management companies, online travel agencies, bricks-and-mortar travel agencies, super apps, meta channels, and perhaps AI agents soon. If you’re making a travel booking via a travel agent or your company is booking a business flight for you, there’s a good chance Amadeus is the plumbing that enables it, given that they have over 50% market share in distribution, which is actually rising.

Distribution perhaps wrongly gets a lot of the investor focus, as the biggest business by profits is actually now the air IT business. This is predominantly the inventory management, reservation management, and departure control system for over 50% of the airlines globally, processing over 2 billion passengers boarded per annum. When you’re making a booking on, say, Lufthansa, Amadeus will process that booking. When you’re boarding a Lufthansa plane, Amadeus will facilitate the information flow all the way to the travel gate.

Finally, Amadeus also enables travelers to make reservations in hotels via its scaling hotel customer-reservation business. Amadeus, in effect, is uniquely positioned right the way across the travel ecosystem, enabling the flow of content, bookings, transactions, and ultimately travelers.

It’s interesting as an industry. I would’ve previously assumed that most airlines built this system internally or had some type of internal software that would do this for them. Can you frame the space as a whole a little bit—how Amadeus fits into this ecosystem, whether there are competitors, and how the technology evolved into a place where they’re such a key player in the industry?

The air IT part of Amadeus is a community-based platform, and what that means is they’re able to share economies of scale with their customers. They can amortize more of their R&D spend across more customers as they grow. That’s why airlines, which are pretty cyclical and pretty capital-intensive businesses, outsource these types of operations to people who can do it better than themselves.

That’s why, typically, when you look at the airline space today—or the travel IT space today—just over 80% of air IT is outsourced to the likes of Amadeus, and only 20% today is actually in-house. That’s the background to the air IT ecosystem and business.

When you think about the market shares, Amadeus has over 50% market share within the air IT space. It processes over 2 billion passengers boarded per annum out of a market size of just over 4 billion, excluding China. The second-biggest company is Sabre, which processes 600 to 700 million passengers boarded, and then the majority is mom-and-pop players and/or in-house providers.

I think I read an anecdote that Lufthansa had pulled away from Amadeus back in the mid-2010s because of its industry power, but then quickly reversed that decision and signed up with them again. It shows an indication of core competency, focus, and understanding where you can outsource and make that effective for your own system as well.

What is the origin story of Amadeus? How long have they been around? They’ve clearly built up this advantageous position in the market. What does that trace back to?

2. Amadeus Was Born From Airlines

The business has been around for a long time, and the history is really important. They were actually created by the airline customers. Distribution systems were created in the 1960s and 1970s to enable airlines to process bookings and get close to their customers, in effect, and this was before the dot-com era.

In the 1960s and 1970s, the airspace was deregulated. There was globalization, and within that, more airlines were created and more agencies were also created. There was a problem with the way the distribution systems worked: because they were owned by the airlines, they gave preference to their own inventory. That was anticompetitive, so regulation came in following that deregulation.

The airlines decided to sell off their distribution systems, and they merged. That’s when Amadeus was actually formed, in 1987, following the merger of the distribution systems of Lufthansa, SAS, Air France, and Iberia at the time. That formed the distribution business of Amadeus today.

In the early 2000s, they created something called Altéa, which was the air IT business. That is the customer reservation system, the inventory control system, and the departure control system that sits behind the distribution system as a whole, enabling the orderly processing of bookings.

Since then, they’ve created the hotel IT business. That was in the 2010s. When I actually started covering the stock over a decade ago, they didn’t really have much revenue or free cash flow there, and that business is going from strength to strength. They also bulked out parts of their air IT and hotel IT businesses via M&A.

It’s interesting to hear about the horizontal expansion across travel. I certainly love businesses that were born from an industry consortium founding something together and then spinning it out on their own. You often see quite a few success stories in that category, with Visa being one of the most obvious examples.

I think you’ve tapped into the segments of the business. Is that how they split out revenue—by distribution, air IT, and then hospitality? Are those the 3 buckets?

There are 3 main divisions. The main one in terms of profitability is now air IT. 15 years ago, that was 20%-ish of group profits. Now it’s up to 50%.

35% of group profits are from the distribution business, even though that gets a lot of the limelight and a lot of the focus. Then 10% is hotel IT and other, which is 10% of group profits.

The hotel IT business is the most embryonic. It has the lowest margins because of that. They’re scaling very quickly, and there’s a lot of investment to enable the scaling of that business. The air IT business is the most profitable, with almost 70% contribution profit margins, which are pretty formidable. The distribution business’s contribution margins are in between hotel IT and air IT.

You mentioned some of their competitors on the air distribution side of things. In air IT, are those the same competitors? You mentioned Sabre, which was a name I came across quite a bit. Is there more competition in that space? Less? Does it look similar to distribution?

The competition is broadly similar. Within the air IT business, Amadeus has 50% market share, and then there’s Sabre, which has just under 20% market share, along with smaller operators and the airlines themselves, which in-house the technology.

If you look at the distribution business, it’s a similar type of shape. Amadeus has about 50% market share, Sabre is a clear number two with about 30% market share, and there’s also Travelport, with 20% market share today. It’s important to note that Amadeus has a lot more scale than its nearest competitors. Travelport, for example, doesn’t actually have an air IT business. It’s just a distribution player. The advantage of having both is enormous.

I have a sense of what’s going on with buying an airline ticket, accounting for that in terms of the transaction, and how that’s actually working through the system. In terms of IT, does it get into accounting? I think I read something about dynamic pricing, which can always drive me crazy in terms of seeing what flight costs do from one week to the next. But could you say a little more about the different ways that Amadeus is serving its customers, whether it’s value-add versus housekeeping?

Depending on which division we’re talking about, they’re doing slightly different things. The distribution business is a network business, and they’re playing a role in content aggregation, taking inventory from the airlines. That can be in traditional EDIFACT form or NDC form, and they provide or supply that inventory to the travel agents. So that’s serving the indirect channel. That indirect channel is approximately 1 billion bookings per annum.

On the air IT side, they’re doing the reservation management for the airline. They’re doing the inventory control for the airline and the departure control system as well. They also have revenue optimization tools, revenue accounting tools, and fare optimizer tools. Really, what they’re trying to do there is enable their airline customers to generate more revenue per passenger and make smart, personalized offers to the passenger. Airline customers can benefit from those initiatives, and that’s clearly key in an industry which, as I said before, is pretty cyclical, quite capital-intensive, and very competitive. Generating that revenue from, say, ancillary income is really key, and Amadeus enables that to take place.

When they’re taking the inventory and bringing it to the travel agents, they’re not actually owning or having that risk sit on their balance sheet in any way. Is that correct, just in terms of the process?

That’s absolutely true. They will create offers in the distribution channel via the global distribution system that they have, and they will basically package those offers up and provide them to the agencies, but there’s nothing on the balance sheet.

AI is something that comes up when talking about any space, particularly within any IT system. What would you say are the implications of AI for Amadeus and for the broader travel vertical?

3. AI Will Not Break Amadeus

There have been some industry discussions about the implications for Amadeus of the potential adoption of AI agents at the top of the travel search funnel, with question marks over whether this will disintermediate the distribution business. I think this is largely a storm in a teacup, and I actually think they’re very well placed at the top of the funnel if travel search becomes AI-agent-led.

I think this for a few reasons. Firstly, you need to remember that they’re the IT backbone for their customers. The customer reservation systems or order management systems are mission-critical and will still be required in an all-AI travel world. This part of Amadeus represents the majority of profit today, at 60% or more of it.

Secondly, would AI agents really want the hassle of setting up the infrastructure to aggregate and orchestrate industry content when players like Amadeus already do this in a very economical way, with economies of scale enabling a very low industry take rate? My personal view here is that AI agents might do things to improve top-of-the-funnel search, thereby potentially competing with online travel agencies, where the industry take rate is incidentally also much higher. But the agents will still need an aggregator to pull together disparate data from the travel ecosystem, with many different technology standards at play between NDC and EDIFACT, for example.

Thirdly, there’s also an airline push toward dynamic and real-time pricing, essentially enabling them to generate more revenue per booking. This will put more demand on AI agents to get the right offers to the customer. With the move to NDC, there’s actually been a big increase in something called the look-to-book ratio on airline websites, which can get expensive, and Amadeus helps to solve this problem.

Fourth, one could argue that the great thing about AI will be more personalization alongside the use of dynamic pricing for the airline, which could lead to a snowball effect for their offer and order management system, Nevio, as well as many other products.

Fifth, a big part of the indirect channel and distribution revenues is from corporate travel management companies, where the complexity of booking is high. This complexity likely means it will be difficult for AI agents to totally displace corporate travel management businesses, and even if they are displaced eventually, the AI agents will need a content aggregator and/or someone to orchestrate the data.

Finally, if agents really do lead to chronic disintermediation of the distribution business, then peers that are much more distribution-heavy, with a much smaller IT offering and a lot of operational and financial leverage, are in deep trouble. So Amadeus can arguably swoop up even more market share across all areas of its business, including air IT. All in, I feel pretty good about AI as it relates to the distribution business and the business model as a whole today.

From a revenue perspective, how much of this is per-booking based on the amount of air travel versus contracted revenue? What’s recurring versus what’s exposed to cyclicality? How do you frame that? How does the company frame that? How do you think about that?

4. Amadeus Monetizes Every Booking

Amadeus is predominantly a transaction-processing business, and the way the revenue model works is a bit akin to Visa. They take a cut of the ticket price every time there’s a booking made via the indirect channel and it goes through the Amadeus plumbing. They’ll take a booking fee from that. Then, every time there’s a booking placed on a dot-com site with the likes of Lufthansa, they’ll take a very small fee from that.

There are pros and cons to that. The con is that when air travel volumes fall, like they did in the pandemic, Amadeus’s revenues really drew down because of that. The pro, and the positive from the airline’s point of view, is that they become a variable cost for the airlines. So when volumes do fall, there’s no cost for the airline. Their take rate from an industry point of view is very low.

On the air IT side, the take rate is about €1 per passenger boarded. On the distribution side, the take rate is about €6. We think there’s a big runway for that to increase over time as they provide more value for their channel and their customers. There’s an enormous untapped pricing opportunity over time, as Amadeus can share in the value creation with its customers by enabling more revenue per passenger to come through for its airline customers.

Do you have any sense of what the airline would theoretically generate from the passenger in terms of revenue, relative to €1 and €6 per passenger?

It depends if you’re doing short-haul or long-haul flights. If you’re doing long haul, ticket prices can be anything north of €1,000, so it’s less than 1% on both sides of the business. If you’re doing short haul, traveling from London to Edinburgh, then maybe you’re a slightly bigger part of the overall cost. But still, we think the value proposition is a very good one from an industry point of view.

How are those fees determined? Is it some specific rate per transaction? What are the pricing mechanics of Amadeus selling into the airlines that use it, or whoever is the end customer?

On the air IT side, typically they go into 10- to 15-year contracts with the airline customer, and it’s pretty modular how they sell their products. If you’re using Amadeus for its passenger service system on the air IT side, you’ll be using the departure control system, the inventory management system, and the reservation management system. You’ll be using those 3 modules. They’re fixed-price in nature, but they’re inflation-linked over time.

That has some positives if you do enter a down cycle. Typically, what happens in a down cycle—pandemics aside—when there’s channel weakness, is that volumes go a little bit negative, but not too negative, because what the airlines do is lower prices. Amadeus doesn’t suffer from lower ticket pricing because it’s fixed-price and linked to inflation.

So at times of industry pain, the revenue streams and the free cash streams are very defensive. If we go to the distribution part of the business, it is slightly different. There’s something called home bookings, and then there’s also something called away bookings.

When you’re doing home bookings via a travel agency, the fees are pretty low. Most people, when they’re making airline bookings, if they’re only traveling locally, will typically go direct to .com. That’s the reason those prices are low.

If you’re doing an away booking—so if you’re traveling from, say, the UK to Australia, and Qantas wants to get its inventory to an agent in the UK to enable them to sell more—Amadeus facilitates that process, and they can therefore charge a higher price for enabling that. The distribution business, a key concern with it is that there’s disintermediation and more people are booking via .com.

But for those away bookings, the role that Amadeus plays is a really important one: enabling inventory to be sold in different countries. That’s why they can charge a higher price for that. Actually, in the distribution business, the majority of revenues today—not necessarily volumes, but revenues—are derived from away bookings, which is a bit of a misconception that we think people have on the stock. I’m not sure they’re quite aware of that revenue mix being predominantly away bookings.

Can you talk me through the net fee model in distribution? How exactly does that work?

That’s actually a good question, Matt, and an important one in the context of the discussion we were just having. An important element of how the distribution business works is that the airline will pay the distribution system company, which will then actually pay the travel seller.

In effect, this means that the gross fees—the €6 which I mentioned earlier—are very different from the net fees, which are €3, and therefore the industry take rate is even lower than what I assumed. Again, this is a really important point that you’re touching on.

I think the interesting thing to pull out here is that, in the evolution to an NDC model, the industry participants first thought that this would be dilutive to the revenue per booking, but actually it wouldn’t be dilutive at all to the net fee per booking. If you look at the contribution profit per unit that Amadeus has made over time, which in effect is a proxy for net fee per booking, you can see that it’s actually steadily increased over time, which is a good sign of system health.

As the NDC model is evolving, the actual net fee model that they have in place—where the airline pays the distribution company, which then pays the travel provider, i.e., the travel agent—has still mostly remained the case. The airline often does not want a direct relationship with the agency because it creates more of a cumbersome process.

So the move to NDC, in all likelihood, probably won’t be dilutive to the revenue-per-booking economics and definitely should not be dilutive to the contribution-profit-per-unit economics. I should also add that this evolution to NDC is a gradual process. The airlines have to be ready to service the channel with NDC, and particularly for the full-service carriers, the rollout has been pretty slow, so they’re ensuring that the channel is ready for them to provide an efficient service in this NDC switch.

You mentioned that there is a concern about disintermediation, or maybe less use of agents and more direct bookings. Is there any data that shows that piece of the market has shrunk materially over time? Or where does it stand today, just in terms of percentage of the market? And if you have historical reference, that would be notable too.

I think approximately 25% of bookings are done in the indirect channel. If you look at Amadeus’s booking trends in the distribution business over time, typically they’ve grown slightly below global passenger volumes. That’s because of the prolific growth in low-cost carriers, which are a cheap form of travel, and the fact that you typically go direct to .com when you’re making those bookings.

The other aspect here is that there’s something called direct connects. Sometimes agents are directly linking APIs with the airlines. That process typically only happens in the home markets, where there’s a lot of volume going through that channel.

When the volume’s a bit more fragmented, and when you’re having to include small inventory loads from airlines in a different country, it becomes very difficult to manage tons of APIs. That’s why you go through the distribution systems that Amadeus, Sabre, and Travelport provide, as opposed to disintermediating them and having direct connects. But that’s another aspect to think about.

Should I assume that there’s a lot of overlap in the distribution and airline IT customer base, in terms of those that use one are usually using both?

I think that’s right. Having both an airline IT and distribution business is a commercial advantage for Amadeus, and to a degree Sabre, in that you have more revenue to spread across similar customers, so you can therefore arguably charge lower prices for one or the other.

The other aspect of this is that there’s a slight halo effect in terms of bookability. If you are the airline IT provider as well as a distribution provider, you know that the inventory is live and real. The bookability can therefore be better for the channel if your distribution system and your airline IT system are the same.

There’s a commercial advantage, and there’s also a potential customer advantage in that you don’t get cached data that’s wrong, where you go to buy an airline flight and it’s actually not there, which can happen.

On the hospitality business, you mentioned airline IT was a much smaller percentage of overall revenue years ago versus where it is today. That has been something that has taken up more revenue mix—a positive mix shift seen from that business. Is the hospitality business a similar growth engine for them? Is that a target market? I think you mentioned it’s 10%, fairly small today, but would you expect that to be materially higher in the future?

5. Hotel IT Drives Expansion

It’s 10% of group profits and 15% of group revenues. They’re providing similar services to what they provide to their airlines in the airline IT business. They’ve essentially created a customer reservation system, which is a community-based platform. They’re trying to win more and more contracts with large-scale hotels.

Similar to the airline IT business, you can amortize that cost across more customers, and you get an economies-of-scale shared model, so more customers want to come to you, and it becomes a self-fulfilling process, as the Air IT Altair proposition has been over time.

The good news is that it’s scaling very well. They won a landmark contract with InterContinental Hotel Group in 2015. Since then, they’ve recently won contracts with both Accor and Marriott, and Ascott Limited. They really have established a scale position within that customer reservation proposition, and they are now the leading hotel IT company globally.

Competition is very fragmented. Their nearest peer was Sabre. They actually were a fourth seller of their hotel IT business because they have balance sheet issues to address. One of their nearest competitors selling its business is great for Amadeus because they’re retreating from the market.

Do they take inventory in a similar way and bring it to travel agents? Is it a similar model in terms of that being a core piece of what they’re doing in the hospitality business?

They also do distribution, but that’s only a smaller part of the hotel business. It’s mainly the reservation system managing inventory and enabling their customers to generate more revenue per room.

If you go onto InterContinental Hotels Group’s website and you’re making a booking in one of their hotels, you get offers: Do you want a marble bath? Do you want a PlayStation? That wasn’t there before. What they’re doing is enabling their customer to generate more revenue per room, and then Amadeus can share in that value creation with the channel.

I’m sure they push the travel insurance option as well. Is that a piece of their offering?

As part of their distribution business, they do have partnerships with travel insurance companies, but it’s only a small part of the business.

In terms of the revenue trajectory, do I assume this is volumes based on air travel plus some pricing mechanism? You’re kind of seeing that as the top-line growth rate. Is there anything else that goes into it? And what does revenue growth tend to look like in a mid-cycle or normalized environment?

6. Amadeus Has Multiple Growth Engines

Typically, travel grows at 1.5 to 2 times GDP. If you look at the volume growth that they’re able to achieve without winning new contracts, it’s typically in the 3% to 4.5% range.

I’m being guarded with that because of geopolitical risk. I think you have to guard for that: American shutdowns recently, the war in the Middle East. In today’s world, that’s just pragmatic. Then there are inflation-linked contracts on all sides of the business. If global inflation rates were 2% to 3%, you should see that come through in the top line.

There is also an upselling process here. With the airline IT business, they’re about to embark upon a potential massive structural growth change in that part of the business. Their airlines want to modernize their retailing platforms and move to order management systems away from legacy passenger service systems.

Why the airlines want to do that is because they want to provide a better service for their passengers. If there’s disruption on their journey, they’re able to reach out to their customers and tell them that there’s going to be disruption and that the flight is going to be delayed, actually before they get to the airport, which currently, because of the cumbersome nature of how the tech stack works for airlines, isn’t possible.

But they also want to sell smarter. They want to personalize the sales process for the customer in order to generate more ancillary income and basically drive higher revenue per booking. If Amadeus can facilitate that transition, they have a product called Nevio, which is in its early innings. It has 4 customers currently signed up to it: Finnair, Saudi, British Airways, and Air France-KLM.

They’re going to be able to charge a higher revenue per booking, and we should see that come through in a better top line as well. It’s also important to note that industry experts think this transition to order management systems will be a mid-teens uplift in revenue per booking. If they get it right and can generate a 5% to 7% revenue uplift per passenger with these new order management systems, which is what the early look is for the airlines that are using Nevio, arguably that’s lowballing it, particularly in the context of only taking €1 per passenger boarded today. That’s their current rent extraction in the Air IT space.

So could that be a 50% uplift? Could it be a 100% uplift in the next 5 to 10 years? Potentially. They deserve to share in that value creation with their customers. Then there’s the market share dynamic. If you track the distribution business and the Air IT business over time, they’ve constantly taken market share from their nearest competitors. That can give another 1%, 1.5%, to 2% to the top line.

When I take all that together, air traffic at 3% to 4%, plus inflation, gets you to 5% to 6%. Revenue per passenger in Air IT can get you another couple of percent. That can get you to 7% to 8%. Then market share, I think, can get you to 8% to 9%. I think it’s a high-single-digit top-line business.

Within that, distribution should remain pretty turgid. Keep growing at 4% to 5%. Grow volumes a little bit less than air traffic volumes, but with a little bit of price. Then Hotel IT, because it’s a lot more embryonic, and they’ve currently got Marriott, Accor, and Ascott Limited migrating onto their platforms in the next year and a half to 2 years, should grow at 15% to 20%, probably for the next couple of years. There’s a lot of growth opportunity ahead of it.

Based on my own personal experiences and the number of altercations that I see on my social media feed happening at airports dealing with the inventory systems, it certainly seems like that’s an area these airlines could benefit from upgrading. What does the timeline look like for something like that? From an execution perspective, where does the confidence lie in their ability to execute on doing that efficiently? Is there a time in their history when they’ve done things like this where the execution has been on par with what’s needed?

I think they’ve executed very well. Amadeus goes about its business quietly but brilliantly and very effectively. From a capital allocation point of view, we really admire that.

If an airline is going to enter an Air IT contract, it takes it seriously because this is the central nervous system of its business. Managing the inventory, managing the reservation systems, and managing the departure control systems needs to be done properly. Revenue is everything for these companies, and managing that effectively is key.

The RFP processes take a long time. They can take years to get across the line. The reference contracts that they have on the Nevio side of the proposition are going to be really key. Listening to what British Airways and Finnair say about the Nevio proposition, now that they’re already wired up to it, and how it’s actually benefiting revenue per booking over time, is going to be key.

Obviously, the industry will listen to those companies, and then you could get a snowball effect because you get FOMO. It could commercially really help these companies that are early adopters. It can take a long time, but I think it’s a bit of push, in terms of Amadeus selling the proposition properly to the airline customers, and a bit of pull. They’ll see some of the early benefits for those companies that are tapped into Nevio, and then you can see a big growth phase play out.

Altair has already been through this. Its original passenger service system snowballed and scaled from the early 2000s through to today. People are still signing up to the Altair proposition, which is its passenger service system—a more legacy system that airlines hooked onto. They’ve done a great job of that, and that’s why they’ve got over 50% market share within the Air IT space.

Just a final point to note is that balance-sheet considerations are really key when airlines are negotiating with air technology companies. If you’re very levered and if you’re losing market share, in an RFP process, that’s a problem. It’s an oligopoly within Air IT, but the number-two player is pretty levered.

Mission-critical things like those systems require some thinking around the financial positioning of your counterparts. Taking it down a level on the margin side of things, I have some understanding that they’re a network business connecting different parties here. What does that look like from a margin perspective? Where have they ranged historically? I know there are differences between the segments, but at a consolidated level, and to the extent that the nuances of each segment are relevant, what would you point to there?

The gross margins are very attractive. They’re in the mid-70s, and they’ve been pretty consistent through time, which is remarkable, really, given how embryonic the Hotel IT business is within that. They’ve been scaling that business for 10 years. There’s been some dilution there from the scaling of that Hotel IT business.

EBITA margins are in the high 20s. They touched 30% a few times. When you’re processing an incremental booking for Lufthansa or Air France, or processing an incremental booking on the distribution side of the business, the incremental cost is next to nothing. The business is very scalable, and that affords nice economics within the business.

They reinvest that when they are scaling, and when there is growth and volume with the airlines, they will reinvest it. If you look at their R&D spend over time, the R&D-to-sales ratio has actually reached new highs more recently, currently at the 22% level, which is very high. If you look back 15 to 20 years ago, that was actually 10%.

They’ve been investing in their Nevio proposition and in their IT. They’ve been investing in NDC within the distribution business. There’s a Microsoft partnership. There’s been a replatforming of the business to make it cloud-enabled and enable them to sell native products, where the cost to serve is going to be dramatically improved because of that.

Despite all of that, and despite an embryonic Hotel IT business, they’re still almost making peak margins at the 30% level. The financial model is a good one, and it’s worth noting as well that the cash conversion is good.

They have a negative working-capital balance. They do capitalize customer implementation. When they’re doing a big IT project, if they’re installing Nevio or installing their Hotel IT proposition with the various customers, they take on that cost themselves, then capitalize that cost. They make up for it when the room bookings or airline bookings come through over a 10- to 15-year view, which is a nice thing for the value chain in that you’re only actually spending once bookings are happening.

You’re not actually spending the money as a hotel provider on that transition. That’s also why you need to be well capitalized, and that is a big differentiator for them versus their competition.

I kind of bucket R&D into capital allocation in a similar vein to CapEx, but what do they do with excess capital in terms of distributions? How do you think about them beyond some of the operational needs from a capital allocation standpoint?

First and foremost, they’ll reinvest in their business organically where they think their return on capital stacks up. You can see that, as I said, in the R&D spend over time.

An interesting statistic there on the R&D spend is that their R&D spend is actually equivalent to the total revenues of the number-three player within distribution, Travelport. Their R&D spend is equivalent to 50% of the revenues of their second-nearest competitor, Sabre. They’re reinvesting a lot of capital to fortify and differentiate themselves versus the nearest competition, and that’s a really important point to make.

They will also do inorganic spend, and I think their track record has been very good here. They bought a business called Navitaire in 2016. That was a passenger service system in the Air IT vertical that overweighted low-cost carriers. Amadeus, because of its history with Air France and Lufthansa, was overweight the full-service carriers and the hybrid carriers, and it had a gap in its portfolio, which was the low-cost carriers. Navitaire was a leading business there. It actually has Ryanair as a major customer.

They also bought a business called TravelClick in the Hotel IT space, and that was to give them more penetration with the independent hotels. So, doing exactly what they do with the likes of Accor, Marriott, Ascott Limited now or soon to be, and InterContinental Hotel Group in the independent channel.

Those larger deals that they’ve done, we think, make a lot of strategic sense, and the return on capital has been good. Their capital structure is very good. Their leverage is less than 1 times today. They degear very quickly because they are very cash generative, with high margins and very good cash conversion.

They do have a sensible dividend policy. They pay out about 30% to 40% of their free cash flow in a dividend, which steadily grows over time, pandemic aside. They’re increasingly doing share buybacks. The multiple we think they trade on is actually pretty low given the growth opportunity.

They started to do share buybacks with their excess cash, which we think is a very sensible form of capital allocation.

On the valuation point, I'm curious to hear how you'd frame it. Much of my research about the business beforehand made it pretty clear they're in this funny spot where they're grouped in with a lot of the airlines or transportation stocks, despite being more IT software-ish in nature. How do you approach valuation? How do you think the market approaches valuation in terms of there being a difference there? Any commentary around that would be interesting to hear.

7. The Market Misprices Amadeus

I absolutely agree with you on that observation. The stock's kind of in no man's land in that it's covered by airline analysts, so we're used to horrible capital intensity and cyclicality, and I think it gets tarred with the wrong brush there. But then for tech analysts, it's almost not cool enough, not AI enough, even though they're doing quite a lot themselves with AI.

But I think because of the pandemic as well, the market is absolutely schizophrenic about the cyclicality, but the pandemic is the wrong anchor because in typical travel downturns, the volume headwinds in global traffic are not that bad. They're 2% to 3%, which is what you saw in the global financial crisis, and that's a very bad recession. For those broad reasons, that constrains the multiple Amadeus trades on.

When we think about valuation, the qualitative side of things is really important, and we think Amadeus has an abundance of untapped pricing power, which it can unleash over time. It's extremely well invested as well, in that its R&D spend is approaching new highs. You could argue that they're about to embark on a harvesting phase, where they've got their product propositions ready. They're getting commercial gains with those new product propositions, and there's increasing evidence of that.

As those gains come through, the R&D spend can start to go down over time. They embark on this financially attractive phase of their history. So we like it from that lens. The cash generation profile of the group is very good. Pandemic aside, the free cash flow per share trends have been in the high single digits.

We think it can be better than that on a forward basis with buybacks, particularly at the share price and this high-teens multiple, which it trades on today. That's also because of these new propositions that are coming through, and we think Navio is a really interesting angle to the investment case. Free cash flow yield today is in the 5% to 6% area, 5.5%. You've got a business with that type of free cash flow yield, which is supported but can grow its free cash flow per share, we think, at low double digits on a risk-adjusted basis. The expected return is quite juicy. We'd be very disappointed if it was less than low double digits.

Certainly, I experienced, as a former sell-side transportation analyst, how any business that gets bucketed in there can have management teams that are not happy if they're not a traditional transportation business and they're covered by those analysts. It seems to be guilty by association. On the risk front, you've detailed competition pretty well in terms of the incumbents. Do they face risks from new technology disintermediating them or some new evolution that would change their positioning? Has that over time popped up in different ways?

This is a technology business, so the innovator's dilemma is front of mind, I think, when you're looking at these types of companies. But I do think the really great thing about them is they are willing to reinvest. When you look at industry changes that have taken place, like the evolution to the New Distribution Capability, which is basically the airlines trying to take control of their own offers as opposed to going through the global distribution systems, where the likes of Amadeus did the offers themselves for those airlines, Amadeus have actually invested in their own proposition.

They now have over 70 airlines hooked up to their NDC proposition. Because of their scale position within the channel, they're able to offer agents both the traditional form of providing inventory via the EDIFACT GDS-type model, but also sell NDC into that channel as well. They're not actually losing share within NDC. They're holding their own.

And within airline IT, there is a potential evolution to new order management systems, or offer-and-order management systems, but they have a Navitaire proposition, which they've invested in over the last decade, so they're ready for that transition. When airlines are making these transitions, it's so mission-critical, and there's going to have to be smart bridging technology while these migrations take place from old passenger service systems to the new.

So Amadeus can manage that at a low price for their airline customers. Technology risk is very large, but I do think they manage it well. That is a reason that they reinvest so heavily in research and development over time. It's also why the gap's increasing between themselves and their peers.

Their market position is so much greater now than their nearest 2 competitors that it's just becoming very hard for them to keep up if these technology evolutions do take place. It's a really big risk potentially for their peers, but potentially not so much for them.

Are there other risks that stand out to you? What would you highlight as a risk to the thesis or to the business that stands out the most?

It would have to be something that was very left field, which was a technology evolution that was very hard to foresee. That's always something that's hard to guard against, but I think that's one of the risks. A big traffic downturn is a risk.

If there was a war and traffic volumes crashed, obviously, being a transaction-processing business and a derivative of traffic volumes, that's going to result in revenue headwinds. But then I could also spin that into a positive because if that happens, arguably they'll take even more market share because they're so well capitalized, and you'd get bankruptcies among their competitors. You would actually take market share even quicker, potentially, in that scenario. So the investment case for a long-term investor is hedged to a degree with that risk.

This has been fascinating. We finish these conversations out with some of the lessons that you can take away. What's been your experience in having broader lessons that you can pull away from Amadeus and potentially apply elsewhere?

The mark of a good business is whether you would hate to compete with that company in question, and I would hate to compete with Amadeus. But customers like to do business with you, and I think they're a channel friend, not a channel foe.

When you find those 2 characteristics together, competitors hate you because you're good and customers love you for the same reason. That's a fantastic cocktail for value creation.

I like it. That is a new one, and I will certainly look for it elsewhere. Ben, this has been a pleasure. Thank you for joining us on Business Breakdowns.

Thanks, Matt.