Latest / Deep Values: Value Investing Deep Dives / AppLovin's Transformation Into An AI Cash Machine
Transcript
- 0:00Welcome back to the Deep Dive. We've got a fascinating stack of
- 0:02documents to get through. Today, we really do.
- 0:04We are looking at a company that is just creating a massive
- 0:09amount of noise in the financial world right now.
- 0:11And honestly, if I told you we were going to spend the next,
- 0:14you know, 20 minutes talking about Applevan, you'd probably
- 0:17look at me funny. You'd think.
- 0:18Wait, the mobile game guys? The ones who make those word
- 0:21puzzles my aunt plays on the bus?
- 0:23And until very recently, you would have been completely
- 0:26right. That was the narrative.
- 0:27There were a gaming studio that, you know, dabbled.
- 0:30Exactly. But looking at this Q32025
- 0:33report, wow, that description isn't just slightly off, it's
- 0:37officially ancient history completely.
- 0:39We are not looking at a gaming studio today, we are looking at
- 0:42what might be one of the most efficient cash generating
- 0:45machines in the entire tech sector.
- 0:47It really is a structural anomaly, and I don't use that
- 0:50word lightly. We're talking about a company
- 0:52that just posted 68% revenue growth and get this an Adjusted
- 0:57EBITDA margin of 82%. 82% that is.
- 1:00That's eye watering. It is.
- 1:02That's not just good software margins, that's practically
- 1:05unheard of. It sounds like a typo.
- 1:07It does, but it's real. It rivals the profitability of
- 1:09the biggest tech giants, you know, like Meta, but they're
- 1:13doing it with a fraction of the headcount.
- 1:14It sort of breaks all the traditional rules of scaling,
- 1:16right? Usually when you get that big
- 1:18you get bloated. Apple of went the other way.
- 1:20Not complete other. Way so that's our mission today.
- 1:22We need to figure out how they pulled this off.
- 1:25How did they pivot from making casual games to building, well
- 1:29the rails of the mobile economy and crucially with the stock
- 1:32trading around 524 dollars, is this a bubble waiting to pop, or
- 1:37is this a genuine quality compounder that's just getting
- 1:39started? It's a fascinating case study in
- 1:42strategic pivoting and really the hidden power of AI
- 1:47infrastructure. OK, before we get into the
- 1:49valuation and the nuts and bolts of their AI engine, because we
- 1:52are going to get technical, we do have a quick tool tip for
- 1:54you, right? Since we are talking about
- 1:56valuations and stock prices, if you are trying to keep track of
- 2:00the total market temperature, you might want to head over to
- 2:03the Apple App Store and download the Buffett Indicator app.
- 2:06It's a really handy tool. It sends you daily alerts on
- 2:09market valuation levels. Just it tells you if things are
- 2:11getting frothy. Exactly.
- 2:13If you're worried about whether the market as a whole is
- 2:15overheating, spotting those bubbles before they burst, it's
- 2:18a great way to just keep an eye on the macro picture.
- 2:20Absolutely. It's useful stuff for context.
- 2:22Now a quick disclaimer before we jump in.
- 2:24Of course, everything we're discussing today is based on
- 2:26publicly available sources, specifically that Q32025 data.
- 2:31This is all for educational purposes.
- 2:32We are not giving you investment advice.
- 2:34Please do your own homework. Always do your own due
- 2:37diligence. We're just exploring the data.
- 2:39That's it. All right, let's unpack this
- 2:40pivot, because looking at the timeline, 2025 was really the
- 2:44year Apple Win decided to stop being one thing and start being
- 2:48something else entirely. Right.
- 2:50This is what we call segment one from player to house.
- 2:52The old app, Loven was known for games like Wordscapes.
- 2:55Sure, I've played. That yeah, they were a content
- 2:57creator. They had studios, artists, level
- 2:59designers. But in 2025, they made a massive
- 3:03move. They completed the strategic
- 3:05divestiture of their legacy apps portfolio, selling it all off to
- 3:09Triple dot Studios. So they just sold the games?
- 3:12Why? I mean, if the games were making
- 3:13money, why dump them? Because they realized they were
- 3:16competing in the ecosystem they wanted to control.
- 3:19If you own the platform that sells ads, but you also make
- 3:23games that compete for those ads, your customers, other game
- 3:26developers, they don't trust you.
- 3:27Right, you're the house and a player at the same table.
- 3:30Exactly. This was the moment they decided
- 3:32to stop competing with developers and start running the
- 3:34infrastructure for them. I love the analogy of the gold
- 3:36rush year. For years, Applevin was digging
- 3:38for gold alongside everyone else, making the games, fighting
- 3:41for users. But then they looked around and
- 3:43said wait. A minute.
- 3:44The real money isn't in finding the gold, it's in selling the
- 3:49shovels. Or in this case, the high tech
- 3:51AI powered shovels that help everyone else find the gold.
- 3:54So by selling the games division they became a pureplay ad tech
- 3:58firm. Right.
- 3:58They aren't distracted by game design, level creation or art
- 4:01assets anymore. They are 100% focused on the
- 4:04software platform. And the financial validation of
- 4:07that decision in Q32025 is just it's staggering.
- 4:12Revenue hit $1.405 billion, which beat estimates.
- 4:16Net income doubled to $836 million.
- 4:20But there's one stat in the report that I just couldn't wrap
- 4:22my head around at first the flow through rate.
- 4:25That is the key metric. They achieved a 95% flow through
- 4:28rate. OK, hold on 95%.
- 4:30Can you break that down? What does that actually mean in
- 4:32plain English for someone who isn't ACFO?
- 4:34It's actually pretty simple, but the implication is huge.
- 4:37It means that for every new dollar of revenue they brought
- 4:39in, $0.95 if it went straight to the bottom line is profit.
- 4:42That is insane. I mean, usually there are costs
- 4:44associated with making more money, right?
- 4:46Exactly. It's unheard of in physical
- 4:48businesses. Think about it.
- 4:50If you sell cars, to sell more cars, you have to buy more
- 4:53steel, hire more workers, build more factories.
- 4:55Your costs go up as your revenue goes up.
- 4:58But here here, because the software is already built and
- 5:01the costs are largely fixed, the new money is almost pure profit.
- 5:05So they don't need to hire a new engineer every time they get a
- 5:07new customer, right? The server costs go up a tiny
- 5:10bit, but that's it. The operational leverage in its
- 5:12purest form. It's almost like printing money,
- 5:15but the engine that is printing that money isn't magic.
- 5:18It's something called Axon 2 Point O.
- 5:21And this is where I want to slow down a bit because AI is a
- 5:23buzzword everyone throws around. Every company claims they have
- 5:26AI now. True, but Axon 2 point O is the
- 5:30real deal. This is their secret weapon.
- 5:31It's a machine learning engine. You can think of it as the brain
- 5:34that connects advertisers to users.
- 5:36OK. But how does this brain actually
- 5:38work? Because Google has AI, Meta has
- 5:40AI. What makes this so different?
- 5:42It comes down to vertical integration.
- 5:44Apple Oven has a vertical integration advantage.
- 5:47They operate what's called a dual sided marketplace.
- 5:49Dual sided. OK, so on the supply side, they
- 5:51have a platform called AXX. OK, let's define that.
- 5:54What is Mac? Max is a mediation platform in
- 5:57the app world if you are a game developer.
- 6:00So you made a puzzle game, you want to show ads to make money.
- 6:04Of course you install the Max SDK.
- 6:06That's a software development kit.
- 6:08Basically a chunk of code into your game.
- 6:11Max then acts as an auctioneer. It calls out to all the
- 6:13different ad networks and says, hey, I have a user here playing
- 6:17this puzzle game who wants to show them an ad.
- 6:19Who will pay the most? And they have a big piece of
- 6:21this market. They see over 60% of the market
- 6:24here. Wow, so they control the auction
- 6:25house where the ad space is sold.
- 6:27They know exactly what everyone is bidding.
- 6:29Exactly. They see the supply, but then on
- 6:31the demand side they have app discovery.
- 6:33This is their own ad network where advertisers come to bid
- 6:35for users. Wait so this sounds like they
- 6:37are playing both sides of the fence?
- 6:39They run the option with Max and they are bidding in their own
- 6:41auction with app discovery. That's where it gets
- 6:44fascinating. This creates what we call
- 6:46information asymmetry. Information asymmetry.
- 6:49Because App Lovin runs the auction and they also bid in the
- 6:52auction, they essentially see everyone else's cards before
- 6:56they play their hand. That feels, well, from a
- 6:59competitor standpoint, that feels incredibly unfair.
- 7:02Structurally, it's a massive Moat.
- 7:04It gives them a privileged view of the entire mobile auction
- 7:07ecosystem and they feed all that data.
- 7:09And we're talking about data from 1.6 billion daily active
- 7:13users into a Exxon. 1.6 billion people.
- 7:16That's a data set rivaling Facebook or Google.
- 7:19It is, and that data scale creates A flywheel.
- 7:22This is the crucial part to understand.
- 7:24The more data they have, the smarter a Exxon gets at
- 7:27predicting which user will click which ad.
- 7:29The smarter egg Exxon gets, the better the ad targeting is.
- 7:33The better the targeting, the more advertisers are willing to
- 7:35spend because they get better results.
- 7:36And the more they spend. The more data app lovin gets,
- 7:39it's a self reinforcing loop. And I read in the notes that
- 7:42once a developer installs their software, the App Lovin SDK,
- 7:45it's really hard to get rid of it.
- 7:47Oh yeah. Technical lock in.
- 7:49It is not a trivial engineering task to RIP out a mediation SDK
- 7:53and replace it with a competitor like Unity or Google AdMob.
- 7:57Why not? If you switch, you risk
- 7:59downtime, you risk bugs, and you risk revenue loss.
- 8:03So once app loven is in your app, they tend to stay in.
- 8:06So they have the data, they have the lock in and they have the
- 8:09AI. Let's talk about the result of
- 8:11all this cash Segment 3, the cash flow machine.
- 8:15We touched on this with the margins, but it's worth digging
- 8:18deeper into the operational leverage.
- 8:19Their costs are largely fixed. They have engineers, they have
- 8:22servers, but they don't need a massive sales force anymore.
- 8:24Why not? Don't you need sales people to
- 8:26sell ads? Not with this model.
- 8:27The platform is becoming self-service.
- 8:29Advertisers plug into app discovery, they set their budget
- 8:31and they let the AI do the work. So as Applavin scales up to
- 8:34billions in revenue, they don't need to hire thousands of new
- 8:37account managers. So as they scale, their costs
- 8:39stay flat. That explains that 82% Adjusted
- 8:44EBITDA margin. Exactly.
- 8:45In Q3 alone, their free cash flow was $1.05 billion.
- 8:50A billion dollars of cold hard cash in three months.
- 8:54In three months. That is a staggering amount of
- 8:57liquidity. What are they doing with all
- 8:59that money? Are they buying competitors?
- 9:01Buying yachts. They're buying themselves, they
- 9:03are aggressively buying back their own stock.
- 9:06They bought back $571,000,000 worth in Q3 alone, OK.
- 9:10And the board just authorized another $3.3 billion for
- 9:13buybacks. That signals a lot of
- 9:14confidence, right? Usually you see companies buy
- 9:17back stock when the price is low.
- 9:18But Applavin's stock has been rocketing up.
- 9:21It has buying back at all time highs.
- 9:23That's a bold move. It usually means they think the
- 9:24market is still wrong. It tells you exactly what
- 9:27management thinks. They believe the stock is still
- 9:29undervalued even at these prices, and that aligns
- 9:32perfectly with the management structure.
- 9:33The CEO, Adam Forrey, owns about 10.9% of the company. 10% which?
- 9:38Is worth nearly $20 billion. That's a serious chunk of skin
- 9:41in the game. And here is the kicker, he takes
- 9:44almost all his compensation in stock.
- 9:47So if the shareholders lose, he loses, If they win, he wins big.
- 9:50Exactly, management alignment is very very high here.
- 9:54They aren't taking huge cash salaries while the stock tanks,
- 9:56they are all in. OK.
- 9:57So the business model is amazing, the cash flow is
- 9:59amazing, the CEO is a line, but and there's always a butt in
- 10:03these deep dives. Let's look at the price tag,
- 10:07Segment 4, the valuation reality check.
- 10:09This is where the rubber meets the road for sure.
- 10:11The stock is trading around $524.00.
- 10:14It is just skyrocketed over the last year.
- 10:16So if I'm a value investor, if I'm looking for a bargain, what
- 10:19does the checklist say? If you are a strict deep value
- 10:22investor like a Benjamin Graham disciple, this stock fails
- 10:25almost every test. Ouch.
- 10:26Give me the bad news. The PE ratio price to earnings
- 10:29is 63.7. For context, Warren Buffett
- 10:32usually prefers something under 15, maybe 20 for A truly great
- 10:35company. So way off.
- 10:37Way off. The price to book ratio is over
- 10:39100 X. 100 times book value. It is incredibly expensive
- 10:43relative to its book value. There is absolutely no margin of
- 10:46safety here. If you look at it through that
- 10:47traditional lens, you are paying a massive premium for future
- 10:52growth. So it's overpriced.
- 10:53It's priced for perfection. It's priced as if this growth is
- 10:56going to continue for a very long time.
- 10:58However, OK, it passes the quality test with flying colors.
- 11:02The return on equity is over 262%.
- 11:05That's elite. It's absolutely elite and the
- 11:07debt levels are safe because they are just paying it down so
- 11:10fast with all that cash. What about the intrinsic value
- 11:13if we do a discounted cash flow analysis, basically trying to
- 11:15calculate what all the future cash is worth today.
- 11:17So if you assume a nominal growth scenario, say 25% growth
- 11:20for the next five years, now 25% is aggressive.
- 11:24But given the AI boom, it's possible right?
- 11:26Under that scenario, the fair value comes out to around $531.
- 11:30Which is basically exactly where it's trading.
- 11:32Today it was at 524. Dollars, right?
- 11:34So the market has officially priced in that massive growth.
- 11:36If they stumble even a little bit, if growth drops to 15%
- 11:40instead of 25%, there is a long way down.
- 11:43So no bargain here. You are paying a premium for a
- 11:46high quality asset. You are not finding a dollar
- 11:49bill selling for $0.50. So essentially, the market knows
- 11:51it's a good company and they're making you pay for it.
- 11:54Speaking of stumbling, let's look at the Dark Clouds segment.
- 11:585 headwinds and risks because it hasn't all been smooth sailing,
- 12:02has it? No, it has not.
- 12:05There was a report from Capital Watch that caused quite a bit of
- 12:07volatility recently. The pig butchering allegation?
- 12:10That sounds terrifying, but I assume we aren't talking about
- 12:13actual livestock, no? Pig butchering refers to a
- 12:16specific type of long con financial scam, often involving
- 12:19cryptocurrency. The report alleged that proceeds
- 12:22from these Ponzi schemes were converging inside the ad
- 12:25ecosystem. The implication was that some of
- 12:28the ads spend on the platform might be, let's just say, dirty
- 12:31money washing through this system.
- 12:32That is a heavy accusation. What did APP Levin say?
- 12:35Management completely denied it. They maintain the allegations
- 12:37are unfounded. But anytime you have allegations
- 12:40of money laundering or scam proceeds in a financial
- 12:42ecosystem, it Spooks investors. It's a reputation risk.
- 12:46A big one. Even if it's not true, the noise
- 12:49can hurt the stock. And then there is the Sword of
- 12:51Damocles that hangs over every mobile company, Apple and
- 12:54Google. Platform risk.
- 12:56This is the biggest structural threat.
- 12:58Applevan lives in their house. They rely on the Apple App Store
- 13:01and Google Play Store to exist. And we saw what happened a few
- 13:04years ago. We did when Apple introduced App
- 13:07Tracking Transparency or ATT. It just nuke the ad industry for
- 13:11a while because it made it so much harder to track users.
- 13:13Right. Suddenly, advertisers were
- 13:15flying blind. Exactly.
- 13:16Now Applev navigated that better than most.
- 13:18That's part of why they're winning now.
- 13:21But if Apple or Google decide to change the privacy rules again?
- 13:25If they decide to restrict data access even further, Akon's
- 13:28vision gets blurry. They are building a castle on
- 13:30rented land. Apple could change the zoning
- 13:33laws tomorrow. Precisely.
- 13:35And finally, there is the AI arms race.
- 13:37Right now, a Exxon 2.0 appears to be best in class for the open
- 13:40Internet, but the barrier to entry for AI is lowering every
- 13:44single day. Can Meta or Google just
- 13:45replicate this? Can they build their own Axon?
- 13:48That's the multibillion dollar question.
- 13:50Can the Giants create a similar optimization tool for the open
- 13:55web and just crush Apple Evans advantage?
- 13:58Hasn't happened yet, but in tech yet is a very dangerous word.
- 14:02So, bringing it all together, what does this all mean?
- 14:05Applevan has transformed from a game maker to an AI
- 14:07infrastructure powerhouse. It prints cash like almost
- 14:11nothing else we've seen. But that price tag?
- 14:14It assumes they will keep winning forever.
- 14:16It's a quality compounder structurally.
- 14:18If you are a growth investor, you look at this and see a
- 14:20company that is capitalizing on the AI revolution perfectly.
- 14:23You see the margins, the buybacks.
- 14:25You see the margins, you see the buybacks and you think this is a
- 14:27winner. But if you are a bargain hunter,
- 14:31you sit on your hands, right? This is for the investor who
- 14:33believes Axon is the future of mobile marketing and is willing
- 14:36to pay full price for the ticket It's.
- 14:37Definitely not for the faint of heart.
- 14:39No high risk, high reward. I want to leave you with a
- 14:41provocative thought to Mull over.
- 14:43We talked about Applov and being the discovery engine for apps,
- 14:46but as AI gets better it drastically lowers the cost of
- 14:50coding. We might see millions of new
- 14:52hyperniche apps created by AI. The market could get absolutely
- 14:56flooded. Exactly.
- 14:57Does that make Applovin's discovery engine even more
- 15:00critical because there is so much noise?
- 15:02Or does AI eventually replace the need for apps entirely?
- 15:07Do we just ask our AI assistant to do things bypassing the App
- 15:11Store and app love and all together?
- 15:12That is the existential question for the next decade, isn't it?
- 15:15If the interface becomes a chatbot, do we still need apps?
- 15:18Something to think about. Remember, always conduct your
- 15:21own due diligence. We are just two voices in your
- 15:23headphones. Absolutely check the numbers
- 15:25yourself. If you enjoyed this deep dive
- 15:27and want to go even further, consider joining the Deep Value
- 15:29Club membership for advanced content.
- 15:32And hey, make sure you are subscribed to Deep Value
- 15:34Investing and Lula VC Comm on Apple and Spotify Podcasts.
- 15:38Thanks for listening. See you on the next deep dive.