AppLovin has about 900 employees, a market value near $100 billion, and operating cash flow of $869 million against $1.4 million of capital spending.
The show paired it with the opposite kind of business: Royalty Pharma, which buys the future sales of drugs other companies develop, publishes its financials in exhaustive detail, and reports a revenue line its own presenter says investors should ignore.
"It's just an algorithm with employees."
The two hosts run The Canadian Investor for the Canadian Investor Podcast Network, and this is the format where each of them spends a day researching one name and defends it to the other. Neither owns the stock he presented.
I listened to the full episode so you can skip it. 45 minutes of audio, 15 minutes of reading.
Here are the 11 takeaways that matter.
🎙️ Hosts: the two co-hosts of The Canadian Investor, the Canadian Investor Podcast Network's show on individual stocks and portfolio construction, each presenting a company he had spent the week researching
📰 Published: 31 August 2026
🔴 YouTube | 🔗 Episode page | ⏱️ 45 min | ✅ Time saved: 30 min
Key Takeaways
Royalty Pharma is Franco-Nevada for drugs, and the parallel holds all the way down to the concentration risk
One Vertex cystic fibrosis drug was 25% of portfolio receipts in the second quarter, almost exactly what Cobre Panama was to Franco-Nevada
Its reported revenue is the wrong number, and so are price-to-earnings and free cash flow
Portfolio receipts were $3.2 billion in 2025 against revenue of $2.3 to $2.4 billion, because accounting spreads the cash over the years it was forecast for
Milestone payments are how a royalty buyer stops a failed trial from costing the full check
The company borrows at under 4% and lends into drug royalties, and could clear its debt in about two and a half years
AppLovin's margins come from owning the auction software inside the games it buys ads on
It sees the traffic and the player data before any rival bidder does
Revenue per employee is more than double Meta's
The short-seller campaign ended with the SEC closing its investigation and taking no action
The growth problem is that AppLovin already has the market
Installs fell 2% year over year while revenue per install rose 58%
The multiple has compressed from 35x forward earnings to 16x, and nobody knows where growth lands
The competition is now the platforms AppLovin depends on
Meta says its AI is good enough to work without the tracking data it lost, and Alphabet is handing developers free tools to build games inside its own ad system
The Trade Desk is the warning both hosts keep coming back to
1. A Royalty Model for Drugs
The episode opened on a company neither host had covered: Royalty Pharma, ticker RPRX.
The business is a drug financier that takes a cut of sales instead of equity. "So, essentially, they finance pharmaceutical drugs in exchange for a percentage of their future sales, which is the royalty"
The parallel he used throughout is Franco-Nevada, which funds mining projects for a royalty or a stream
There are three revenue lines. Third-party royalties on approved or late-stage therapies bought from someone other than the developer, which are most of the portfolio; "synthetic royalties" created directly with the developer; and a small amount of other funding — debt, direct equity, launch and development capital in exchange for fixed long-term payments
The cost structure is what makes it a royalty company rather than a pharmaceutical one. "So what's good for them is they have no manufacturing costs, no salespeople"
Like a mine, a drug patent runs out, so the company has to keep finding new deals
The portfolio holds over 35 commercially approved products and 19 development-stage therapies
The appeal is diversification inside one ticker. He said it gives exposure to pharmaceuticals without depending on a single blockbuster, and called it the most compelling alternative to an exchange-traded fund he has come across
2. Betting on the Underwriter
The other host pushed back immediately: this has to be riskier than a metals streamer.
The answer was yes and no, and it turns on what you are buying. An investor is betting on the company's ability to judge approval probability, market size, market share, pricing, patent life and legal challenges — the same expertise bet as with Franco-Nevada or Wheaton, applied to drugs rather than ore
The host raising the objection put it in the terms he understood. A miner can say how much gold is in the ground; a biotech in phase one cannot, and a lot of micro and nano-cap developers go bust when a drug fails
"Yeah, at the end of the day, they're making a bet, right — it's a calculated bet, but it's still a bet, and they're making a bunch of different bets."
The structure that limits the damage is the milestone payment. On a $500 million deal the company might pay $100 million up front, $200 million only if phase three completes, and $200 million only on FDA approval
"So that's how they mitigate risk — they're not on the hook for the whole financing deal unless it actually goes through and they can profit from the drug."
Neither host thought the developer owes anything back if the drug fails; the milestone structure is what stands in for that
The scale of the market is why the model exists. Royalty transactions reached $10 billion across the sector in 2025 and industry-wide research and development is expected to exceed $1 trillion over the next decade; Royalty Pharma deployed $2.6 billion into new deals in 2025
3. Why P/E Is Useless Here
The longest stretch of the segment was about an accounting quirk, and the presenter treated it as the thing an investor has to understand before buying.
Reported revenue and cash received are different numbers, and the gap is structural. Portfolio receipts were $3.2 billion in 2025 against revenue of $2.3 or $2.4 billion
The mechanism is that the revenue line follows the original forecast, not the year's cash. A $1 billion payment expected to return $2 billion gets booked at roughly $100 million a year; if the drug throws off $200 million in cash that year, the revenue line still shows the smaller figure, and the estimate is readjusted later
"But still, you'll see that revenues are consistently lower than portfolio receipts, and portfolio receipts are better at actually looking at the actual cash coming in."
Two standard ratios therefore do not work. "So I would kind of disregard price-to-earnings." and "Same thing for price-to-free-cash-flow, or free cash flow — not useful at all, because they have zero capex."
What he does instead is build the metric by hand from the cash flow statement: cash from operating activities, less purchases of investments, adjusting for items such as short-term Treasuries. The capital spending for this business is the royalty purchase itself
The generalizable point he drew is about not applying one screen to every business. "Like cash flow is not useful for banks."
The other host's first instinct was a life insurer, which books a policy's profit over its life rather than at the point of sale
One oddity he flagged: the company's own cash flow statement starts with portfolio receipts rather than net income
4. Debt at Under 4%
The spread is the business. The company borrows and lends the money out at a higher rate, funding deals from existing royalty cash and debt
"So it's very manageable — it would make even some governments enviable in terms of that rate." The weighted average interest rate is just below 4%
On adjusted EBITDA, which uses portfolio receipts rather than revenue, the debt would take about two and a half years to clear if every dollar went to it, and interest is covered about 10 times
The dividend yields about 1.5%, though he cautioned that the payout ratio has to be measured against the cash metric rather than reported earnings
A cost change he thought worth mentioning: the company was externally managed and paid a management company to run it, and bought that arrangement out last year. It took a short-term cost hit and now runs everything internally
5. One Drug, 25% of Receipts
The risk list closed the segment, and he mapped each item onto the mining analogy.
Capital deployment risk. The whole case rests on the people making the allocation decisions being right
Regulatory approval risk. A drug the FDA does not approve is a sunk cost, and he noted one in the most recent report that failed a stage trial
Technological risk, which he tied specifically to AI. A better drug arriving inside the patent window would erode years of royalty revenue that had been counted on
Concentration risk is the one with a number on it. "And concentration risk — in their most recent earnings release, the Vertex cystic fibrosis drug accounted for 25% of their portfolio receipts in Q2 — one drug."
He put that at almost exactly the share Cobre Panama was of Franco-Nevada's revenue
The other host's verdict: "Yeah, it's definitely one I'm going to put on the radar, because I didn't even know a royalty-pharma company like this existed"
His only pharmaceutical exposure outside index funds is Intuitive Surgical, which he pointed out is the same sector but a different industry
6. An Algorithm With Employees
The second half went to AppLovin, ticker APP on the Nasdaq, which the presenter called one of the most talked-about and most contested stocks in the market.
The business is arbitrage on mobile game advertising. A game developer offers a price per download; AppLovin buys ad slots inside other games, uses an AI engine to pick the players most likely to download, and keeps the difference. Pay $1 for the slot against a $5 price, and $4 is the company's
That is where the 85% EBITDA margins come from, and the reason it can price the trade is that it owns the software running the auctions inside the games
It sees the traffic, the demographics and the behavior of everyone playing, so it knows which impression is worth bidding for before a rival does
The capital intensity is close to zero. "So they had operating cash flow of 869 million, and they spent 1.4 million on property, plant and equipment."
No inventory, no factories, no stores — the system runs on AI computing
Revenue per head is the statistic that startled him. "So the company generates more than 7 million in revenue per employee." Against the other advertising companies he compared it to: "Meta was 2.8 million per employee, Alphabet 2.2 million, and Trade Desk 760 million."
"It's just an algorithm with employees."
The other host's reaction was that it sounds terrifying and extremely profitable at once, and that his own first word for the business was "fragile"
The stock could be bought at about $10 in 2022, ran to $715 at the end of 2025, and now trades around 60% below that, in the low $300s
7. The SEC Closed Its Inquiry
The company drew a wave of short reports last year, which he summarized rather than relitigated.
The allegations ranged widely: outright fraud in the form of downloads to fake devices billed to advertisers, fake click-throughs, back-end deals with phone carriers to auto-install apps, and tracking children and using that data
The one he takes seriously is a terms-of-service problem rather than a fraud one. The claim that AppLovin built user profiles with data-collection methods that violate Apple's, Android's and Meta's terms of service — because being removed from those platforms would be a business-ending event
The irony he pointed out is that the company's rise came from the opposite condition. When Apple shut off ad tracking, AppLovin's system could run without that data, which is when it made its money
The resolution was recent and clean. "The reason I wanted to mention the short reports is this just came out a few weeks ago — the SEC did an investigation and pretty much closed it with zero action."
He noted that some reports were taken down or amended after the company involved lawyers, and that a lot of the material seemed far-fetched
8. Market Share Caps Growth
The valuation looks cheap and the reason is deceleration. "So it's gone from 35x expected earnings to just 16x, and this is a company growing at a 30%-plus clip."
The 50%, 60% and 70% growth rates in the past are not the guide; the company itself guides to 30%, and that is the number he would use for 2027
The cause is arithmetic rather than execution. He believes over 50% of mobile game advertising spend already runs through AppLovin's systems, and a company with half the market cannot double its share
The evidence is in the mix of the growth. "And you can kind of tell this because installs fell 2% year-over-year — that's downloads overall — while revenue per install increased 58%."
Market capture, the first engine, is close to finished. What is left is charging more for the share already held
That is why he thinks the market cannot price it. The question is where the deceleration stops — 30%, 20%, 15% or 10% — and nothing in the disclosure settles it
9. E-Commerce Is Engine Two
The second growth engine is the same advertising system pointed at physical goods: a seller pays when the system finds a buyer who purchases
It has existed for a couple of years — he put the start around 2024 — and fully launched in June 2026, and the company says it is going well
What he cannot check is who is actually spending. From his reading it is concentrated in a small number of very large advertisers, and the company discloses no advertiser count and no revenue split
The disclosure gap is the contrast that struck both hosts. He went looking for an investor presentation and found the bare minimum — financial results and consolidated statements
Against Royalty Pharma, which breaks out each royalty, the portfolio receipts and the money going into new deals
"I'm actually pretty shocked they don't have an investor presentation or anything, because I don't want to say this is a complicated company, but it's one that a lot of people probably wouldn't be able to wrap their head around how it works."
10. Meta and Alphabet Circle
The structural problem is that a business printing cash at no cost attracts competition, and here the competitors are the platforms underneath it
Meta is the direct threat. It gave up on the space when Apple ended ad tracking, and now says its AI is good enough to work with untracked traffic again
"And I like this company, but I would think you'd lose against Meta, in my opinion." Even with a current advantage, Meta has the money, though he allowed it may keep pouring that into AI instead
Alphabet's threat is upstream of the ad auction. It is giving people tools to build their own games with an AI chat interface, and a game built inside Alphabet's environment will almost certainly carry Alphabet's ad system rather than AppLovin's software
"But I would imagine there's a reason they're building these tools for free."
The second-order risk is what easy game-building does to the games themselves. Mobile games have historically been durable assets — he named Clash of Clans — and if anyone can generate one at a desk, they become disposable, which lowers what a developer will pay per download
11. The Trade Desk Warning
On price he is unambiguous: "So, last thing would be valuation, and then just a quick summary — they're the cheapest name in ad tech by quite a wide margin." It is far cheaper than Meta and growing much faster
On the run-up he is equally blunt. He called the stock at over $700 insane for a company whose market share was largely captured and whose moat could realistically be attacked
The comparison that stops him is two ad-tech companies with superior systems that were destroyed anyway — The Trade Desk, down more than 95% from its high, and Acuity Ads, now Illumin
"It makes me hesitate a bit — as soon as the growth story broke in either of those names, the stock price absolutely cratered."
The near-term test is already scheduled. The company missed guidance last quarter because improvements to Axon, the underlying platform, came late; because it owns so much of the market, it grows by billing customers more, and it bills more by improving the platform
Management says the improvements arrive next quarter. Another miss, he said, would probably not be good
The range he sees is unusually wide. Growth holding at 30% could take the stock back above $500, and further if e-commerce works; growth falling to the mid-teens sends it lower
"So I think it's a very interesting, high-quality, high-margin business that you have no idea what's going to happen with over the next five years — very difficult to value."
Bonus Insights
On whether he would buy it, the AppLovin presenter came down on speculative rather than core. "I think the ceiling is high — higher than the floor right now — but the floor is not completely priced in."
His warning is aimed at the argument circulating on X — 70%-plus growth at 16 times earnings — which he says misreads both halves: the growth is guided to be cut in half, and forward earnings estimates can be revised down
The other host preferred the floor to the ceiling and said he would stick with Royalty Pharma, while agreeing that AppLovin's upside is the larger of the two
AI cuts both ways for a drug-royalty portfolio. It threatens existing royalties if a better drug arrives inside the patent window, and helps the newer investments
The example given was a melanoma vaccine developed with heavy use of AI — pulling a thousand mutations out of a piece of skin cancer and designing cells to fight them, work he said would have taken a doctor years the patient did not have
Royalty Pharma has royalties on drugs from large pharmaceutical companies as well as small ones, which the presenter compared to a miner such as BHP selling a royalty to offset upfront risk
Both hosts asked listeners whether they want more of these deep dives, noting that each one costs about a day of research, and said the next episode returns to news and earnings
The two companies end up as a study in opposite risk shapes: one whose cash is visible but whose reported earnings are misleading, and one whose earnings are pristine but whose future depends on a market it has already won.
Products, Companies & Tools Mentioned
Royalty Pharma (Buys future drug sales for cash up front; over 35 approved products and 19 in development, with portfolio receipts of $3.2B in 2025)
AppLovin (Mobile game ad arbitrage with 85% EBITDA margins, about 900 employees and roughly $100B of market value)
Franco-Nevada and Wheaton Precious Metals (The mining royalty and streaming companies used as the working analogy for Royalty Pharma, down to the concentration risk of Cobre Panama)
Vertex Pharmaceuticals (Its cystic fibrosis drug was 25% of Royalty Pharma's portfolio receipts in the second quarter — the concentration risk in one line)
Intuitive Surgical (The only healthcare name one host holds outside index funds, and he notes it is not pharmaceuticals)
Meta (The competitor he thinks AppLovin would lose to: it left the space when Apple ended tracking and now says its AI works without that data)
Alphabet (Giving developers free AI game-building tools, which would carry Alphabet's ad system rather than AppLovin's)
Apple (Ended ad tracking in 2021, which is what created AppLovin's opening in the first place)
The Trade Desk and Illumin (Two ad-tech companies with superior systems whose stocks collapsed once the growth story broke — the precedent that makes him hesitate)
BHP (Cited as an example of a large company selling a royalty to offset upfront risk)
Fiscal.ai (The platform used on air to pull portfolio receipts and build the custom cash-flow metric)
Clash of Clans (The example of a durable mobile game, and what easy AI game-building would put at risk)
If this was worth your time, send it to someone closer to the industry than you are.
Get the latest market chatter as it happens: