Latest / Deep Values: Value Investing Deep Dives / Why 2026 Profits and Revenue Decoupled? #MSFT #AMZN #BABA #TCEHY
Transcript
- 0:00You know, normally when you sit down with a stack of mid year
- 0:04financial reports, there's a there's a certain rhythm to it,
- 0:08yeah, like an expectation of almost engineering level
- 0:11precision. Right, you look at the balance.
- 0:13Sheet and you see the top line showing revenue climbing.
- 0:16You glance down at the bottom line to see profit tracking
- 0:19right alongside it. Exactly.
- 0:21And the analyst community collectively nods like there it
- 0:24is a healthy, functioning company.
- 0:27Yeah, the machine is working exactly as designed.
- 0:29Up as good, down as bad. It's clean.
- 0:32It's a very comforting narrative, honestly.
- 0:34I mean, we are fundamentally wired to like things that are
- 0:37visible and predictable. Oh, for sure.
- 0:39We want things neatly categorized into columns that
- 0:41just, you know, balance out at the end of the quarter.
- 0:44It makes logical binary sense. But then, and This is why we're
- 0:48here, you and I started tearing into this specific stack of 2026
- 0:53mid year financials from like the global tech and
- 0:56manufacturing heavyweights and honestly.
- 0:58It's a mess. It really is that traditional
- 1:00accounting machine. Looks like someone just threw a
- 1:02massive wrench into the gears. Yeah, we are looking at a
- 1:05financial landscape that is incredibly murky right now.
- 1:08Very. We've got companies pulling in
- 1:10billions more in revenue than last year, yet they are like
- 1:14hemorrhaging cash, right? And on the flip side, we have
- 1:17companies selling significantly fewer produce than they did a
- 1:20year ago, but somehow posting record-breaking.
- 1:23Just eye watering profits. It completely defies the
- 1:26traditional rules of reading a balance sheet.
- 1:28Totally. The standard diagnostic tools
- 1:30that investors have relied on for decades feel like they've
- 1:33been, I don't know, temporarily suspended.
- 1:36We are navigating through absolute Muddy Waters here.
- 1:40And that is exactly our mission for you.
- 1:41On this deep dive, we are tearing into a massive stack of
- 1:452026 mid year financial reports covering over 20 global and
- 1:51Chinese heavyweights. It's a big stack.
- 1:54It is a massive stack, and our goal today isn't just to read
- 1:57you the numbers, and if you wanted a spreadsheet, you
- 1:59wouldn't be listening to us. Our goal is to uncover the
- 2:02hidden operational mechanics and the actual stories driving those
- 2:06numbers. We're going to pull apart the
- 2:08massive AI capital expenditures. Investigate this fascinating
- 2:12pivot from companies chasing pure scale to suddenly chasing
- 2:15pure profit. Which is a wild shift.
- 2:17Wild and we'll explore the highly counterintuitive ways
- 2:20these giants are surviving some truly brutal macroeconomic
- 2:23cycles. But before we unpack all of
- 2:26this, a quick note to access more advanced researches, follow
- 2:30Guru's latest movement and leverage our latest deep values
- 2:33agents. Welcome to download Deep Values
- 2:36APP from App Store to track stock market temperature.
- 2:39Welcome to download iOS app, Buffett Indicator from Apple App
- 2:42Store or Bubble Watch from Google Play.
- 2:45Links are in the description. Yep, check those out.
- 2:47And a quick disclaimer, all analysis today is based on
- 2:50publicly available sources and may be subject to revisions or
- 2:53differing interpretations. All contents from this channel
- 2:57are for research and entertainment purposes only and
- 2:59do not constitute any investment or financial advice or service.
- 3:03Please conduct your own due diligence before making
- 3:05investment or financial decisions.
- 3:07Always do your own research focus.
- 3:09All right, housekeeping done, let's set the stage here.
- 3:12Looking at this mountain of paper you've brought in, if we
- 3:15had to stamp a single like overarching theme on to 2026 so
- 3:19far, what is the defining characteristic?
- 3:23If I had to distill this entire stack of reports down to a
- 3:27single concept, I definitely call 2026 the year of financial
- 3:31decoupling. Financial decoupling, OK, break
- 3:33that down for me. What exactly is decoupling from
- 3:35what? Well, historically, revenue and
- 3:37profit hold hands, right? They walk down the street
- 3:39together. If you sell more widgets, you
- 3:40make more money. It's simple.
- 3:42But in 2026, they're telling 2 entirely divergent stories.
- 3:46They've completely decoupled. And as we go through these
- 3:49companies today, you'll see this decoupling is being driven
- 3:52primarily by three massive market distorting forces, which
- 3:57are first, the sheer unprecedented scale of AI
- 4:01investments. I mean, companies are throwing
- 4:03historical amounts of cash at infrastructure.
- 4:06Right, which completely warps their cash flow statements.
- 4:08Exactly. Second, we're seeing wild
- 4:11foreign exchange rate fluctuations, basically the
- 4:14invisible hand of currency markets and that's wiping out
- 4:17massive operational gains in an instant.
- 4:20Ouch. Yeah.
- 4:20And 3rd, we are witnessing these structural business pivots where
- 4:24legacy companies are fundamentally altering how they
- 4:27make money. They're choosing to prioritize
- 4:29per unit margins and aggressively sacrificing total
- 4:32volume. So if you're just skimming the
- 4:34headlines, the numbers on the surface might look like a total
- 4:37disaster, like plunging cash flow, dropping sales volume.
- 4:40But the underlying machinery might actually be going through
- 4:42a massive upgrade. Exactly.
- 4:44Or conversely, a great revenue quarter might be hiding a
- 4:48structural collapse. Right.
- 4:50OK, let's get right into the thick of it, because there is no
- 4:53better place to see this decoupling than the absolute
- 4:56biggest global theme right now, which is artificial
- 5:00intelligence. Oh, absolutely.
- 5:01The numbers we're seeing in the AI infrastructure space are just
- 5:05staggering, but they reveal a massive cash flow paradox.
- 5:10I mean, we are talking about companies earning billions yet
- 5:13simultaneously spending billions at a rate that almost defies
- 5:17comprehension. It really does.
- 5:18Let's. Look at the undisputed kingpin
- 5:20of this whole movement, NVIDIA. Oh, NVIDIA.
- 5:23The financials for NVIDIA are frankly difficult to
- 5:26contextualize, even for seasoned analysts who've been doing this
- 5:29for decades. Let's just look at their Q2
- 5:32numbers. Revenue came in at $96.22
- 5:35billion. Wow, that is a 105.85% year over
- 5:41year increase and their net profit 59.68 billion.
- 5:45Wait. Just pause on that for a second.
- 5:47Nearly $60 billion in pure net profit in a single three month
- 5:51window. It's insane.
- 5:53That's larger than the entire GDP of several countries
- 5:56generated in 90 days. It is historic, and the primary
- 6:00engine behind this, unsurprisingly, is their data
- 6:03center revenue, Right. That division alone accounts for
- 6:0689 billion of their total revenue, which is up 117%.
- 6:11But the number that truly stops you in your tracks, the metric
- 6:15that really demands explanation, is their gross margin.
- 6:19Oh yeah, I saw this. They are holding firm at an
- 6:22insane 75%. OK, here's where it gets really
- 6:25interesting and where I have to push back a little on the
- 6:27revailing narrative deal for it. Maintaining a 75 gross margin on
- 6:31physical hardware suggests an absolute total lack of pricing
- 6:36pressure, right? It means customers are taking
- 6:38whatever price NVIDIA dictates. But someone is footing that
- 6:41bill, right? The tech giants are handing over
- 6:44cash hand over fist. Yeah.
- 6:45Are the cloud providers actually seeing a tangible return on this
- 6:50investment or are they just hoarding chips and some sort of
- 6:52panic driven digital arms race? That's the $1,000,000 question.
- 6:56Because economic gravity dictates that a 75% margin
- 7:00should attract competitors who will undercut you immediately.
- 7:03You're touching on the exact anxiety rippling through the
- 7:05market right now. Like how is this sustainable?
- 7:08Yeah, how is it? The answer is that NVIDIA is no
- 7:11longer just selling a stand alone GPU, even if that's what
- 7:15the headlines focus on. They have successfully evolved
- 7:18into a full stack AI infrastructure platform.
- 7:22When a cloud provider buys from NVIDIA, they aren't just buying
- 7:25the chip. What else are they getting?
- 7:27They are buying the incredibly complex switches, the high speed
- 7:30networking equipment, and crucially, the proprietary
- 7:34software layers that make all these chips talk to each other
- 7:37efficiently. So it's an ecosystem lock in,
- 7:40much like how Apple locks you in with iOS and iCloud, but on an
- 7:45enterprise infrastructure level. Exactly.
- 7:46But it goes even deeper than that.
- 7:48They're rapid technological iteration is acting as a massive
- 7:51Moat against commoditization. I mean, they moved from the
- 7:55highly anticipated Blackwell architecture to their newest
- 7:57generations at a speed that is leaving competitors just gasping
- 8:01for air. Yeah.
- 8:02The pace is unreal. By the time arrival reverse
- 8:04engineers or matches the current generation's performance, NVIDIA
- 8:07has already shipped the next iteration.
- 8:10Furthermore, they've cultivated deep, mutually beneficial
- 8:14ecosystem ties with the major cloud providers.
- 8:17It's not just a transactional vendor client dynamic.
- 8:21No, it's a symbiotic relationship where Nvidia's
- 8:24software stack is deeply embedded in how these cloud
- 8:27providers optimize their own data centers.
- 8:29But even with that symbiosis, 75% volumes mean the cloud
- 8:33providers buying this hardware must be feeling a massive
- 8:36financial squeeze. Oh, absolutely.
- 8:38Which? Brings us perfectly to Amazon
- 8:40and specifically their cloud computing division, AWS.
- 8:44Because if you want to know if this 75% margin NVIDIA tax can
- 8:47actually be monetized by the people paying it, Amazon is the
- 8:50proof of concept. And Amazon's numbers provide the
- 8:53exact proof the market was desperately waiting to see.
- 8:55Let's hear it. Total revenue for Amazon crossed
- 8:58the $200 billion mark. AWS specifically brought in 42.2
- 9:03billion, a very healthy 37% increase, and the operating
- 9:08profit for AWS was 16.6 billion of 64%.
- 9:13Their margins for AWS hit a record 39.4%.
- 9:17But here is the massive paradox for you.
- 9:21Despite AWS practically printing money at a 39% margin, Amazon's
- 9:25overall free cash flow fell off an absolute Cliff.
- 9:29Last year. They were sitting on a +18.2
- 9:33billion in free cash flow. This year -7.6 billion.
- 9:38Yeah, quite the swing. A nearly $26 billion swing into
- 9:42the red. Why?
- 9:43Where did all the cash go? Capital expenditures complex to
- 9:47run these massive generative AI models, to buy those NVIDIA
- 9:50chips with the 75% margins and to build the physical data
- 9:54centers to house them, right? Amazon spent 54.2 billion in
- 9:58just one quarter. That's insane.
- 10:00And they announced plans to spend 220 billion for the full
- 10:03year. $220 billion on infrastructure in a single year,
- 10:07that is mind bending number. It is.
- 10:09So if you're an investor looking at a -7.6 billion free cash
- 10:13flow, what does this actually mean?
- 10:16Like, is Amazon engaging in a reckless cash incinerating arms
- 10:20race? Or is this simply the new cost
- 10:23of admission for doing business in tech in 2026?
- 10:26I would definitely argue it's the latter, and the underlying
- 10:29math supports that. It proves AI can be monetized,
- 10:32but it requires an iron stomach for massive upfront
- 10:35infrastructure costs. OK, layout the math.
- 10:37Look closely at that AWS operating profit.
- 10:41AWS only accounts for about 21% of Amazon's total corporate
- 10:44revenue, right? Yeah, yet it generates A
- 10:46staggering 61% of the entire group's operating profit.
- 10:50Wow. Furthermore, their AI related
- 10:52services and their own in house custom chips are already
- 10:55generating over 25 billion in annualized revenue.
- 10:58So that cash flow hit isn't a failure of their business model.
- 11:01Not at all. It is the deliberate
- 11:03construction of an impenetrable strategic Moat.
- 11:07By dropping 220 billion in a single year, Amazon is
- 11:11effectively ensuring that there are perhaps only three or four
- 11:14companies on the entire planet that possess the capital
- 11:17required to even attempt to compete with them.
- 11:19They were literally pricing out the competition.
- 11:23It's a game of high stakes poker, and Amazon just shoved
- 11:26their entire stack of chips into the middle of the table, daring
- 11:29anyone else to call. Well said.
- 11:31Speaking of high stakes, let's pivot to Microsoft because
- 11:34they're playing a very different and highly controversial game
- 11:38with their accounting number 0. Yes, they are.
- 11:41On the surface, Microsoft's report is glowing.
- 11:44Revenue hit 90.01 billion, up 18%.
- 11:48Their cloud division is up 27% and Copilot, their AI assistant
- 11:53integration, hit 30 million paid seats.
- 11:56That is a massive milestone. It really is.
- 11:58But just like Amazon, their free cash flow drops significantly,
- 12:02down 23% to 19.6 or billion, right?
- 12:06The AI build out is universally expensive.
- 12:09There is no escaping the CapEx reality.
- 12:12But what is truly fascinating about Microsoft's report is the
- 12:15accounting adjustment they made to cushion the blow of these
- 12:18expenses on their quarterly earnings.
- 12:21Right. This is the part of the report
- 12:22that had analysts arguing all week.
- 12:24They fundamentally changed their accounting rules regarding the
- 12:28physical lifespan of their servers, correct?
- 12:31That is correct. They extended the depreciation
- 12:34life of their data center equipment from 15 years to 25
- 12:37years. They changed the estimated
- 12:39useful life of those assets, yes.
- 12:41OK, here's an analogy for you. It feels exactly like
- 12:44refinancing your 15 year mortgage into a 25 year
- 12:47mortgage. Yeah, that's a good way to look
- 12:48at. It your monthly payment drops
- 12:50significantly which makes it look like you have way more
- 12:53disposable cat flow today. You look richer on paper this
- 12:56month, but the actual house still cost the exact same
- 13:00amount. You're just taking the pain of
- 13:02paying for it and spreading it out over an extra decade.
- 13:05So is Microsoft is doing financial engineering here?
- 13:09Are they artificially inflating their earnings per share, their
- 13:12EPS, just to keep Wall Street happy while they burn cash on
- 13:15data centers? It's a highly debated move for
- 13:18sure. Technically yes, it absolutely
- 13:20boosts their short term EPS by extending the life to 25 years.
- 13:25They are recording a much smaller depreciation expense
- 13:28each quarter on their income statement.
- 13:30Critics are aggressively arguing that this lowers the quality of
- 13:34their earnings because it masks the true immediate cost of this
- 13:38AI transition. Yeah, it hides the pain.
- 13:40But Microsoft's management vehemently defends the move.
- 13:44They argue that modern data centers, the physical buildings,
- 13:47the cooling systems, the power routing are structurally
- 13:50superior now and genuinely last longer, making 25 years a more
- 13:55accurate, faithful reflection of their useful life cycle.
- 13:58I mean, I see the logic that it certainly feels convenient to
- 14:00realize your buildings last 10 years longer right at the exact
- 14:03moment your caital expenditures hit all time highs.
- 14:06Very convenient timing. But if we dig a layer deeper AST
- 14:10the counting tricks, the real story for Microsoft's long term
- 14:13profitability isn't the physical buildings, it's the silicon
- 14:17inside. Them precisely, the depreciation
- 14:19debate is a sideshow. The central narrative for
- 14:22Microsoft is their aggressive push into in house custom
- 14:26silicon, specifically their Maya 200 line of chips.
- 14:30Because the NVIDIA tax. Exactly.
- 14:32Think back to Nvidia's 75% margin.
- 14:35Microsoft is writing those massive checks to NVIDIA and
- 14:38they realize that to truly win this era, they need to keep that
- 14:42margin for themselves. Makes sense.
- 14:44If Microsoft can successfully transition a significant portion
- 14:47of their internal predictable AI workloads like running Copilot
- 14:50for those 30 million paid users, onto their own Maya 200 ships,
- 14:54the economics completely flipped.
- 14:56If they can drop the cost per token of running AI by 30% using
- 15:00their own hardware instead of paying the NVIDIA premium, their
- 15:03return on investment skyrockets. The capital they are deploying
- 15:06becomes radically more efficient.
- 15:08So to summarize the US tech landscape, NVIDIA, Amazon,
- 15:13Microsoft, they are all engaged in a war of brute force
- 15:17computational scale. Build the biggest data centers,
- 15:20buy or design the most powerful chips, and rent that sheer
- 15:23computational horsepower out to the world.
- 15:25That's the playbook. But this is where the global
- 15:27Divergent gets truly fascinating.
- 15:30When we flip the stack and look at the mid year reports from the
- 15:33Chinese tech giants, we see a remarkably different philosophy.
- 15:37Completely different. They aren't trying to out
- 15:39compute the US giants in a raw hardware race.
- 15:42They are taking a deeply pragmatic approach.
- 15:44They are choosing integration over pure scale.
- 15:47It is a profound, strategic, divergent While the US giants
- 15:51are pouring billions into building the widest, fastest
- 15:53digital highway, the Chinese tech ecosystem has realized that
- 15:57they don't necessarily need to own the asphalt right.
- 16:00Instead, they are hyper focused on building the most efficient
- 16:02cars, the best delivery services, and the stickiest
- 16:05applications that will run on that highway.
- 16:07That's a great analogy. They are aggressively
- 16:09integrating AI into their massive pre-existing digital
- 16:13ecosystems and focusing on commercializing application
- 16:16programming interfaces, or APIs. Let's start with Tencent because
- 16:20their numbers basically illustrate this.
- 16:22Their Q2 revenue was up a steady 11%.
- 16:25Their non IFRS net profit was up 9%.
- 16:30But here is the detail buried in the notes.
- 16:32If you actually exclude the costs associated with launching
- 16:36their new AI products, their core profit would have been up
- 16:3919%. So AI is costing them a lot,
- 16:41right? Now, exactly right now, AI is
- 16:44acting as a massive drag on their bottom line.
- 16:46Furthermore, their Q2 CapEx hit 52.8 billion RMBA staggering
- 16:52176% increase, which completely wiped out their cash generation
- 16:56and turn their free cash flow negative.
- 16:58Yeah, that's a tough pill to swallow for investors.
- 17:00It is if you are a $0.10 shareholder.
- 17:02We have to be asking like, are they just stubbornly trying to
- 17:04build a Chinese clone of Open AI from scratch and lighting
- 17:08billions on fire to do it? The short answer is no, and that
- 17:11assumption completely misses the genius of their long term
- 17:14strategy. OK, explain.
- 17:15Tenson is not interested in building a standalone monolithic
- 17:20disrupts the world general AI. Their strategy is deeply
- 17:24pragmatic. They are building what they call
- 17:26AI Plus to deepen and fortify the Moat of their existing
- 17:30ecosystem. You have to remember the context
- 17:33here. WeChat has 1.4 billion active
- 17:37users. It isn't just a messaging app,
- 17:40it is the fundamental operating system of daily life in China.
- 17:44Right. It handles your communications,
- 17:46your payments, your mini programs, your gaming, your ride
- 17:48hailing. It's an ecosystem of absolute
- 17:51daily necessity. Exactly $0.10 fundamental bet is
- 17:54that the ultimate victor in the AI race will not necessarily be
- 17:58the company with the absolute highest raw IQ or the most
- 18:01computationally dense foundational model.
- 18:04Who will it be them? The winner will be the model
- 18:06that possesses the most complete holistic user context.
- 18:09Oh, that's interesting. As time goes on on, the gap in
- 18:11raw intelligence between the top tier models will inevitably
- 18:14shrink, right Foundational intelligence will slowly
- 18:16commoditize. Yeah, everyone will eventually
- 18:18catch up. Right.
- 18:20So when that happens, the true differentiator becomes
- 18:22personalization. Because Tencent has over a
- 18:25decade of social graphs, reading habits, payment behaviors, and
- 18:29transaction histories, their integrated AI can be infinitely
- 18:33more helpful and intuitive to a user than a brilliant but
- 18:36generic stand alone model. Wow.
- 18:38Yeah, they aren't trying to replace their core business with
- 18:41a new AI business. They are using AI to turbocharge
- 18:45the WeChat ecosystem, making it utterly indispensable.
- 18:49That is a brilliant distinction. It's context over raw IQ.
- 18:53Like a genius assistant who knows everything about you is
- 18:56way more useful than a genius assistant who knows everything
- 18:59about quantum physics but doesn't know what you want for
- 19:00line. Exactly.
- 19:01Well put. OK.
- 19:02Let's look at Alibaba because their financial report is, well,
- 19:05it's a tough read. It is rough.
- 19:07Revenue came in at 269 billion RMB, which is up 9%, but that is
- 19:12where the good news stops. Their Adjusted EBITDA, which is
- 19:15a measure of their core operational profitability fell
- 19:1830%. Their net profit completely
- 19:20plummeted, down 75% and to top it off, they posted a negative
- 19:25free cash flow of 44.7 billion RMB.
- 19:30It looks really bad on paper. It does what?
- 19:32Where is all this cash going? Are they just losing the
- 19:34e-commerce war? It looks disastrous at first
- 19:37glance for sure, but Alibaba is actively choosing to invest
- 19:41through the pain. We are watching a company
- 19:44undergo A fundamental structural metamorphosis.
- 19:48They are transitioning from being a relatively asset light
- 19:51e-commerce platform connecting buyers and sellers into a highly
- 19:55capital intensive heavy asset AI infrastructure player.
- 19:59Oh wow. The cash they are burning is
- 20:01primarily going into two distinct areas.
- 20:03First, defending their core retail turf with massive
- 20:06investments in Taobao and their fresh Shippo grocery lines.
- 20:08Basically fighting off intense domestic competition.
- 20:11But second. And more importantly for our
- 20:13discussion is AI infrastructure where they deployed 6.7 billion
- 20:18RMB in CapEx just in the single quarter.
- 20:20So they are intentionally bleeding out their e-commerce
- 20:23profits to fund this massive pivot into infrastructure.
- 20:28That is a terrifying tightrope to walk.
- 20:30It takes nerves of steel. Are there any indicators in this
- 20:33report that it's actually working?
- 20:36Yes, there is a very bright silver lining.
- 20:39If you look at Alibaba Cloud, their cloud division is
- 20:42experiencing its fastest growth in 22 quarters, generating 48.4
- 20:48billion RMB, which is a 45% increase.
- 20:51OK. That's substantial.
- 20:52And crucially, unlike many competitors who are subsidizing
- 20:55their cloud growth with massive losses, Alibaba Cloud is
- 20:59actually profitable. Their EBITDA margin for the
- 21:02cloud division hit 12%. Oh nice.
- 21:04What they are doing is calculated, they are absorbed,
- 21:06curbing massive margin hits on the e-commerce side to ensure
- 21:10that they establish absolute dominance over the foundational
- 21:12compute layer in China. They want to be the AWS of the
- 21:15East. Got it?
- 21:16It is a painful transition for shareholders right now, but they
- 21:19believe it is an existential necessity for the next decade.
- 21:22All right, let's move from the giant ecosystems to a pureplay
- 21:26AI company that is absolutely turning heads to the developer
- 21:29community right now. Zipu AI.
- 21:32Their numbers are wild and they completely challenge a major
- 21:35Silicon Valley narrative. They really do.
- 21:37In the first half of the year, Zipu's revenue hit 954,000,000
- 21:42RMB, which is up 400%. Incredible growth.
- 21:45But look at where it came from. Their API revenue, which is the
- 21:49money they make when outside developers pay to plug their own
- 21:52apps into Zipu's AI model, hit 825,000,000 RMB.
- 21:58Unbelievable. That is a 2736% increase.
- 22:02API revenue shifted from being just 1/4 of their business to
- 22:05making up 86.5% of their total revenue.
- 22:08It is a textbook explosion of API adoption.
- 22:10Developers are just flocking to their ecosystem really.
- 22:13However, this hyper growth comes at a very steep, very visible
- 22:17cost. Zipu reported a massive net loss
- 22:20of 1.96 billion RMB and that is driven almost entirely by their
- 22:25aggressive R&D spending, which came in at 2.1 than 3 billion
- 22:28RMB. OK, I have to stop you there,
- 22:30because this is the piece of data that I find most
- 22:32fascinating. Earlier this year, Jeepu raised
- 22:37their API prices by 83%. They did.
- 22:40Yet despite nearly doubling the cost to use their product, the
- 22:44number of API calls they processed jumped by 400%.
- 22:48Doesn't that completely destroy the prevailing narrative we've
- 22:51been fed that AI models are just racing to the bottom on price?
- 22:55Like, for months analysts have warned that AI is a commodity
- 22:59and whoever offers the cheapest inference tokens will win the
- 23:02market. But Zipu raised their prices and
- 23:05saw a forex explosion in volume. It is a fantastic observation,
- 23:09and yes, it proves that the race to the bottom narrative is
- 23:11fundamentally flawed or at the very least, highly premature.
- 23:15Zipu's numbers demonstrate that they possess genuine pricing
- 23:18power in a crowded market. Which is rare.
- 23:20Extremely. What it tells us is that not all
- 23:23AI models are created equal. Zipu's models are proving to be
- 23:26genuinely superior for complex, specific developer tasks, and
- 23:31developers are highly rational actors.
- 23:33You know they will gladly pay an 83% premium if the model is more
- 23:38reliable, if it hallucinates less, or if it requires less
- 23:41prompt engineering to get the desired result.
- 23:43If it saves them time, it saves them money.
- 23:45Exactly. And to give you a sense of their
- 23:48technical velocity, Zipu recently released a new
- 23:51iteration of their model that overtook DeepSeek, which had
- 23:54held the number one bought on the major benchmarks for 56
- 23:58days. Zipu dethroned them in just six
- 24:01days. Six days.
- 24:03That iteration speed is terrifying.
- 24:05It's a relentless pace. It means a competitor can
- 24:08achieve A breakthrough and less than a week later they are
- 24:11obsolete again. Exactly.
- 24:12Now concerning their profitability, they are
- 24:14absolutely burning cash right now based on their filings.
- 24:17Calculations show they're burning roughly 3.2 RMB for
- 24:21everyone RMB they earn in revenue.
- 24:23Ouch. But this is a calculated land
- 24:25grab. They are aggressively capturing
- 24:27market share and, more importantly, securing developer
- 24:30mindshare. Lock them in now, right?
- 24:33The strategy is clear, locking the developers now while the
- 24:35technology is nascent as the underlying cost of computing
- 24:39these tokens, the inference cost naturally drops over the next
- 24:42few years due to a better hardware and software
- 24:44optimization. That massive, deeply entrenched
- 24:47API revenue base will transition into a very sticky, incredibly
- 24:52high margin business. So it's the classic tech
- 24:55playbook. Build the addiction, secure the
- 24:57developer ecosystem and worry about the margins later.
- 25:00Exactly. All right, so we've spent a lot
- 25:02of time talking about software, AP IS and the digital ether.
- 25:06But none of this exists without physical reality.
- 25:09You cannot run these massive AI models without highly advanced
- 25:13memory chips, and you certainly can't power those data centers
- 25:16without astronomical amounts of electricity and next generation
- 25:19batteries. Let's move from the abstract to
- 25:22the physical. Let's look at section three of
- 25:24our stack powering the future hardware and energy.
- 25:26And we have to start with Chinksin Memory because their
- 25:28mid year report is nothing short of a miraculous turn around.
- 25:31Miraculous is exactly the right word.
- 25:34The data from Chinksin Memory's half year report is stunning.
- 25:38Let's review the top line. H1 revenue hit 150.3 billion
- 25:43RMB. That represents an 873% year
- 25:48over year increase and their net profit landed at 77.6 billion
- 25:54RMB, which is a massive rebound from a net loss in the same
- 25:58period just last year. I.
- 26:00Need to make sure you heard that right. 873% revenue growth and
- 26:04swinging from a loss to 77.6 billion in profit.
- 26:08It's. Real.
- 26:09How does a hardware manufacturer achieve that kind of velocity in
- 26:12six months? It was a perfect confluence of
- 26:14positive factors. They aggressively expanded their
- 26:17wafer capacity right as demand spiked.
- 26:19OK, Crucially, they significantly improved their
- 26:22manufacturing yield rates, meaning fewer defective chips
- 26:25per silicon wafer. That's.
- 26:26Huge for margins. Massive Simultaneously, global
- 26:29unit prices for DRAM memory surged, and they highly
- 26:32optimized their product mix toward premium offerings.
- 26:35OK, so firing on all cylinders. Basically.
- 26:39But out of all those factors, the single most staggering
- 26:42figure in their report is their gross margin.
- 26:44It hit an incredible 84.8%. 84.8% margin on a physical piece
- 26:50of hardware that is higher than many software as a service
- 26:53companies. It is.
- 26:55That essentially rivals Nvidia's gross margin, but they are doing
- 26:59it with memory chips. It does rival NVIDIA, and it
- 27:01underscores A deeply important shift in the geopolitical tech
- 27:05landscape. This report represents a true
- 27:07watershed moment for domestic Chinese technology.
- 27:10How so? For years, the established
- 27:12accepted narrative was that Chinese memory manufacturers
- 27:15were relegated to the budget tier.
- 27:17Right. Yeah, Volume over quality.
- 27:18Exactly. They could produce high volumes
- 27:21for cheap low end devices, but they couldn't compete at the
- 27:24bleeding edge. Well, Jenksen shattered that
- 27:27narrative. They successfully mass produced
- 27:29LPDDR 6 memory chips and secured placement in the Xiaomi 18 Fold,
- 27:34which is a premium high end flagship device requiring
- 27:37intense thermal management and efficiency.
- 27:40So they aren't just the good enough cheap alternative
- 27:42anymore, they are being chosen for the highest tier consumer
- 27:46tech. Exactly.
- 27:47And perhaps more impressive than the placement itself is the
- 27:50speed at which they executed it. The global standard for LPDDR 6
- 27:55was only finalized recently, right?
- 27:57Jenksen went from participating in the standard setting phase to
- 28:00architectural iteration through rigorous client validation, and
- 28:04all the way to commercial mass production in the span of just
- 28:07one year. That is an unbelievable timeline
- 28:09for hardware engineering. It is.
- 28:11It proves unequivocally that domestic memory fabrication can
- 28:14compete at the absolute highest global standards.
- 28:17It completely rewrites the supply chain risk models for
- 28:20mobile and edge AI hardware over the next decade.
- 28:23Incredible. OK, Speaking of the physical
- 28:24components that power our world, let's talk about batteries,
- 28:28CATL. The Behemoth.
- 28:30They are the undeniable behemoth in the EV battery space.
- 28:34Their numbers were remarkably strong.
- 28:36Revenue hit 276.9 billion RMB, which is up 55%.
- 28:41Profit was 43.3 billion RMB, up 42%.
- 28:46But honestly, in a sea of massive numbers, the revenue
- 28:49wasn't the headline grabber. The headline was what they
- 28:52decided to do with their cash. Yes, the strategic capital
- 28:56allocation is the real story here.
- 28:58CATL announced a massive share buyback program, committing
- 29:03between 20 billion and 40 billion RMB to repurchase their
- 29:07own stock. It's a lot of cash.
- 29:09But here is the critical distinction.
- 29:11All of those repurchase shares are earmarked for cancellation.
- 29:15Oh, they are not buying them back to hold in the Treasury and
- 29:18they are not using them to fund employee stock option pools,
- 29:22which would simply dilute the shareholder base all over again,
- 29:24right? They are buying the shares and
- 29:26destroying them, permanently reducing the share count and
- 29:29instantly increasing the underlying value and earnings
- 29:32per share for every existing shareholder.
- 29:34That is a massive flex of corporate confidence.
- 29:38It's essentially management looking at the market and saying
- 29:40like, we generate so much cash and we think our stock is so
- 29:44undervalued that the absolute best investment we can make
- 29:46right now is simply buying our own company.
- 29:49It is the ultimate flex and it reflects a highly rational,
- 29:53mature growth strategy. Look at their operational
- 29:56efficiency. They reported an astounding 95%
- 30:00capacity utilization rate across their global factories. 95%,
- 30:04yeah. In a global EV industry that has
- 30:06been plagued for the last 18 months by intense fields of
- 30:09structural overcapacity and idle factories, achieving 95%
- 30:14utilization is phenomenally efficient.
- 30:17Yeah, no wasted space there. Furthermore, they are
- 30:20successfully diversifying their revenue base away from just
- 30:23passenger EVs energy storage. The massive grid scale battery
- 30:28used to stabilize renewable energy networks now accounts for
- 30:31a full third of their total sale.
- 30:33That's a huge pivot. And they announced they're
- 30:35beginning the commercialization phase of sodium ion batteries, a
- 30:39technological leap that will significantly reduce their
- 30:42reliance on volatile global lithium prices.
- 30:45So if I'm understanding this correctly, while their
- 30:47competitors are slashing prices to the bone, fighting a brutal
- 30:50price war just to maintain factory throughput and win
- 30:53scraps of market share, CATL is completely opting out of that
- 30:56race to the bottom. Precisely.
- 30:59They are refusing to engage in a destructive price war.
- 31:03They are hyper focused on providing technological value,
- 31:06defending their premium margins and returning massive amounts of
- 31:10cash to their shareholders. It is a textbook example of how
- 31:14a dominant market leader operates in a maturing industry.
- 31:18All right, let's trace the energy demand all the way back
- 31:20to the source. Let's look at the raw power
- 31:23required to run the factories and the data centers.
- 31:25Yangtze Power. Yeah, yes.
- 31:27I like to call them the cash flow Leviathan.
- 31:30They reported net profit of 13% and they are consistently
- 31:33generating over 30 billion RMB and free cash flow annually.
- 31:37They are an absolute financial machine.
- 31:39Their cash generation is incredibly reliable.
- 31:42I was looking at their business model to understand how they can
- 31:44be so consistent, and really they're basically operating a
- 31:47giant toll booth on the Yangtze River.
- 31:49That's a great way to put. It the water flows down the
- 31:51river, it spins their massive hydroelectric turbines, and they
- 31:55just sit there and print money. They don't have to mine for
- 31:57fuel. They don't have to buy coal.
- 31:59It seems like the ultimate passive income business on a
- 32:02geopolitical scale. It is an incredibly strong,
- 32:06enviable foundational base, yes. But for anyone analyzing this
- 32:10company, there is a layer of operational nuance that you must
- 32:14understand. Their baseline revenue is not
- 32:17entirely within their control. What do you mean?
- 32:20Their entire fiscal year performance is heavily dictated
- 32:23by the volume of water flow, specifically during the third
- 32:27quarter, the wet season. The weather.
- 32:29Right. If there is a severe drought,
- 32:31their power generation drops and revenues dip.
- 32:34Conversely, if there are heavy floods, revenues spike.
- 32:36But they also have to manage infrastructure risks.
- 32:39So despite all the advanced engineering, they are
- 32:41fundamentally at the mercy of the weather.
- 32:43Completely. And they recognize that
- 32:45vulnerability. To combat that volatility and
- 32:48secure future growth, their core strategic expansion relies
- 32:51heavily on something called pumped hydro storage.
- 32:54Can you explain the mechanics of pumped hydro for me?
- 32:56I hear it mentioned a lot in green energy reports, but how
- 32:59does it actually generate a return?
- 33:01Sure. Think of pumped hydro as a
- 33:03massive GR geography based battery.
- 33:05It relies on having two reservoirs of water, one at the
- 33:08bottom of a mountain and one at the top.
- 33:10When electricity on the National Grid is cheap and abundant, For
- 33:13example in the middle of the night when demand is low but the
- 33:16wind is blowing hard and wind turbines are spinning, Yanks
- 33:21Power uses that cheap surplus electricity to literally pump
- 33:25millions of gallons of water from the bottom reservoir up the
- 33:28mountain into the top reservoir. So they use power to move the
- 33:31water. Up exactly then, Then, during
- 33:34the peak daylight hours, when factories and cities are drawing
- 33:37massive amounts of power and electricity prices are at their
- 33:40highest, they open the gates. They let the water flow back
- 33:44down the mountain through their turbines, generating high value
- 33:47electricity that they sell back to the grid.
- 33:49That is brilliant. It's the most basic rule of
- 33:51trading. Buy low, sell high.
- 33:54But instead of trading stocks, they are trading gravity and
- 33:56water. Exactly.
- 33:57It's energy arbitrage. However, while the physics are
- 34:00sound, the economics are currently facing headwinds.
- 34:04As the National Grid integrates higher percentages of
- 34:06intermittent wind and solar power, the regulatory frameworks
- 34:10dictating how these pump storage facilities are compensated are
- 34:13constantly evolving. That makes sense.
- 34:15Furthermore, there is a looming risk of regional overcapacity.
- 34:19If too many power companies build pump storage plants
- 34:22simultaneously to capture these subsidies, the arbitrage spread
- 34:25will collapse, right? So while their river based hydro
- 34:29business is an absolute cash machine, their future growth
- 34:32vector in pumped storage is significantly more complex and
- 34:35carries real regulatory risk. OK, we've covered the abstract
- 34:39AI models, the advanced memory chips, the batteries, and the
- 34:42raw power grid. Let's bring this deep dive
- 34:45squarely down to earth and look at the physical economy.
- 34:48Well, let's do it. How are the companies making
- 34:50physical goods, cars, metals, houses dealing with the reality
- 34:54of global economic cycles? Let's look at BYD, the EV giant,
- 34:59because frankly, reading their media report gave me financial
- 35:02whiplash. Byd's report is a perfect
- 35:04example of why you cannot just read the headline numbers.
- 35:07It requires very careful methodical parsing.
- 35:10On the surface, the top line metrics looked deeply concerning
- 35:14for a growth stock. Their H1 auto sales volume
- 35:17actually dropped 15.7%, moving exactly 1.8085 million vehicles.
- 35:26Correspondingly, their total revenue was down 7.1%.
- 35:30Right. If you stop reading there, you
- 35:32think the BYD growth story is dead.
- 35:35They sold fewer cars and they made less revenue.
- 35:38But you kept reading. I did.
- 35:39I looked at their balance sheet, specifically their contract
- 35:42liabilities for those listening, contract liabilities in the auto
- 35:45industry essentially represent the cash deposits customers have
- 35:48put down to order a car that hasn't been billed or delivered
- 35:50yet, right? And that specific line item
- 35:52jumped 63%. So consumer demand is clearly
- 35:56there, they just haven't delivered the metal yet.
- 35:58Exactly. But here is the big reveal, the
- 36:00single metric that completely broke my brain when I read it.
- 36:04Despite delivering 15% fewer cars, DYD Score operating profit
- 36:08per car surged by a massive 60%, hitting 8543 RMB per vehicle.
- 36:15And the underlying operational performance is even stronger
- 36:18than that. If you strip out the financial
- 36:20impact of their foreign exchange losses, their actual operational
- 36:24profit per car was an astonishing 9300 R MB.
- 36:29I need you to explain the physics of this to me.
- 36:31How is that possible? How do you sell significantly
- 36:35fewer cars, yet make drastically more money on every single unit
- 36:39that rolls off the line? It seems impossible.
- 36:41It contradicts every traditional rule of automotive economies of
- 36:44scale I've ever learned. The old rule is that you have to
- 36:47pump out millions of identical low margin cars just to cover
- 36:51your massive factory overhead and scrape together a profit.
- 36:55How did BYD reverse the math? Because they executed A
- 36:58masterful deliberate strategic pivot, what we are seeing in
- 37:02this report is BYD successfully transitioning away from their
- 37:05historical scale first strategy into a highly optimized profit
- 37:09first strategy. OK, walk me through that.
- 37:12In their early days, they desperately needed volume.
- 37:15They had to just the massive capital expenditure of building
- 37:18their factories and they had to flood the streets with their
- 37:21cars just to build consumer brand awareness and trust.
- 37:25They sold cheap cars at razor thin margins to win market
- 37:29share. Right, get the cars on the road.
- 37:31But they have achieved that scale now.
- 37:34Now they are relentlessly optimizing their product mix.
- 37:38They are aggressively pushing a higher margin premium vehicles
- 37:42like their Yang Wang and Feng Shing Bao brands.
- 37:45Moving up market. Furthermore, they are leveraging
- 37:47their unparalleled vertical integration.
- 37:50I mean, they manufacture their own batteries, their own
- 37:53semiconductor chips, their own seats to squeeze every single
- 37:56drop of operational efficiency out of the assembly line.
- 37:59They don't need to sell the most cars anymore, They just need to
- 38:01sell the right cars. OK.
- 38:03That explains the incredible margin expansion.
- 38:05But you briefly mentioned foreign exchange losses.
- 38:08They took a massive multibillion dollar hit there, didn't they?
- 38:11They did. They were hammered by a 4.88
- 38:15billion RMB foreign exchange loss in this period.
- 38:18Almost 5 billion RMB wiped out simply because currencies
- 38:21fluctuated. Exactly.
- 38:23And this highlights their next great challenge.
- 38:25As BYD rapidly expands globally, particularly as they try to
- 38:30navigate the complex geopolitical minefield of new
- 38:33European Union tariffs, and as they deploy capital to build new
- 38:37factories overseas in places like Hungary and Brazil, hedging
- 38:41against these violent currency swings will become their biggest
- 38:44operational headache. That sounds like a nightmare.
- 38:47They have proven they can beat legacy automakers in engineering
- 38:50and manufacturing. Now they have to prove they can
- 38:53fight the global currency markets.
- 38:54It is a completely different, much more sophisticated weight
- 38:57class. Speaking of fighting global
- 38:59macro, our markets, let's look at the resources sector, Z Gen.
- 39:02Mining and Tianky Lithium. Comparing these two reports is
- 39:06truly a tale of two resources, and it perfectly illustrates how
- 39:09commodity cycles dictate destiny.
- 39:11It really is a stark contrast. Let's start with Z Gen.
- 39:14Mining. Their revenue was up a
- 39:15respectable 16%, but their net profit exploded up 68%.
- 39:20Why is there such a massive disparity between top line
- 39:23growth and bottom line profit? The answer boils down to 2
- 39:27words. Profit leverage.
- 39:30The economics of the mining industry are defined by
- 39:32incredibly high, inflexible fixed costs.
- 39:36To get an ounce of metal you have to buy the billion dollar
- 39:39heavy machinery, you have to blast the rock, you have to
- 39:42build the roads and processing facilities, and you have to
- 39:45employ thousands of miners regardless of what the metal
- 39:48actually sells for on the open market.
- 39:50But once you have sold enough metal to cover those baseline
- 39:53fixed costs, the financial dynamics change radically.
- 39:57Because the costs to extract the next oz is essentially 0, the
- 40:01profit margins expand exponentially.
- 40:04Right now, global goal and copper prices are sitting at
- 40:07historic multiyear highs. So because the price of gold is
- 40:10so high, every extra dollar they make in revenue from selling it
- 40:14doesn't cost them a single extra cent in overhead?
- 40:16Exactly. It bypasses the expenses
- 40:18entirely and drop straight to the bottom line profit.
- 40:21It's pure leverage. Furthermore, Zegen's management
- 40:24has been brilliantly aggressive during the cycle.
- 40:27They've been actively acquiring new high quality mines globally.
- 40:30Mergers and acquisitions are driving their underlying volume
- 40:33growth. Smart.
- 40:35Yet amazingly, despite spending capital to buy up these massive
- 40:38assets, they generated so much cash from high commodity prices
- 40:42that they managed to drop their overall corporate debt ratio
- 40:45below 50% for the first time since 2012.
- 40:49It is an absolute master class in capital allocation during a
- 40:52commodity supercycle. OK, so Z Gen. is riding the wave
- 40:55perfectly, but what about Tianki Lithium?
- 40:59Their numbers look even Wilder on paper, but the reality is
- 41:02much scarier. They reported revenue up 150%
- 41:06and profit up 49 X 49 times higher than the previous year.
- 41:12Yeah, we have to be very careful not to be blinded by that 49 X
- 41:15multiplier. It's a statistical anomaly
- 41:17caused by a phenomenally low, almost non existent profit base
- 41:21in the previous year's comparison period.
- 41:23OK, that makes sense. But if you look past the
- 41:25headline multiplier, what is truly fascinating about Tianki
- 41:28is their Q2 operational performance.
- 41:30In the second quarter, their revenue rose 38%, even though
- 41:34global lithium prices only rose 12% in that exact same time
- 41:37frame. Wait, if the price of the
- 41:39commodity they sell only went up 12%, how did their revenue jump
- 41:4338%? Did they find a premium market?
- 41:46No, they didn't find a premium market.
- 41:48They found volume. They successfully and
- 41:51aggressively increased their raw production and their sales
- 41:54volume to offset the sluggish price action.
- 41:58They pumped significantly more physical lithium into the market
- 42:01to make up for the fact that the price wasn't where they wanted
- 42:03it to be. That seems risky.
- 42:05It is a dangerous game. With Q3 data showing that global
- 42:10lithium prices are falling once again due to EV slowdown fears.
- 42:13Tianki is entirely at the mercy of the macro cycle.
- 42:17Unlike Zejan, who is comfortably riding the tailwinds of high
- 42:21copper and gold prices, Tianki is desperately trying to out
- 42:24produce a falling price curve. They have to dig faster and
- 42:27faster just to maintain their revenue.
- 42:29It is a much more precarious, exhausting operational position.
- 42:33Which perfectly transitions us into looking at companies
- 42:35fighting for their lives. If you want to talk about
- 42:38precarious positions, we need to talk about real estate,
- 42:40agriculture and global shipping. Let's dig in.
- 42:43How does a massive corporation survive when its entire
- 42:46underlying industry is sitting at the absolute suffocating
- 42:49bottom of a brutal economic cycle?
- 42:52Let's look at vonke 1 of Chinas largest, most historically blue
- 42:55chip real estate developers. Yeah, because looking at this
- 42:58report, it looks like a financial bloodbath.
- 43:00It has undeniably been an incredibly painful systemic
- 43:03period for the entire Chinese real estate sector, and Vonke is
- 43:07not immune. They reported total revenue of
- 43:1070 billion RMB, which represents a massive 30% drop year over
- 43:15year. And more alarmingly, they posted
- 43:18a net loss of 14.95 billion RMB. Almost 15 billion RMB and pure
- 43:24losses in half a year. Yes, the accounting losses are
- 43:27severe. But, and this is a massive
- 43:29structurally vital, but for anyone analyzing their survival
- 43:32prospects, their operating cash flow actually turned positive.
- 43:35OK, we need to unpack this slowly.
- 43:37How in the world is a the company declare a 15 billion RMB
- 43:40loss on their income statement but simultaneously generate
- 43:43positive cash flow on their bank accounts?
- 43:45It requires understanding how real estate accounting works
- 43:48within China's entirely new regulatory paradigm.
- 43:52Historically, the entire sector relied on a high leverage
- 43:54presale model. Developers would market a
- 43:57blueprint, collect full payment from buyers before a single
- 44:01shovel hit the Burt, and use that cash to buy more land.
- 44:04Right, spinning the plates. Exactly.
- 44:06But government policies have forcefully dismantled that
- 44:09system to protect consumers. The industry has been forced to
- 44:13pivot to a high capital completed home model where
- 44:16developers must actually finish construction and deliver the
- 44:19physical keys before they can recognize the bulk of the
- 44:21revenue. So you have to finish the house
- 44:23first. Right.
- 44:24And Vankari is surviving this agonizing transition by
- 44:27ruthlessly shedding non core assets to raise capital, tightly
- 44:31controlling construction outflows and crucially, securing
- 44:34vital financial support and credit lines backed by the
- 44:36Shenzhen government. The 15 billion RMB loss you see
- 44:40on the books largely reflects massive non cash asset
- 44:44impairments. They are writing down the stated
- 44:46value of their land banks to reflect new lower market
- 44:49realities I see. But from a pure operational
- 44:52standpoint, by aggressively collecting on completed projects
- 44:56and pausing land buys, they actually brought more hard cash
- 44:59in the door than they sent out this quarter.
- 45:01So by managing the actual cash in the bank, the bleeding has
- 45:05stopped. I would say the arterial
- 45:06bleeding has been tourniqueted but the patient is still in the
- 45:10ICU. The positive cash flow buys them
- 45:12critical breathing room, but it absolutely does not not solve
- 45:15their structural debt mountain. They still have a mountain of
- 45:18debt. They still have roughly 180
- 45:20billion RMB and short term debt that they urgently need to roll
- 45:25over, refinance or repay in the next 12 months.
- 45:28The positive cash flow proves their core operations can still
- 45:31function, but they are walking a high wire act in a hurricane.
- 45:35It is a harrowing position to be in.
- 45:37Let's look at another industry famous for its brutal Boone and
- 45:40bust cycles, agriculture, specifically pork production.
- 45:44Mui Wan is a giant in this space and they posted a highly
- 45:48concerning H1 loss of 6 billion RMB.
- 45:51The pork cycle in China is notoriously vicious.
- 45:54It oscillates wildly based on herd sizes and feed costs.
- 45:58When market prices crash below the cost of breeding, literally
- 46:02every producer in the country loses money on every pig they
- 46:06sell. Just a terrible place to be.
- 46:07Exactly. But what makes MU Yuan's Mid
- 46:09Year reports so fascinating is the discrepancy between their
- 46:13accounting and their cash. Despite posting a 6 billion RMB
- 46:18accounting loss, their actual cash outflow, the money that
- 46:23left the business was only 2.2 billion RMB.
- 46:26Again, why the massive difference between the stated
- 46:29loss and the cash loss? Primarily depreciation on their
- 46:32massive farming infrastructure and other non cash accounting
- 46:35expenses. But the real story isn't the
- 46:37accounting mechanics, it's what me Yun's management actively did
- 46:40while the market was crashing and they were losing money.
- 46:42I did, They do. They didn't just sit on their
- 46:43hands, freeze operations and pray for pig prices to magically
- 46:46recover. They aggressively used the cycle
- 46:49Bottom to restructure. They methodically closed down
- 46:53their oldest, most inefficient farms.
- 46:55They poured capital into upgrading their genetic breeding
- 46:58programs to produce healthier, faster growing livestock.
- 47:02And most innovatively, they partnered with tech giants like
- 47:05Huawei and Alibaba to implement highly advanced sensor driven
- 47:10smart digital farming systems across their facilities.
- 47:14I love this. It's like they use the downtime
- 47:16to install a massive system wide software and hardware update for
- 47:20their entire biological infrastructure.
- 47:22That is exactly what they did, and the operational result of
- 47:25that upgrade is stunning. Their core breeding cost per
- 47:28kilogram of poor dropped to just 11.5 RMB.
- 47:32That is an industry leading level of baseline efficiency.
- 47:35Wow. So think about what happens
- 47:37next. They survived the brutal winter
- 47:39by completely rebuilding A vastly superior, more efficient
- 47:42furnace. When the pork cycle inevitably
- 47:44turns, when supply tightens and market prices rise back to 15 or
- 47:4816 RMB per kilo, MU Yuan is primed to see their profits
- 47:52absolutely explode because their baseline operational costs are
- 47:55now so incredibly low. They're ready for the rebound.
- 47:58They use the crisis to forge an insurmountable competitive
- 48:01advantage. That is a master class in cycle
- 48:03management. Survive the winter by building a
- 48:05better furnace. Definitely.
- 48:07All right, let's. Move from the farms to the
- 48:08oceans. Segasco, the global shipping and
- 48:11logistics giant. Their headline profit dropped
- 48:1423%. If you're casually tracking
- 48:18global shipping rates, that number doesn't seem to make
- 48:20sense. On paper, a 23% profit drop
- 48:24looks like a significant miss, but again, the headline is a
- 48:29mirage. We have to look at the financial
- 48:31line items. OK, break it down.
- 48:33In this period, Costco suffered a massive 2.45 billion RMB hit,
- 48:38specifically from foreign exchange losses and a
- 48:41corresponding reduction in interest income.
- 48:44This was largely driven by the sudden appreciation of the RMB
- 48:47against the currencies they conduct trade in.
- 48:50If you strip away that currency fluctuation, if you add back the
- 48:522.45 billion RMB currency loss, they actually had a fantastic
- 48:57operational quarter, right? That completely syncs up with
- 49:00the rising global freight rates we've been seeing due to the Red
- 49:03Sea shipping diversions and the lingering supply chain
- 49:05bottlenecks. Correct.
- 49:07The core underlying shipping business is booming.
- 49:10They are moving boxes at very healthy margins, but the much
- 49:14bigger, more consequential story for long term investors is
- 49:18Costco's ongoing structural transformation.
- 49:21Transformation into what? The maritime shipping industry
- 49:23has historically been viewed as a wildly volatile, boom and
- 49:27bust, highly cyclical sector, right?
- 49:29You make a fortune one year and you go bankrupt the next.
- 49:32Yeah, totally. But Costco's management is
- 49:34actively transitioning the company into a predictable,
- 49:37highly reliable dividend machine.
- 49:41They are currently paying out nearly 50% of their massive net
- 49:44profits directly back to shareholders as dividends and
- 49:47they just launched a new 10 billion plus RMB share buyback
- 49:51program. They are sending a very clear,
- 49:54very loud signal to the global capital markets.
- 49:57We aren't just a speculative gamble on next month's spot
- 49:59freight rates anymore. We are a mature, stable, high
- 50:02yield cash generator. Which is exactly what
- 50:05institutional investors are desperate for in a choppy,
- 50:09uncertain macro market. Absolutely.
- 50:12OK, let's bring our focus back to the everyday domestic
- 50:14reality. Let's look at the consumer
- 50:17pivot. Are people upgrading or are they
- 50:20down grading? Because the domestic consumer
- 50:22market in China is undeniably tight right now.
- 50:25It's. Very tight.
- 50:25People are watching their wallets, they're delaying big
- 50:28purchases, and consumer confidence is fragile.
- 50:32So companies selling consumer goods are faced with a stark
- 50:35binary choice. Do you aggressively defend your
- 50:38premium pricing and risk losing market share?
- 50:40Or do you accept that the domestic market is tapped out
- 50:43and aggressively expand overseas?
- 50:45Right fight or flight? Let's look at 2 domestic giants
- 50:48that chose defense, Yeley Dairy and Luzu Lo Xiao, the premium
- 50:53liquor brand. They both chose to hold the
- 50:54line, and their reports offer a fascinating contrasting case
- 50:58study and how to preserve brand equity in a downturn.
- 51:01Let's examine Yeley first. OK, so the broader dairy market
- 51:05in China is stagnant. It is currently growing at a
- 51:08sluggish rate of about 1% annually.
- 51:11Yet in that exact same environment, Yeley somehow
- 51:14managed to grow their core revenue by 4.13%.
- 51:19How is that mathematically possible?
- 51:21How do you grow 4 times faster than the overall market when the
- 51:24population isn't magically drinking more milk?
- 51:27You must be stealing that growth directly from your competitors.
- 51:30Exactly. It is a story of ruthless
- 51:33operational efficiency and aggressive market share.
- 51:36Consolidation yearly recognized the market was flat, so they
- 51:39looked inward. They systematically audited
- 51:42their sales network and permanently cut their weakest
- 51:45underperforming distributors by 6%, trimming the fat.
- 51:48By removing the dead weight, they drastically increase the
- 51:50velocity and efficiency of their remaining logistics network.
- 51:54Furthermore, they took a painful 2.45 billion RMB asset
- 51:58impairment hit this quarter. Why do that voluntarily?
- 52:02They chose to take a massive write down to clean up their
- 52:04balance sheet, writing off outdated goodwill and clearing
- 52:07about slow moving older inventory.
- 52:10By taking the pain now and streamlining operations, they
- 52:14are demonstrating absolute market dominance.
- 52:17They are systematically squeezing out revenue growth by
- 52:20simply out executing every other dairy company in a stagnant
- 52:24environment. They're just suffocating the
- 52:26weaker players. Yeah, OK, but what about Luzu
- 52:30Lao Zhao? Because if you look at their
- 52:32numbers for a premium luxury liquor brand, it looks
- 52:36absolutely terrifying. The revenue crashed down 36% and
- 52:41their net profit fell right alongside at down 43%.
- 52:44By any normal metric, that is a catastrophic quarter.
- 52:47It lists disastrous yes, until you look past the revenue drop
- 52:50and focus entirely on their gross margin.
- 52:53Despite their top line revenue literally cratering by more than
- 52:561/3, their gross profit margins remain at an astronomical
- 52:59gravity defying 90.7%. Wait, a 90.7% margin?
- 53:03They're basically bottling and selling and liquid gold.
- 53:07Essentially, yes, the cost of the liquid and the glass is
- 53:09negligible compared to the retail price.
- 53:11But what that 90.7% margin tells us is something profound about
- 53:15corporate strategy, which is it means that Lucy Lauerjao's
- 53:18management looked at the weak consumer demand and they
- 53:20actively, deliberately chose to sacrifice massive amounts of
- 53:24sales volume in order to fiercely defend their premium
- 53:28price point. But why take a 43% hit to
- 53:31profit? Why wouldn't you just authorize
- 53:33like a 10 or 15% discount? Wouldn't dropping the price just
- 53:37a little bit stimulate demand, keep the sales flowing and
- 53:41smooth out the quarterly earnings?
- 53:42Because in the hyper competitive world of premium luxury liquor,
- 53:46the price is the brand. The high price tag is the
- 53:49primary feature that signals exclusivity and status to the
- 53:52consumer. If you panic during an economic
- 53:55dip and drop the price of your luxury good, you send a
- 53:58devastating signal to your core demographic that your product is
- 54:01no longer rare, no longer exclusive, and no longer
- 54:05valuable as a status symbol. Wow, yeah, I guess you can't
- 54:09undo that. Once you discount luxury, you
- 54:12destroy the brand equity and you almost never get it back.
- 54:15Consumers will forever anchor to that lower price.
- 54:18Luzzo Lorjo's executive team decided they would much rather
- 54:21sell significantly fewer bottles but maintain their massive 90%
- 54:26margin, then offer a discount and permanently ruin a luxury
- 54:30brand image that took them decades and billions of RMB to
- 54:33build. That's a huge gamble.
- 54:35It is an incredibly painful short term financial defense
- 54:39mechanism deployed to protect their long term existential
- 54:43survival. That makes total ruthless sense.
- 54:45You cannot unring the discount bell.
- 54:47Once the consumer views you as cheap, cheap, you are cheap
- 54:49forever. Exactly.
- 54:51OK, so Yeely and Luzzu Lodge out, dug trenches and defended
- 54:54their domestic turf. Let's look at the companies that
- 54:57realize the domestic market was a ceiling and chose the entirely
- 55:00opposite path. Aggressive global expansion.
- 55:03I'm looking at the white goods home appliance giants Medea Hair
- 55:07and Gray. The financial reports from the
- 55:09appliance sector clearly, unequivocally show that the
- 55:12ceiling for domestic volume growth in China has been hit.
- 55:14The equation is simple. The real estate market is
- 55:18slowing down, meaning fewer new apartments are being built,
- 55:21which directly translates to fewer consumers needing to buy
- 55:24new washing machines, refrigerators and central air
- 55:27conditioner. It's a direct pipeline.
- 55:29Exactly so. Faced with that reality, the
- 55:31future survival for these giants requires succeeding in two
- 55:35distinct strategic arenas. First, outbound strategies,
- 55:39which means aggressively building factories and capturing
- 55:42sales volume in overseas markets.
- 55:44And 2nd AI smart homes, which means creating an interconnected
- 55:49proprietary digital ecosystem of appliances to justify charging
- 55:53premium prices to the domestic customers they already have.
- 55:57And looking at the revenue breakdown, Madea seems to be
- 56:00absolutely dominating that outbound global race.
- 56:02Undeniably, Madea was the only one of the big three legacy
- 56:06manufacturers to actually grow their overall corporate revenue
- 56:08this quarter, up 3.55% to 260 billion RMB.
- 56:14And the sole driver of that growth was their overseas
- 56:17expansion. Their international revenue hit
- 56:19113 billion RMB. That means 43% of their total
- 56:25global revenue is now generated entirely outside of China.
- 56:29They're no longer just a Chinese manufacturer exporting goods.
- 56:33They are rapidly transforming into a truly global,
- 56:36decentralized multinational corporation.
- 56:39Precisely. They are building localized
- 56:41supply chains in target countries now contrast that with
- 56:44GREE. GREE actually posted the highest
- 56:47net profit margin of the three, largely due to their absolute
- 56:50historical dominance in the highly profitable air
- 56:52conditioning sector. But their overseas footprint is
- 56:55significantly smaller than Medea's.
- 56:57Because they are so tethered to the domestic market, their top
- 57:00line revenue growth potential is severely limited moving forward.
- 57:04They are highly profitable today, but media is building the
- 57:07volume pipeline for tomorrow. OK, let's pivot from washing
- 57:10machines to a very different kind of consumer product, IP,
- 57:14entertainment and culture. We are looking at Pop Mart and
- 57:17China literature, and both of these reports prove that in the
- 57:21modern economy, intellectual property is the new gold it.
- 57:25Truly is. Let's start with Pop Mart.
- 57:27They are famous for their blind box designer toys, the vinyl
- 57:30collectibles. Their domestic revenue was up a
- 57:33very impressive 47%, but surprisingly, their overseas
- 57:37revenue actually dipped by 11%, right.
- 57:40If, as we just discussed with Media Outbound is the required
- 57:44future for Chinese brands, why is Pop Mart seemingly failing to
- 57:47capture international consumers? I would strongly argue it is not
- 57:51a failure. It is a highly deliberate,
- 57:53highly calculated strategic pivot.
- 57:56Popmart is actively in the process of completely
- 57:58restructuring their overseas retail model.
- 58:02How so? Previously their initial
- 58:04international push relied heavily on low cost, small
- 58:07footprint outlet style stores, kiosks or third party
- 58:10distributors. It was a scattergut approach to
- 58:12test the waters. But they realized that to build
- 58:15a true lasting global lifestyle brand, they needed prestige.
- 58:20So they are intentionally pulling back from those low tier
- 58:23channels. They accepted a short term hit
- 58:25to their overseas revenue so they could reallocate that
- 58:28capital to launch massive, highly experiential premium
- 58:32flagship stores and some of the most expensive prime real estate
- 58:36in the world, like their massive new locations in New York City.
- 58:39I see, so they are essentially executing the exact same
- 58:43strategy as Lizu Laujiao, but instead of using price to
- 58:46protect their brand, they're using premium retail real
- 58:49estate. Exactly.
- 58:50They want to be seen next to global fashion brands, not in
- 58:53the discount kiosk. They're protecting the premium
- 58:56feel of the IP. Spot on.
- 58:58And what about China literature? Their trajectory is fascinating.
- 59:01Yeah, they essentially started as a massive reading app for
- 59:03serialized user generated web novels.
- 59:06Right. Yeah.
- 59:06It was their genesis, yes, But their latest mid year report
- 59:09shows a company that is rapidly and successfully realizing that
- 59:13the written text itself is no longer their final or most
- 59:16valuable product. If you look at the breakdown,
- 59:19their core reading revenue, the money they make from users
- 59:23paying to read chapters on their app is actually dropping.
- 59:26Why is that? And the reasons are obvious that
- 59:28there is a flood of free reading alternatives online.
- 59:31And more pressingly, the explosion of short form video
- 59:35dramas on platforms like Do In and TikTok is aggressively
- 59:39cannibalizing their users free time.
- 59:41So their core historical business model is slowly dying.
- 59:46How are they surviving? Because while the Reading
- 59:49business is slowly shrinking, their IP commercial
- 59:51actualization business is absolutely exploding.
- 59:54Their revenue derived from IP licensing, copyrights and
- 59:58adaptations jumped by 40% year over year.
- 1:00:01That segment now makes up nearly half their total corporate
- 1:00:04revenue. That's a massive shift.
- 1:00:06They recognize the shift in consumer behavior and adapted.
- 1:00:09They are no longer just a digital reading app.
- 1:00:11They have transformed themselves into a massive, highly efficient
- 1:00:14intellectual property pipeline. They use the reading app
- 1:00:17essentially as a massive crowd sourced testing ground.
- 1:00:21They take the most popular data proven web novels and they
- 1:00:24package and license those stories to be turned into highly
- 1:00:28profitable viral short video dramas, long form television
- 1:00:32shows, anime series and physical merchandise.
- 1:00:36That is a brilliant evolution. They essentially outsource the
- 1:00:39risk of writing and testing ideas to millions of amateur
- 1:00:42authors. When a story goes viral for free
- 1:00:45on the app, they swoop in, take the IP and make millions selling
- 1:00:49the television and merchandising rights.
- 1:00:51They own the foundational source code for the next generation of
- 1:00:55Chinese entertainment. It's an incredibly powerful
- 1:00:57model. Right as we near the bottom of
- 1:00:59this massive stack of reports, let's look at Section 7.
- 1:01:02I call these the slow burners. These are industries that
- 1:01:06require massive eye watering, upfront capital investments,
- 1:01:09years of complex infrastructure building, and demand absolute
- 1:01:14unwavering patience from their investors before they ever see a
- 1:01:16payout. Definitely.
- 1:01:18Let's start with SF Express, the dominant logistics giant.
- 1:01:21SF Express reported a slight dip in their overall net profit, but
- 1:01:25as analysts we have to immediately adjust for context.
- 1:01:28That dip was largely A statistical illusion caused by a
- 1:01:31one off 600 million RMB asset sale they recorded in the same
- 1:01:36period last year, which skewed the year over year comparison.
- 1:01:39Yeah, OK. Good catch.
- 1:01:41The real, highly consequential story hidden deep within their
- 1:01:44segment reporting is their international and supply chain
- 1:01:47revenue division, which surged by a massive 46%.
- 1:01:51By 46% growth in international shipping in just six months.
- 1:01:56What is driving that? Is that just a reflection of the
- 1:01:58massive volume of cheap small packages being shipped globally
- 1:02:03by e-commerce giants like Timu and Shine?
- 1:02:05Indirectly, yes, it is tied to that macro trend.
- 1:02:08But the specific mechanism driving SF Express's revenue
- 1:02:12surge is a fundamental shift in how those e-commerce giants
- 1:02:15operate. OK, tell me more.
- 1:02:16Initially, the cross-border e-commerce model relied on
- 1:02:19shipping millions of individual small parcels directly from
- 1:02:23factories in China across the ocean to consumers in the US or
- 1:02:26Europe. But that model is slow, highly
- 1:02:30inefficient and increasingly targeted by punitive
- 1:02:33international tariffs and customs taxes.
- 1:02:35Right. It's getting harder to do that.
- 1:02:37Exactly. The Chinese ecommerce giants
- 1:02:39recognize this vulnerability and are actively shifting their
- 1:02:43logistics strategy. Instead of small parcels, they
- 1:02:46are moving to a localized fulfillment model.
- 1:02:48Warehouses. Yep, they ship massive bulk
- 1:02:51containers via ocean freight to giant overseas warehouses
- 1:02:56located within the destination countries and then execute the
- 1:02:59final mile delivery locally. Oh wow.
- 1:03:01SF Express management saw this strategic shift coming years
- 1:03:04ago. They deployed massive amounts of
- 1:03:06capital and painstakingly build over 2.75 million square meters
- 1:03:11of overseas warehouse space to capture this exact evolution in
- 1:03:15the global supply chain. That is the definition of a slow
- 1:03:18burn. They spent years and billions of
- 1:03:20RMB building the physical infrastructure in foreign
- 1:03:23countries before their core e-commerce clients even fully
- 1:03:27realized they needed to change their business model.
- 1:03:30But now that the shift is happening, SF Express is
- 1:03:33standing there with the exact solution ready to capture the
- 1:03:35margin. Foresight pays off.
- 1:03:38And finally, to close out our deep dive, let's look at the
- 1:03:41ultimate slow burn industry, pharmaceuticals, specifically
- 1:03:45the queso. Biotech and pharmaceuticals are
- 1:03:47the purest definition of a slow burn.
- 1:03:50The capital required is immense, the timeline stretch over
- 1:03:53decades, and the binary risk of failure is ever present.
- 1:03:57A quesos mid year revenue actually slightly missed
- 1:03:59consensus market expectations. But if you are analyzing A queso
- 1:04:03based purely on current quarterly revenue, you are
- 1:04:06entirely missing the point of the company.
- 1:04:08You have to understand the sheer audacious scale of what they are
- 1:04:12attempting to achieve on a global level.
- 1:04:14They are attempting to execute the hardest, most complex
- 1:04:17maneuver in the entire biotech sector.
- 1:04:20They are taking a novel Chinese developed oncology drug known as
- 1:04:24AK112 and they are taking it global.
- 1:04:26That's ambitious. Very They are currently running
- 1:04:29massive, incredibly expensive phase three clinical trials
- 1:04:33globally intending to go head to head in the market against
- 1:04:37Merck's Keytruda, which is arguably one of the most
- 1:04:39successful, best selling cancer drugs in the history of
- 1:04:42medicine. It's a literal David versus
- 1:04:44Goliath scenario played out in FDA trial data.
- 1:04:47It is exactly that, and because of the massive stakes, the
- 1:04:50global financial markets are currently pricing a case of
- 1:04:52stock not on their current relatively small commercial
- 1:04:55revenue, but entirely on the future statistical probability
- 1:04:58of success for those specific clinical trial data readouts.
- 1:05:01Oh man, that is stressful. It requires a level of extreme
- 1:05:05iron willed investor patients that most sectors simply do not
- 1:05:09tolerate because the reality of biotech is binary.
- 1:05:12If the phase three data is superior, if they prove they can
- 1:05:15beat Keytruda, the financial upside is almost astronomically
- 1:05:19high. They own the market.
- 1:05:21They will capture a massive slice of a multibillion dollar
- 1:05:24global oncology market, but if the data fails to meet
- 1:05:28endpoints, the floor drops out entirely and billions in R&D are
- 1:05:33essentially wiped out overnight. Wow.
- 1:05:35OK, let's take a breath. We've covered a truly staggering
- 1:05:38amount of ground today. Looking back over this massive
- 1:05:41stack of 2026 reports, we've seen an incredible, almost
- 1:05:45dizzying diversity of corporate survival and growth strategies.
- 1:05:48We really we saw Amazon willingly burning 10s of
- 1:05:51billions in free cash flow just to build a massive AI hardware
- 1:05:55mode that they hope no competitor can ever cross.
- 1:05:57We saw BYD completely ignoring a massive drop in overall sales
- 1:06:01volume, choosing instead of focus ruthlessly on expanding
- 1:06:04their per car profitability through vertical integration.
- 1:06:07We watched Vanke fighting for its very corporate life,
- 1:06:10navigating A collapsing real estate paradigm through strict,
- 1:06:14painful cash flow management and asset sales.
- 1:06:18And we saw Luzu Lauchiao stubbornly refusing to drop
- 1:06:21their prices by a single RMB, accepting a massive revenue hit
- 1:06:26just to protect the untouchable soul and equity of their luxury
- 1:06:29brand. It is a fascinating cross
- 1:06:32section of global commerce, and what it ultimately proves to
- 1:06:35investors is that in a highly complex, rapidly shifting
- 1:06:38macroeconomic environment, there is absolutely no single dogmatic
- 1:06:43right way to run a business. The playbook changes based on
- 1:06:47your position in the cycle. But as I review all these
- 1:06:49different strategies, there is one shared, almost hidden
- 1:06:52vulnerability that runs like a fault line through almost all of
- 1:06:55these multinational reports. I thought you might have caught
- 1:06:57a common thread. What's the hidden risk?
- 1:06:59If we step back and connect us to the broader geopolitical big
- 1:07:02picture, I want you to think about this dynamic.
- 1:07:04We spend so much of our time and we spent most of this deep dive
- 1:07:08hyper analyzing management decisions.
- 1:07:09We dissect revenue growth, we debate the merits of new product
- 1:07:12launches, we track AI technological breakthroughs, and
- 1:07:15we applaud operational efficiency.
- 1:07:18But if you look closely at the fine print and the reports from
- 1:07:21BYD, from Costco and from Popmart, in all three of those
- 1:07:24cases, a massive, highly material chunk of their hard
- 1:07:28earned operational profitability was simply wiped out by foreign
- 1:07:31exchange fluctuations. So I leave you with this final
- 1:07:35thought to Mull over in a hyper connected, highly globalized
- 1:07:392026 where capital and goods cross borders at lightspeed is
- 1:07:44the biggest, most uncontrollable risk to a multinational
- 1:07:47company's fundamental success. No longer the brilliance of its
- 1:07:51competitors, but rather the invisible, unpredictable hand of
- 1:07:54the global currency markets. That is a deeply unsettling yet
- 1:07:58vital thought to end on. You can spend a decade
- 1:08:01engineering the most efficient supply chain on Earth, You can
- 1:08:03build the absolute best electric vehicle in the world, and a
- 1:08:06central bank adjusting an interest rate on the complete
- 1:08:08opposite side of the planet can wipe out your entire quarterly
- 1:08:12profit margin before you've even finished your morning coffee.
- 1:08:15Thank you for joining us on this deep dive.
- 1:08:17Keep questioning the headlines, keep looking for the hidden
- 1:08:20operational mechanics driving the spreadsheets, and we will
- 1:08:23see you next time.