Speaker A
Every financial channel on this platform is telling you the same story right now. The market had a rough week, but nothing to panic about. Dow down 0.7%.
Prof. Jiang Xueqin reveals why the current AI-driven market risks a structural collapse worse than a recession.
Key Takeaways
- AI-driven market growth is artificially inflated by closed-loop financing among major tech firms.
- The current downturn is structurally worse than a normal recession due to this circularity trap.
- Cheaper Chinese AI alternatives put downward pressure on US AI companies' pricing and growth.
- Gold and mining stocks are emerging as safer hedges compared to Bitcoin in this uncertain environment.
- Investors need to monitor their portfolios for exposure to these vulnerable tech companies.
What the video covers
- The apparent stability of major US tech indices masks a deeper structural problem in the AI investment ecosystem.
- Five major companies are caught in a circular financing loop, inflating growth metrics without real external revenue.
- This circularity trap resembles the dotcom bubble's vendor financing dynamics that led to a collapse.
- Chipmakers and hardware suppliers are signaling market concerns as hyperscaler capital expenditures may decline.
- Chinese AI labs are releasing cost-effective models that threaten US AI companies' pricing power and revenue growth.
- Half a trillion dollars was spent on AI infrastructure last year, but cheaper alternatives challenge its sustainability.
- Institutional capital is quietly shifting towards gold and mining stocks as a hedge, while Bitcoin behaves as a risk asset.
- The market drop is modest but hides underlying vulnerabilities in the tech sector's financial structure.
- Prof. Jiang predicts a structural downturn with gold and mining equities outperforming Bitcoin over the next 1-3 years.
- Investors should be cautious of concentrated exposure to companies involved in the circular financing loop.
Chapters
- 00:00Market Overview and Initial Warning
- 00:54Thesis Introduction: The Circularity Trap
- 01:57Tech Earnings and Market Reaction
- 02:58Explaining the Circularity Trap
- 03:56Historical Parallel: Dotcom Bubble
- 04:56Chipmakers and Market Signals
- 05:56Chinese AI Competition and Pricing Pressure
- 06:56Capital Movements: Gold vs Bitcoin
- 08:51Counterarguments and Debate
- 09:55Predictions and Investment Advice
Full Transcript — Download SRT & Markdown
Speaker A
S&P down 1.3%. NASDAQ down 3.1%. Still close to all-time highs. Move along. Nothing to see here. That reading is not just shallow. It is dangerously wrong because underneath those calm headline numbers, something is breaking. Not in the index,
Speaker A
in the machinery that has been holding the index up. And I want you to hold on to one number while I walk you through this. Half a trillion dollars. That is roughly what five companies spent last year building something that the market
Speaker A
for the first time is starting to openly question. By the end of this video, you will understand exactly why that number is about to become a much bigger problem and why what is coming is not a normal recession. It is something structurally
Speaker A
worse. Here is my thesis. I'm going to show you three things today. First, that the companies driving American economic growth right now are financing each other in a closed loop that has almost no outside validation. And closed loops
Speaker A
always break the same way. Second, that the six-month lead everyone keeps repeating about American AI dominance is a comforting story that does not survive contact with how businesses actually buy software. Third, that the smartest money in the market already started moving out
Speaker A
the back door quietly last week while everyone was watching the front door. On this channel, we do not do panic and we do not do cheerleading. We follow the logic of power. We follow the incentives. And when you strip away the
Speaker A
emotion and the headlines, the picture always becomes clear. Let me start with the facts because you need the ground beneath your feet before I take you into the analysis. Alphabet reported earnings on Thursday. The headline was a capital
Speaker A
expenditure number that came in above what investors had priced in. Alphabet stock fell 10% in a single week. Oracle fell 7.9% on the week and is now down 41% for the year. Meta fell 7.3%. Amazon fell 6.8%.
Speaker A
Microsoft fell 2.7% on the week and is now down 19.3% year-to-date, edging toward bear market territory. Two Musk Link companies took even harder hits.
Speaker A
SpaceX fell 7.7% on the week, closing 15% below its own IPO price and 49% below its post-listing peak. Tesla dropped 18% in a week and is now 35% below its all-time high. Between those two stocks alone, Elon Musk's paper
Speaker A
losses ran close to 100 billion in 7 days. The surface reading is simple. Big tech had a bad earnings week. But that explanation is too small for what is actually going on. To understand why, you first have to understand how these
Speaker A
companies are really paying for what they are building. I want to give this a name because once you see it, you cannot unsee it. I call the circularity trap.
Speaker A
Here is the simple version. Imagine five friends who all agree to write each other checks. Friend A pays friend B for services. Friend B uses that money to buy equipment from friend C. Friend C uses the profit to invest in friend A's
Speaker A
company. On paper, every single friend shows revenue growth. Every friend can point to a healthy balance sheet, but no new money ever actually entered the group. It is the same $5 moving in a circle getting counted as growth every
Speaker A
time it changes hands. That is what is happening across the AI buildout right now. Hyperscalers are pouring capital into chip makers and infrastructure vendors. Those vendors extend financing back to the hyperscalers' customers, take equity stakes, or sign multi-billion
Speaker A
dollar compute deals with each other. Nvidia's results look extraordinary, partly because the companies buying its chips are themselves funded directly or indirectly by capital that traces back to the same small circle of players.
Speaker A
This is not fraud. It is not illegal. It is simply a system where growth is being measured inside a loop instead of against the outside world, against actual paying customers, actual profit, actual demand that exists independent of the
Speaker A
companies financing each other into existence. Let that sit for a second. Now, here is what almost nobody is telling you. This is not the first time this exact pattern has happened. At the height of the dotcom buildout, telecom
Speaker A
and equipment companies extended enormous vendor financing to their own customers. So those customers could keep buying gear. Cisco Systems became one of the most valuable companies on Earth on the back of exactly this dynamic. When the customers who owed Cisco money went
Speaker A
bankrupt, Cisco's earnings collapsed with them because the growth had never been external. It had been circular the entire time. Geography is where the truth always lives. And the AI buildout has its own geography. A map of capital flowing in tight loops between a handful
Speaker A
of American firms, all betting the same bet, all financing the same customers, all exposed to the same failure point simultaneously. Layer two of the evidence is where it gets more urgent.
Speaker A
This week, chip makers that supply the hyperscalers also fell. Taiwan Semiconductor was down 1.2% on the week.
Speaker A
SK Hynix fell 4.6%. These are the companies actually manufacturing the hardware and their stock reaction tells you the market is starting to price in a scenario where hyperscaler capex does not simply plateau, it gets cut. When capex gets cut, the chip makers feel it
Speaker A
before anyone else because they are the first domino, not the last. Layer three is the connective tissue and this is the part that ties the whole thesis together. At the exact moment, American hyperscalers are being questioned on the
Speaker A
size of their AI spending, Chinese AI labs are releasing models. Moonshots Kimmy K3 is the one making the rounds right now that compete credibly with American frontier models at a fraction of the cost. This is not a coincidence
Speaker A
of timing. It is the second half of the same story. American AI companies need enormous revenue growth to justify their capex. Cheaper Chinese alternatives put a ceiling on how much American companies can charge. You cannot simultaneously spend at record levels and be forced to
Speaker A
cut your prices. Something in that equation gives. I want you to memorize these two numbers. Half a trillion dollars in AI infrastructure spending last year and a real working cheap alternative now sitting one API call away from every business customer on
Speaker A
Earth. Those two numbers cannot coexist peacefully. One of them has to move. Now, let me take you deeper because this is where it gets genuinely alarming.
Speaker A
Look at where money quietly moved last week while everyone was staring at tech stocks. Gold rose about 1%. Not dramatic, but notable because it rose even as bond yields and oil prices climbed, which normally pulls gold down.
Speaker A
Silver rose 2.4%. Gold mining stocks, which carry leverage to the metal, jumped even harder, the GDX up 5.6%, the GDXJ up 5.8%, both far outpacing the 1% move in gold itself.
Speaker A
That kind of leverage move in miners against a modest move in the metal is historically a signature of a bottom forming of larger capital quietly repositioning before the crowd notices.
Speaker A
Bitcoin meanwhile did the opposite. It stayed essentially flat, down close to 1% on the week, continuing a year in which it is down nearly 29%. It did not follow gold higher. It behaved like a risk asset, not a hedge. That is not the what, that is
Speaker A
the why. Bitcoin acting like a risk asset, while gold acts like a hedge tells you which one institutional capital actually trusts when the story gets uncertain and is not the one most retail investors think of as digital gold. I want to present the strongest
Speaker A
possible argument for the other side because I owe you the real debate, not a simplified version. The strongest counterargument goes like this. Markets fell 3%, not 30. The NASDAQ is still up 7 to 12% for the year. AI is a genuinely
Speaker A
transformative technology, not a fad. And every serious analyst, including plenty of skeptics, thinks its long-term impact will exceed the internet's. The co
Speaker A
years of aggressive investment without real distress. This is not 1999. These are not unprofitable startups burning venture capital. This is Alphabet, Microsoft, Amazon, companies with actual cash generating businesses funding an experiment they can afford to lose money on for a long time. That argument is
Speaker A
genuinely strong. And I want to be honest with you, I could be wrong about the timeline, but the steelman describes the balance sheets accurately. It does not describe the financing structure. A company can be profitable in its core
Speaker A
business and still be building a second capital intensive business on a circular foundation that eventually has to justify itself against external independent demand, not the demand of its own financing partners. Amazon's profitability in retail did not save Amazon Web Services from needing real
Speaker A
customers eventually. It just gave Amazon more runway to find them. The question was never whether these companies can survive a slowdown. The question is whether the specific spending on AI infrastructure ever produces a return that matches its cost.
Speaker A
And the steelman has no answer for that question. It only has an answer for whether the companies themselves will survive, which is a very different question. So here are my three predictions. Write these down.
Speaker A
Prediction one, shortterm. Within the next 90 days, at least one major hyperscaler will publicly walk back or moderate its AI capital expenditure guidance for the coming year. Framed as efficiency or discipline rather than retreat. Watch the language. Companies
Speaker A
never announce they overbuilt. They announce they are optimizing. Prediction two, medium-term. Over the next two to six months, a Chinese openweight or lowcost model will trigger at least one sharp single day selloff across American AI exposed stocks. Similar shape to the
Speaker A
deepseek shock earlier in this cycle. The trigger will not be a superior model. It will be a good enough model at a radically lower price because most business buyers do not need the best model. They need the cheapest model that
Speaker A
clears the bar. Prediction three, structural. Over the next 1 to three years, gold and mining equities will meaningfully outperform Bitcoin as the preferred hedge against this specific kind of uncertainty. AIdriven capital destruction combined with a Federal Reserve caught between inflation and a
Speaker A
slowing economy. Watch what institutions do, not what they say. Wall Street will keep telling retail investors to buy Bitcoin while quietly building larger positions in gold and gold miners themselves. That gap between rhetoric and action is a signal. There are three
Speaker A
things I want you watching over the coming weeks. First, watch hyperscaler earnings calls for the word efficiency replacing the word growth when executives talk about AI infrastructure.
Speaker A
That word swap is the tell before the guidance cut becomes official. Second, watch the spread between gold and Bitcoin because if that divergence widens further, it confirms that institutional capital has already made its decision about which asset actually
Speaker A
protects wealth in this environment. Third, and this is the one that affects your own household directly, watch your own exposure. If your retirement account, your 401k or your investment portfolio is concentrated in a handful of companies driving the circular
Speaker A
financing structure, you are carrying risk you may not have consciously chosen. And if you want the 47 specific strategies for hedging your wealth, cutting your inflation exposure, and reading the five signals that tell you the truth before any headline does, I
Speaker A
put everything into the crisis preparation blueprint. is the personal finance chapter that the geopolitical and market analysis never includes. The link is in the description and the pin comment. Every empire, every boom, every technological revolution in history has
Speaker A
believed in the moment that this time the old rules of financing and return did not apply. The railroads believed it. The dot companies believed it. And today, some of the most capable executives on the planet believe it again. Not because they are foolish, but
Speaker A
because inside a closed financing loop, the numbers genuinely do look extraordinary. Right up until the moment, external reality is required to validate them. The most dangerous kind of downturn is not the one that arrives with a crash everyone can see on a
Speaker A
single red day. It is the one that arrives quietly, company by company, guidance, cut by guidance cut. While the headline index stays close enough to its all-time high that nobody sounds the alarm until it is already well underway.
Speaker A
This story is going to move fast and I will be tracking every guidance call, every model release out of China and every move in gold relative to Bitcoin.
Speaker A
Tell me in the comments which of these three predictions you think is most likely to happen first. I read your responses and your questions shape the next analysis. I am Professor Jang Shu Chin. Thank you for watching.
Topics:AI marketcircular financing looptech stocksrecession riskchipmakershyperscalersChinese AIgold hedgeBitcoin riskmarket analysis
![[YTP] Jordan Peterson: Rat King — Transcript](https://i.ytimg.com/vi/0qUoznGIe1M/maxresdefault.jpg)
![Nick & Charlie ‣ their story [+heartstopper forever] — Transcript](https://i.ytimg.com/vi/p4TiW1TZM3w/maxresdefault.jpg)
