paceXThe scariest thing this time isn't that the valuation is nearing $2 trillion, but that the market has started throwing the earnings report into the trash. Yep, you heard that right—completely stopped looking at it. A company with revenue still under $20 billion and losses not fully erased has a valuation of$1.75trillion. Honestly, that's outrageous. The rockets, Starlink,IPO, and semiconductor rebound you see today are only the surface. The real main theme is that Wall Street is using space,AI, chip manufacturing, and industrialAIto reprice the distribution of wealth over the next decade.
Today we're looking back at the big green candle on June 11. To be honest, this market really gave the bears a beating today. Yesterday it looked like the end of the world—tech sold off, semis sold off, AI names sold off—and then today, boom, a giant green candle just blasted back. The bears had just raised the knife, and the bulls slapped them right back.
But don't get excited yet. Really, don't. The most annoying thing about this kind of market isn't that it rises or falls; it's that the script changes every day. Yesterday it was Middle East risk, oil pressure, inflation pressure, Fed pressure. Today Trump says he will cancel strikes on Iran, and the market instantly switches from panic mode to champagne mode. This isn't an investment market—it's basically a giant emotional rave.
Today the Dow, S&P, and Nasdaq all rebounded, and semiconductors were even crazier: PHLX Semiconductor Index rose nearly8% in one day. But a rebound doesn't mean safety, and a bounce doesn't mean reversal. A lot of people see one big green candle and yell that the bull market is back. Let me tell you: people like that lost money last year, and they'll keep losing money this year.
Because mature trading is about whether the move has follow-through. Bulls can't just rely on one big green candle. They need a second, a third, and at least a pullback that doesn't immediately get broken. Bears are the same. You can't short based on one red candle and say the crash has started. Right now this market is the classic high-volatility range: it looks like a bull market when it rises, a bear market when it falls, and it slaps your face back and forth.
Let's start with today's hottest theme, SpaceX. This thing has changed flavor and become the market's sentiment tester.SpaceX IPO is priced at$135 per share, raising about$75 billion, for a valuation of roughly$1.75trillion. It's already standing at the door of the super-megacaps.
The question is: this company made about$18.7 billion in revenue in 2025 and was still loss-making. So where does this valuation come from? It comes from the future story. Starlink is indeed strong: satellite internet is no longer a PowerPoint story; it has users, revenue, and room for cash flow improvement. Rocket reusability is also real hard tech.
But, well, here's the key point. Can Starlink and the rocket business support a valuation of hundreds of billions? That's debatable. But what supports the rest of that trillion-dollar premium? Space data centers, a million-satellite network, future space compute, lunar manufacturing—stories that sound great now but don't yet have a real commercial loop.
Wall Street loves this game. In a bull market, dreams get priced in; in a bear market, reality gets discounted. In a bull market, one line about a trillion-dollar future market makes retail investors' eyes light up. In a bear market, if cash flow is short by even a dime, the stock can be cut in half. If the dream doesn't land, it eventually becomes scrap; if valuation isn't realized, the bagholders go to the moon.
Some bears are directly attacking the stock, saying SpaceX's valuation is already badly detached from fundamentals. That sounds harsh, but it's not entirely unreasonable. On a revenue multiple basis, SpaceX is already so expensive that it would need many consecutive years of perfect execution to justify it. Any growth company with an ultra-high valuation can't just be "pretty good"—it has to hit every step right, with explosive growth, real profits, improving cash flow, and a story that never breaks.
And the concept of space data centers does have huge imagination. Solar power, space cooling, bypassing ground-based power bottlenecks—sounds sexy. But once you actually run the numbers, it's not so romantic. Launch costs, maintenance costs, insurance costs, orbital safety, hardware failures, data-transmission latency, and regulatory issues—every single one is money. If a terrestrial data center breaks, you can send an engineer. If space hardware breaks, what then?
So yes, SpaceX is great, and Musk has really pulled off things others thought impossible. But investing and worshipping are two different things. Liking the person and deciding what price to buy the stock at are worlds apart. Retail investors must distinguish between a great company and a price worth paying.
Next, let's talk about Musk'sTerafab. This is tied directly to SpaceX's valuation. Recently Musk is supposed to talk about Terafab at an ASML event—his super chip-manufacturing plan. What is ASML? Simply put, it's the core equipment company that can't be avoided in advanced chip manufacturing, especially EUV lithography machines. Without it, advanced process chips are basically not fun to play with.
Musk's logic is very clear. In the future Tesla will need chips, Optimus robots will need chips, and if SpaceX really wants to do space compute, it will need massive amounts of chips. The problem is that if all the core chips have to be supplied by others, your schedule gets controlled by someone else. So he wants to keep part of that key capability in his own hands. In plain terms: he doesn't want to queue, doesn't want to be choked, and doesn't want to keep reading the supply chain's face.
That's a fierce idea. Really fierce. But fierce doesn't mean easy. Chip manufacturing is not buying a few machines, building a factory, and shouting that human civilization's future is done. There are equipment lead times, yield ramp-up, engineering teams, supply chains, water and electricity, chemicals, and manufacturing know-how. TSMC wasn't forged in a day, and Intel won't turn around overnight just by throwing money at it. If even the old hands can stumble in advanced nodes, why would a newcomer glide through effortlessly?
So I'll treat the Terafab line as a long-term narrative. It gives ASML, semiconductor equipment, advanced packaging, power, clean rooms, and materials plenty of room for imagination. But remember: imagination doesn't equal immediate realization. The market loves to price a ten-year story into today, and then ask tomorrow why nobody's made money yet. Wall Street really can do anything.
Then there's Bezos. I need to correct a point that is easy to misread. Many headlines make it sound like Bezos directly poured$41 billion into AI. To be precise, Prometheus is now valued at around$41 billion, and Bezos personally is not putting in that much cash. That's a big difference. Clickbait loves to write valuation as if it's the actual investment amount, making it look like the billionaire hauled a safe onto the table.
But the thing itself still matters. Prometheus is heading in a very different direction: industrial AI, physical-world AI, with the goal of redesigning the design and manufacturing process for complex products like jet engines, medical devices, and consumer electronics. In other words, AI is not just writing articles or drawing images; it's entering the real world, the factory floor, engineering design, and product testing.
Because in the past, AI mainly changed the flow of information: text, images, code, search, customer service. In the next stage, if AI is going to keep expanding, it must enter the physical world. Manufacturing, robotics, aerospace, medicine, energy—those are the places that are truly slow, expensive, and complex. Whoever can break through those efficiency barriers is not just changing tools anymore.
Bezos's line is a bit like Musk's. Musk is moving from space, cars, and robots toward chips and AI compute. Bezos is moving from e-commerce, cloud computing, logistics, and Blue Origin toward industrial AI and physical manufacturing. The top billionaires are now competing for the next industrial foundation. Compute, energy, manufacturing, space, robots—put those words together, and you get the main theme of the next decade.
Okay, back to the market. Today's green candle was beautiful, but I'll say it again: don't let one green candle change your worldview. The Nasdaq surged, semis bounced violently, and many stocks look like they've found a floor. But what matters in trading? It's whether you get follow-through. If the Nasdaq holds above today's breakout zone on the next pullback and keeps moving up, then the bulls really do regain control. But if tomorrow brings another big red candle, today's green candle may just be short covering. Both explanations are possible, so don't play prophet. Until the market confirms, don't treat guesses as conclusions.
The Dow is the same. Yesterday it looked like it broke down; today it bounced back. That's the classic feature of a range-bound market. Yesterday made the bears excited, today made the bulls excited, and the ones really making money are the ones not moving around randomly.
Semiconductors were the strongest today. They fell yesterday, then rebounded just as hard today. SOX jumped nearly8%. The problem is that semis are now very crowded. The AI story is still there, orders are still there, but valuations are high and positioning is heavy. When a street is full of people, you can't run when there's a fire, and you also rise fast when money comes in. That's the double-edged nature of crowded trades.
Equipment stocks like AMAT and ASML were strong today. As long as AI capex doesn't stop, the equipment chain has support. But equipment stocks fear delayed orders most. If customers cut CapEx, stock prices get hit immediately, and then the market starts asking about returns, which gets even more uncomfortable. Right now the biggest issue in the AI chain is that demand exists, but everyone is starting to wonder when all this burning of money will actually turn into profits.
Micron and SanDisk in the memory chain are also worth watching. HBM and AI server demand have tightened the supply-demand balance again. As long as NVIDIA's next platform keeps moving forward, the memory chain still has a story. But don't chase memory stocks blindly either—they rise fast and can get crushed just as fast. If you want to trade them, wait for a right-side confirmation; a pullback to a key support that doesn't break and then turns up is a much better entry.
NVIDIA rebounded today, and that's key. It's still the overall leader of the AI chain. As long as NVIDIA doesn't break, the faith in AI is still there. But the market expects too much from it—you have to keep beating expectations, keep proving Blackwell, Rubin, HBM supply, and customer CapEx are all fine. If even one thing isn't perfect, the market will use valuation against it.
Oracle today is a good example. The earnings didn't look bad, and the cloud business still has an AI story, but the stock still fell. Because the market has already shifted from revenue growth to asking about capital expenditures and free cash flow. In the past people just asked whether you had AI. Now they ask whether AI makes you money. Microsoft's relative weakness today shows the same thing: the AI main theme is still there, but the inside is starting to diverge. Those with strong cash flow, reasonable valuations, and clear CapEx returns will be kept; those who only tell a vision story without a clear math equation will eventually get hit.
So I want you to remember one core judgment from today's episode: U.S. stocks still have opportunity, and AI is far from over. But from here on, the market will get more and more selective. In the past, as long as you touched AI, you rose. Now you must prove that AI can turn into revenue, revenue into profit, and profit into CapEx coverage. That process will be brutal. Only a few will pass.
One common mistake is seeing a big green candle and thinking you missed it, so you have to chase immediately. Your account turns red, your brain heats up, and it feels like if you don't get on the train today, Wall Street won't play with you tomorrow. Honestly, the market never lacks opportunities; what it lacks is your patience to wait for a comfortable entry.
The second mistake is equating a good company with a good stock. SpaceX is good, NVIDIA is good, ASML is good—of course they are. But if you buy a good company too expensively, you'll still feel pain for a very long time. You're not buying a company business card; you're buying future return. If the price is wrong, even belief can't save you.
The third mistake, the most deadly, is treating other people's dreams as your own principal. Wall Street talks about space, AI, robots, and trillion-dollar markets, and you get excited and go all in. In the end, they're selling the story and you're paying real cash. The money you saved with hard work ends up becoming liquidity for someone else's exit—that's the truly painful part.
The answer is simple. Don't guess tops, don't guess bottoms, and don't go all-in on a single candle. Split the stocks you want to buy into three categories: core leaders, supply-chain beneficiaries, and pure concept speculation. Wait for pullbacks on core leaders, look at earnings for supply-chain names, and use only small positions for pure concepts. If you don't understand it, don't touch it. Chasing the hotspot is worse than waiting for a pullback; buying dreams is worse than checking the books.
Finally, here's the key summary. Today's U.S. stock rebound shows the bulls are not dead, and the AI main theme is not over either. Semis, memory, equipment, and AI server chains remain the most elastic parts of the market.SpaceX IPO will continue to drive the narratives of the space economy, satellite internet, chip manufacturing, and AI compute.Bezos's Prometheus reminds us that the next stage of AI is not just software; it is moving into manufacturing and the physical world.
But the risks are clear too. First, SpaceX's valuation is too high, and post-IPO volatility will definitely be large. Second, AI capex has entered a period of market scrutiny; whoever burns money without a clear return will be punished. Third, the Middle East and oil prices are still affecting inflation expectations, and the Fed won't suddenly turn dovish because of one day's rally. Fourth, semiconductor positioning is too crowded—what rises fast can also fall fast.
So my strategy is simple. If you already hold core AI leaders, don't get euphoric after one big up day, and don't panic after one big down day. Watch the trend, the quality of pullbacks, the volume, and whether there's a second green candle. If you're entering fresh, don't chase the most heated move. Wait for divergence, wait for a retest, and wait for the market to sort out the companies with real earnings.
If you think today's episode made SpaceX, Terafab, Prometheus, semiconductors, and the broader market logic clear, please give A Quan Talks Finance a like and save it. Really, I'm not being polite—under this kind of market, understanding it once today may save you from chasing a high next time, which could save you a lot of wasted money. And feel free to tell me in the comments whether you think SpaceX's valuation is the start of a new tech revolution or an epic bagholder event. See you in the comments.
