저자: 矽基先知SiliconOracle · 외국인
요약 보기
AI 반도체 랠리가 집단 환상이라며 경고. 마이클 버리가 반도체 지수 풋옵션으로 8개월 내 40% 하락에 베팅했다고 소개. 마이크론은 HBM 점유율 10% 미만에 경기순환주이며 800~1000달러 분할 익절, 500달러 이하 매수 대기를 제시. AMD는 PER 143으로 440달러 이상 추격매수 금지, 350달러 손절선 제시. 코닝은 PER 100 이상으로 PER 35 이하 전까지 추가매수 자제. 인텔은 10년 연속 실패로 흑자 전환 전까지 매수 금지.
다듬은 글
You are currently participating in the largest collective illusion in human history. In the past three months, Micron has risen by 40%, AMD by 80%, and the market value of the entire semiconductor sector has increased by more than $1 trillion. Every day when you open your financial software, all you see are red candlestick charts, and all you hear is that the AI revolution has just begun and HBM is the next oil.
If you miss this , you'll miss an era. Your friends on social media are showing off their doubled returns, your colleagues are discussing which chip stocks will rise next at lunch , and even your relative who knows nothing about stocks is asking if you should buy some Nvidia. Then you start to feel anxious, you start to regret not getting on board sooner, and you start to convince yourself that maybe this time it's really different.
You open your trading software, your finger hovers over the buy button, and your heart races. In this video , I'm here to pour cold water on you. After watching this video, you won't instantly become a billionaire , but at least you won't be the last one to pay the bill in this frenzy. You will see the true valuation of four hot chip stocks that have been hyped up to the sky , and you will know when to stop.
When should you observe and wait? The most important thing is understanding why the most frenzied market moments are often the most dangerous. Tell me now: are you already making money in AI chip stocks and hesitating about whether to take profits , or are you watching others profit from the sidelines, tormented by market sentiment and unable to sleep, ready to jump in at any moment ?
Type your answer in the comments. While everyone is immersed in the dream of getting rich quick with AI chips, Michael Barry, the man who made over $2 billion shorting the entire subprime mortgage market in 2008, wrote a sentence on his Substack: "The current market is so outrageous it's like jumping on sharks." In layman's terms, this drama has gone too far ; even the most basic logic is ignored.
More importantly, he has already staked real money, buying a large number of put options on the semiconductor index, betting that the entire semiconductor sector will plummet by 40% in the next eight months. What does 40% mean? It means a sector with a current market capitalization of $1 trillion could disappear into thin air.
$400 billion is equivalent to the entire market capitalization of Tesla evaporating within 8 months. Many people say Michael Berry is old and outdated ; he doesn't understand AI and will miss out on this greatest technological revolution in human history. Others say he will lose badly this time, just like when he shorted Tesla before.
But have you ever considered this: would someone who saw through the subprime bubble when everyone believed housing prices would never fall, someone who could endure two years of floating losses and withstand all pressure until the collapse, just bet hundreds of millions of dollars on a new technology because he doesn't understand it ?
Of course, he doesn't understand how to design chips, nor does he understand what HBM or KUDA are. But he understands one thing: this thing is more important than all the technical details combined. He understands human nature. He understands that when everyone starts to believe this time is different, it's often the most similar time.
He understands that when stock prices are too far ahead of profits, gravity will eventually take effect. He understands that the market's greatest strength is making you... To generate a very sincere belief at the wrong time —let's start with Micron, the hottest topic right now. The narrative surrounding Micron in the market is practically perfect.
HBM is the lifeblood of Nvidia GPUs; without HBM, even the most advanced GPU is just scrap metal . Micron is one of only three companies in the world capable of producing HBM ; the other two are Samsung and SK Hynix. With the explosion in AI demand, HBM prices have skyrocketed, and Micron's gross margin has jumped from negative last year to over 50% now.
Doesn't that sound like a perfect growth stock story? An undervalued tech giant has finally ushered in its own era, with the next ten years a golden age? But I have to tell you a truth that most people don't want to hear: Micron has never been a growth stock. It wasn't, isn't , and won't be in the future.
It's a complete cyclical stock. What is a cyclical stock ? It's a stock whose performance fluctuates like a rollercoaster with changes in industry supply and demand. When supply is insufficient , prices soar and profits skyrocket ; when supply exceeds demand, prices plummet and losses accumulate. Over the past 30 years...
Micron's stock price has experienced at least six such cycles in recent years , each time plummeting from its peak to its lowest point and then rebounding. This time will be no different. Remember this number: Micron's current HBM market share is less than 10% . Yes, you heard right. Over 90% of the global HBM market share is held by Samsung and SK Hynix.
Micron is a latecomer; its HBM production capacity won't truly be released until the second half of this year. At that time , Samsung and SK Hynix's production capacity will also double simultaneously. More importantly, there are no insurmountable technological barriers in the memory industry . With enough money and time, any company can manufacture it.
Once production capacity increases, a price war will immediately erupt. This is an inherent rule of the industry, without exception. The market's current valuation of Micron is based on its ability to maintain a gross margin of over 50% for the next five years . However, history tells us that Micron's long-term average gross margin is only around 20%.
Every time its gross margin exceeds 40%, a price war inevitably follows, and the gross margin drops rapidly. The price of Micron shares is rapidly falling back below 20%, even turning negative. Therefore, buying Micron now is essentially betting on a miracle that has never happened before : betting that there won't be a price war in the memory industry , betting that Micron can maintain a 50% gross margin, and betting that a cyclical stock has suddenly become a growth stock.
I'm not saying Micron will collapse immediately. In the short term, AI demand is indeed still high, and the HBM shortage may continue for several quarters. But in the long term, gravity will definitely play its role. If you already hold Micron shares, my advice is clear: start taking profits in batches at $800, reduce your position by half at $900, and sell everything at $1,000.
Don't be greedy; don't try to sell at the highest point. No one can sell at the highest point every time. Being able to leave the bubble with a profit is already a win for 90% of people. If you don't have a position yet, don't chase the price . Be patient and wait for it to fall back below $500 before considering buying.
Remember, in cyclical stocks, patience is more important than anything else. You'd rather miss an opportunity than buy the wrong one. Next, let's talk about AMD. If Micron packaged a cyclical stock as a growth stock, then AMD has turned a growth stock into a bubble stock. AMD's current price-to-earnings ratio (P/E ratio) is 143.
This means that at current profit levels, it would take 143 years to recoup the initial investment. Many would argue that you can't use a static P/E ratio to measure growth stocks, claiming that AMD's M00 and the upcoming MiZero will steal a large number of AI chip orders from Nvidia, leading to explosive profit growth in the future .
True, AMD is indeed an excellent company ; its M00 boasts strong performance and has secured orders from major clients. It's the only company in the world that can truly compete with Nvidia in the GPU market . However, the problem is that the market has already priced in all these expectations , and even more, into the stock price.
The current stock price already foreshadows AMD's perfect performance over the next three years, assuming MiZero can steal 30% of Nvidia's market share and that AI chip prices won't fall. The stock price drop is based on the assumptions that AMD's gross margin can remain above 70% and that no new competitors will emerge.
If any of these assumptions are wrong, the stock price will plummet. In other words, AMD's current margin for error is zero. It must deliver perfect financial reports every quarter for the next three years, with every product exceeding market expectations. There can be no mistakes. If even one quarter's performance falls short of expectations or one product is delayed, the stock price will immediately plunge by more than 20%.
This is like walking a tightrope with no safety net. You might earn a little more if you make it , but if you wobble even slightly, you'll be shattered. I've seen too many examples like this. In 2021, Tesla's price-to-earnings ratio exceeded 100, and everyone believed it would dominate the entire automotive industry.
What happened? The following year, its stock price fell by 75%. In 2022, Nvidia's P/E ratio exceeded 80, and then its stock price fell by 60%. History always repeats itself , only with different stories each time. So, if you haven't bought AMD yet, don't chase it above $440. At that price, there's no margin of safety.
If you already hold it, you can continue to hold, but you must set a stop-loss. Once the stock price falls below $350, don't hesitate to exit immediately. Remember, a good company doesn't equal a good stock. Even the best company will lose a lot of money if you buy it at a high price. Now let's talk about Corning.
Corning has the strongest logic among these four stocks . It's deeply tied to Nvidia. Nvidia's DGX supercomputer uses all Corning fiber. With the explosive growth of AI data centers, the demand for high-speed fiber will also grow exponentially . This logic is sound; it's not a story , it's a fact that's happening right now.
But the problem is still the same old problem: price. Corning's current P/E ratio has exceeded 100 , while its average P/E ratio over the past 10 years was only 28. That means the current stock price is more than 3.5 times its historical average valuation. Let me do a simple calculation for you. Assuming Corning's net profit grows by 30% annually for the next five years —a very optimistic assumption— its net profit would be 3.7 times its current value in five years.
If its price-to-earnings ratio (P/E ratio) returns to its historical average of 28 times by then, its share price will be roughly the same as it is now. In other words, if you buy Corning now and hold it for five years, the best-case scenario is breaking even. However, any unexpected events could lead to losses.
This is like knowing a restaurant will definitely earn 1 million next year , but someone wants to sell it to you for 3.5 million. Would you buy it? Of course not, because even if it really does earn 1 million annually, it would take three and a half years to recoup your investment. This doesn't even account for the various risks that might arise.
But people in the stock market do this every day because they believe an even bigger fool will buy the restaurant from them for 4 million. This is the "big fool theory. " If you currently hold Corning, you can continue to hold it because its logic is indeed very strong. It might even rise in the short term, but don't add to your position.
Wait until its P/E ratio falls below 35 before considering adding to your position. Finally, let's talk about Intel . Intel is the most dangerous of these four stocks. The current narrative surrounding Intel in the market is this: Intel is transforming itself to become the world's largest chip foundry.
It has already secured orders from Apple and will likely receive orders from Amazon and Microsoft in the future. Once its 18A process achieves mass production, it will once again become the leader in the semiconductor industry. This narrative sounds wonderful, right? A former giant falling from grace, enduring hardship, and ultimately making a triumphant return – this is a story everyone loves to hear.
But I have to tell you a harsh truth: Intel has failed for 10 consecutive years. Its 10nm process was delayed by 5 years, its 7nm process by 3 years, and now its 18A process has also been delayed once. Its management has changed repeatedly, but it has never truly solved the problems. Intel's EPS is still negative , meaning it's still losing money , but its stock price has risen by 60% in the past three months.
Buying Intel now is essentially a bet: you're betting that a management team that has failed for 10 consecutive years can suddenly become the industry leader in the next four years. You're betting that its 18A process can achieve mass production on time and outperform TSMC . It's a gamble that Intel can snatch a large number of foundry orders from TSMC.
You're betting that Intel can achieve the current market-expected profits by 2028. In other words , the money you're spending now is buying Intel's 2028 performance . This is no longer an investment ; it's gambling , and gambling with an extremely low probability of success. I'm not saying Intel will definitely not succeed; it might , but the probability is extremely low.
Most people only see how much the stock price will rise after it succeeds , but they don't see how much the stock price will fall after it fails. If Intel's transformation fails again, its stock price will fall back to below $20 or even lower. So, if you haven't bought Intel yet, don't touch it. Wait until it has two consecutive quarters of positive EPS , until its 18A process is truly in mass production , until it gets real, large-scale foundry orders , and then consider whether to buy.
Before that, all the rises are just hype driven by building momentum. At this point, you might ask me , if you're so bearish on semiconductors , why are these stocks still rising ? Why are so many people buying them? The answer is simple: because the current market is no longer driven by fundamentals; it's driven by sentiment, narratives, and FOMO.
FOMO (Fear of Missing Out) is a genetic flaw ingrained in humans. Watching others make money is more painful than losing money yourself. You'd rather lose with everyone else than watch others profit alone. This is why bubbles always seem to grow bigger and bigger— because everyone thinks they won't be the last fool, everyone thinks they can escape before the crash.
But history tells us that 99% of people don't escape. When the bubble is at its most insane , that's when you're most greedy. You watch the stock price rise day by day, you feel you sold too early, you regret it , you rush back in, and then the crash happens. I've seen too many people like this. They don't dare buy at the bottom, they start to observe when it's up 20%, they start to get excited when it's up 50%, and when it's up 100%, they finally convince themselves it's a long-term opportunity and rush in to buy.
Then they get trapped—trapped for five, ten , or even a lifetime. Peter Lynch famously said that the essence of investing isn't seizing every opportunity , but avoiding being killed by any mistake. Most people misunderstand this; they think investing is about seizing every upward trend and missing every opportunity.
All investments may fail , but what truly determines your investment success or failure isn't how many opportunities you seize , but whether you avoid fatal mistakes. If you can avoid even one major crash, you've already won. For most people , as long as you keep your snowball rolling and don't let it break, time will eventually reward you handsomely.
The secret to a snowball rolling isn't rolling fast , but preventing it from breaking. I hope you can keep this in mind. Many people think that if they don't buy AI chip stocks now, they'll miss an era. But actually, you won't miss any era . Historically, no era has ever given you only one chance to get on board.
During the internet revolution, if you missed Microsoft, you still had Google; if you missed Google, you still had Amazon; if you missed Amazon , you still had Apple. Even if you bought these companies after the dot-com bubble burst in 2000, you would have already made tens or even hundreds of times your initial investment.
The AI revolution is the same; it won't end in a year, it will last for decades . Countless companies will rise from this revolution, and countless others will fall. You have enough time to research , wait, and find companies truly worth holding long-term. You don't need to wait until everyone else is going crazy.
Rushing in to grab chips, what you need to do is stay calm and patient , wait for the bubble to burst , and wait for truly valuable companies to fall to reasonable prices before slowly buying in. Michael Berry isn't always right; he did lose a lot of money shorting Tesla , but he's always right about one thing: he never buys when the market is at its most frenzied.
He's always fearful when others are greedy and greedy when others are fearful. This time will be no different . He's suppressing the semiconductor crash now, which doesn't guarantee a win , but he can afford to lose . Even if he's wrong this time , he'll only lose a small portion of his assets . But if he's right, he'll make a fortune again.
And what about you? If you put all your money into AI chip stocks now , if you leverage it, if you bet your entire fortune, what will happen if Michael Berry is right ? You 'll lose everything. That's the biggest difference between you. He's betting with profits, while you're betting with your principal.
So stay clear-headed. Don't let the market frenzy cloud your judgment. Don't let others' profits make you lose your rationality. Don't mistake luck for ability, and don't mistake cycles for growth. If you've already made money in this frenzy, congratulations! It's time to consider taking profits. Secure your gains and don't be greedy.
Don't try to earn the last penny; the last penny is often the most expensive. If you're still outside the market, don't rush . Be patient. The market never lacks opportunities ; what it lacks is patience and cash. Wait for the bubble to burst and for bloodshed to occur—that's when you should truly be greedy.
Remember, the market never closes. It will open tomorrow, the day after, and next year. You will always have opportunities. But if you lose all your capital, you will never have another chance. If you've heard this far and are starting to realize that you might have been led astray by narratives, then please follow me.
I will continue to dissect this market logic that is most easily fooled by ordinary people, helping you find truly valuable signals in this noisy world. Finally, I want to ask you one more question: When the entire market is telling you that this time is different, have you ever considered that it 's actually the same every time?
Finally, a reminder: the content of this video is for information sharing and personal opinion discussion only and does not constitute any professional investment advice. If it involves financial, legal, or tax issues, please consult relevant professionals. Everyone's financial situation, risk tolerance, and judgment framework are different. Please make rational decisions based on your own circumstances.
Comments 0