원문 보기
On April 8 of this year, with oil prices soaring due to the war and the Dow Jones weeks away from a correction, Tom Lee went on CNBC and said three words: "The bottom has been reached." Six months later, on October 6, the S&P 500 index closed at a record high of 7,818 points. This is the man who advised clients to buy during the pandemic in March 2020 and who predicted a 2023 bull market when almost everyone on Wall Street was predicting a recession.
Bloomberg recognized his forecast for 2023 as the most accurate in the industry. So, on Thursday, he appeared again on CNBC's Closing Bell program. The Nasdaq was falling, the Dow Jones was down more than 6% from its high, and AI stocks were being actively sold off. His message: "We are approaching the final stages of the AI stock sell-off." Playing the tech sector lower has become the consensus strategy , Treasury yields , which were approaching 5%, have fallen, and the Fed is changing its rhetoric from “ hawkish” to potentially “dovey.
” When asked directly if he would buy it, he replied, "Yes." And he named exactly where: " bottlenecks" of AI, memory, Nvidia, big tech companies. So today we'll look at six stocks that relate to this forecast. Stay with us until the sixth stock—it's a $6 stock that's seen revenue grow 13-fold year-over-year, the company has received over $270 million in new orders since July, and its stock price has still fallen.
But remember, this video is for educational purposes only. This is not financial advice. Always do your own research and be aware of the risks before investing. With that in mind, let 's begin. The number one stock, CoreWeave, ticker CWV (note: corrected for context), is currently priced at $158.74.
CoreWeave is an AI landlord. It builds data centers equipped with Nvidia chips and rents out computing power. Arguments in favor: Revenue in the second quarter reached $582 million , up 454% from a year earlier. Adjusted EBITDA became positive and amounted to $236 million . Profitability 41%. The revenue forecast for this year is over $3 billion, and at the end of June the company had $37.5 billion in contracts outstanding, secured by Microsoft for $ 17.4 billion and Meta for $27 billion.
And that's why she's on this list. The stock has fallen more than 30% this month, and almost none of it has to do with news from the company itself. It's about CoreWeave. CoreWeave bonds started yielding around 15%, and the market sold off all the cloud AI companies at the same time. Nvidia's own convertible notes have coupons of only 0.5% and 4.5%.
Different balances, but the same punishment. Arguments against. This business operates on borrowed funds. Nvidia plans to spend $20 billion to $25 billion in capital expenditures this year on about $3 billion in revenue. So the price of each new loan is now formed taking into account Coreweave. Short interest is about 20% of the free float, and two clients provide about 4/5 of this order book.
If any of them hesitate, the stock will also fall. Stock number two, Marvel Technology, ticker MRVL, is currently worth $228.51. Marvell is developing two things that AI data centers cannot function without: custom-made chips for giants like Amazon and Microsoft, and connectivity chips for transferring data between processors.
Arguments in favor: Revenue in the second quarter hit a record high of $2.74 billion, up 37% . Then, on October 6, at its investor day, Marvell raised its fiscal 2028 revenue target to $20 billion. Less than a year ago, that goal was $13 billion. This is the fifth increase in 10 months, and the company has set a 2031 target of $70 billion to $90 billion, up from about $8 billion last year.
Arguments against: To reach $70 billion in revenue, it must grow by more than 50% per year for five consecutive years. The stock has more than doubled since January, so much of that is already priced in. When Marvell beat forecasts in August , the stock still fell 12% in 2 days. And custom chips mean dependence on a handful of giant customers , who can always decide to develop their own.
And before I move on to the third choice, I want to take a minute to talk about what has been a real revelation for many viewers of this channel. If you like these videos and want to dive deeper— like what exactly I buy, when I buy, and when I get out—you should check out Zane Capital Intelligence. It's where I share my most compelling ideas about the stocks I'm most confident in, backed up by high-probability setups.
As a member, you get my best recommendations, real-time buy and sell alerts so you're never left guessing, and a monthly portfolio status report so you can see how everything is performing. No "water" or unnecessary noise. If you're serious about your portfolio and want to get an edge, the link is in the description below, so feel free to jump on over.
Stock number three, Micron Technology, ticker MU, is currently worth $877.55. Micron makes memory —the chips that store data for AI models— and memory is the very “ bottleneck” that Tom Lee pointed out. Arguments in favor: Micron's revenue in fiscal year 2026 was $133 billion. A year earlier, it was $37 billion .
Three and a half times in 1 year. The fourth quarter alone was $54.2 billion, up 379% with a gross margin of 87%, and the pace is not slowing down. Micron is forecasting a record $ 61.5 billion in revenue next quarter, says more than 75% of this year's production has already been sold out, and expects an even bigger supply shortage in 2027 and 2028 .
The " bear" argument: memory is the most cyclical segment of the chip industry. When prices change, they change quickly, and 87% margins are an open invitation for new capacity, particularly from China. Micron itself plans capital expenditures of about $25 billion for the first half of this fiscal year alone, almost as much as for all of last year.
And when the company beat forecasts in September , the stock barely budged, and has since fallen nearly 18% . The market is already asking when the peak will come. Stock number four, Lumentum, ticker LITE, is now at $756.17. Lumentum makes lasers and optical components that enable AI data centers to transmit data using light instead of copper.
As AI models get bigger, chips need to communicate with each other more often, and that's Lumentum's whole business. The bulls' argument: revenue for fiscal year 2026 was $3 billion, an increase of 83%. The fourth quarter alone was over $ 1 billion, up 109%, and the company is forecasting $ 1.25 billion this quarter, up 130% year-over- year, hitting its targets a quarter early.
Its optical switch business is expected to report first-quarter revenue of more than $100 million, and the company has a multi-million dollar order for co- packaged optics due in the first half of 2027 . Its market value grew from $6.6 billion to $63.6 billion in just one fiscal year. The " bears" argument: that's the problem.
The stock delivered a 762% total return last fiscal year, meaning any fluctuation is severely punished. Its biggest clients are a handful of hyperscalers, and the report is due on November 5th, so you're buying right before the exam. By the way, if this review was useful to you , click the "like" button and subscribe.
This genuinely helps the channel and means you won't miss the next video. There are two stocks left, and the next one just made the largest offering by a foreign company in the US in history. Stock number five, SK Hynix, ticker SKHY, is now at $137.02. SK Hynix is the world's largest manufacturer of high- bandwidth memory, the specialized memory that goes into Nvidia's AI chips, with about 56% of the market.
In July, it went public on the Nasdaq at a price of $149, raising $26.5 billion—the largest offering of a foreign company in the United States in history. Arguments for. Revenue for the second quarter was 79 trillion won, or about $55 billion , up 257% year-on- year, with an operating margin of 76% .
Production of HBM 4 chips for Nvidia's next platform, Vera Rubin, is gaining momentum. Third- quarter results are due next week, with expected operating profit of about 77 trillion won, the sixth consecutive record, and shares trading at less than five times next year's expected profit. Arguments against.
The price is now below the IPO price, with shares down about 24% in a month. When the company reported in July, revenue and profit missed forecasts, and Korean stocks fell nearly 10% in one day. Cyclical risks, like those at Micron, apply here, and Tom Lee's argument works both ways. Memory is a "bottleneck", which is why stocks have become the object of massive short positions.
Now you may be wondering, when is the best time to buy these stocks ? That's what my free notification service is for. Behind the scenes, I determine precise entry points and price targets for each stock I cover, and you will receive a daily alert when the price reaches the entry or exit. It's completely free, and the registration link is in the description below.
And the sixth stock mentioned at the beginning— Ondas, ticker ONDS, is now at $6.20. Ondas is not an AI stock, and it is not the company that Tom Lee mentioned. It produces drones, counter- drone systems, and loitering munitions for armies. It's here because it was sold along with AI stocks, and it's fallen by more than a third this year , even though the order book has only been growing.
Arguments " for". Revenue for the second quarter was $83.8 million , more than 13 times higher than in the same quarter last year. The forecast for the year was raised to $525–550 million. As of the end of June, the order book stood at $757 million, and since then Ondas has received new orders worth more than $270 million , including a $56 million European ammunition contract.
At the end of June, the company had $1.4 billion in cash, and this month it launched a defense AI project with Palantir. Arguments against. Much of this growth is driven by businesses that Ondas bought, not created. The company reported a $ 51 million adjusted EBITDA loss in the quarter and spent $325 million on two new acquisitions over the summer.
The company is expected to reach profitability only by the end of 2027. High growth, but not safe. So, we have the result. Six stocks, one call from a person who predicted six months ago that the bottom had already been reached. Thank you very much for watching, see you in the next video.
Comments 0