저자: JerryRomine · 외국인
요약 보기
TSMC가 실적 발표 후 3.13% 하락했지만, 매출 1.134조(전년비 35.1%↑), 순이익 58.3%↑, 매출총이익률 66.2%, 영업이익률 58.1%, 다음 분기 가이던스 390~402억 달러 등 실적은 강하다고 주장. HPC 비중 61%, 7나노 이하 74%로 AI 수요가 견조하며 하락은 투자자 공포에 따른 오판이라는 논리. 엔비디아·AMD·브로드컴·마이크론·버티브·마벨·ASML 등 AI 밸류체인 전반에 긍정적 파급효과 언급. 340 부근 POC 지지선에서 분할매수(DCA) 관심 표명. 과거 실적 해석과 장기 낙관이 중심이며 명시적 1~3개월 방향 예측은 제한적.
다듬은 글
On Thursday, Taiwan Semiconductor reported earnings before the market opened. The stock finished down 3.13% on the day, and I think a lot of investors just got baited into the wrong conclusion. Because if the single most important manufacturing choke point in AI drops after earnings, most people assume something broke.
Demand must be slowing. AI must be cooling off. Tech must be running out of gas. But, what if the exact opposite happened? What if one of the most important earnings reports in the entire market just told you the AI build-out is still alive, still strong, and Wall Street was too busy following the candles notice.
That's what we're talking about today, because this report wasn't normal. And if you don't understand what Taiwan Semiconductor just said through the numbers, you may be missing one of the biggest signals in all of AI and tech right now. . Right now, investors are nervous. The market's volatile. War headlines are everywhere.
People are tired, and a lot of investors are just looking for the next excuse to panic. So, when a giant semiconductor stock closes red after earnings, the easy move is to assume the smart money knows something. That's what people do when they're emotional. They borrow conviction from price action because they don't have enough conviction in the business.
And that's how people get smoked. They react to the move, ignore the report, and let the crowd do their thinking for them. Because if you looked at Thursday, and your first thought was, "Wow, Taiwan Semiconductor must have had a bad quarter." Then you missed the report entirely. Now, this is where this gets bigger than just one company.
Taiwan Semiconductor isn't just another chip stock. This is one of the biggest bottlenecks in the entire AI chain. There are analyst estimates out there that put Taiwan Semiconductor at around 90% share in AI data center chip production. That's massive. So, with this company still growing hard, still guiding higher, and still showing strength where it counts, that's not just bullish for one ticker.
That's a read-through for the whole tech and AI build-out. Think about it like this. If the company making the most advanced chips for the biggest player in the world still seeing strong demand, then maybe the bigger story isn't fear. Maybe the bigger story is that the foundation underneath AI is still a lot stronger than people want to admit.
But, the crowd did what the crowd always does. They saw a red day. They saw macro fear. They saw war headlines. They saw uncertainty. And instead of asking what the business said, they just assumed the market had already decided the answer. That kind of thinking is lazy. It's emotional. And in this market, that's the kind of thinking that gets people chopped up.
Because price action and business quality aren't the same thing. Sometimes they line up. Sometimes they don't. And on Thursday, I think the market was telling you more about investor nerves than it was telling you about Taiwan Semiconductor. And once you actually get into the report, the whole thing starts looking even more ridiculous.
Let's make this real simple. If the AI build-out was cracking, if demand was slowing, if all the spending was finally hitting the wall, then the one company you'd expect to slow is Taiwan Semiconductor. This is the bottleneck. This is the place where a lot of the truth has to pass through. So, if the business was soft, you'd expect weaker demand, weaker margins, weaker guidance, or at least some sign that the machine was slowing down.
That's not what happened. What happened is the business put up numbers that say strength. The business put up numbers that say the part of the market tied most closely to advanced AI compute is still pushing hard. And that's why Thursday's market reaction is such a trap. The stock looked weak for a day.
The business looks strong. Now, let's get into the proof, because this is where the story really stops being debatable. Last quarter, Taiwan Semiconductor reported 1.134 trillion in revenue. That was up 35.1% year-over-year, and 8.4% from the prior quarter. Net income rose 58.3% year-over-year. Gross margin came in at 66.2%, and operating margin 58.1%.
Those aren't shaky numbers. Those aren't good enough numbers. Those are monster numbers. And then it gets even better. Taiwan Semiconductor guided the current quarter from 39 billion to 40.2 billion in revenue, and said full year revenue is expected to grow above 30% in the US. So, let's be honest here.
If a company's putting up revenue growth like that, earnings growth like that, and margins that strong, plus guidance, what exactly are people bearish about? This is where a lot of investors lose the plot. They look at the red candle and suddenly forget how to read a business. And I'll say it even more directly.
If this report came out and the stock was green 5%, people would be calling it incredible. But, because it closed red on Thursday, suddenly people want to act confused. That's headline chasing. That's not analysis. This is the difference between watching stocks and actually evaluating a business. If you want to see how I apply that process in real time, that's what Patreon is for.
It goes deeper into the entries, the risk, and the decision-making behind the moves. The link's in the description. Now, let's talk about where the strength actually showed up. Here's what I think Wall Street missed. It wasn't just that Taiwan Semiconductor grew, it's where the strength showed up. HPC, which stands for high-performance computing, made up 61% of revenue.
That includes the heavy compute side of the world, like AI accelerators, data center chips, and the kind of advanced infrastructure that powers this whole AI race. That was up from 55% in the prior quarter. Then, look at the node mix. 7 nanometer and below represented 74% of wafer revenue. That means the center of gravity keeps moving toward the hardest, the most advanced part of the stack.
That's a huge deal. Because if the part of the business tied most closely to advanced AI compute is getting stronger, then the market reaction on Thursday starts looking dumb, not cautious. Dumb. This is where people get fooled in markets like this. They think price is intelligence. They think the first move is the smart move.
A lot of the time, it's not. A lot of the time, it's just nerves, fast money, and people dumping before they even understand what they own. So, once you see that, the next question becomes obvious. If Taiwan Semiconductor is still this strong, who else wins? Here's my take. Taiwan Semiconductor had a phenomenal earnings report.
Not a decent report, not a better than feared report. A phenomenal report. The growth was there. The margins were there. The guidance was there. And the business mix kept leaning harder into the exact part of the market that matters most for AI and advanced tech. And that's why I think this quarter was massive for the whole sector.
Because if the biggest manufacturing choke point in AI is still seeing this kind of demand and this kind of growth, then the bottom line's simple. The AI and tech build-out remains strong, regardless of what the headlines say. People can argue about valuations. They can argue about macro. They can argue about war.
Fine. But, the proof is still in the numbers. Now, let's cover the bull and bear case. The bull case is obvious. Strong revenue growth, strong earnings growth, elite gross margin, elite operating margin, strong forward guidance, more exposure to high-performance computing, more exposure to advanced nodes.
This business is getting stronger in the exact part of the market where strength counts most. The bear case is that this is still a volatile name in a very noisy world. You've got geopolitical risk. You've got overseas fab cost. You've got a premium valuation. And inventory days rose to 80 from 74 as newer nodes ramp.
So, I don't look at the bear case and think the thesis broke. I look at the bear case and think that's the price of admission. If you want to own one of the most important businesses in the world, you're going to have to deal with some noise along the way. Now, let's widen this out, because this is where the story gets really interesting.
If Taiwan Semiconductor is the primary bottleneck in advanced AI chip manufacturing, then it becomes the kingmaker. And when the kingmaker puts up a report like this, you need to start asking who else benefits if that signal is true. The most obvious name is Nvidia. If leading-edge demand is still this strong at the foundry level, Nvidia is still in a beautiful spot because it's AI accelerators sit right in the middle of that demand.
Then, you've got AMD. Same logic. AI GPUs, server CPUs, advanced manufacturing, all of that ties back into a healthy high-end foundry environment. Broadcom also benefits, because if hyperscalers are still spending, custom silicon and AI networking still matter. Then, I'd look at names like Micron and Vertiv.
If AI servers keep ramping, Micron stays well positioned through memory and high-bandwidth memory demand. And if the compute build-out is still real, Vertiv stays in a strong lane because power, cooling, and infrastructure don't stop mattering just because people got scared by a red day. You can even widen it further to names like Marvell and ASML.
Marvell has the connectivity and custom silicon angle. ASML remains one of the foundational picks and shovels names behind advanced node manufacturing. That's why I call Taiwan Semiconductor the kingmaker. If this company is still telling the truth about strong AI demand, a lot of other names down the chain are still very well positioned.
And that brings me to where I stand. I'm as bullish as ever on Taiwan Semiconductor, and this report reaffirmed my thesis. I'm as bullish as ever on Taiwan Semiconductor, and this report reaffirmed my thesis. I hate giving out hard target numbers in a market like this, because things can change fast, especially when the macro backdrop is messy and emotions are running high.
But, if Taiwan Semiconductor pulls back near 340, that's the area where I'd be interested in DCA. And the reason is simple. On the chart, that zone stands out because that's where the point of control, or POC, sits. The POC is just the price area where the most volume traded, and that often acts like a strong support or battleground level.
So, to me, Thursday was a perfect example of how investors get baited into the wrong conclusion. They saw the move and ignored the message. They followed the crowd and missed the business. And that may be the biggest lesson in this whole video. The risk isn't always the stock that drops on good news.
Sometimes the risk is not being the kind of investor who can't tell the difference between fear and facts. If the most important bottleneck in AI is still this strong, then Wall Street may not be ready for what that means next. And the people waiting for the whole AI and tech build-out to quietly die may be waiting a lot longer than they think.
And this is exactly where most investors stay stuck. They see a move, they feel pressure, and they let the market make the decision for them. Inside my Patreon, people are actually doing the opposite. They're applying a process and executing with structure. Just this week, we had real wins like Spectre locking in $39,899, Weeks Havoc with $4,927, and Steven with $1,620.
Those aren't guesses. Those are decisions backed by a framework. And again, this isn't about chasing someone else's trade. That's exactly what I don't want you doing. This is about understanding how those decisions get made in the first place. In my Patreon, you see how it actually gets used in real time.
Entries, exits, risk, position sizing, and how to stay disciplined when the markets gets noisy. And when you join, there's an onboarding that shows you exactly where to start so you're not guessing. If you're serious about improving how you operate in this market, join me on Patreon. The link is in the description.
If you stayed all the way through this one, drop no headline panic in the comments. Tribe check. Calm investors see things emotional investors never will.
Comments 0