Today semiconductors really took a hard hit, and the entire AI infrastructure line is under pressure, with many stocks falling more than 5% during the session.

No one feels good in this kind of market. Watching your account keep dropping is a normal reaction to feel uncomfortable. I feel the same way today, so don't think it's because you're not rational enough.
Don't doubt yourself or wonder if something is wrong with you, questioning why you didn't take profits when you should have, or cut losses when you should have. What I want to say is that this really isn't just your problem.
During this period, as long as you were long, life has probably been tough. The frustration and pain you're experiencing right now are being felt by many others in the market at this very moment. You are not alone in holding on.
Whenever I encounter a time like this, I remind myself of one sentence: One day or even one week's stock price does not define an industry cycle that lasts several years.
Over the past two years, AI has risen very quickly. Valuations have gone through several rounds of corrections, and I think that's very normal, even inevitable.
Almost every time there's a correction, the market starts discussing: "The AI trade is over." "The bubble has burst." "This time is really different." But looking back, these voices have appeared many times before, yet the AI industry hasn't stopped; it has still moved forward step by step to where it is today.
At least up to today, I haven't seen anything that would make me overturn my original judgment. Cloud vendors are continuing to expand AI capital expenditures, the speed of enterprise AI deployment is still increasing, the commercialization of leading model companies like OpenAI and Anthropic is accelerating, token consumption for large models is growing rapidly, data centers are still being built, and demand for AI infrastructure components like storage, networking, optical interconnects, and advanced packaging hasn't suddenly disappeared because of the past two days' decline.
So in my view, this is a round of valuation correction and mid-term adjustment, not the end of the AI supercycle. AI is still in a bull market, and I don't see signs of a bear market yet.
My view on this market is bearish in the short term, but the long-term logic hasn't changed at all. Until the end of next year, I am firmly a die-hard bull. After the end of next year, if real signs of a bear market appear, I will seriously consider exiting.
Based on this judgment, whether I buy common stocks or LEAP Calls, I deliberately choose an expiration date of January 2028, because I truly believe this trend can last until then—it's not just talk.
Let me give an example of a BTC bull market washout, hoping it can inspire you.
If you look back at that history, during the BTC bull market, you would see 30%, 40%, or even 50% pullbacks occurring many times. Almost every time there was a decline, some people thought the bull market was over.
Everyone has summarized the four-year cycle theory and knows it. This theory has been confirmed several times. If you buy and sell according to this theory every time, you will almost never make a mistake. But the real difficulty is in holding your chips through wave after wave of large fluctuations—very, very few people can do that. Because what works against us is human nature. Even if you know the answer, you will still be severely affected by the emotions in between.
Going back to AI, if you originally bought these companies because you believed AI would be a multi-year industry trend, then don't easily overturn your original investment logic because of a period of volatility.
Market sentiment is indeed very poor right now, especially in the AI infrastructure sector, with more and more pessimistic voices. But the market has always been like this—sentiment often moves faster than fundamentals. The more people start doubting the entire industry, the more you need to ask yourself: Has sentiment changed, or has the fundamental logic changed?
The truly hard part is sticking to the judgment you made through research when the market is at its most pessimistic.
Of course, this doesn't mean you should hold every stock forever. If the fundamentals really change, sell when you should. If the logic is disproven, you should re-evaluate—that's what rational investing looks like. But if only sentiment is changing and valuations are correcting, while the company's long-term logic hasn't actually changed, I won't easily overturn my original judgment because of a few days or even weeks of decline.
In the future, there will likely be many more such adjustments, possibly even more severe than this one. I don't think this is the last time either. But when we look back before the end of next year, or at the latest before the first half of 2028, we will find that those who truly made money from this industry trend are not necessarily the ones who perfectly timed every top and bottom, but those who didn't get shaken off the train time and again before the trend truly changed.
A true major trend does not go up in a straight line. Rising requires new capital, new capital requires more reasonable prices, and chips need to change hands repeatedly before the trend can continue. This process is painful—I know, and I understand the torment everyone is feeling right now. But if you believe in industry cycles, then patience itself is part of investing.
I hope everyone can rest early and not rush to draw conclusions about an entire AI cycle in a single day. No one is bearing this alone—I am also in the market, enduring it with everyone. Even if we don't hold until the end of 2027, at least we should hold until the end of 2026.