Shyon

For me, I am more interested in the memory trade than SG banks right now. $Micron Tech(MU.US) saying more than 75% of its 2027 output is already committed gives me more confidence that the memory cycle is being supported by real AI infrastructure demand, rather than just short-term excitement. With MU and MUU already in my portfolio, I am looking for the follow-through to continue, but I still prefer adding on pullbacks instead of chasing a strong move.

At the same time, tonight’s payroll report could bring another layer of volatility, especially with Treasury yields still elevated. I would rather see how yields and the broader market react first before adding aggressively. For me, the plan remains simple: stay patient, collect quality semiconductor exposure on weakness, and avoid letting one strong session push me into FOMO.

I think the $Micron Tech(MU.US) reaction is a good reminder that strong earnings do not always mean an immediate stock rally. The numbers were impressive, but expectations were already extremely high, so even a major beat can be largely priced in. For me, the more important question is whether AI-driven memory demand can keep accelerating into the next few quarters, especially with HBM and data-center demand supporting pricing. I would rather watch the next guidance updates and demand visibility than chase the initial earnings reaction.

I am still constructive on the longer-term AI memory cycle, but I would not assume every strong result will translate into an instant upside move. With Treasury yields still above 5.2%, valuation and macro conditions can easily offset good company-specific news. I see this as a situation where patience matters: if the fundamentals continue improving while the stock consolidates, I would be more comfortable accumulating gradually rather than chasing a flat or sudden breakout.

For me, $Micron Tech(MU.US) guidance matters more this week. The memory cycle is becoming much more than a traditional cycle story, with strong AI server demand, tight DRAM supply and growing HBM demand supporting the longer-term outlook. I am especially watching whether Micron can raise its revenue, margin and HBM outlook rather than simply beat the current quarter.

PCE still matters because a hotter inflation print could keep Treasury yields elevated and pressure high-growth tech stocks. But for my portfolio, I think Micron’s forward guidance could have a more direct impact on the semiconductor names I am holding and watching. I would rather focus on the fundamentals and demand outlook than chase the immediate post-earnings move.

My vote is Up 4.3% to 8.6%. I am moderately bullish because $Micron Tech(MU.US) enters earnings with strong AI-driven memory demand, tight supply and solid pricing power. I will be watching DRAM/NAND pricing, HBM demand and FY2027 guidance more than the headline EPS beat.

The risk is that MU has already rallied sharply this year, so expectations are high and a “sell the news” reaction is possible. Still, if Micron confirms strong HBM demand, improving margins and continued memory tightness, I think the stock can deliver a positive move within my +4.3% to +8.6% range.

I would not rush into an Anthropic IPO at a $2 trillion valuation. The revenue growth is impressive, with 2025 revenue reaching nearly $4.6 billion, but the scale of spending is what makes me cautious. Reuters reports that the $42 billion net loss included about $34 billion of accounting charges, so it is not the same as burning $42 billion of cash, but the company still had more than $8 billion in operating losses and plans around $518 billion in future cloud, computing and infrastructure commitments.

For me, this is a case where I would rather watch first than chase the IPO on day one. I am bullish on the long-term AI infrastructure theme, but a $2 trillion valuation already prices in an enormous amount of future growth. I would want to see how quickly Anthropic can turn its explosive revenue growth into sustainable margins and cash flow before building a meaningful position. At this valuation, patience matters more than FOMO for me.

For me, I am watching $Micron Tech(MU.US) most closely this week ahead of the September 30 earnings. The recent strength in AI infrastructure continues to support the memory cycle, but I do not want to chase the stock purely on expectations. I am looking for evidence that HBM, DRAM and NAND demand remains strong, while pricing and margins continue to improve. Earnings guidance will be especially important because it can tell us whether the current AI memory demand is sustainable or already priced in.

I am still comfortable holding and gradually adding on pullbacks rather than making a large move before earnings. With Treasury yields staying elevated and October hike expectations still in focus, volatility could remain high even if Micron delivers solid numbers. My approach is simple: follow the fundamentals, respect the technical levels, and let the earnings confirm the thesis instead of chasing the excitement.

For me, the Anthropic-Akamai deal is another reminder that AI infrastructure spending is still expanding even as the cost of capital rises. I am still constructive on the long-term AI story, especially semiconductors and compute infrastructure, but I am becoming more selective about valuation and leverage. A 5.5% 30-year Treasury yield changes the opportunity cost significantly, so I would rather accumulate quality AI names gradually on pullbacks than chase a strong move after the headlines.

I am also watching PCE and the jobs report closely because higher-for-longer rates could create more volatility for growth stocks and REITs. I would not rush to make a major portfolio shift based on one rate move. My approach remains simple: keep some cash available, continue DCA into companies I understand, and use weakness to build positions instead of FOMO buying strength. AI demand can remain strong while multiples still compress, so I think both the growth story and the cost of money need to be respected.

For me, this looks like more than a short-term rotation. With the 10-year Treasury above 5%, investors are becoming more selective toward companies with heavy capex and distant cash flows. That explains why semiconductors can face pressure even while long-term AI demand remains strong.

I also think software deserves more attention as AI monetization becomes easier to measure. Companies like PLTR and MSFT can potentially turn AI into higher revenue and free cash flow without the same infrastructure burden. The key question is shifting from whether AI makes money to which part of the AI stack captures the profits.

Personally, I would not simply sell SOXX and buy IGV after one sharp move. I still believe chips are essential to AI, so I prefer accumulating quality names gradually while watching rates, credit spreads, and earnings. For me, cash flow will increasingly separate the winners from the stories.

I would not rush to trim growth stocks just because the 10-year yield moved above 5%. Higher yields clearly put pressure on high-valuation growth stocks, but I see this more as a reason to be selective and manage position sizes rather than exit the AI trade completely.

For me, a sharp pullback can actually create better opportunities to DCA into companies where the long-term fundamentals remain strong. I would rather buy gradually on weakness than chase when prices are running, while keeping enough cash to handle further volatility. The key is whether earnings and AI demand continue to justify the valuations.

For me, the $Sandisk(SNDK.US) story is interesting, but I would not chase the stock simply because Wall Street raised the target. AI memory demand is clearly a strong long-term theme, and SanDisk is benefiting from the growth in data-center storage, HBM and AI infrastructure. At the same time, memory is still a cyclical business, so a very high price target also comes with very high expectations. I would rather watch earnings, pricing and demand trends than follow a target price blindly.

I also understand why Burry is taking the opposite side, especially with memory and semiconductor valuations having moved so strongly. Personally, I am more comfortable staying invested in the semiconductor theme through pullbacks and gradual accumulation rather than trying to predict the exact top. For me, the key question is not whether SanDisk can rally further, but whether AI-driven memory demand can remain strong enough to support the current cycle.