The most interesting takeaway from this earnings season is that AI demand isn’t slowing—the market is simply becoming more selective. CRWV and NBIS were rewarded because their numbers show real demand: accelerating revenue, massive backlogs, and improving profitability. SMCI also benefited because AI demand is translating directly into stronger revenue and margins. Meanwhile, COHR and CBRS tell the other side of the story. COHR delivered a strong quarter, but after a huge run-up, a simple beat was no longer enough. CBRS had impressive future commitments, yet investors focused on weak hardware revenue. That tells us where the market is heading: AI stories are cheap. AI earnings are valuable. Going forward, I’d focus less on who has the most exciting AI narrative and more on who can convert

The STI rally looks impressive, but I think the easy money has already been made. A 24% YTD gain is difficult to ignore, especially when banks have been doing much of the heavy lifting. Yangzijiang’s record earnings show that this isn’t purely a liquidity-driven rally — real earnings are supporting parts of the market. But that also raises the bar. After such a strong run, valuation expansion alone probably won’t be enough. The next leg needs stronger profits, dividends and guidance. I’d be more selective here: banks for cash flow, shipbuilders like Yangzijiang for earnings momentum, and REITs if the rate environment becomes friendlier. My biggest takeaway: the STI may still have upside, but 2026 is shifting from an index-buying market to a stock-picking market.

I think the biggest takeaway is that political risk is becoming a real business cost. Apple hiring a Washington veteran is more than a PR move. With tariffs, China policy, regulation and a new CEO coming in, Apple clearly wants to reduce the risk of getting caught on the wrong side of Washington. As for the midterm election, history is encouraging, but I wouldn’t blindly bet on a post-election rally. The S&P 500 has already run hard this year, so some of that optimism may already be priced in. For me, the better strategy is simple: don’t try to predict who wins Congress. Watch earnings, margins and cash flow. If election uncertainty creates a correction in high-quality companies, that could be a better opportunity than chasing the rally. Washington can change overnight. Great business

8,000 Is Possible, But Earnings Must Deliver I’m bullish on the S&P 500 reaching 8,000, but the reason isn’t valuation—it’s earnings. Goldman’s thesis is straightforward: EPS growth, especially from AI infrastructure, can push the index higher without requiring a much higher P/E. Cooling inflation also gives the Fed room to stay patient. But there’s one big catch: expectations are already very high. If AI spending keeps translating into real profits, 8,000 looks achievable. If oil spikes, the Fed turns hawkish, or AI earnings disappoint, the rally could lose momentum quickly. So I wouldn’t chase the headline target. Watch earnings, inflation and Fed policy. If those three stay friendly, the bull market still has fuel. @Capital_Insigh

I think Nvidia’s $500B financing push is more genius than gamble — at least for now. Jensen Huang is effectively bringing Wall Street capital into the AI infrastructure boom without putting the entire burden on Nvidia’s own balance sheet. More financing means customers can build more data centers, buy more GPUs, and accelerate AI deployment. That creates a powerful cycle: capital → infrastructure → Nvidia chips → AI revenue. But the risk is obvious. If AI data centers struggle to generate enough returns, leverage could work in reverse, putting pressure on lenders, infrastructure valuations and eventually Nvidia’s growth expectations. So I wouldn’t call this a circular bubble yet. I’d call it a massive bet on AI economics. My view: Jensen may have found a brilliant way to scale AI demand —

My take: B — Too early. $SpaceX(SPCX)$ is clearly trying to become more than a rocket company. Grok Bot puts it directly into the enterprise AI agent race, while the potential $60B Cursor acquisition could give it a powerful AI software platform. But the market may be getting ahead of itself. The real test isn't whether SpaceX can launch an AI product — it’s whether Grok Bot and Cursor can generate meaningful revenue and eventually justify the massive AI spending. The recent rebound shows investors are excited, but the pullback is a reminder that the AI thesis still needs proof. If Cursor closes smoothly and enterprise adoption takes off, $SPCX could get a major AI re-rating. For now, I’m watching the numbers rather than chasing the hype.

I think Singapore’s market revival is real, but it still needs to prove itself. The STI hitting a record high, stronger trading volumes and SGX’s improving results all point to a genuine recovery in investor confidence. The biggest positive is the IPO pipeline, especially the growing presence of technology, healthcare and advanced manufacturing companies. But I wouldn’t get too excited about the “50 IPOs” headline yet. The real test is what happens after listing. If new companies can attract institutional investors, build liquidity and trade above their IPO prices, confidence in SGX will improve significantly. So my view is cautiously bullish: Singapore may be entering a new market cycle, but the next 12–18 months will determine whether this is a lasting revival or simply another short-te

I think this rebound is about more than just strong earnings AI demand clearly hasn’t disappeared—the bottleneck is shifting from GPUs to power, cooling, networking and data-center capacity. CoreWeave’s $104B+ backlog and Super Micro’s raised FY2027 guidance show customers are still spending aggressively. That said, I wouldn’t chase the after-hours spike. CRWV has huge growth potential but also massive capital and financing risks, while SMCI’s improving margins make it particularly interesting. My take: If CRWV and SMCI can hold their gains during regular trading, this could be the early stage of an AI infrastructure recovery rather than just a short-lived earnings bounce. @Tiger_comments [暗中观察]

I’d say JPMorgan’s $6,000 gold call is aggressive, but the underlying thesis is worth watching. Gold’s rally is becoming more than a safe-haven trade. Central-bank diversification, geopolitical uncertainty and concerns over fiscal stability are creating structural demand. Even if central-bank buying slows temporarily, the broader trend hasn’t necessarily changed. The biggest risk is a hawkish Fed and rising real yields, which could trigger another sharp pullback. My take: I wouldn’t chase gold after such a huge run, but I also wouldn’t underestimate the structural bull case. The key question now isn’t whether gold can hit $6,000 — it’s whether the market is already pricing too much of that optimism. @Capital_Insights [得意]

I’m still cautiously bullish on Singapore equities. This rally looks more structural than just a National Day boost, with DBS, OCBC and UOB delivering strong earnings, while SGX benefits from rising market activity. The key point is that banks are no longer relying purely on net interest margins. Wealth management, fees and trading income are becoming increasingly important as rates come down. That said, after a 20%+ rally, I wouldn’t chase aggressively at current levels. The STI now needs earnings and dividends to catch up with the valuation. My view: 5,400–5,700 could be a consolidation zone. If bank earnings remain strong and capital continues flowing into Singapore, a break above 5,700 could open the door to 6,000. I’m choosing D — Holding Steady. I’d rather collect dividends and wait

I’d pick B) Cybersecurity. AI is creating a huge productivity wave, but it’s also expanding the attack surface across cloud, identity, data and AI agents. As enterprises deploy more AI, cybersecurity becomes less of a discretionary expense and more of a necessity. Among the names, PANW would be my top pick. Its platform-consolidation strategy, strong ARR growth and exposure to multiple areas of enterprise security give it a compelling long-term setup. CRWD is also attractive, especially with its strong platform ecosystem and recurring revenue model. That said, both stocks have already rerated significantly, so I wouldn’t blindly chase new highs. I think cybersecurity still has legs, but the next phase will need to be supported by earnings, cash flow and sustainable growth—not just the AI

AI Infra: Which One Stands Out? $AMD$, $LITE$ and $CRWV$ represent three different layers of AI infrastructure: compute, optical connectivity and AI cloud. AMD’s Q2 was strong, with revenue up 50% YoY and Data Center revenue more than doubling. Despite the post-earnings pullback, the bigger story is whether Helios and its next-gen AI systems can drive another growth cycle. For Lumentum, the key is 1.6T optics, EML supply and OCS adoption. If these accelerate, the optical bottleneck could become a major earnings catalyst. CoreWeave offers the fastest growth, but also the biggest risk. Investors need to watch its $99B backlog, margins, CapEx and debt load closely. My take: AMD for long-term strength, LITE for the optical bottleneck, CRWV for high-growth/high-risk exposure. The AI boom is st

Orchard Road isn’t just Singapore’s luxury shopping belt — it’s also a battleground for S-REIT investors. If I had to pick, CICT and OUE REIT stand out to me. 🏆 CICT is the quality play. Its Paragon acquisition further strengthens its luxury retail exposure, while its scale and diversified portfolio offer better stability. 💰 OUE REIT is the value play. At around 0.66× P/NAV, the discount looks attractive, especially with 1H 2026 DPU jumping 28.6% YoY. 📈 Lendlease REIT is my turnaround candidate, while Starhill offers the higher-yield angle. Suntec is interesting for MICE and office recovery, but its higher gearing deserves attention. My ranking: CICT for quality, OUE REIT for value, Lendlease for growth. @AI_FocusedTrader [胜利]

The AI infra story is becoming less about “who wins AI” and more about where the bottleneck is. AMD = compute, Lumentum = optical connectivity, CoreWeave = GPU capacity. All three benefit from the same AI spending cycle, but the risks are different. AMD has the strongest platform, but expectations are already high. Lumentum’s 1.6T optics and OCS ramp could be the next major catalyst, while CoreWeave offers huge growth but carries much higher capex and balance-sheet risk. For me, LITE has the most interesting bottleneck story, AMD the strongest fundamentals, and CRWV the highest-risk/highest-reward profile. @WallStreet_Tiger [贱笑]

My favorite stock from this week’s list is $RKLB.  I think Rocket Lab has an interesting long-term story, not only because of its launch business, but also its Space Systems segment and the potential of Neutron. The valuation is not cheap and execution risk is still high, but if the company continues to deliver, I believe the upside could be substantial. I also like $CSCO as a more established choice. AI data centers are creating strong demand for networking infrastructure, so I’ll be watching its EPS, revenue growth and management guidance closely. For dividends, $IBM stands out to me because it offers a combination of income and exposure to AI/software growth. Overall, I don’t think investors should focus only on whether EPS beats estimates. Guidance, margins, cash flow and future

The market is showing a clear divergence: AI monetization remains strong, while traditional consumer tech is starting to show signs of weakness. OpenAI’s accelerating revenue and Microsoft’s strong cloud growth suggest that AI demand is becoming a real business rather than just a hype story. However, Qualcomm’s cautious outlook raises concerns about smartphone demand and Apple’s supply chain. For me, the key question is whether AI earnings can keep growing fast enough to justify current valuations, especially with interest rates remaining elevated. I’m watching earnings, capex and cash flow closely — the companies turning AI spending into real profits should have the strongest long-term advantage. @MillionaireTiger [思考]