I'd be careful treating Rocket Lab or AST SpaceMobile as "SpaceX proxies". The bull case is straightforward: a successful SpaceX IPO could bring massive attention and fresh capital into the space sector, lifting related names through sentiment alone. That's what many traders are betting on. The bear case is that expectations may already be priced in. If investors can finally buy SpaceX directly, capital could rotate out of RKLB, ASTS, Virgin Galactic and Redwire rather than into them. History is full of "sell-the-news" events following highly anticipated listings. Between the two, RKLB has a clearer business model today with launch services, spacecraft systems, and growing government contracts. ASTS offers larger potential upside if its direct-to-cell network succeeds, but execution risk r

A $1.3T wipeout grabs headlines, but it doesn't automatically mean the AI story is broken. The market had become extremely crowded, valuations were stretched, and stronger-than-expected jobs data reduced hopes for near-term rate cuts. That combination was enough to trigger a sharp repricing. The key question is whether AI demand is slowing. So far, hyperscaler capex, data centre buildouts, and AI infrastructure spending remain intact. If earnings and spending plans hold up, this may prove to be a valuation reset rather than the start of a fundamental downturn. As for the SpaceX IPO, capital could rotate temporarily, but long-term liquidity is driven far more by monetary policy and corporate earnings than a single listing. Personally, I would be far more interested in buying quality names a

A $1.3T wipeout grabs headlines, but it doesn't automatically mean the AI story is broken. The market had become extremely crowded, valuations were stretched, and stronger-than-expected jobs data reduced hopes for near-term rate cuts. That combination was enough to trigger a sharp repricing. The key question is whether AI demand is slowing. So far, hyperscaler capex, data centre buildouts, and AI infrastructure spending remain intact. If earnings and spending plans hold up, this may prove to be a valuation reset rather than the start of a fundamental downturn. As for the SpaceX IPO, capital could rotate temporarily, but long-term liquidity is driven far more by monetary policy and corporate earnings than a single listing. Personally, I would be far more interested in buying quality names a

A 16% weekly decline is painful, but not unusual for Bitcoin. The more important question is whether this is a temporary sentiment shock or a change in the broader liquidity regime. If the selling is primarily driven by concerns over Michael Saylor and Strategy reducing exposure, confidence can recover once the market digests the news. However, if liquidity is tightening, rate-cut expectations are fading, and risk assets broadly weaken, Bitcoin could face further pressure. The AI vs Bitcoin pair trade is also worth watching. If funds are long semiconductors and short BTC, a sharp AI correction may force position unwinds that could actually benefit Bitcoin. Pair trades do not always mean both sides fall together. For long-term investors, buying gradually into weakness often makes more sense

Bitcoin's new low doesn't automatically mean liquidity is tightening. Crypto is often the first asset sold during risk-off periods, and recent weakness may reflect deleveraging and sentiment more than a macro liquidity shock. If Friday's nonfarm payrolls come in near 60k, rate-cut expectations could strengthen as growth concerns rise. However, higher oil prices complicate the picture by keeping inflation risks alive. The Fed may find it harder to cut aggressively if energy-driven inflation reaccelerates. As for Iran and oil, I think the market is pricing in a limited conflict, not a major supply disruption. That's why equities remain relatively resilient. The real black swan would be a prolonged escalation that pushes oil above US$100 and keeps it there. My base case: this is a growth sca

I'd be interested in SpaceX, but not necessarily at IPO pricing. SpaceX has real businesses: launch services, defence contracts, and Starlink. Unlike many hyped IPOs, it already generates substantial revenue. The question isn't whether it's a great company, but whether the valuation already assumes years of perfect execution. The xAI angle is where I'm more cautious. AI revenue growing 100x sounds exciting, but Grok remains behind leading models, and profitability is still distant. Investors may be paying today for cash flows that are many years away. My approach: if the IPO opens at a reasonable premium, I'd consider a starter position. If it surges 50-100% on day one, I'd rather wait for earnings and lock-up expiries. SpaceX could eventually justify a massive valuation, but even excepti

A 16% weekly decline is painful, but not unusual by Bitcoin standards. The more important question is whether this is a sentiment shock or a structural change in the investment case. If Strategy's sale marks a genuine shift away from its long-standing accumulation strategy, confidence could remain fragile in the near term. However, Bitcoin's long-term trajectory has historically been driven more by liquidity conditions, institutional adoption, ETF flows, and macro policy than by any single holder. The AI-vs-Bitcoin pair trade is interesting. If funds have been long semis and short BTC, a semiconductor pullback could force some profit-taking on both sides, creating additional volatility. That does not automatically make Bitcoin bullish, but it does suggest the recent weakness may not be ent

$Broadcom(AVGO)$  At this stage, it looks more like a healthy reset than the start of an AI bull market reversal. Broadcom's selloff was driven less by weak fundamentals and more by expectations running ahead of reality. When a stock rallies aggressively into earnings, even strong results can disappoint if guidance does not materially exceed what investors have already priced in. The fact that NVIDIA rose while Broadcom fell suggests investors are becoming more selective rather than abandoning the AI theme altogether. The Dow reaching new highs while semiconductors pull back also points to sector rotation, not necessarily a market-wide risk-off event. After an extended AI-led rally, it is normal for capital to rotate into financials, industri

For long-term investors, a 3-4% pullback in NVIDIA is not especially meaningful. The core thesis remains intact: hyperscaler AI spending is still growing, TSMC continues to report strong demand, and NVIDIA remains the dominant supplier of AI accelerators. That said, the market is no longer pricing NVIDIA as a cyclical chip company. Expectations are extremely high. Broadcom's reaction shows that even strong results can trigger selloffs if guidance fails to exceed lofty forecasts. My approach would be gradual accumulation rather than trying to time the bottom. Geopolitical headlines and profit-taking could create further volatility, but waiting for complete clarity often means buying at higher prices later. The key question is not whether NVDA can grow, but whether growth can continue to out

At this stage, I view it more as a pause than a definitive end to the bull run. The market is facing three simultaneous headwinds: geopolitical tensions (oil and inflation risk), interest-rate uncertainty, and extremely high expectations for AI-related stocks. When valuations are stretched, even strong earnings can trigger selloffs, as seen with AVGO and several software names. However, one weak ADP report does not necessarily signal a major economic slowdown, and the broader AI capex cycle remains intact. Companies continue to spend aggressively on data centres, chips, networking, and power infrastructure. That fundamental driver has not materially changed. For investors who are already appropriately allocated, I would be cautious about aggressive trimming solely because of a single risk-

$Broadcom(AVGO)$  Broadcom's selloff looks more like an expectations reset than a breakdown of the core thesis. AI revenue still grew strongly, and the custom ASIC opportunity with hyperscalers remains one of the most compelling themes in AI infrastructure. The VMware acquisition also continues to strengthen Broadcom's software cash flow profile. The issue is valuation. After a huge rally and record highs, investors were expecting not just strong results, but accelerating guidance. When expectations become extreme, even good earnings can trigger profit-taking. Whether AVGO falls below $400 depends more on market sentiment than fundamentals. If AI-related names continue derating and investors rotate away from high-multiple growth stocks, a mov

A 3.6% pullback in NVIDIA is hardly unusual after its massive run. The key question is whether the investment thesis has changed. So far, AI infrastructure spending remains strong, hyperscalers continue investing aggressively, and TSMC's demand commentary suggests the underlying trend is intact. That said, I would not rush to chase every dip. Geopolitical tensions can create further volatility, and semiconductor stocks have become crowded trades. A deeper correction of 10-15% would offer a more attractive risk-reward profile than a routine 3-4% pullback. For long-term investors, gradual accumulation still makes sense. For those with limited cash, patience may be rewarded given stretched valuations and elevated market expectations. The AI compute story remains compelling, but sentiment can

A 3.8% drop alone would not be enough for me to change my view on Alphabet. The key issue is whether the market believes AI capex is creating long-term value faster than it is consuming capital. If Alphabet can maintain strong growth in Search, Cloud, YouTube and AI services, a temporary increase in spending is easier to justify. What matters more than the share issuance headline is: Whether AI revenue growth accelerates. Whether Google Cloud margins remain healthy. Whether Gemini and AI-powered Search monetise successfully. Whether capex growth eventually slows relative to revenue growth. For a long-term investor, I would become increasingly interested if the pullback expanded into the 10-15% range from recent highs, assuming fundamentals remained intact. At that point, sentiment may be p

$Marvell Technology(MRVL)$  I would not chase MRVL after a 32% one-day surge, even if the long-term ASIC story remains compelling. The bullish case is clear: Hyperscalers increasingly want custom AI chips to complement, not replace, NVIDIA GPUs. Marvell is a major beneficiary through custom silicon, networking, interconnects, and optical infrastructure. Jensen Huang's public endorsement adds credibility to the ASIC/XPU growth narrative. Alongside Broadcom, MRVL is one of the purest ways to play the custom AI silicon trend. The risk is valuation and expectations. A single CEO endorsement does not create a trillion-dollar market cap overnight. After a 30%+ gap-up, the stock is pricing in years of future growth. Even great companies can deliver

Yes, Intel and AMD can still hold most desktop dominance in the near term, but their moat is clearly weakening. NVIDIA’s threat is strongest in premium AI PCs, creator laptops, workstations, and developer machines, not ordinary office desktops yet. RTX Spark/Arm needs Windows software compatibility, OEM scale, pricing, battery proof, enterprise support, and gaming/app optimisation before it can truly replace x86 broadly.  For Intel, the risk is bigger: its desktop moat is already under pressure from weak execution, AMD competition, and ARM momentum. For AMD, the threat is less existential because it has stronger x86 performance credibility and can still ride AI PCs with Ryzen + Radeon/NPU. DSX is separate but important. It strengthens NVIDIA’s AI infrastructure ecosystem by helping bu

If I already owned Micron, I would hold, but trim gradually rather than exit outright. The bullish case remains intact: HBM demand is still supply-constrained. AI servers require far more memory than traditional servers. Memory has evolved from a commodity into a strategic AI bottleneck. Micron is one of the few companies capable of supplying leading-edge HBM at scale. However, several warning signs are emerging: A trillion-dollar valuation leaves less room for execution mistakes. Expectations are becoming extremely optimistic. Capacity expansions from competitors could eventually reduce scarcity premiums. "AI stock outperforming NVIDIA" narratives often appear late in a cycle. History shows that semiconductor leaders rarely move in straight lines. Even during strong secular upcycles, 20-3

If forced to choose between the two today, I would favour the NVIDIA-ARM alliance over Intel. The reason is not simply AI hype. ARM benefits from a relatively attractive business model. Every successful ARM-based PC, server, smartphone or edge AI device potentially expands royalty revenue without ARM having to manufacture chips itself. NVIDIA's commitment gives ARM another avenue for growth beyond mobile and data centres. Intel, meanwhile, faces multiple challenges simultaneously: Pressure from ARM-based PCs. Competition from AMD in x86. Massive capital expenditure requirements for foundry ambitions. The need to prove its AI strategy can generate meaningful growth. That said, Intel is now becoming a classic turnaround story. If management executes well, manufacturing improves, and enterpri

Yes, Intel and AMD can still hold desktop dominance near term, but the moat is now clearly weaker. NVIDIA’s RTX Spark is not just a chip launch. It combines Arm CPU, Blackwell GPU, CUDA/RTX software, Windows AI agents, and OEM support from Dell, HP, Lenovo, Asus and others. That attacks the PC market through AI performance, not traditional CPU benchmarks.  But x86 still has major defences: app compatibility, enterprise fleets, gaming support, existing supply chains, and years of OEM optimisation. Most users still buy PCs for price, reliability and compatibility, not local AI agents. My view: Intel is more exposed than AMD. AMD has stronger execution and GPU/CPU credibility. Intel’s moat depends heavily on legacy x86 share, manufacturing recovery and enterprise stickiness. So, not an i

One of Singapore's strengths is how different cultures and faiths are celebrated side by side. Wishing everyone a meaningful Hari Raya Haji and Vesak Day. I'll be keeping things simple this long weekend: spending time with family, enjoying some good food, and taking the opportunity to recharge. Looking forward to exploring a few local spots as well. Have a wonderful and peaceful holiday, everyone!

NVIDIA's move is strategically important, but I would not declare the x86 moat broken yet. The bigger story is not the PC chip itself. It is NVIDIA extending its ecosystem from AI training to AI inference, robotics, digital twins and now client PCs. The new DSX vision complements platforms such as NVIDIA Omniverse by allowing companies to simulate AI factories before deploying real hardware. For PCs, NVIDIA faces three hurdles: Software compatibility: x86 still dominates enterprise Windows workloads. OEM relationships: Intel and Advanced Micro Devices have decades-long partnerships with PC makers. Enterprise inertia: Businesses refresh PCs slowly and value compatibility over cutting-edge AI features. However, NVIDIA's advantage is that AI PCs may shift the battleground from CPU performance