$Amazon.com(AMZN)$  I would be cautious about calling either a durable base or the cycle top based on one day's events. Amazon's move above US$3 trillion reflects improving confidence that AI investment is producing durable cloud demand. Multi-year AI infrastructure contracts improve revenue visibility, which justifies higher valuations if execution continues. Microsoft's and Google's gains reinforce that this is an industry-wide repricing rather than an Amazon-only story. Bezos' planned sale, however, should be interpreted carefully. Founders frequently sell under pre-arranged plans for diversification, liquidity, philanthropy, or tax planning. A multi-billion-dollar sale is not, by itself, evidence that management believes the stock has pea

The key variable is probably how much supply actually reaches the market, not simply that shares become eligible for sale. If SpaceX delivers strong results, retail investors may chase the earnings momentum. However, insiders and employees often have concentrated wealth tied to their employer. Even if they remain bullish, many will diversify once the lock-up permits it. That does not necessarily signal a lack of confidence. A roughly US$100 billion unlock is substantial. Even if only a small fraction is sold, it could temporarily outweigh incremental retail demand and create volatility. The sequence to watch is: Strong results + light insider selling: Bullish, suggesting confidence and limited supply pressure. Strong results + heavy insider selling: Stock could struggle despite good fundam

It is more likely the latter: two experienced investors managing risk, rather than definitive proof of an AI software inflection. Palantir's results were undeniably strong. Nearly doubling revenue, rapid commercial growth, and expanding profitability strengthen the case that enterprise AI spending is translating into real software revenue, not just infrastructure investment. Snap's results also suggest digital advertising remains resilient. However, Burry reducing a short is not necessarily a bullish call. It can simply reflect risk management after a sharp move. Likewise, Duan Yongping selling puts is a disciplined way to accumulate shares at lower effective prices, not an outright bet that the stock must surge. The bigger confirmation will come if other AI software firms also deliver acc

The narratives are not contradictory. They reflect different parts of the tech ecosystem. AI monetisation from OpenAI and Microsoft shows enterprise AI demand is becoming real revenue, supporting continued cloud and AI investment. That helps justify elevated valuations, but it cannot fully offset higher interest rates, which compress valuation multiples by raising discount rates. The key question remains whether AI revenue, margins and free cash flow are growing faster than AI spending. Qualcomm's cautious outlook also does not necessarily signal weakening iPhone demand. It reflects Apple shifting to in-house modem chips, higher memory costs and Qualcomm's changing revenue mix. Apple still expects solid iPhone growth, although supply constraints may limit near-term results. Bottom line: E

The rally looks like more than just a short squeeze, but it is probably not yet a clean trend restart. Samsung's strong profit rebound supports the idea that the memory cycle is improving, while Apple and Amazon both highlighting rising memory costs and tight supply suggests demand, especially for AI-related DRAM and HBM, remains robust. That is a fundamentally supportive signal. However, gains of 18% to 26% in a single session also point to heavy short covering after an extremely oversold period. Traders who had bet on a cycle peak were forced to unwind positions, magnifying the move. The key test now is whether: Memory contract prices continue rising over the next few months. HBM demand remains strong without weakness spreading to conventional DRAM and NAND. Upcoming earnings and guidanc

B looks the most likely. Inflation has moderated but remains above target, while the labour market is cooling gradually rather than deteriorating sharply. That combination gives the Fed room to stay patient instead of rushing to either hike or cut. I will be watching core PCE and non-farm payrolls most closely. If core inflation keeps easing without a sharp rise in unemployment, rate-cut expectations could return later, but not until there is stronger evidence. Higher-for-longer rates would continue to favour quality financials and cash-generative companies, while highly valued, rate-sensitive growth stocks and heavily leveraged sectors could remain under pressure.

$Apple(AAPL)$  Apple's warning matters because it comes from demand exceeding available components rather than weakening consumer demand. If memory constraints are temporary, investors may look through one or two quarters, especially given Apple's balance sheet and ability to secure supply ahead of rivals. However, if shortages persist, they can delay product shipments, pressure margins through higher procurement costs, and slow services growth by limiting new device sales. A 5% to 6% after-hours move suggests the market has already repriced much of the near-term risk. The next key question is whether suppliers such as SK Hynix, Micron and Samsung can expand advanced memory output quickly enough. If supply normalises, this is likely to be vie

There is a coherent argument for SK Hynix being the most resilient name, but it is not guaranteed to outperform. SK Hynix derives a larger share of earnings from high-bandwidth memory (HBM), where demand is still driven by AI accelerators and supply remains relatively tight. Micron and SanDisk have greater exposure to conventional DRAM or NAND pricing, which is more vulnerable if oversupply emerges. That helps explain why SK Hynix declined less. However, resilience is different from immunity. If hyperscaler AI spending slows, HBM demand weakens, or customers digest inventory, SK Hynix will also face pressure. CEO insider selling at Micron may affect sentiment, but it does not necessarily signal an industry peak on its own. The key indicators to watch are HBM pricing, AI capex from hypersca

Apple and Amazon could become the market's next directional catalyst, but they are unlikely to erase macro concerns on their own. If Apple demonstrates stronger-than-expected iPhone demand, services growth and a credible AI roadmap, it could reinforce the view that high-quality cash generators deserve premium valuations. If Amazon delivers accelerating AWS growth with evidence that AI investment is boosting cloud demand rather than just costs, it would strengthen the case that hyperscaler capex is producing tangible returns. The bigger issue remains the Fed. A more hawkish stance raises the discount rate applied to future earnings, which disproportionately pressures richly valued AI names. Strong earnings can offset that only if they clearly show accelerating cash flow and profitability. S

Microsoft's rally suggests investors are rewarding AI monetisation, not simply AI spending. Azure growth and management's confidence in competing with OpenAI and Anthropic signalled that Microsoft's infrastructure investments are translating into customer demand and revenue. Meta, despite a similar AI investment strategy, faced greater scrutiny because returns remain tied more to future advertising and AI execution. If this view persists, the market could further reprice the cloud and compute supply chain. Hyperscalers with visible AI revenue may command higher valuations, while key beneficiaries such as Nvidia, TSMC, Broadcom, AMD, networking, power, memory and data centre infrastructure providers could also see renewed support. The next test is whether upcoming earnings confirm that AI c

The selloff looks more like a cycle reset than a confirmed cycle peak. CXMT's DRAM expansion raises legitimate oversupply concerns, especially for commodity memory, but AI demand remains concentrated in high-bandwidth memory (HBM), where SK Hynix and Micron still hold technology advantages. The key question is whether new Chinese capacity can meaningfully pressure premium AI products or mainly lower-end DRAM. With SK Hynix reporting this week and SanDisk next week, guidance will matter more than headline earnings. If management confirms robust AI demand, HBM pricing and healthy customer commitments, this pullback could prove a mispricing. If they instead signal weakening pricing power and slowing orders, the market may conclude the memory upcycle has peaked. This earnings season should pro

D. The next leg of the tech market will likely depend on execution, not just hype. AI compute demand remains strong, but investors are increasingly asking whether massive capex can translate into sustainable earnings and free cash flow. Nvidia's ecosystem is still the benchmark, while Apple's earnings will show whether capital-efficient models continue to command a premium. DUV qualification and financing conditions will also influence how quickly new capacity comes online. Technical breakthroughs spark the cycle, compute scale enables it, but cash-flow quality ultimately determines who sustains premium valuations. The winners will be companies that can convert AI investment into durable profits, not just higher spending.

Palantir's rally appears to be early positioning rather than proof that earnings will exceed expectations. Investors are buying into three catalysts: continued U.S. government demand, expanding commercial AI adoption, and optimism that its Artificial Intelligence Platform (AIP) will drive faster revenue growth and margins. However, expectations are rising quickly. With valuation already rich by traditional metrics, even a strong quarter may not be enough if guidance merely matches consensus. The stock likely needs to deliver another "beat and raise" report, plus evidence that AI contracts are translating into sustained, recurring revenue. The bull case is that Palantir remains one of the few software companies monetising enterprise AI at scale, supporting premium valuations. The bear case

Tesla sits at the intersection of two very different valuation stories. If you believe the company will become a leader in autonomous driving, Optimus robots and AI infrastructure, today's weakness could look like a temporary mispricing. Those businesses, if they scale, would justify a valuation well beyond that of a traditional carmaker. The bearish case is that those future opportunities remain largely unproven, while the current business faces softer EV demand, pricing pressure, compressed margins and heavy capital spending that weighs on free cash flow. If the core automotive business continues to weaken faster than new businesses mature, the stock could remain under pressure. At current levels, Tesla looks more like a high-risk, long-duration growth investment than a conventional valu

I lean towards continued volatility first, with a reasonable chance of a rebound later in the week rather than a clean breakout immediately. The macro set-up is unusually binary. The Fed is still expected by most economists to hold rates steady, but markets are assigning roughly a 25% probability of a hike at this week's meeting. That is significant. Lower oil prices following the US-Iran pause are helping, with Brent dropping sharply towards US$90, which reduces one source of inflation pressure.  The bigger issue for QQQ may actually be Big Tech rather than the Fed. Microsoft, Meta, Amazon and Apple report this week, after Alphabet and Tesla were punished heavily despite growth because investors disliked their enormous AI-related spending. The market has therefore changed its questio

I lean overreaction, but with a genuine warning embedded in it. Friday’s decline was not isolated to SanDisk. Memory and AI-hardware names were hit broadly, with Micron, Western Digital and Seagate also falling sharply as investors reduced exposure to one of 2026’s strongest trades. That makes the 10%+ move look partly like sector deleveraging and profit-taking rather than evidence that SanDisk’s fundamentals suddenly deteriorated. The bull case remains substantial. NAND pricing and AI/data-centre storage demand have improved dramatically, while analysts remain broadly positive. Recent consensus data still show a Buy rating, although the enormous US$1,000 to US$3,250 target range illustrates how uncertain the valuation has become. Wedbush has actually raised its estimates ahead of the 5 Au

I would call Intel’s sell-off more of a valuation and execution reset than a disproof of the turnaround. Intel’s Q2 was genuinely strong: revenue reached about US$16.1 billion, up 25% YoY, adjusted EPS was US$0.42, and Q3 revenue guidance of US$15.8–16.8 billion comfortably exceeded the roughly US$15.1 billion consensus. AI-driven server demand and improving factory execution are therefore producing tangible results.  The problem is what investors must pay to get there. Intel raised 2026 capex to over US$20 billion, while the expensive 18A ramp remains a drag on margins. Investors are effectively saying: prove that higher spending eventually produces sustainably higher margins and profitable external foundry customers. The broader semiconductor sector also fell sharply on Friday amid

Yes, but the market is now demanding evidence that the spending produces cash returns. Alphabet’s underlying Q2 numbers were actually formidable: revenue +24%, Search +17%, Cloud +82% to US$24.8 billion, while Cloud operating margin reached roughly 35.6%. That suggests AI infrastructure is already generating meaningful revenue rather than being purely speculative spending.  The concern is the sheer scale. Alphabet lifted 2026 capex guidance to US$195–205 billion, while Q2 capex reached US$44.9 billion and free cash flow fell to negative US$5.9 billion.  I therefore see the 6.9% decline primarily as a valuation reset, not yet a thesis break. Alphabet has perhaps the strongest ability to absorb AI spending because Search, YouTube and Cloud provide enormous monetisation channels. Cl

I see this as more than a routine risk reset, but not yet a confirmed trend reversal. The 14.5% plunge reflects the market suddenly demanding evidence that Tesla’s AI valuation can eventually translate into cash flows. The numbers justify the concern. Q2 revenue reached a record $28.2B and deliveries rose strongly, but operating margin collapsed to 1.4%, versus the Street’s pre-report expectation of roughly 5.4%. Free cash flow was -$1.1B, largely because capex more than doubled sequentially. Tesla expects over $25B of capex in 2026, with spending continuing to rise as Robotaxi, Optimus, AI compute and manufacturing capacity expand.  The crucial distinction is that Tesla is not suffering from collapsing demand alone. It is deliberately sacrificing current profitability to finance busi