This Week’s Market Fixed-Point Observation (November 30, 2025)

Observation period: Sunday, November 23 – Saturday, November 29, 2025

【Overview】
This week, we reviewed global market liquidity, AI-bubble–related developments, and key investor sentiment indicators.
This report is designed as a weekly fixed-point observation to detect early warning signs of a bubble burst, by monitoring seven major categories on a recurring basis.

【Observation Categories】

1. Macro Financial Environment

  • U.S. 10-year Treasury yield / real yield (TIPS)

  • Balance sheet trends of major central banks (Fed, ECB, BoJ, etc.)

  • Policy rates and yield curve steepness/inversion

  • High-yield spread (HY OAS)

  • TED spread / liquidity indicators

  • Signs of policy shifts at major central banks

[Assessment] Stable

[Summary]
This week the U.S. 10-year yield traded around 4.0% (slightly down from the low 4.1% range the previous week), while the dollar index weakened, marking the worst weekly performance in about four months. This points to somewhat easier conditions on both the rates and FX fronts.
High-yield OAS remained around just above 3%, near the bottom of its year-to-date range, indicating that credit markets are still in a “risk-on” stance and that financial conditions are looser than policy rates alone would suggest.
On the Fed side, some major houses such as JPMorgan are now factoring in a 25bp rate cut at the December FOMC, but within the Fed itself both hawkish and dovish voices remain, so the rate-cut scenario is “strongly expected but far from guaranteed.”
There have been no reports of acute stress in TED spreads or dollar funding that would resemble a “pre–financial crisis” systemic event.
Taken together, macro financial conditions sit in a “moderately easy and broadly stable zone,” though this very looseness is also allowing leverage to build up, creating fertile ground for a potential bubble later on.

2. Capital Flows & Credit Conditions

  • FINRA margin debt (trend)

  • MMF balances and ETF flows (especially AI / tech)

  • HY and corporate bond issuance / credit spreads

  • Credit card debt and delinquency trends

  • Hedge fund leverage

[Assessment] Caution

[Summary]
FINRA margin debt reached a record high of about USD 1.18 trillion as of October, up roughly 40% year-on-year, a pattern reminiscent of the pre-dotcom and pre–Global Financial Crisis phases and widely read as a tactical “euphoria signal.”
Meanwhile, MMF balances have risen back to around USD 7.57 trillion, near all-time highs, meaning that capital is pouring into both high-yielding cash instruments and risk assets, reflecting a “barbell-style” risk-taking pattern.
HY spreads remain near the lower end of their year-to-date range, suggesting that credit risk is hardly priced in, even though history shows that periods of surging margin debt tend to be followed—often with a lag—by sharp widening in credit spreads.
AI and tech ETFs (QQQ, SOXX, etc.) have corrected from their October highs; this week saw only a mild rebound, with flows now reflecting a tug-of-war between “buying the dip” and “de-risking.”
Overall, the combination of elevated leverage and compressed credit spreads points to a “fragile stability” that could easily become a trigger point for a bubble unwind, hence the “Caution” assessment.

3. Corporate Earnings & Fundamentals

  • S&P 500 EPS growth / guidance revisions

  • Earnings trends for AI sector leaders (NVIDIA, TSMC, ASML, etc.)

  • CapEx and investment trends of hyperscalers

  • Gap between revenue growth and share-price performance

  • Changes in operating and net profit margins

[Assessment] Attention

[Summary]
Earnings season has largely passed, so new large surprises were limited this week. However, the latest numbers show that NVIDIA’s revenue and profit growth remain exceptionally strong, with data center–driven AI demand continuing to lead results (e.g., Q3 FY26 revenue up about 62% YoY).
Across the hyperscalers, AI infrastructure investment has continued, but some names—such as Google—have produced guidance that sparked talk of a potential CapEx peak-out, making the market highly sensitive to changes in the slope of AI-related CapEx.
On the macro side, U.S. manufacturing PMI fell to 51.9 in November, with rising inventories and slowing new orders suggesting future downside risks to production, while services PMI remained firm at 55, indicating that the slowdown is so far concentrated in manufacturing.
For the AI sector, prices have corrected somewhat but earnings have not yet collapsed, so we still have a backdrop of “high growth and high expectations,” and no obvious direct trigger for a collapse has appeared.
In short, fundamentals remain “broadly solid but with growing risk of a gap versus expectations,” so the category is rated “Attention.”

4. Sector Valuation

  • Relative performance: Semiconductor index (SOX) vs S&P 500

  • Valuation heat levels (PER, PBR, PSR), especially for AI names

  • Valuation gap between real-economy sectors (energy, utilities, materials) and tech

  • Capital flows into tech ETFs (QQQ, SMH, SOXX)

[Assessment] Caution

[Summary]
The SOX index and key semiconductor ETFs (e.g., SOXX) have pulled back from their October highs but still sit near the upper end of their post-COVID range; their relative performance vs the S&P 500 is approaching dotcom-era extremes.
Several research pieces highlight that AI-related stocks’ valuations (PSR, PER) have climbed above historical norms and exceed levels implied by aggregate measures such as the Buffett indicator, making “AI bubble concerns” a mainstream theme rather than a fringe view.
Meanwhile, real-economy sectors such as energy, materials, and utilities continue to trade at discounts relative to their earnings levels, so the gap between “AI premium” and old-economy sectors remains wide.
Recent corrections have compressed some multiples, but the market is increasingly aware of a structure where “expected growth rate > actual AI adoption pace,” which means that, from a valuation standpoint, the market is still very much in “bubble-like territory.”

(For reference, NVIDIA is widely viewed as a bellwether for the AI bubble; its price chart can be checked on your broker or financial data provider.)

5. Real Economy & Supply–Demand Data

  • Semiconductor shipments / inventory-to-sales ratio (SI ratio)

  • GPU and HBM price trends

  • Global manufacturing PMI / ISM indices

  • Export data for China and South Korea (electronics / semiconductors)

  • Power supply–demand and logistics bottlenecks

[Assessment] Attention

[Summary]
In the U.S., manufacturing PMI stands at 51.9—still in expansion territory but at a four-month low, with accumulating inventories and slowing new orders pointing to potential downside risks to production. By contrast, services PMI at 55 suggests the overall economy is still expanding moderately.
China’s manufacturing PMI remains around 49 and has been below 50 for eight consecutive months, underlining continued weakness in industrial activity and acting as a drag on the global manufacturing cycle.
In sharp contrast, South Korea’s exports in early November rose around 8% YoY, with semiconductors up more than 20%, confirming that AI-related high-tech demand is driving East Asian exports.
Public statistics on GPU/HBM spot prices and SI ratios are limited, and no decisive new data emerged this week. However, on the policy side, rising data-center electricity demand and the need to strengthen power grids continue to be a topic of debate.
Overall, “strong AI-related high-tech demand” vs “weakness in traditional manufacturing centered on China” is the key tension, leaving global growth in a mildly expansionary but uneven state.

6. Information & Sentiment

  • VIX (Volatility Index)

  • Put–call ratio

  • Bull–bear indicators

  • Fear & Greed Index

  • AAII Sentiment Survey

  • Google Trends for “AI,” “NVIDIA,” “generative AI”

  • IPO volume / first-day returns

  • Share of “Buy” ratings among analysts

[Assessment] Attention

[Summary]
The VIX spiked into the high 20s around November 20, but then fell back to the mid-16s by the end of this week, suggesting the peak in fear-driven volatility has passed, though the index remains above the ultra-low levels seen at the start of the year.
AAII sentiment shows bears still outnumber bulls (around 32% bullish vs 43% bearish), and the CNN Fear & Greed Index has dropped into 24 (Extreme Fear), indicating an environment of heightened anxiety.
At the same time, the S&P 500 and Nasdaq have only corrected about 4–7% from their October highs and are still viewed by many as richly valued; thus we are in a rare configuration where “prices remain high while sentiment is fearful.”
News flow and search trends related to AI remain elevated, with an increasing number of reports and articles focusing on “AI bubble” and “AI valuation risk,” which shows that the dominant market narrative is now “AI overheating and its sustainability.”
In short, sentiment has swung from euphoria to rapid fear and is now in a choppy post-shock adjustment phase—less like the intoxication that precedes a crash and more like an unstable cooling process.

7. Meta Elements (Narratives & Expectations)

  • Tone of policy and regulatory news around AI and tech

  • Shifts in investment themes (e.g., “generative AI → power → edge AI”)

  • Changes in tone of management and major investors

[Assessment] Caution

[Summary]
This week saw a surge of official reports and media pieces that explicitly address an “AI bubble” narrative: for example, the UK’s OBR (Office for Budget Responsibility) modelled scenarios in which an AI share-price collapse worsens fiscal dynamics, and multiple reports highlighted that AI valuations and the Buffett indicator are at historically elevated levels.
On the policy front, tensions rose over federal vs state authority in U.S. AI regulation, with many state attorneys general arguing that the federal government should not pre-empt state law; in Europe, phased-in and adjusted AI regulation is being discussed. The core political question is: “AI should be promoted—but who bears the risk?”
Geopolitically, reports that the U.S. government may allow NVIDIA’s H200 to be sold to China have fueled speculation about a partial tech détente in AI semiconductors, even as export-control risks remain.
In terms of investment themes, the story is shifting from “pure generative AI” to a broader ecosystem that includes power infrastructure, data-center REITs, networking equipment, and edge AI, while skepticism over the ROI of AI investments is gaining more voice. This makes the overarching narrative more complex and more polarized.
In sum, AI enthusiasm remains strong, but market participants are increasingly conscious that this may in fact be a bubble, which is why the meta-narrative category is rated “Caution.”

Overall Assessment

🟠 Caution (Structural Imbalance)

  • State of the AI Bubble
    While AI-related companies’ fundamentals remain strong, the combination of record-high margin debt, excessively tight HY spreads, and historically stretched valuation metrics suggests we are in a “fragile bubble” that has shifted from pure overheating into a correction phase. A clear collapse trigger has not yet emerged, but the structural disconnect is undeniable.

  • Global Market Stress Conditions
    On the surface, global markets look comfortable—10-year yields around 4%, a softer dollar, lower VIX, and tight HY spreads all point to low visible stress. However, leverage and valuations are piling up beneath the surface, leaving markets highly vulnerable if a shock hits.

Key Takeaways for This Week (3–5 items)

  1. Under the seemingly benign mix of “lower yields × weaker dollar × low HY spreads,” margin debt has reached new record highs, indicating that leverage-driven risk-taking has intensified further.

  2. The VIX has fallen sharply from the high 20s back to the mid-16s, while AAII and CNN Fear & Greed signal “fear,” creating the unusual combination of subdued price volatility with deteriorating investor psychology.

  3. South Korean semiconductor exports are growing more than 20% YoY, while China’s manufacturing PMI is expected to remain below 50 for an eighth consecutive month, highlighting a sharp bifurcation between “AI-driven high-tech growth” and “legacy manufacturing weakness.”

  4. Institutions such as the OBR are starting to quantify the fiscal and macro impact of a potential AI bubble burst, elevating AI from just an equity theme to a macro policy and fiscal risk issue.

  5. While the U.S. and EU move ahead with AI and digital regulation, the U.S. is also exploring the possibility of allowing NVIDIA’s H200 exports to China, turning the “AI promotion vs regulation and national-security” tug-of-war into a direct driver of volatility in AI-related stocks.

Events & Indicators to Watch Next Week (3–6 items)

  1. U.S. ISM Manufacturing Index / PMI (November)

    • Focus: Whether inventory buildup and weak new orders persist, or whether signs of bottoming emerge.

    • Implications:

      • Upside surprise → Hawkish / Risk-on (easing growth fears, reinforcing soft-landing narrative)

      • Downside surprise → Dovish / Risk-off (growth slowdown concerns, potential drag on AI hardware demand)

  2. Euro Area CPI & Unemployment (Flash November)

    • Focus: Whether Middle East risks and FX effects are translating into renewed inflation pressures.

    • Implications:

      • Disinflation with rising unemployment → Dovish / Risk-on (earlier ECB rate-cut expectations)

      • Sticky inflation with flat unemployment → Hawkish / Risk-off (rate cuts pushed out)

  3. China’s November PMIs (Official & Private)

    • Focus: Whether the economy can break out of an eight-month streak below 50, or weak momentum persists.

    • Implications:

      • Back above 50 / improvement → Risk-on (hopes for a bottoming in the global manufacturing cycle)

      • Further deterioration → Risk-off (headwind for commodities and cyclical stocks)

  4. Comments from NVIDIA and other AI-related firms at tech/AI conferences (e.g., UBS Global Tech & AI Conference)

    • Focus: The pace of AI infrastructure investment, commentary on customer ROI, and inventory conditions.

    • Implications:

      • Continued CapEx and strong demand tone → Risk-on (re-rating of AI valuations)

      • Slowing investment or emphasis on ROI uncertainty → Risk-off (renewed pressure on AI-bubble names)

  5. Fed Speakers & Expectations Ahead of the December 9–10 FOMC

    • Focus: How officials frame the likelihood of a December rate cut.

    • Implications:

      • Treating a cut as almost a done deal → Dovish / Risk-on (support for lower yields and weaker dollar)

      • Emphasizing “data dependence” and keeping cuts in doubt → Hawkish / Risk-off (rebound in long yields and the dollar)

  6. Movement in the CNN Fear & Greed Index and AAII Sentiment

    • Focus: Whether we see a rebound from Extreme Fear, or if anxiety becomes entrenched.

    • Implications:

      • Rebound in the indicators (less fear) → Risk-on (scope for a relief rally in tech and AI stocks)

      • Persistent Extreme Fear → Risk-off (continued outflows and valuation compression driven by sentiment)

【Tags】
Economic News / Market Observation / AI Bubble / Investment Analysis / Macro Trends

【Notes】
This report was prepared with the assistance of ChatGPT (GPT-5.1 Thinking).
While the content is based on publicly available information deemed reliable, some data rely on publication timing and media reports, so
there is a possibility of errors or inaccuracies.
Please make your final investment decisions at your own discretion and risk.

Comments

Post a Comment

Popular posts from this blog

Today’s Macro & Geopolitical News (2025-12-29)

United States With the year-end shortened week underway, focus shifts to the “FOMC minutes” and fine-tuning expectations for early-2026 cuts (markets are prone to think in terms of the “late stage” of the easing cycle). The policy rate is reported in the 3.50–3.75% range; the near-term tug-of-war remains between “how the slowdown unfolds” and the risk of inflation re-accelerating. In housing, pending home sales and house-price indicators tend to drive headlines; the key question is how much elevated rates continue to restrain demand. Equities, after strong year-to-date gains, can be dominated by rebalancing flows; thinner liquidity can amplify day-to-day swings. Europe For the covered date (overseas: 12/28), fresh headlines on major data/monetary policy are limited (news flow typically thins into year-end). Rates sit at elevated levels—Germany 10Y around 2.86% and UK 10Y around 4.51%—reflecting a tug-of-war between disinflation and fiscal/supply dynamics. With holiday-thin trading, bot...
Read more

Today’s Economic & Geopolitical News (2026-01-02)

United States Early-year positioning is again clustering around “AI + data centers,” with sovereign-style capital tending to favor U.S. assets tied to digital infrastructure and AI—supportive for U.S. risk assets at the margin. At the same time, diversification pressure remains a medium-term theme: newer pools of sovereign capital often originate in emerging markets, which can create periodic rotations away from a single-country concentration. Holiday-thinned liquidity can make price action more position-driven than news-driven; the market can swing between “Fed path pricing” and political headlines. In the near term, attention tends to tilt from “growth prints” to the combination of capital cost (long yields) and liquidity (risk appetite) . Europe Bulgaria’s entry into the euro area underscores continued expansion of the currency union: it can reinforce institutional stability, while reviving scrutiny on fiscal rules and cross-country divergence. As an integration milestone, it can a...
Read more

Today’s Economic & Geopolitical News (2025-12-31)

United States The Fed cut the policy rate by 25 bps to 3.50–3.75% at the December meeting, but the minutes highlighted a wider range of views—making further cuts in 2026 more likely to be gradual and data-selective. Some observers note that a government shutdown can disrupt the continuity of official statistics, pushing market attention toward early-year labor and inflation prints. Housing data suggest cooling: FHFA home prices (Oct) rose +1.7% YoY, implying that improving “price stability” could complement lower rates in supporting activity. Market implication: a tug-of-war between “insurance against slowdown” and “sticky inflation” likely persists; if long yields don’t fall cleanly, equity upside tends to become more selective. Europe European equities were supported by banks and resources, with the STOXX 600 pushing toward the 592 area—showing resilience even in thin year-end trading. The ECB remains centered on a 2.00% deposit rate (in a hold phase), balancing disinflation against ...
Read more