【大市睇路】11月先收油？下一個機會在哪？@jlawstock

Market review and outlook

The review uses market data through October 23. Broad-index performance was mixed, while sector results varied. Utilities, consumer-related stocks and energy were among the stronger areas discussed. Cybersecurity and software also reportedly outperformed, while some AI hardware and other technology names lagged.

At the time of the presentation, the speaker said the market had already made substantial gains—approximately 10% for the S&P 500 and close to 20% for technology stocks. These figures were approximate.

Upcoming earnings from Microsoft, Alphabet/Google, Amazon and Apple, along with results from NVIDIA and AMD, were identified as important tests of whether AI-related growth can meet high market expectations.

The speaker’s near-term stance was cautious, citing elevated expectations and uncertainty around elections, economic data and corporate results. They discussed adjusting positions or waiting for a possible correction rather than chasing recent gains. A possible 15% pullback was mentioned as a scenario, not a forecast.

Midterm-election seasonality and November events

A historical review covering 1986–2020 suggested that the fourth quarter of U.S. midterm-election years was positive in most of the periods examined. The presentation also showed generally positive average index returns over the 1-, 3-, 6- and 12-month periods after elections, with greater volatility around the election itself.

The speaker cautioned that the sample is limited and that election outcomes and economic conditions differ from cycle to cycle. Historical patterns are not a guarantee.

November watch items included the Federal Reserve meeting, U.S. GDP and labor-market data, and large-cap technology earnings. The speaker said markets appeared to expect no rate increase at the upcoming Fed meeting, while emphasizing the uncertainty.

The presentation also noted that U.S. daylight-saving time changes in November could affect market-session times for viewers outside the U.S.

Interest rates, bonds and equity valuations

The speaker argued that stock investors should monitor bond markets because Treasury yields affect discount rates, financing costs and the pricing of risk.

A simplified explanation of the 10-year U.S. Treasury yield, described as around 5%, was:

• Nominal yield ≈ real yield + expected inflation.
• Corporate-bond yields may also include additional compensation for credit/default risk and liquidity risk.

The presentation discussed how large corporate capital-spending plans and debt issuance—including AI-related investment—may increase financing demand and put pressure on yields.

Historical comparisons were used to challenge the idea that rising yields automatically mean falling stocks:

• In the longer-period sample beginning in 1970, equities reportedly rose in about 60% of months when bond yields rose.
• In a more recent sample described as extending to 2026, equities reportedly rose in nearly 75% of yield-rising months.
• The speaker contrasted these results with the high-inflation 1970s, when equities rose in fewer than half of the months with rising yields.

The main caution was to examine why yields are rising. Higher yields driven by growth may have different implications from yields rising because of inflation, fiscal stress or a broader risk shock.

The speaker compared bond yields with equity earnings yields and valuation multiples, arguing that bonds could look attractive relative to stocks if expected equity growth fails to materialize. They raised the possibility of shifting some long-term capital toward bonds, but did not present this as a blanket recommendation.

CoreWeave: AI infrastructure and leverage risks

CoreWeave (CRWV) was presented as a case study in the financing risks of AI infrastructure. Its business rents GPU computing capacity, including NVIDIA equipment, using debt-financed investment and long-term customer contracts.

The presentation described high customer concentration: the largest customer was said to account for about 70% of second-quarter revenue. Contracts were described as generally lasting two to three years, with cancellation possible after 90 days under some arrangements.

CoreWeave was contrasted with larger technology companies such as Amazon, Google and Microsoft. AI infrastructure is a much smaller part of those companies’ broader businesses, while CoreWeave is more concentrated and leveraged.

The speaker discussed project-specific SPVs (special-purpose vehicles) used to finance assets and contracts. This is a common structure in project finance, but it does not remove the underlying business and repayment risks.

Credit-market figures cited

A CoreWeave bond with a 9.625% coupon was said to be trading around $80 per $100 face value, implying a yield near 15% under the assumptions discussed. The presentation also referred to yields rising from roughly 9% to 16% and elevated CDS/default-risk pricing. These figures were presented as market indications, not certainty about an actual default.

The speaker said market pricing implied roughly a 10% chance of bankruptcy before the end of 2027, while stressing that such estimates depend on assumptions.

Risks and scenarios

Risks highlighted included:

• High leverage and interest expense
• Customer concentration
• Continued GPU spending
• The resale or residual value of older GPUs
• Whether AI demand and rental pricing can sustain projected growth

In a potentially favorable scenario, revenue and utilization continue to grow, financing costs ease and customer demand remains strong. The speaker said it could take two or three more quarters of results to judge the business’s cash generation and funding needs more clearly.

The speaker compared current AI infrastructure financing with the 2000 technology bubble and the 2008 financial crisis, but argued that the situations are not directly equivalent. The key concern was not only CoreWeave’s debt, but whether problems at AI infrastructure providers could spread to their customers or undermine confidence in AI investment.

Biotech and personalized cancer medicine

The presentation discussed Moderna (MRNA) and a personalized cancer-vaccine program developed with Merck. It described favorable clinical findings but said the market had expected especially strong results. The stock reportedly fell sharply when the reported benefit did not exceed those expectations.

The speaker emphasized that the clinical evidence, regulatory process and commercial prospects still need to be assessed. A possible approval timeline around 2027 was mentioned, but was not presented as assured.

Risks included high expectations already reflected in share prices, trial results that may not generalize across cancer types, manufacturing complexity and cost, and the possibility that regulators do not approve a treatment.

Cancer diagnostics and treatment ecosystem

The presentation organized the ecosystem into four stages:

1. Screening and sample collection, including blood-based testing
2. Genetic sequencing and tumor analysis to identify relevant mutations or biomarkers
3. Treatment development, including personalized vaccines and cell therapies
4. Post-treatment monitoring for residual disease or recurrence

Companies and assets mentioned included:

• Illumina (ILMN): Genetic-sequencing equipment and related tools
• Personalis (PSNL): Cancer genomics and monitoring-related work
• Twist Bioscience (TWST): DNA-synthesis and research tools
• Veeva Systems (VEEV): Software and data systems used in pharmaceutical development and clinical trials
• Tempus AI (TEM): Medical data and AI-related services
• Gilead Sciences (GILD): Established HIV business and oncology/cell-therapy programs
• ARK Genomic Revolution ETF (ARKG) and biotech ETFs generally

The speaker noted that some biotech and genomics companies had outperformed parts of the AI-hardware sector during the year, but stressed that individual biotech stocks can be highly volatile. Clinical-trial and regulatory outcomes can cause large, sudden price moves.

The speaker favored understanding the industry’s underlying drivers and suggested that ETFs can diversify company-specific risk. This was framed as a general approach, not a recommendation to buy a particular fund or stock.

Methodologies and frameworks shared

• Assess bond yields: Separate real yields from inflation expectations, then account for credit and liquidity premiums when evaluating corporate debt.
• Interpret rising yields in context: Compare periods and identify the cause of a yield move rather than assuming it is automatically bearish for equities.
• Stress-test a leveraged company: Examine debt costs, customer concentration, contract terms, collateral and asset resale value; then consider optimistic, neutral and adverse outcomes.
• Map biotech exposure across the value chain: Screening and sample collection → genomic analysis → treatment → post-treatment monitoring.
• Manage biotech risk: Consider diversified exposure through ETFs rather than relying entirely on a single company whose valuation may depend on trial or regulatory outcomes.

Disclosures and cautions

The video description states that the views and strategies are the presenter’s personal opinions, that the presenter is not a securities-licensed professional, and that the discussion is not an offer, inducement, invitation, investment recommendation or financial advice. Securities can rise or fall in value, including to zero, and investors may lose money.

The presentation also promoted the channel’s trading courses and membership program.

Presenters and sources

J Law introduced the video. The detailed sections were presented by an unnamed member of J Law’s internal research team. No specific external analyst or research publication is clearly identifiable in the subtitles; the discussion refers generally to historical market data, company results and bond-market pricing.