November 2025 Market Commentary

Markets in the month of November, much like the weather, were marked by volatile swings. November started with a selloff as markets wrestled with elevated uncertainty, especially around monetary-policy expectations for the Federal Reserve and concerns about stretched valuations in AI and tech stocks. However, markets staged a rally late in the month, recovering much of the losses. By month-end, breadth improved meaningfully. After falling over 4% at one point, the S&P 500 managed a slight gain of 0.13% by the end of the month. The Dow Jones Industrial Average, a price-weighted index comprised of 30 “blue-chip” companies often drawn from financials, industrials, consumer goods/services, and other more value- or cyclical-type sectors, ended the month with a gain of 0.32%, boosted by industrials. The technology-heavy Nasdaq 100 Composite was unable to recover from November’s mid-month drop, ending the month with a disappointing -1.5%, breaking the 7-month streak of gains.

More broadly, 8 out of 11 S&P 500 sectors ended the month in positive territory, underscoring the breadth of the recovery. The top-performing sector was health care, which rose 9.3% as investors looked for stability and value in a segment that had been a consistent laggard in 2025. The underperformers included technology, consumer discretionary, and Industrials, returning -4.4%, -2.4%, and -4.4%, respectively. The bottom three sectors all experienced headwinds. Technology suffered from AI-bubble woes and heightened valuations of mega-cap AI companies. Consumer Discretionary suffered from the lowest consumer confidence since April 2025. Finally, Industrials, one of the top-performing sectors year-to-date, experienced uncertainties surrounding a potential U.S. Supreme Court Ruling against President Trump’s authority to install tariffs.

Fixed Income markets showed positive signs with yields falling and yield curves steepening. U.S. government bonds had a positive month, as the drop in yields helped lift bond prices and fixed-income total return indices. For example, the 10-year U.S. Treasury yield declined over the month. One weekly report shows the 10-year yield moving from around 4.08% to 4.01%. The yield curve steepened modestly — shorter-term yields fell (or rose less), while long-term yields held firmer or declined more slowly. As a result, the bond market generally outperformed, as U.S. bonds were among the top fixed-income performers in November. Markets' increasing prices in potential rate cuts by the Federal Reserve at the December 10th FOMC meeting were driven by signs of softening in the labor market and weaker macro data, which ultimately boosted demand for Treasuries and pushed yields down. With interest-rate policy expected to ease, longer-duration bonds became more attractive, therefore supporting fixed-income valuations.

Additionally, supply in some parts of the credit market, including corporate, high-yield, and emerging markets, was heavy. This constrained spread tightening and made credit sectors underperform relative to Treasuries. AI capex using debt issuance came under scrutiny following reports of NVIDIA promising to buy any excess data center capacity at a premium, from hyperscalers and other pure-play data center providers. This artificially boosts the creditworthiness of these unprofitable companies, like CoreWeave, which are relying heavily on debt to fuel rapid expansion.

Declining yields are expected to have a 25bp rate cut in December. This is a modest steepening of the yield curve, as fixed-income is looking more attractive for income-oriented investors. Intermediate-/long-duration Treasuries and tax-exempt munis may offer good value, assuming further rate cuts or stable monetary conditions. Credit remains challenged with high supply, and cautious sentiment may limit near-term upside for riskier corporate bonds — at least until clarity on economic growth and rate path improves. Due to elevated issuance and fiscal deficits, long-term rates may stay volatile. Fixed-income investors should be mindful of duration risk, credit quality, and spread sensitivity.

The leading source of headlines and uncertainty this month stemmed from growing concerns about the control of capital expenditure tied to the AI buildout. The AI industry operates through a tightly interconnected ecosystem in which each participant plays a distinct role, while remaining highly dependent on the others. At the front of the supply chain is NVIDIA. This is the dominant designer of advanced graphics processing units (GPUs), which provide the raw computing power for training and running AI models. In 2025, NVIDIA controls roughly 92% to 94% of all global GPU shipments. This level of market share places it at the center of the industry’s growth narrative.

Importantly, NVIDIA does not operate large AI systems itself. Instead, NVIDIA supplies the hardware and software that power them. The firms that run the massive systems are the hyperscalers. These include Amazon (AWS), Microsoft (Azure), Google (GCP), and Meta, whose cloud platforms host the bulk of the world’s AI workloads. Hyperscalers purchase enormous volumes of NVIDIA chips and build software, networking, and developer-facing tools around them, effectively becoming the “factories” where AI models are trained, tuned, and deployed. Their diversified revenue streams also cushion business risk. For example, in Q3 of 2025, Amazon generated $67.4 billion from its e-commerce operations versus $33 billion from AWS (including both AI and non-AI software services). This demonstrated the breadth of their business models even as AI infrastructure becomes an increasingly important growth driver.

Supporting both upstream chipmakers and downstream cloud operators are the data center providers. Companies such as Equinix, Digital Realty, CoreWeave, and a range of colocation and specialized infrastructure operators. These firms do not design chips or run cloud platforms; instead, they build, lease, and maintain the specialized physical facilities required to house AI hardware. Modern data centers are capital-intensive projects, typically costing $600–$1,100 per square foot or $7–$12 million per megawatt (MW) of IT-ready power. Total construction budgets often reach into the hundreds of millions—or even billions—for large-scale deployments, with electrical infrastructure alone accounting for 40–45% of costs and cooling systems another 15–20%. AI-optimized facilities are even more expensive due to higher rack density, extreme power requirements, and advanced thermal systems needed to support GPU-heavy workloads.

One strategic adaptation has come from CoreWeave. This adaptation has accelerated its buildout by leasing or acquiring former bitcoin-mining facilities and converting them into GPU-dense AI data centers. Many crypto-mining operations already possess high-power electrical connections, cooling infrastructure, and racking layouts. As mining economics weakened, CoreWeave capitalized on these legacy assets—such as Galaxy Digital’s Helios campus and Core Scientific facilities—retrofitting them with NVIDIA-powered clusters, upgraded power and cooling systems, to meet AI performance requirements. Even with repurposing, the costs remain enormous, forcing data center operators to rely heavily on capital markets through a mix of equity and debt issuance.

To ensure adequate capacity and to secure its own growth trajectory, NVIDIA has increasingly struck creative agreements with data center operators. In September 2025, NVIDIA and CoreWeave signed a new order under their existing services contract, in which NVIDIA committed to purchase up to $6.3 billion of cloud-computing capacity from CoreWeave. If CoreWeave is unable to fully lease its GPU clusters to other customers, NVIDIA will step in and buy the unused portion—effectively guaranteeing utilization. While NVIDIA has no need for data center capacity, this structure provides CoreWeave with a demand backstop, improving its creditworthiness, lowering borrowing costs, and allowing it to continue building aggressively without fearing short-term fluctuations in AI demand.

At the application layer sits OpenAI, which occupies a central role. Developers and operators of advanced AI models do not manufacture chips or own hyperscale data centers. Instead, OpenAI depends on partners for computing capacity, primarily Microsoft Azure, which provides tens of thousands of NVIDIA GPUs under a multi-billion-dollar arrangement. When demand surges, OpenAI can also draw on specialized GPU cloud providers like CoreWeave. In essence, OpenAI builds the AI “software engines”, which are large-scale models, such as GPT-5 and Sora-2. It delivers them through APIs but relies heavily on NVIDIA for hardware, hyperscalers, and data center operators for the infrastructure that brings its models to life on a global scale.

In simple terms: NVIDIA makes the engines, hyperscalers assemble and operate the vehicles, and data center providers build the roads and garages that make the entire system run. OpenAI is the driver, determining how the system is used and where it goes. Each plays a separate role, yet no one can scale without the others. This underscores why AI has become one of the most capital-intensive and collaborative sectors in today’s economy.

Finally, the U.S. government’s growing involvement in the AI economy has also raised questions about how much of the sector’s expansion the private market can finance on its own. In October, the federal government acquired a 10% equity stake in Intel, followed by a $1 billion Department of Energy partnership with AMD to build two advanced supercomputers for research in nuclear energy, oncology, and national security. With nearly 40% of U.S. GDP growth year-to-date attributed to AI-related investment, the trajectory of AI infrastructure spending—and the sustainability of its financing—remains one of the most critical macroeconomic trends to watch in the years ahead.

With the federal government shutdown halting all major economic releases for October, investors were left without official updates on employment or inflation, with the most recent data still dating back to September 2025. The Bureau of Labor Statistics confirmed it will not retroactively publish the missing October reports, creating an unusual information gap for analysts. As a result, markets increasingly leaned on alternative indicators throughout November, including real-time “nowcasting” models such as those from the Federal Reserve Bank of Cleveland. These estimates pointed to inflation remaining elevated, with headline PCE and CPI running in the 2.9%–3.0% year-over-year range. Core inflation was projected to hold steady as well, underscoring persistent underlying price pressures and highlighting that disinflation progress has slowed even as it has improved meaningfully compared to earlier in the year. Despite tariffs adding upward pressure to prices, inflation has nonetheless moved substantially closer to the Federal Reserve’s 2% target over the course of 2025, though not yet enough for policymakers to declare victory.

While official inflation and labor data for October remained unavailable due to the federal government shutdown, we did receive an important gauge of consumer conditions through the University of Michigan Consumer Sentiment survey. The headline sentiment index declined modestly to 53.3 in October from 55.1 in September. Although the drop was small, the reading remains well below year-ago levels. This underscores how much consumer confidence has weakened over the past twelve months. Survey commentary highlighted that inflation continues to weigh heavily on household outlooks, while concerns about the labor market and overall economic prospects persist. Notably, inflation expectations ticked up slightly, indicating that many consumers still anticipate elevated prices in the near term.

For markets, this matters because consumer sentiment often serves as an early signal for near-term consumer spending, which is central to U.S. economic growth. The relative stability of the index suggests consumption is unlikely to collapse, but the subdued level—and lingering anxiety about prices—argues against a strong rebound in spending anytime soon. Taken together, October’s reading reinforces a picture of an economy that is holding steady but not accelerating. Hence, resilient enough to avoid a sharp downturn, yet not strong enough to dispel downside risks.

Looking ahead, December’s Market Commentary should include a more complete set of economic data, as post-shutdown releases catch up with their usual schedule. The key uncertainty now lies with the Federal Reserve’s December FOMC meeting. More importantly, whether the absence of October data proves significant enough to influence the expected 25 bp rate cut, or whether policymakers proceed based on broader trends rather than missing monthly inputs.

As we move into December and approach year-end, several major economic and policy-driven events stand poised to shape market direction. This is particularly through their impact on monetary-policy expectations, growth sentiment, and overall risk appetite. The upcoming release of the core Personal Consumption Expenditures (PCE) index, the Federal Reserve’s preferred inflation gauge, will almost certainly draw outsized attention. A softer reading could reinforce expectations for a near-term rate cut, while a hotter-than-expected print may stall easing hopes and introduce renewed volatility. With official data releases still catching up after recent disruptions, a backlog of key reports — including ISM and PMI surveys, labor-market data such as ADP employment and weekly jobless claims, and several consumer-sentiment indicators — will roll out over the coming weeks. Together, these data points will help clarify whether the economy is maintaining its resilience or slipping into a slower growth trajectory.

Markets are increasingly pricing in a December rate cut by the Federal Reserve, with trader-implied probabilities now well above 80%. Should the Federal Reserve deliver, or signal a clear and confident path toward easing, risk assets may find support into year-end. Conversely, any hawkish surprises or even a more cautious tone could temper investor enthusiasm at a time when economic visibility remains somewhat clouded.

Given current pricing, macro conditions, and updated earnings expectations, Wall Street forecasts generally center around the S&P 500, as it is finishing the year near the 7,000 level. This suggests modest single-digit gains for December, which is not quite the classic “Santa Claus Rally” often highlighted by analysts, but still a constructive finish given ongoing labor-market uncertainty and historically low consumer confidence. The “Santa Claus Rally” refers to the tendency for markets to rise during the last five trading days of the year and the first two of the new year, a pattern identified by Yale Hirsch in 1972. Since 1950, the S&P 500 has averaged a 1.3% gain over these seven days, with positive returns nearly 80% of the time. While often attributed to behavioral dynamics rather than fundamentals, several structural factors help explain the trend. This includes lighter institutional trading as many professionals take holiday leave, increased participation from retail investors, potential reinvestment of year-end bonuses, and capital flows from tax-loss harvesting being reversed once the calendar turns.

Regardless of seasonal tendencies, we remain firmly data-dependent. As the pace of economic recovery slows and uncertainty around the growth outlook persists, our focus remains on monitoring incoming indicators and positioning portfolios to capture opportunities, while managing the risks that 2026 may bring. Third-quarter earnings were exceptionally strong, with S&P 500 earnings rising 13.5% year-over-year, which was well above the 7.9% expectation entering the season. Should the anticipated 7.5% earnings growth materialize in the fourth quarter, total 2025 earnings growth would reach a solid 12%. Early Wall Street estimates for 2026 project earnings growth of roughly 14.2%, with mid-year S&P 500 price targets in the 7,300 to 7,500 range under a more bullish scenario, in which monetary easing continues, and earnings momentum—along with the benefits of AI adoption—broadens beyond the technology sector.

Notably, November marked the first month since July in which the equal-weight S&P 500 outperformed its cap-weighted counterpart. This development supports our cautiously optimistic view that the advantages of AI investment, productivity gains, and modernization efforts may begin to diffuse across a wider set of industries and market caps, rather than remaining concentrated among mega-cap technology names.

The AI buildout remains the dominant theme in equity markets, and we continue to emphasize exposure to the “picks and shovels” of this transformation. Utilities, electrical-equipment manufacturers, critical-materials suppliers, and selected hyperscalers are well-positioned to benefit and may present compelling opportunities during market pullbacks. Additionally, the Federal Reserve’s conclusion of its quantitative-tightening program on December 1 and its shift into a more neutral policy stance—paired with expectations of lower rates—enhance the outlook for traditionally rate-sensitive and debt-heavy sectors such as Real Estate.

** This commentary is provided for informational and educational purposes only and should not be construed as investment advice, an offer, or a solicitation to buy or sell any security. The views expressed are based on current market conditions and are subject to change without notice. Past performance is not indicative of future results. Index performance is shown for illustrative purposes only; indices are unmanaged, do not reflect fees or expenses, and are not directly investable. Any forward-looking statements are not guarantees of future performance and involve risks, uncertainties, and assumptions. Investors should consult with a qualified financial professional before making any investment decisions.

***Some or all portions of this commentary were prepared with the assistance of artificial intelligence (“AI”) and large language models. The information generated by these tools has not been independently verified, and while reasonable care has been taken to ensure accuracy, Stonemark Wealth Management does not guarantee the completeness or reliability of any AI-assisted content. This material is provided for informational purposes only and should not be construed as individualized investment advice.