Nasdaq Hits New High, Igniting Expectations for U.S. Earnings Season! Citi Predicts Nearly 90% of Tech Stocks to Deliver "Earnings Surprises," with Nvidia and AMD Leading the Beat Estimates List

Title context: Nasdaq Hits New High, Igniting Expectations for U.S. Earnings Season! Citi Predicts Nearly 90% of Tech Stocks to Deliver "Earnings Surprises," with Nvidia and AMD Leading the Beat Estimates List Text: After the Nasdaq Composite Index surged to a record high in one move, and after Nvidia, the "AI chip superpower" that carries significant weight in both the Nasdaq and the S&P 500, hit a record high, the U.S. third-quarter earnings season is about to kick off in a big way. Global investors' enthusiasm is heating up once again for U.S. corporate revenue and earnings to beat expectations, especially for tech giants closely tied to AI computing power to exceed analysts' consensus estimates on both revenue and profit. On October 6, the Nasdaq Composite Index, which covers the world's hottest tech stocks including Nvidia, AMD, and Micron, rose 0.45% to 27,599.79 points, setting a record closing high in tandem with the benchmark S&P 500, with the Nasdaq even hitting an all-time intraday high; the Philadelphia Semiconductor Index, known as the "global barometer for AI computing power and the semiconductor sector," rebounded about 4.7% cumulatively from the end of September, getting closer to setting a record high. Although it pulled back the next day, global stock market focus has gradually shifted to whether the core performance data of U.S. tech companies during earnings season can beat expectations and support further bull market gains above the new highs. The latest quantitative earnings-season research from Wall Street financial giant Citigroup provides concrete evidence for this earnings-driven thesis (i.e., tech companies leading S&P and Nasdaq constituents in beating revenue and earnings expectations): Citigroup's proprietary model predicts that 66.2% of Russell 1000 constituents will post positive earnings surprises (i.e., earnings exceeding consensus expectations) and related positive stock price return trajectories (i.e., Citigroup's model also predicts a positive direction for related stock price returns), significantly above the already-strong 60.9% last quarter and reaching the highest level since the fourth quarter of 2021; the information technology, healthcare, and industrials sectors in the U.S. market lead with 88.1%, 75.9%, and 73.8% respectively in terms of the proportion of positive earnings surprise predictions and positive stock price returns. Citigroup also emphasized that investment opportunities are further skewed toward large-cap companies focused on AI computing power infrastructure themes such as Nvidia, AMD, Intel, Texas Instruments, Applied Materials, and KLA, while the next round of bull market rotation is expected to spread first to healthcare and industrials. Meanwhile, Citigroup's strategist team said that Wall Street analysts' consensus estimate for S&P 500 third-quarter earnings growth has been significantly revised up from 26.7% at the start of the quarter to 29.6%, with energy and AI technology themes being the main sources of incremental upward revisions, and all 11 sectors are expected to achieve year-over-year earnings growth. Citigroup also combined earnings forecasts with fundamental stock ratings and long-position crowding to screen out candidates such as Nvidia, AMD, Intel, Applied Materials, and KLA that have both "positive earnings-season surprise predictions" and the most optimistic "Buy" stock rating, and noted that healthcare and industrials also offer broad opportunities for earnings delivery. The most investment-significant judgment in Citigroup's latest research report is that overall U.S. market earnings support remains concentrated in a handful of tech leaders closely tied to the AI computing infrastructure boom, but opportunities to beat market expectations are rapidly spreading and penetrating into healthcare and industrials; the benchmark for excess alpha performance in the next phase will depend more on specific earnings growth and the degree of delivery, how much optimism investors had already priced in, and whether bullish positioning has become overly crowded. The AI super bull market currently sweeping global stock markets appears to be looking for an "earnings baton-passing window" that is, when valuation compression driven by 10-year Treasury yields repeatedly hitting multi-decade highs causes the denominator side of DCF to expand, as long as earnings per share continue to grow, the stock market does not need to rely on P/E ratios rising back to elevated levels to advance; if interest rate/Treasury yield pressure subsequently eases, valuation stability could add further upside room. Another Wall Street financial giant, JPMorgan, said its strategist team favors repositioning opportunities created by easing position crowding, falling valuations, and earnings resilience, with a particular preference for semiconductors. According to the institution's latest research report, since June, forward 12-month EPS forecasts for semiconductors have been revised up by about 30% to 40%, while global software sector earnings expectations have lacked corresponding improvement; the institution's latest cited capex outlook for hyperscale cloud providers is approximately $950 billion in 2026, about $1.4 trillion in 2027, and at least about $3 trillion by 2030, and it expects AI-related revenue growth to show explosive expansion in 2027. The most core change brought by the recently global-viral Meta Muse AI agent and the "AGI-comparable" OpenAI Astra large model/AI agent can be described as a single user instruction triggering sustained, multi-stage computational work. A research, coding, or office task may sequentially include planning, retrieval, reading documents, calling tools, executing code, checking results, and correcting errors, with multiple steps requiring repeated model calls, and complex tasks may also employ parallel exploration and verification. Such nearly endless and increasingly complex AI workloads will accelerate the transmission of growth opportunities to the complete AI inference workload-focused computing system beyond GPUs. From the perspective of AI inference system architecture, more complex AI tasks led by Meta Muse often include longer contexts, multi-turn model calls, tool execution, and result verification: prefill needs to process input, decode continuously generates output, and the key-value cache (KV Cache) consumes more memory as context and concurrency scale expand, requiring coordinated improvement in compute throughput, memory bandwidth, and capacity. Therefore, as frontier agents such as Meta Muse further ignite AI computing power demand, the core inference for the AI computing power industry is that GPUs and TPUs handle model computation, high-performance data center CPUs handle tool execution and task orchestration, and HBM, server DRAM, storage, high-performance network infrastructure, and data center optical interconnect components jointly support efficient data transport and state management; ultimately, a complete and increasingly massive AI computing power server cluster will be needed to deliver continuously running services. After the Nasdaq Surges to a New High: Tech Leads Citigroup's "Earnings Surprise" List, Healthcare and Industrials Poised to Join the Super Bull Market Rotation Citigroup's strategist team said the positive signal for U.S. earnings season is reflected in the continued expansion of the "positive forecast" list, while the advantages of large-cap companies such as Nvidia, AMD, and Broadcom are further strengthened. Citigroup expects the proportion of Russell 1000 companies with positive earnings surprise and positive return candidates to rise from 60.9% to 66.2%, a significant increase of 5.3 percentage points; divided into five groups by market cap, the largest-cap group's positive forecast proportion reaches 77.0%, followed by 68.3%, 66.7%, and 60.7%, while the smallest-cap group has only 46.6%, down 3.4 percentage points from last quarter. From a sector change perspective, healthcare's positive forecast proportion increased by about 18 percentage points from last quarter, materials and consumer discretionary each rose about 10 percentage points, and industrials rose about 9 percentage points; consumer staples and utilities each fell about 6 percentage points, with utilities' negative forecast proportion reaching 72.2%. Citigroup said these model calculations mean the market is forming an earnings structure of "tech maintaining its lead, healthcare and industrials improving, and large companies prevailing," rather than all sectors strengthening simultaneously. Therefore, for market-cap-weighted benchmark indices, large-cap earnings delivery has a stronger index-driving effect; improvements in healthcare and industrials provide a crucial foundation for upward momentum to expand from a handful of AI computing power-related tech companies to more industries. The earnings growth rate and ability to beat expectations of U.S.-listed companies jointly determine the investment appeal of earnings season, and the contrast between healthcare and energy sectors is especially instructive for global stock market investment strategy. The expected S&P 500 third-quarter earnings growth is 29.6%, up 2.9 percentage points from the start of the quarter; among this, energy earnings growth expectations were revised up from 79.3% to 118.6%, tech sector earnings growth was revised up from 57.1% to 65.0%, communication services is expected to grow 51.5%, and industrials 14.5%. However, Citigroup's forecast data and model calculations show that healthcare sector earnings growth expectations were revised down from 8.7% to 5.1%, yet its positive surprise candidate proportion reached 75.9%; energy, despite having the highest expected earnings growth, has a positive surprise candidate proportion of only 48.1%. The underlying logic of Citigroup's quantitative model is that stock price reaction depends on the gap between actual results and prior expectations: after expectations were revised down, healthcare is more likely to see companies whose fundamentals perform better than the market previously judged; energy needs to clear an already significantly raised earnings bar. Citigroup's 12-month Wall Street top analyst earnings expectation revision indicator for the entire U.S. market remains near the year-to-date high set in early July, indicating that analysts' upward earnings revision momentum remains strong, but the third-quarter 29.6% growth expectation is significantly below the 52.4% already reported for the second quarter. Therefore, the positive change this quarter is mainly upward earnings expectation revisions and broader positive surprise coverage. Citigroup's model seeks to find the combination of "results better than expected and able to stock price gains." The report uses standardized unexpected earnings SUE = the difference between actual quarterly EPS and the mean analyst consensus estimate, divided by the standard deviation of analyst forecasts to measure earnings surprises, and notes that in historical samples about 38% of companies have shown earnings surprise direction opposite to stock price return direction, which explains why "earnings beating expectations" can still be accompanied by stock price declines. To improve the targeting of earnings-event screening, Citigroup uses logistic regression, calibrating the model with data from the past 24 quarters, incorporating the past four quarters' SUE, the surprise direction of the prior two quarters, nine-week earnings expectation momentum, total return relative to the market over the past 20 trading days, analyst consensus ratings and nine-week rating changes, and excluding specific cases where earnings surprise and stock price reaction directions are opposite. Citigroup's strategist team investment logic is to simultaneously seek continuity in earnings performance, improvement in analyst expectations, and market price confirmation. Therefore, the 88.1% figure for the tech sector specifically measures the proportion of companies within the sector classified by the model as "positive forecast" candidates, reflecting the breadth of positive signal coverage. From Strong AI Chip Orders to Sector Relay: Seeking the "Expectation Gap Dividend" of Earnings Season It is understood that Citigroup's "positive candidate" list shows that the semiconductor sector's positive list covers the most core areas of the AI computing infrastructure boom namely AI GPU/AI ASIC and data center CPU AI computing chips and also covers analog and power management, manufacturing equipment, testing, and materials, providing a complete semiconductor supply chain perspective for observing strong tech sector earnings beats. Nvidia, AMD, Intel, Texas Instruments, Analog Devices, Monolithic Power Systems, Microchip Technology, Applied Materials, KLA, Teradyne, Entegris, and MKS all simultaneously satisfy "model-predicted positive earnings surprise and positive return" and Citigroup's fundamental "Buy" rating. From an engineering and industry logic perspective, these companies respectively correspond to the most core computing power supply, chip power supply and signal processing, semiconductor manufacturing processes and control including etching and thin-film deposition, advanced semiconductor packaging and testing, and material purity: AI computing infrastructure must be converted into usable computing power, requiring the entire supply chain to jointly complete delivery, while different companies' earnings elasticity depends on product mix, capacity utilization, and cost structure. This positive forecast list compiled by Citigroup, covering multiple U.S.-listed companies, also includes chip companies serving traditional industrials and other end markets, thus providing significant candidates for observing the semiconductor upcycle's substantial expansion into more segmented investment sectors/areas. Related positive candidates also include optical communications companies namely Coherent and Lumentum, leaders in the U.S. market's data center optical interconnect supply chain as well as Western Digital, a leading U.S. storage device company, and software companies such as Cadence and ServiceNow. The prices provided in the report are all as of October 6, 2026, without specific target prices or expected gains. The significance of healthcare and broader industrials lies in providing earnings surprises and positive investment returns from different sources for earnings season, extending investment opportunities from the hot computing power trade to broader corporate operational improvements. Citigroup's strategist team said healthcare positive buy candidates include Johnson & Johnson, Abbott, Intuitive Surgical, Danaher, Thermo Fisher, Dexcom, Edwards Lifesciences, and Medtronic, covering pharmaceuticals, medical devices, surgical Siasun Robot&Automation, diagnostics, and life science tools; industrial candidates include GE Aerospace, RTX, Caterpillar, as well as Vertiv, Eaton, Quanta Services, Carrier, Emerson, and Rockwell Automation. Derived from a supply chain perspective, companies such as Vertiv and Eaton provide specific targets for observing the global AI computing infrastructure investment landscape's acceleration and transmission into power supply and distribution, thermal management, and data center infrastructure revenue; aerospace, machinery, and automation companies represent other earnings sources in the industrials sector. Tech and industrials therefore have business links through AI infrastructure investment, while healthcare provides an earnings storyline driven by its own operating performance and expectation adjustments. Citigroup's model calculations show that some healthcare candidates also have significantly lower crowding than hot computing power stocks, for example Medtronic at 0.170, Intuitive Surgical at 0.266, and Dexcom at 0.287, while Vertiv reaches 0.970 and GE Aerospace reaches 0.980. This makes the healthcare sector's combination of "significantly increased positive forecasts + lower trading crowding for some companies" an incremental opportunity worth watching in Citigroup's research report. Another truly substantive "positive" stock selection difference in earnings season also depends on whether good news has already been fully traded, so Citigroup places crowding at the same importance as earnings forecasts. Among the Magnificent Seven U.S. tech giants, Google parent Alphabet, Microsoft, Apple, and Nvidia received positive model-calculated signals, while Tesla, Meta, and Amazon received negative signals; among them, Meta and Amazon still have Buy fundamental ratings, reflecting the stark difference between Citigroup's model-calculated short-term earnings judgments and medium-to-long-term investment ratings. Meta's long crowding is 0.955, Nvidia's is 0.876, Microsoft's is 0.684, and Apple's is 0.671; Microsoft's crowding rose from the start of the quarter, while Tesla's fell significantly. Citigroup accordingly emphasized that companies with more crowded trades are more vulnerable to concentrated adjustment pressure if negative surprises occur; low-crowding companies that deliver major positive earnings surprises may attract more new investors. Citigroup's strategist team therefore specifically listed low-crowding positive buy candidates such as GoDaddy, BD, Leidos, and Procter & Gamble, as well as high-crowding negative neutral or sell candidates such as Phillips 66, Vornado, and PPL; the subsequent negative list also includes neutral-rated Qualcomm and AbbVie, as well as sell/high-risk-rated Moderna and The Trade Desk. The earnings-season investment path given by Citigroup's research is centered on the strategy of strong earnings beats from AI computing power-related tech leaders, while seeking new return sources from healthcare, industrials, and low-crowding positive candidates, and using an expectation-gap benchmark strategy to differentiate opportunities among different companies within the same industry/sector.