Nasdaq Hits New High, Igniting Earnings Season Expectations! Citi Predicts Nearly 90% of Tech Stocks Will Deliver “Earnings Surprises”, with Nvidia and AMD Leading the Outperformance List

According to Zhitong Finance APP, following the Nasdaq Composite Index reaching a historic high and the “AI chip superpower” Nvidia, which holds heavy weights in the Nasdaq and S&P 500 Index, hitting an all-time high, the U.S. stock Q3 earnings season is about to kick off. Global investors' enthusiasm for U.S. corporate revenue and profit exceeding expectations, especially revenue and profit of technology giants closely linked to AI computing power surpassing consensus analyst expectations, is heating up once again.

On October 6, the Nasdaq Composite Index, which covers the world’s hottest tech stocks like Nvidia, AMD, and Micron, rose 0.45% to 27,599.79 points, hitting a new all-time closing high in sync with the S&P 500 Index—a benchmark for U.S. equities. In intraday trading, the Nasdaq also set a new historical high; the Philadelphia Semiconductor Index, seen as a barometer for global AI computing and semiconductor sectors, has rallied approximately 4.7% since the end of September and is edging ever closer to its historical peak. Although there was a pullback the next day, the global equity market's focus has clearly shifted to U.S. tech companies' core earnings data during earnings season—whether exceeding expectations can further support a bull market above historic highs.

Wall Street banking giant Citigroup's latest quantitative research for earnings season provides concrete support for this profit narrative (i.e., tech companies leading S&P and Nasdaq constituents to revenue and profit outperformance): Citi’s proprietary model predicts that 66.2% of Russell 1000 constituents will post positive earnings surprises (i.e., profits beating market consensus) and corresponding positive share price returns (i.e., Citi’s model also predicts positive directionality for these stocks)—significantly higher than the already robust 60.9% last quarter, and the highest since Q4 2021. In the U.S. market, the information technology, healthcare, and industrial sectors lead with positive earnings surprise and positive price return estimates of 88.1%, 75.9%, and 73.8% respectively. Citi emphasizes investment opportunities increasingly favor large-cap companies focused on AI infrastructure themes such as Nvidia, AMD, Intel, Texas Instruments, Applied Materials, and KLA. The next bull market rotation is likely to first spread to healthcare and industrials.

Meanwhile, Citigroup’s strategists note that the consensus estimate for S&P 500 EPS growth in Q3 has been significantly revised up from 26.7% at the beginning of the quarter to 29.6%, with energy and AI tech being the main sources of upward revision; all 11 sectors are expected to achieve year-over-year profit growth. Citi has also combined earnings forecasts with stock fundamental ratings and “long trade crowding” to screen for “positive earnings surprises + ‘buy’” candidates including Nvidia, AMD, Intel, Applied Materials, KLA, etc., noting that there are also broad performance opportunities in healthcare and industrials.

The most investment-significant conclusion of Citigroup’s latest report is that overall profit support in the U.S. equity market remains concentrated in a handful of tech leaders deeply tied to the AI infrastructure boom, but opportunities to exceed market expectations are rapidly expanding to the healthcare and industrial sectors; the next stage of excess alpha will be determined more by the extent of actual profit delivery, what investors had already priced in, and whether market bullish positioning is overcrowded.

The current AI-driven global equity bull market appears to be seeking a “profit handoff window”—that is, as 10Y U.S. Treasury yields repeatedly reach 20-year highs and drive valuation compression (DCF denominator expansion), as long as EPS continues to rise, the stock market does not need to rely on P/E ratios rebounding to new highs to support gains; if rate/yield pressure later eases, valuation stability may boost upside potential again.

Another Wall Street giant, JPMorgan, states that its strategist team favors the reset opportunity shaped by de-crowding, falling valuations, and profit resilience, with a particular preference for semiconductors. According to the institution’s latest report, since June, semiconductor’s next-12-month EPS forecasts were raised ~30-40%, while global software sector expectations saw no similar improvement; its latest reference for capital expenditure forecasts from hyperscale cloud providers is: about $950 billion in 2026, ~$1.4 trillion in 2027, at least ~$3 trillion by 2030, and AI-driven revenue growth will explode by 2027.

The recent global sensation—Meta’s Muse AI agents and OpenAI’s Astra large model/AI agent, on par with AGI—usher the key shift where one user command can trigger a sustained, multi-stage computational workflow. A research, programming, or office task may involve planning, retrieval, document reading, tool invocation, code execution, result checking, and error correcting, with multiple steps requiring renewed model invocation. Complex tasks may involve parallel exploration and verification, and such nearly endless and escalating AI workloads will accelerate growth opportunities toward the full AI inference load-focused compute system, extending beyond the GPU itself.

From the perspective of AI inference system architecture, more complex Meta Muse-led AI tasks often involve longer context, multi-round model calls, tool execution, and result verification: prefill processes inputs, decode continuously generates outputs, and KV cache grows in memory footprint with increasing context and concurrency. Compute throughput, memory bandwidth, and capacity need collaborative upgrades. Therefore, as Muse and similar agents further ignite AI compute demand, for core AI compute industry logic, GPUs and TPUs handle model computation, high-performance datacenter CPUs run tool execution and workflow management, HBM, server DRAM, storage, high-speed network infrastructure, and optical interconnects jointly ensure data movement and state management; ultimately, a full and ever-larger cluster of AI compute servers is needed for sustainable service delivery.

After Nasdaq’s Historic High: Tech Leads Citi’s “Earnings Surprise” Big List, Healthcare and Industrials Set to Join the Super Bull Market Rotation

Citi’s strategist team notes that the positive signal of earnings season is in the consistently broadening “positive prediction” list, with large-caps such as Nvidia, AMD, and Broadcom further strengthening their advantage. Citi expects the share of positive earnings surprise and positive return candidates among Russell 1000 companies to rise from 60.9% to 66.2%, a clear 5.3pp increase; split by market cap quintile, the top quintile’s positive forecast share hits 77.0%, followed by 68.3%, 66.7%, 60.7%, and just 46.6% for the smallest, which fell 3.4pp vs. last quarter.

Sector-wise, healthcare’s positive forecast share is up ~18pp from last quarter, materials and consumer discretionary each up ~10pp, industrials up ~9pp; consumer staples and utilities are down ~6pp each, with negative forecasts for utilities as high as 72.2%. Citi observes that these model data suggest the market is formulating an earnings mix of ‘tech remains in the lead, healthcare and industrials improve, large caps outperform’, not a uniform sector boom. Thus, for market cap-weighted indices, large cap profit delivery carries more index impact; improvement in healthcare and industrials provides the essential foundation for upside momentum to expand beyond the few AI-linked tech firms.

The profit growth pace and ability to beat expectations among U.S.-listed companies jointly decides the season's investment appeal; healthcare versus energy especially offers valuable insight for global equities strategies. S&P 500 Q3 profit growth is expected at 29.6% YoY, 2.9pp above quarter-beginning; energy’s profit growth forecast rises from 79.3% to 118.6%, tech from 57.1% to 65.0%, communication services up 51.5%, industrials 14.5%.

However, Citigroup’s projections show healthcare’s profit growth forecast cut from 8.7% to 5.1%, but its positive surprise candidate share hits 75.9%. For energy, despite the highest growth pace, only 48.1% of candidates are expected positive. Citi quant model logic is that share price reactions depend on the gap between actual results and previous expectations: after downgrades, healthcare sees more firms outpace the market’s prior view; for energy, the bar for surprise is already much higher.

Citi’s U.S. market 12-month profit revision index, compiled from Wall Street’s top analysts, remains near the annual high set in early July, signaling rising analyst optimism. However, the 29.6% QoQ profit growth estimate for Q3 still lags well behind Q2’s 52.4%, so the main positives this season come from forecast upgrades and expanded positive surprise coverage.

Citi’s model seeks “better-than-expected earnings that actually drive stock gains.” The report uses standardized unexpected earnings—SUE = difference between actual quarterly EPS and consensus, divided by the analysts’ forecast SD—to measure surprise, noting that historically about 38% of companies had a surprise direction opposite their share price return, explaining why “beats” can still accompany share price falls.

To improve event selection, Citi uses logistic regression, calibrates on 24 past quarters, tracks the last 4 quarters’ SUE, two prior quarter surprise directions, 9-week earnings momentum, last 20 trading days’ market-relative total return, analyst consensus and 9-week rating changes, and excludes cases where earnings surprise and price move diverge. Citi’s investment logic seeks streaks in profits, improved analyst sentiment, and market price confirmation. Thus, technology’s 88.1% figure specifically reflects the model-predicted proportion of “positive” companies, indicating broad coverage of positive signals.

From Surging AI Chip Orders to Sector Rotation: Seeking the Earnings “Expectation Gap Dividend” of Earnings Season

Citi’s “positive candidates” semiconductor list covers the heart of the AI compute infrastructure—AI GPU/ASIC and datacenter CPU compute chips—as well as analog and power, equipment, testing, and materials, providing a complete view of the tech profit chain’s outperformance.

Nvidia, AMD, Intel, Texas Instruments, Analog Devices, ACM Research, Microchip Technology, Applied Materials, KLA, Teradyne, Entegris, and MKS all meet “model-predicted positive surprise & return” and Citi fundamental “Buy” rating. Industrially, these firms correspond to compute supply, chip power + signal handling, fabrication controls (etching, ALD, etc.), advanced packaging/testing, and materials purity: converting AI infrastructure to usable compute power requires the whole chain to deliver, and each company’s profit elasticity depends on product mix, capacity utilization, and cost structure.

This positive prediction list also covers chipmakers servicing traditional industrial/other end markets, offering notable candidates for observing profit expansion into further sub-sectors. The list also includes leading U.S. data center optical interconnect firms—Coherent, Lumentum—storage giant Western Digital, plus software names like Cadence and ServiceNow. All stock prices are as of October 6, 2026; no target prices or expected returns are provided.

The significance of healthcare and broader industrials lies in providing different sources of earnings surprises and positive investment returns during earnings season, allowing opportunities to extend from hot compute trades to broader enterprise operational improvement. Citi strategists list positive buy candidates in healthcare including Johnson & Johnson, Abbott, Intuitive Surgical, Danaher, Thermo Fisher, Dexcom, Edwards Lifesciences, and Medtronic—covering drugs, devices, surgical robots, diagnostics, and life science tools; for industrials: GE Aerospace, RTX, Caterpillar, Vertiv, Eaton, Quanta Services, Carrier, Emerson, and Rockwell Automation.

From a value chain viewpoint, Vertiv and Eaton allow us to track the acceleration of AI infrastructure investment toward power/thermal/data center revenue; aerospace, machinery, and automation firms stand for other industrial profit sources. Thus, technology and industrials share ties relating to AI infrastructure, while healthcare brings an earnings theme driven by business improvement and guidance upgrades. Citi quant data shows “crowding” is much lower in some healthcare candidates than compute stocks—e.g. Medtronic at 0.170, Intuitive Surgical at 0.266, Dexcom at 0.287, versus Vertiv at 0.970, GE Aerospace at 0.980. Therefore, a combination of “rising positive forecasts + low crowding” makes medical picks notable incremental opportunities in Citi’s study.

A truly substantive “positive” stock selection difference this earnings season hinges on whether good news is already fully priced—and so Citi weighs crowding as strongly as earnings forecasts. Among the Magnificent Seven, Alphabet, Microsoft, Apple, and Nvidia get positive model signals; Tesla, Meta, and Amazon are negative. Meta and Amazon retain Buy ratings, evidencing that Citi's model-based short-term earnings read can differ sharply from medium-term investing ratings. Meta’s long signal crowding is 0.955, Nvidia’s 0.876, Microsoft’s 0.684, Apple’s 0.671; Microsoft’s crowding is up from quarter-start, Tesla’s is down significantly. According to Citi, crowded stocks are more prone to sharp corrections on negative surprises; less crowded stocks with positive earnings surprises can attract more incremental investors.

Thus, Citi lists separate low-crowding positive bets—GoDaddy, Becton Dickinson, Leidos, Procter & Gamble—and high-crowding neutral/sell bets including Phillips 66, Vornado, PPL; Qualcomm, AbbVie (neutral), Moderna, the Trade Desk (sell/high risk) are also flagged. Citi’s stated playbook: strong outperformance from the AI compute/tech leaders at the core, while seeking new returns in healthcare, industrials, and low-crowding positives, and distinguishing intra-sector opportunities by expectation gap strategies.