The S&P 500 Index is on track for its strongest earnings growth in five years, with 83% of companies reporting results that exceeded analyst expectations. Following an 18% rebound from its war-induced low, the index is poised for its eighth consecutive week of gains.
According to Zhitong Finance APP, market expectations that the United States and Iran are poised to reach a long-term peace agreement—mediated by Pakistan and converting the previously fragile ceasefire into a durable peace—have bolstered bullish sentiment on Wall Street. This increasingly optimistic outlook has propelled the S&P 500, one of the benchmark U.S. equity indices, toward its longest weekly winning streak since December 2023. In the view of Wall Street analysts, the AI-driven mega bull market led by the Magnificent Seven (Mag 7)—the seven largest U.S. tech giants—and key players across the AI computing infrastructure supply chain is far from over; rather, it is transitioning from an ‘AI valuation narrative’ to a more robust phase characterized by ‘AI profit realization, easing geopolitical risks, and broader profitability across the AI computing value chain.’
The so-called “Magnificent Seven” (also referred to as the “Mag 7”)—widely recognized by Wall Street analysts as comprising Apple, Microsoft, Alphabet (Google’s parent company), Tesla, NVIDIA, Amazon, and Meta Platforms (Facebook’s parent company)—have been the primary engine behind the repeated record highs of the S&P 500 Index. Top-tier investment institutions on Wall Street view this group as the portfolio best positioned to deliver substantial returns to investors amid what is arguably the most transformative technological shift since the dawn of the internet era.
Analysts believe that the ongoing wave of AI infrastructure investment and the strong earnings growth driven by AI applications among the Magnificent Seven will remain a key driver of overall EPS expansion for the S&P 500 over the long term. Moreover, according to senior market strategists at major Wall Street institutions such as Morgan Stanley and Nomura, the central theme of the global AI-driven bull market—including U.S. equities—is evolving from a singular focus on ‘surging demand for NVIDIA’s AI GPUs’ to a broader ‘bullish wave encompassing the entire AI computing infrastructure ecosystem,’ which will continue to support the AI-led supercycle in U.S. and global equities.
The clearest example lies in NVIDIA’s latest earnings report, which vividly illustrates that the global surge in AI infrastructure investment is far from subsiding. The momentum is expanding beyond AI GPUs and AI ASICs to include data center CPUs, high-performance networking infrastructure, enterprise-grade HBM/DRAM/NAND memory, full-stack server clusters, AI super factories, and large-scale enterprise AI cloud computing systems. On Wall Street, bullish sentiment toward NVIDIA—the undisputed global leader in AI—is intensifying, with the average analyst price target implying a potential market capitalization exceeding $7 trillion.
Consequently, the AI bull market’s ‘pie’ is expanding from core GPU assets to encompass the entire hardware stack facing supply constraints. While NVIDIA remains the pricing anchor defining the tone of the AI bull market, components such as PCBs, MLCCs, ABF substrates, ODM services, liquid cooling systems, power supplies, silicon wafers, CMP consumables, photoresists, glass substrates, SOI/InP materials, data center optical interconnect equipment, optical modules, and advanced packaging tools could all undergo a new round of valuation re-rating.
Having rebounded 18% from the lows triggered by the Iran conflict, the S&P 500 is on track to close this Friday with its eighth consecutive weekly gain—the longest such streak since December 2023. However, U.S. two-year Treasury yields have risen after Federal Reserve Governor Christopher Waller stated that the next move in Fed monetary policy is equally likely to be a rate hike or a cut. Money markets have fully priced in one rate hike by the end of 2026, reflecting a hawkish outlook.
U.S. crude oil prices have experienced significant volatility as traders attempt to assess when energy flows through the Strait of Hormuz will fully normalize. Nearly three months after the outbreak of renewed Middle Eastern geopolitical tensions, Tehran is now considering a new proposal submitted by the United States. Pakistan announced that its Chief of Army Staff is traveling to the Iranian capital, signaling positive progress in negotiations aimed at ending the conflict. Media reports indicate that a Qatari negotiating team, coordinated with the U.S., has arrived in Tehran to facilitate a long-term ceasefire agreement.
U.S. Secretary of State Rubio stated that talks with Iran have seen ‘some positive progress,’ but added, ‘we’re not there yet.’ The United Arab Emirates has recently taken a more active role in pushing for an end to the conflict, joining Saudi Arabia and Qatar in advocating for U.S.-Iran peace negotiations rather than initiating another round of hostilities.
According to Craig Johnson, senior analyst at Piper Sandler, investors are effectively overlooking macroeconomic headwinds and are positively pricing in rhetoric surrounding the prospect of peace—a trend that has consistently provided tailwinds for equities. He remarked, ‘Global equity markets are exhibiting a hope-driven rally.’
Mag 7 Ignites Earnings Engine! U.S. Stocks Poised for Longest Weekly Winning Streak Since Late 2023
As the bull market momentum driven by AI computing infrastructure investment spreads across most technology sectors of the U.S. corporate landscape, the S&P 500 is on track to deliver its strongest earnings growth cycle in five years. According to the latest data from Bloomberg Intelligence, approximately 93% of companies in the index have already reported results, with 83% significantly surpassing analyst expectations—the highest proportion since 2021.
Beyond healthcare, earnings growth has a broad base. Overall, the Magnificent Seven (Mag 7) remain the primary drivers of earnings expansion in the S&P 500, though strong performances across the energy and technology sectors have overshadowed weakening consumer sentiment linked to higher oil prices stemming from the Iran conflict. According to data compiled by Bloomberg Intelligence, the communication services and consumer discretionary sectors delivered the largest positive earnings surprises, while materials and industrials also exceeded expectations.
Bloomberg Intelligence senior analysts Nathaniel Welnhofer and Christopher Cain stated, ‘If cyclical and non-AI sectors begin contributing robust growth while NVIDIA and the AI-related computing infrastructure complex continue generating earnings, 2026 may not resemble a late-cycle bull market slowdown but rather a replay of the post-pandemic profit boom seen globally in 2021.’
For the full year, gains in the S&P 500 are still expected to be concentrated in a few sectors centered around technology stocks, with upward revisions in earnings expectations for the energy, materials, and technology sectors likely to drive stronger index performance.
As shown in the chart above, full-year earnings growth expectations for the S&P 500 have been consistently revised upward—led by the energy, technology, and materials sectors, which are driving the overall earnings-per-share (EPS) growth trajectory for the index in 2026. Note: Data as of May 15.
Bloomberg Intelligence noted that the bulk of S&P 500 earnings growth continues to come from large-cap tech firms like the Mag 7, with the AI computing theme viewed as a key reason why the index’s earnings growth could exceed 20% in 2026.
The S&P 500, in which tech giants such as NVIDIA, Apple, and Microsoft account for approximately 40% of the index weight, has seen significant upward revisions to both price targets and EPS estimates from Wall Street analysts following the first quarter, with over 90% of its constituent companies beating earnings expectations for four consecutive quarters.
NVIDIA’s results and outlook both exceeded Wall Street consensus estimates, prompting analysts to further raise their earnings forecasts. For the world’s largest company by market capitalization, adjusted profits are now projected by analysts to grow by approximately 84% this year, up from an initial forecast of 64% at the beginning of the year.
Citi raised its full-year earnings forecast after noting that NVIDIA management anticipates sales could surpass $1 trillion by 2027—a projection that excludes contributions from newer revenue sources such as Groq LPX systems and Vera processors. As shown in the chart above, NVIDIA’s adjusted earnings expectations for fiscal year 2027 have been revised upward. Note: Data as of May 22.
According to Tajinder Dhillon, Head of Earnings Research at LSEG, earnings growth from the Magnificent Seven will remain a key driver of overall EPS expansion for the S&P 500 index. Their growth trajectory is expected to stabilize heading into 2027, although the group is still projected to significantly outperform the broader S&P 500 (.SPX). The chart below displays quarterly earnings growth forecasts by portfolio segment—highlighting the critical role of robust profit growth from the Magnificent Seven in supporting U.S. equity earnings fundamentals.
This latest trend further reinforces bullish investors’ core view that market leadership will remain concentrated among mega-cap tech giants—and that this concentration is fundamentally driven, rather than being a sign of an impending ‘AI bubble’ collapse fueled by speculative euphoria.
Moreover, the U.S. energy sector has seen the largest upward revision in earnings expectations due to the Iran conflict. Analysts now project earnings growth of 61% for the year, significantly higher than the 7.6% forecast at the start of the year. Exxon Mobil and Chevron both reported stronger-than-expected first-quarter results, as higher oil and gas prices offset production disruptions related to the conflict. Together with ConocoPhillips, they are the primary drivers of sector growth, accounting for roughly half of the sub-index’s weight.
Bank of Nova Scotia analyst Brandon Bingham stated in a report that energy companies are generally optimistic about next year, citing strong project backlogs, steady growth, and potential for further development.
Materials stocks have become the third driver of the index, benefiting from rising prices and tightening supply. Citi analyst Patrick Cunningham noted that Middle East-related disruptions have had limited impact, primarily exerting upward pressure on raw material costs, and added that demand trends appear stable, with companies maintaining a cautiously constructive outlook.
Cunningham stated that paint manufacturers Sherwin-Williams Co., PPG Industries Inc., and Axalta Coating Systems Ltd. all anticipate cost inflation, although they still expect modest volume growth in the second half of the year. Chemical giant Dow Inc. remains cautious but sees potential for upside if pricing improves further.
Bloomberg Intelligence indicated that industrial gas producers such as Air Products and Chemicals Inc. and Linde Plc could benefit, as the global helium market has tightened following the shutdown of a Qatar-based plant in March, which took approximately one-third of supply offline.
The AI super bull market has entered a phase of ‘full-chain revaluation’: from Vera Rubin racks to glass substrates, GPUs are no longer the sole focal point of the AI compute narrative! The entire AI compute value chain beyond GPUs is now taking off.
As AI agents gain global traction, investment focus in AI computing power is shifting from a ‘point solution race centered on AI GPUs’ to a ‘full-stack computing infrastructure driven by AI agents.’ Consequently, the next wave of excess alpha returns will no longer accrue solely to top-tier leaders in AI GPUs or AI ASICs but will systematically broaden across the entire AI infrastructure stack—including data center CPUs, DRAM/NAND/HBM memory, AI-optimized PCBs, liquid cooling systems, optical interconnects for data centers, ABF substrates/glass substrates, and a wide range of foundry services.
On April 30, three cloud computing supergiants—Microsoft, Google, and Amazon—reported stellar earnings on the same night, highlighting the unexpectedly explosive growth of their cloud businesses driven by the AI wave, prompting Wall Street to reevaluate the commercial returns of AI. The latest research report from Morgan Stanley’s analyst team predicts that the combined capital expenditures of the five hyperscale tech giants (Amazon, Google, Meta, Microsoft, Oracle) will reach approximately 800 billion US dollars in 2026 and are expected to exceed 1.1 trillion US dollars in 2027, marking another upward revision from the previous forecast of 950 billion US dollars.
Morgan Stanley analysts emphasized that the core logic behind these massive capital outlays is: first, make heavy investments and build capacity; then recover returns through scalable commercial revenue and ROIC based on AI compute resources. The surge in cloud computing backlogs and AI application tokens provides direct evidence that this model works. The unexpectedly rapid expansion of these tech giants’ cloud businesses has led Wall Street to reassess AI’s commercial returns and adopt an increasingly positive investment sentiment toward the entire AI infrastructure supply chain.
Wall Street’s consensus price target for NVIDIA already prices it as a $7 trillion company; the most aggressive aggregated targets push it toward the $12 trillion mark. The core rationale behind these high-profile, named upside targets is NVIDIA’s evolution from a GPU leader into an ‘end-to-end AI infrastructure platform’—with GPUs, CPUs, networking, rack-scale systems, software ecosystems, and capital returns collectively underpinning its valuation expansion. Analysts expect global spending on AI-accelerated computing infrastructure to reach $3–4 trillion by 2030.
Thus, the central theme of the AI super bull market is evolving from ‘NVIDIA GPU-driven single-point breakout’ to ‘full-chain revaluation of AI infrastructure.’ In the earlier phase, the market primarily focused on GPUs, HBM, cloud capex, and NVIDIA’s earnings elasticity. However, according to Morgan Stanley’s research report, the Vera Rubin BOM breakdown reveals to investors that the value leap in next-generation AI racks is not driven by GPUs alone—components such as PCBs, MLCCs, ABF substrates, power systems, liquid cooling, and ODM assembly and testing are all simultaneously contributing to higher content value.
Morgan Stanley estimates that the price of the Vera Rubin VR200 NVL72 rack is approximately $7.8 million—nearly double that of the GB300 rack at around $3.99 million. Within this, PCB content value has increased by 233% compared to the GB300, MLCCs by 182%, and ABF substrates by 82%. This indicates that the AI super bull market is entering its second phase: ‘component content expansion.’
From an engineering perspective, Rubin represents not merely a GPU generational upgrade but a leap in rack-level system complexity. As NVLink/NVSwitch, ConnectX, BlueField DPUs, midplanes, switch trays, SOCAMM/high-bandwidth memory, and liquid-cooled power supplies are integrated into a single rack, the value chain naturally expands beyond GPU chips to include high-speed copper interconnects, data center optical interconnect systems, power delivery, thermal management, substrates, and high-layer-count PCBs. Consequently, the GPU’s share of the BOM has declined from approximately 65% in the GB200 to about 51% in the VR200, though the absolute dollar value of GPUs has risen from roughly $2.52 million to approximately $3.96 million. This shift marks the most critical evolution in the AI supply chain: while market leaders remain strong, previously overlooked ‘supporting components’ are now being repriced as irreplaceable performance bottlenecks.
A Nomura research report highlights an even more fundamental structural shift in AI data centers: AI is transforming semiconductor growth logic from ‘process node scaling’ toward ‘structural innovation + material substitution + advanced packaging + optoelectronic interconnects.’ Historically, capital markets focused on Taiwan Semiconductor’s advanced nodes, NVIDIA GPUs, and ASML lithography tools. However, as the industry moves into the 2nm, 1.4nm, and post-Moore eras, technologies such as backside power delivery, GAA/cFET transistors, SoIC hybrid bonding, wafer-bonded NAND, glass-core substrates, photonic SOI, metal-oxide photoresists, and InP/silicon photonics for on-chip I/O are becoming key determinants of AI chip energy efficiency, bandwidth, yield, and cost.
This precisely explains why the ‘AI super bull market’ has not ended despite elevated valuations among the Magnificent Seven, but is instead transitioning from a valuation-driven narrative to one grounded in earnings realization and broadening supply chain participation. NVIDIA, Microsoft, Google, Meta, and Amazon continue to ramp up AI capex, generating systemic demand for AI server racks, advanced packaging, HBM, PCBs, MLCCs, ABF substrates, optical modules, liquid cooling, power systems, and data center power infrastructure. Unlike in 2023–2024, the market is no longer solely betting on ‘who has the strongest model or GPU,’ but is actively identifying every bottleneck within AI infrastructure—whether constrained by power density, interconnect bandwidth, packaging yield, or supply shortages—and re-pricing those segments accordingly.
Glass core substrates and optical interconnects in data centers warrant particular attention. As AI chip packages grow larger, I/O counts increase, and signal speeds rise, traditional organic ABF substrates and first-generation silicon interposers face challenges related to warpage, thermal expansion, signal loss, and large-area planarity. Consequently, glass core substrates are emerging as a leading candidate for next-generation advanced packaging. According to TrendForce, Taiwan Semiconductor, Samsung, and Rapidus are advancing glass interposer/glass substrate solutions, while SK Absolics aims to achieve mass production of glass substrates by 2026. IDTechEx further notes that surging demand from AI and high-performance computing (HPC) is compelling packaging stacks to handle higher currents, more I/Os, and faster signal rates, thereby transitioning glass core substrates from a niche technology toward commercialization.
Optical interconnects represent an inevitable outcome as AI data centers evolve from ‘compute stacking’ to confronting ‘network bottlenecks.’ With the near-unbounded scaling of training and inference clusters, data traffic between GPUs, across racks, and among data centers has surged dramatically. Electrical interconnects are increasingly constrained by distance, power consumption, and bandwidth density, rapidly elevating the importance of 1.6T/3.2T optical modules, silicon photonics, co-packaged optics (CPO), photonic SOI, and InP lasers.
In AI superclusters such as TPU and GPU computing power groups, there is a core commonality: the need for extremely high bandwidth, ultra-low latency, and highly energy-efficient intra-data-center interconnects. Traditional copper cables or electronic switching solutions cannot meet requirements when scaled to thousands or even tens of thousands of chips due to explosive increases in power consumption and thermal losses. Optical interconnect technologies (primarily including co-packaged optics (CPO), silicon photonic switches, and optical circuit switching (OCS)), as well as more advanced silicon photonics-level optical I/O technologies, use optical signals instead of electrical ones, significantly improving bandwidth density and energy efficiency while reducing latency and power consumption in large-scale AI training/inference networks. This demand for higher optical interconnect capacity is a shared feature of GPU and TPU clusters.
In other words, the next phase of the AI super bull market will be defined not merely by faster chips, but by an integrated ecosystem comprising chips, advanced packaging systems, network infrastructure, optical interconnect systems, materials, liquid cooling solutions, PCBs, MLCCs, ABF/glass substrates, rack-scale compute clusters, and power systems—collectively shaping what NVIDIA CEO Jensen Huang has termed ‘AI factory economics.’