AI spending shake-up: chipmakers overtake tech titans as market leaders amid sector jockeying

The market leadership landscape is undergoing a pronounced rotation from the concentrated dominance of mega-cap tech giants, notably the 'Mag 7' such as Microsoft and Meta, toward semiconductor chip makers and smaller-cap sectors. This shift is driven by evolving AI-related capital expenditure dynamics, with semiconductor companies like Micron, Intel, and AMD collectively adding nearly $2 trillion in market value in a single quarter and now representing about 20% of the S&P 500—quadruple their weight from six years ago. As Anu Ganthi observes, this broadening leadership is reflected in improved market breadth, with over 60% of stocks outperforming the S&P 500 in June, including defensive sectors like healthcare, signaling a diversification beyond traditional mega-cap tech dominance.

Alongside the semiconductor surge, financials and healthcare sectors are gaining significant prominence, complementing the AI-driven tech leadership and reflecting a broader sectoral shift. The Russell 2000’s robust 22% gain in the first half of 2026—the best since 1991—underscores the sustained outperformance of small and mid-cap stocks, bolstered by elevated Treasury yields and banking sector strength. This rotation is not merely size-driven but sector-driven, as banks and healthcare firms increasingly benefit from AI infrastructure demand and defensive positioning amid geopolitical uncertainties, as highlighted by investors’ moves into these areas during renewed market volatility.

Within the mega-cap tech cohort, leadership is becoming more nuanced, with the market penalizing the heaviest AI capital spenders like Microsoft and Meta—down roughly 22% and 14% respectively—while rewarding firms such as Alphabet, Apple, and Nvidia that balance innovation with more measured expenditure. This reflects investor preference for cash-flow durability over momentum, as Howard Marks notes, where the aversion to the build phase of AI infrastructure spending obscures the long-term growth potential. Meanwhile, companies like Meta are pivoting toward custom AI chip production to reduce costs, and Apple is integrating Google's Gemini AI, illustrating a dynamic, evolving leadership landscape within tech itself.

Despite recent capital rotations, AI infrastructure demand remains robust and multifaceted, fueling growth across semiconductor, memory, and cloud infrastructure sectors. Nvidia is projected to maintain a dominant 90% share of AI training workloads through 2028, while memory makers like Micron and Samsung report historically high margins supported by massive investments, including South Korea’s half-trillion-dollar commitment to memory capacity. However, valuation disparities have prompted investors to rotate from stretched semiconductor multiples back into hyperscalers and AI infrastructure beneficiaries, reflecting a maturing AI trade that prioritizes sustainable cash flows and broad sectoral participation—from data center REITs surging 36% year-to-date to cybersecurity firms capitalizing on AI-driven demand spikes.

The narrow market leadership concentrated in a handful of AI-related mega-cap stocks, such as Nvidia, Meta, Alphabet, and Microsoft, has significantly heightened concentration risk, complicating active managers’ efforts to outperform benchmarks. With the top 10 S&P 500 stocks now commanding roughly 40% of the index—surpassing even the dot-com peak—many active managers find themselves underweight these giants while holding more concentrated portfolios, making it difficult to capture the skewed market gains driven by chipmakers and AI infrastructure leaders. As Anu Ganthi explains, this positive skewness underscores why diversification and index-based approaches gain appeal in such an environment, as narrow breadth means only about 24% of stocks beat the S&P 500, squeezing active stock pickers’ room to maneuver.

Active managers grapple with the challenge of timing and valuation amid volatile AI-driven market dynamics, where strong earnings growth does not always translate into sustained price appreciation. Samsung’s 19-fold surge in operating earnings, followed by a stock decline after a significant run-up, exemplifies the difficulty in balancing position sizing and entry points when robust fundamentals are already priced in. Similarly, Nvidia’s dramatic loss of nearly $1 trillion in market value over two months, despite controlling 97% of the AI data center chip market and trading at its cheapest forward multiple since 2019, highlights the disconnect between earnings strength and market sentiment. This volatility forces managers to discern persistent secular trends from transient earnings surprises, focusing on long-term shifts in market or execution rather than short-term noise.

The concentrated gains in AI mega-cap stocks, coupled with investor conviction in a narrow set of names like Tesla and AMD, further complicate active management by limiting diversification and amplifying valuation concerns. SoFi investors, for instance, allocate about a quarter of their holdings to the AI super cycle, often adding to high-conviction positions during downturns, which challenges managers trying to time entries amid strong sentiment-driven rallies. Meanwhile, emerging names like Rocket Lab signal a subtle rotation as investors hunt for the next AI cycle, but these shifts are deliberate rather than random, underscoring the difficulty in anticipating leadership changes within a tightly clustered market. As Jim Cramer notes, market moves often reflect shifting sentiment more than fundamental updates, making it harder for active managers to interpret mixed signals from related sectors such as Dell and Micron, which declined despite solid earnings.

Amid this narrow AI-driven leadership, active managers face the dual challenge of balancing optimism fueled by bullish analyst reports and high-profile endorsements against cautious views on AI spending returns and valuation risks. While companies like Microsoft and Alphabet continue to generate returns above their cost of capital, skepticism remains—highlighted by Warren Buffett’s endorsement of Alphabet contrasted with doubts about Amazon’s AI investments yielding tangible returns. This divergence creates uncertainty in investor positioning, as managers must weigh positive momentum against potential overvaluation and mixed signals, emphasizing the critical need for selectivity and disciplined portfolio construction in the second half of 2026.

Market breadth has notably improved, with nearly 350 stocks showing year-over-year gains and about half of the index delivering double-digit returns across various market caps, signaling a healthier and more diversified earnings landscape beyond the mega-cap tech giants. However, this broad strength masks uneven performance within smaller caps, where only 18% of the smallest 50 stocks are outperforming, suggesting sector-specific dynamics rather than uniform gains. This nuanced picture underscores a market driven more by stock-specific factors than broad sector moves, reflecting the dispersed nature of AI-related earnings contributions across industries.

Despite robust earnings growth outpacing market returns—32% versus 22% over the past year—forward earnings expectations have surged so rapidly that portfolio managers like Lance Roberts warn of a looming risk that reality may fall short, potentially triggering downward revisions and price corrections. Interestingly, S&P 500 valuations have declined even as fundamentals improve, with rising profit margins bolstered by AI-driven efficiencies, indicating that strong earnings are currently justifying market prices rather than inflating multiples. Simultaneously, the Mag 7 mega-cap tech stocks are underperforming, reducing market concentration and fostering a more balanced breadth that supports sustainable valuation dynamics.

The tech sector’s extraordinary outperformance—peaking at a 29 percentage point lead over the S&P 500 in early June, a six-standard deviation event—has since reversed, aligning with historical patterns where such extremes precede mean reversion and relative underperformance lasting several months. This pullback is driven more by Federal Reserve rate hike concerns than valuation fears, as the current hawkish tone from Fed Chair Warsh echoes the 2000 hiking cycle that pressured high-multiple growth stocks. Yet, this rotation is expected to broaden market participation, with healthcare and financials gaining relative favor, allowing tech to pause without a sharp absolute decline, preserving the long-term secular growth narrative for large-cap US tech.

AI-driven capital expenditure is reshaping earnings and valuation dynamics within the semiconductor and hyperscale tech sectors, as companies like SK Hynix and Micron secure multi-year deals to mitigate memory chip cyclicality amid persistent supply shortages. While semiconductor firms tied to AI infrastructure, particularly those supplying specialized memory like HBM for Nvidia’s GPUs, are seeing surging free cash flow—projected at $430 billion combined for Nvidia, Micron, Broadcom, and Applied Materials—the hyperscale giants such as Amazon, Alphabet, and Microsoft face contrasting free cash flow pressures, with their combined cash flow turning negative for the first time. This divergence highlights valuation and risk disparities within AI sectors, compounded by the concentration of hyperscaler spending largely recirculating within the tech ecosystem, which poses potential risks to market breadth and earnings sustainability if AI demand growth slows.

The AI infrastructure boom extends well beyond core semiconductor manufacturing, encompassing critical power infrastructure and data center buildouts essential for sustaining AI’s digital expansion. Eaton Corporation exemplifies this trend, with record backlogs stretching into 2028 driven by surging demand for grid modernization and cooling solutions, underscoring the multi-year investment theme linked to AI-driven power shortages and infrastructure upgrades. As Dan Ives highlights, data centers are becoming the "hearts and lungs" of AI deployment, acting as the physical backbone supporting diverse enterprise applications and signaling a shift in capital expenditure toward tangible, scalable infrastructure.

Capital expenditure in AI infrastructure is reshaping sectoral capital flows and valuations, with semiconductor manufacturers like Intel and Tower Semiconductor aggressively expanding production capacity to meet long-term AI demand, while hyperscalers such as Amazon and Microsoft absorb heavy AI-related investments that push their free cash flow into negative territory. This dynamic creates a tightly interconnected ecosystem where hyperscaler spending circulates back into chipmakers and data center operators, generating record free cash flow—estimated at $430 billion combined for key chipmakers—yet also exposing multiple balance sheets to synchronized risk if demand falters.

The maturation of AI infrastructure is evident as hyperscalers and neocloud providers now generate $1.19 in revenue for every dollar of depreciated AI infrastructure, marking a profitable and sustainable segment at the top of the stack. However, this profitability is uneven: while four cash-generative giants invest $725 billion with strong balance sheets, smaller neocloud firms face heightened risk due to debt burdens and rapid chip obsolescence. This nuanced landscape challenges simplistic bubble narratives, especially as falling costs and widespread adoption undermine bearish arguments, reinforcing the economic rationale behind sustained AI infrastructure spending.

AI-related capital expenditure has become a major driver of U.S. economic growth, accounting for over a quarter of GDP expansion and representing nearly 8% of GDP—surpassing tech spending levels seen during the Dot Com bubble. This surge reflects a fundamental shift in capital allocation within the tech sector, where hardware investments in servers, storage, and memory are crowding out enterprise software budgets, as noted by IBM’s CEO. Meanwhile, despite competitive pressures from China, the U.S. maintains clear leadership in AI infrastructure investment, anchored by hyperscalers and Nvidia’s dominant chip technology, fueling an ongoing arms race in AI capabilities and infrastructure buildout.