Wall Street is commencing the second half of 2026 with a renewed, yet complex, sense of confidence. Over the past six months, global capital markets surged with a seeming disregard for all headwinds: escalating Middle East geopolitical conflicts first doubled international oil prices before a dramatic plunge, while the Federal Reserve's interest rate expectations underwent the most severe hawkish-dovish reversal in years. Yet, diversified portfolios of stocks, bonds, and commodities have just delivered their strongest first-half returns since 2021.
This was no straightforward, broad-based rally. If the core theme for capital markets in the first half of 2026 was "surviving in a jungle of black swans taking flight," then the keyword for the second half becomes "coexisting with what survived." A higher valuation baseline, persistently elevated borrowing costs, and the disruptive wave of artificial intelligence (AI) spreading from virtual large models to the physical world are the three gravitational realities Wall Street must now confront.
In nearly all mid-year outlook reports, sell-side institutions are betting that the market can coexist with these three forces. However, behind the record earnings growth and AI-driven enthusiasm pushing the S&P 500 to new highs, a seismic shift in core buyer strategies is quietly deconstructing the fundamental belief underpinning the past two years' AI bull market.
Stephen Dover and Larry Hatheway of the Franklin Templeton Institute stated: "The global economy and financial markets have performed better than many expected. If one core word could summarize the outlook for 2026, it is 'resilience.'"
Core Thesis Reversal: From "Absolute Scarcity of Computing Power" to "Explosive Monetization Pressure"
The long bull run in global tech stocks over the past two years was built on an unshakable "belief": computing power would always be insufficient; those holding chips and high-bandwidth memory (HBM) held the ticket to the future. However, as the second half of 2026 begins, the first cracks have appeared in this idol.
The Epicenter: Meta's Capital Expenditure and the "Sub-landlord's" Self-Rescue
The shockwave originated from mega-buyer Meta. In its latest capital expenditure guidance, Meta significantly raised its 2026 AI-related CapEx forecast to a historic extreme of $125 to $145 billion. This colossal, near-insane figure instantly absorbed the surplus capacity of the global wafer foundry and high-end memory supply chains. But this comes at a cost. Such intensive capital expenditure has led to a severe compression of Meta's once-envied free cash flow (FCF). Relying solely on the organic growth of its consumer advertising business and AI models cannot fill this terrifying funding gap in the short term. The pressure to monetize has reached a critical point at the board level.
To find a "safety cushion" for its massive investment, Mark Zuckerberg hinted at the May shareholder meeting that external companies were willing to pay a premium for Meta's computing power almost weekly. Subsequently, Meta's long-prepared compute resale plan, "Meta Compute," was officially launched—transforming Meta from the biggest buyer of AI chips into the market's largest "bulk compute sub-landlord."
Raphael Thuin, Head of Capital Markets Strategies at Tikehau Capital in Paris, warned: "Record earnings growth and AI-driven enthusiasm have pushed risk assets to new highs. Nevertheless, the second half is unlikely to simply replay the first. Given that parts of the AI 'picks and shovels' trade appear overextended, any shift in the narrative around computing demand could rapidly reshape the market landscape." Meta's move to resell compute power is precisely the "narrative shift" Thuin described. It mercilessly shatters the myth of perpetual compute shortages, exposing the volcano-like eruption of monetization pressure beneath this multi-hundred-billion-dollar infrastructure investment.
Supply Chain Bullwhip Effect: Extreme 87% Concentration and the Software Sell-off
When Meta began renting out excess compute, it triggered a disastrous domino effect in financial markets. Wall Street suddenly realized that traditional industrial cycle laws had never disappeared in the AI era. The latest statistics from Barclays reveal the fragile structure of the first-half bull market.
Alexander Altmann of Barclays' Equity Tactical Strategy noted that during the first-half frenzy: Semiconductor and computer hardware companies contributed approximately 87% of the S&P 500's gains for the first half. The semiconductor market experienced its second-largest historical gain, with some chip and memory stocks soaring by triple digits. Meanwhile, due to excessive capital drainage, the software market suffered its sixth-largest historical sell-off.
This extreme narrowness in market breadth means the entire market's fate is tied to the semiconductor hardware cycle. Barclays analyst Anshul Gupta clearly stated: "As funds shifted from AI hyperscale cloud providers to AI-enabling chip companies, the market fervor transmitted to the semiconductor sector, driving a massive surge in the sector."
The "Magnificent Seven" Become the "Lagging Seven": $2 Trillion Evaporated in June
This shift in AI investment logic is particularly evident in the "Magnificent Seven." For the entire first half, the total return of the "Magnificent Seven" fell by about 2%, even lagging behind UK government bonds. At the close on June 30, the Roundhill Magnificent Seven ETF (MAGS) tracking the tech giants was down about 2.5% for the first half, while the S&P 500 Equal Weight Index rose over 16% in the same period.
June was a particularly "black June"—the "Magnificent Seven" lost about $2 trillion in market value in a single month. Microsoft fell 17%, its largest monthly drop since December 2000; Amazon fell 12%; Meta fell 11%. Philip Jagd, Head of Equities at Danish pension fund Sampension, bluntly stated the "Magnificent Seven" had become the "Lag Seven." He noted that this underperformance against the broader market in the first half "has only happened once in the past 15 years—in 2022, which was a major 'risk-off' year driven by rate hikes."
The root cause: These companies, once famous for their abundant cash flows, are now funneling most of their funds into the AI arms race. Torsten Slok, Chief Economist at Apollo Global Management, pointed out that the free cash flow of Meta, Alphabet, Microsoft, and Amazon has declined significantly from its 2024 peak to 2026. Microsoft has already informed investors that 2026 capital expenditure could reach $190 billion.
As Jagd analyzed: "Tax and accounting rules mean the earnings of these companies do not reflect the huge costs generated by data centers—costs that have not yet significantly driven sales. This puts pressure on the companies' free cash flow; they issue debt or equity, and the future return on invested capital will decline. The market has now noticed this."
Mid-Bull Market Pause or Approaching the End? Second Half Outlook: Consensus and Divergence on Wall Street
Looking ahead to the second half, Wall Street overall remains optimistic, but divergences are widening. According to summaries from Yardeni Research and others, the median year-end S&P 500 target price from 19 investment banks and research institutions is 7,850 points. Citigroup has raised its 2026 S&P 500 target from 7,700 to 8,100 points; Royal Bank of Canada raised its target to 8,150 points; Goldman Sachs expects the S&P 500 to reach 8,000 points by the end of 2026; JPMorgan raised its target to 7,800 points; Standard Chartered set a target of 7,950 points for the S&P 500 by mid-2027.
However, bearish voices are equally notable. Bank of America maintains its year-end 2026 S&P 500 target of 7,100 points, suggesting the index could fall about 5% from current levels. BofA warns that the primary drivers of the previous bull market—accommodative central bank policies, accelerating corporate earnings growth, accelerating buybacks, inflows from individual and institutional investors—are reversing.
BofA analysts predict the Federal Reserve will raise interest rates three times this year, given "sticky US inflation and a tightening labor market." Raphael Thuin of Tikehau Capital reiterated: "Record earnings growth and AI-driven enthusiasm have pushed risk assets to new highs. Nevertheless, the second half is unlikely to simply replay the first. Given that parts of the AI 'picks and shovels' trade appear overextended, any shift in the narrative around computing demand could rapidly reshape the market landscape."
Alexander Altmann of Barclays wrote: "I firmly believe the second half of 2026 will be just as eventful as the first. Consider this: three major geopolitical conflicts occurred in the past six months, the semiconductor market saw its second-largest historical gain, and the software market suffered its sixth-largest historical sell-off."
Investment Focus: Paradigm Shift from "AI Giants" to "AI Enablers"
Institutions like BlackRock and Invesco believe the focus of the AI trade is no longer just buying the largest tech companies but is expanding into the real economy, semiconductors, memory, power grids, data centers, and industrial infrastructure. Meanwhile, rising bond yields are increasingly seen as an opportunity rather than a challenge, reigniting yield strategies in short-term bonds and high-quality credit.
The BlackRock Investment Institute team stated: "We maintain an overweight on U.S. equities and are focused on bottleneck investment opportunities in AI growth." Bank of America prefers focusing on companies that benefit from the AI investment cycle rather than directly funding it. "During a CapEx boom, which industries should one hold? Choose cyclical industries that invest in manufacturing but generate cash flow, rather than long-term growth companies that need constant financing to maintain competitiveness."
Three Tail Risks Not to Be Ignored in the Second Half
Although most sell-side institutions (like Morgan Stanley and Franklin Templeton) maintain a constructive view for the second half, expecting inventories to rise, business confidence to strengthen, and AI spending to broaden beyond hyperscale tech companies, Barclays and Morgan Stanley have also issued warnings. The second half of 2026 is destined to be eventful. Strategists remind global investors to closely monitor the following three micro and macro landmines during the upcoming Q3 earnings season and Q4 election period.
Risk Point One: Sequential Stagnation in Hyperscaler CapEx
If Microsoft or Google revise down their 2027 capital expenditure forecasts in upcoming quarterly reports due to falling cloud business margins from declining compute rental prices, chip stocks, whose declines are still within a controllable range, could face a second major round of "Davis double-kill."
Risk Point Two: Resurgent Stubborn Inflation and Forced Central Bank Tightening
Both Morgan Stanley and JPMorgan mentioned in their mid-year outlooks that current supply chain constraints are not limited to chips but also include energy and electricity. If AI's demand for grid power triggers a second global surge in energy prices, strong non-tech economic growth could keep core inflation persistently high, forcing global central banks to further tighten monetary policy by the end of 2026.
Risk Point Three: The Political Cycle and Midterm Election Window
As the U.S. midterm elections approach, geopolitical uncertainty and potential trade policy disruptions will directly test the current high-flying U.S. stock market, which already severely lacks valuation error tolerance. Marvin Loh, Senior Macro Strategist at State Street Bank in Boston, concluded the current frenzy and panic with a philosophical remark: "AI has proven capable of soothing a world facing dual uncertainties of geopolitics and monetary policy. While the liquidity for AI construction seems infinite, we can say with certainty it is not, and this will likely test the availability and cost of capital in the second half."
Comments