A team led by Michael Hartnett, a veteran strategist at Bank of America often called Wall Street's most accurate forecaster, has released a report suggesting that strong Republican performance in the midterm elections and the Texas gubernatorial race could drive further gains in U.S. stocks, particularly AI-linked semiconductor shares. Hartnett's team wrote that if President Donald Trump's Republican Party performs strongly on November 3 and Greg Abbott, a key political ally, secures re-election as Texas governor, it would particularly benefit global semiconductor stocks tied to artificial intelligence.

Midterm elections are long viewed by Wall Street as a major risk event affecting the trajectory of U.S. and global stock market bull runs and bullish sentiment, with market volatility historically rising sharply in the months before the vote. The shift from an "AI liquidation storm" to a "bubbly 2027" AI super-bull scenario appears to take just three weeks—a testament to how quickly Wall Street's outlook can change. Strategists now see the July "AI crowded trade liquidation" as more of an extreme leverage and speculative position washout than a reversal in the AI computing infrastructure trend. The Philadelphia Semiconductor Index (SOX) fell nearly 29% from its June 22 high to a July 29 low, but has rebounded about 20% in roughly three weeks, re-entering a technical bull market.

In recent semiconductor trading, Coherent, a key AI optical interconnect name, has bounced about 57% from its July 29 low. SanDisk, a U.S. NAND storage giant, jumped 13.7% in a single day after its August 13 investor day, while another U.S. memory heavyweight, Micron, rose nearly 5%. This signals that the AI computing power rally is broadening from pure GPU stocks to include NAND/DRAM/HBM memory, high-speed optical communications, and network interconnects. South Korea, a barometer for AI computing power, has amplified this semiconductor rebound. The benchmark KOSPI index rose 11.5% this week, ending a seven-week losing streak, with Samsung Electronics and SK Hynix up 19% and 16% weekly, respectively. From a July 30 close of 5,593.56, the KOSPI has rebounded about 24.75% to 6,977.94 by August 14. More critically, South Korea's single-stock leveraged ETF assets plunged from about $50 billion to $17 billion, with JPMorgan estimating that hedge fund deleveraging is about 90% complete. Foreign net buying of around 3 trillion won on August 14 suggests the global semiconductor market is shifting from "forced selling" to a "risk-on" mode for institutions like hedge funds. If low-semiconductor-position funds continue to chase, it could form a positive feedback loop of "rise-cover-short squeeze-further rise."

Midterm Election Trading Heats Up: Republican Senate and Texas Control Could Push AI Bull Run to "Bubbly 2027"

Hartnett's Bank of America team stated in their report: "If Trump can hold the Senate and Abbott can hold Texas, expect stock markets—especially AI-related semiconductor stocks—to surge all the way, potentially pushing the market toward a bubbly 2027." Conversely, if Democrats win the Senate and defeat Abbott, U.S. stocks could face a decline of at least 10%, with the dollar and bond yields also falling sharply by year-end. The Texas gubernatorial race between Abbott and Democratic challenger Gina Hinojosa is seen by Bank of America strategists as a "referendum on cost of living and AI data centers." Texas already hosts 335 data centers, with another 247 in the planning stages. Abbott's pro-business stance, including a temporary halt to data center expansion, reflects growing "election anxiety" over rising energy costs and grid electricity price risks for U.S. consumers.

This month, a strong semiconductor-led tech rally pushed the S&P 500 to record highs, driven by investors buying the dip in AI computing stocks since late July and second-quarter corporate earnings surpassing already elevated analyst expectations. The S&P 500's rise has struggled to keep pace with rising earnings forecasts, as analysts boost projections amid surging AI demand and data center spending. According to Bloomberg Intelligence, second-quarter profit growth for S&P 500 components is now expected at 32%, up from 23% forecast before the earnings season began, with 93% of tech companies beating estimates. Hartnett's team believes the "door is wide open for bulls to take risk higher," citing surging earnings, a $10 trillion increase in U.S. wealth by 2026, and consensus estimates that AI infrastructure capital expenditure will exceed $1 trillion by 2027. They note: "The only constraints are bonds (yield spikes), voters (rising socialism), and the fact that virtually all investors are already significantly positioned for a market rise."

Chris Caso, a strategist at Wolfe Research, noted that the Philadelphia Semiconductor Index roughly tripled in three months before falling about 25% from its peak. He views the recent weakness as a reset of expectations after a massive rally. Caso expects AI chip demand to outstrip supply at least until 2028, arguing that concerns about a slowdown in hyperscaler capital spending have not materialized, and competition around AI agents leaves cloud giants with "no choice but to invest." Morgan Stanley analyst Brian Nowak and his team recently raised their 2027/2028 capital expenditure forecasts for the five largest global hyperscalers (Meta, Amazon, Microsoft, Google, SpaceX) to about $1.2 trillion and $1.4 trillion, respectively. They also hiked their 2026 U.S. tech giant CapEx estimate to $805 billion from $433 billion a year ago.

Post-Leveraging AI Bull Run Re-Ignited: Semiconductor Rally to 2027?

The midterm elections act more as a "valuation and risk premium switch" for the global semiconductor super-cycle than the engine driving AI demand itself. Hartnett's team argues that if Republicans hold the Senate and Abbott holds Texas, the market will interpret it as a continued favorable environment for AI data centers, energy infrastructure, and tax and business regulation. This is especially relevant given Texas's 335 existing data centers and 247 planned projects, which would reduce the tail risk of political constraints on AI CapEx due to electricity prices and grid pressure. Conversely, if Democrats win the Senate and Texas governorship, Bank of America sees a potential 10%+ drop in U.S. stocks. However, from an AI engineering and semiconductor supply chain perspective, the first-order variable determining whether the semiconductor super-cycle continues into 2027 remains hyperscaler CapEx, actual GPU/ASIC cluster deployment, HBM and enterprise NAND supply-demand, data center optical interconnect and network expansion, and real-world AI inference token demand. Recent evidence strongly suggests growing AI fundamental prospects. Bank of America sees hyperscaler CapEx heading toward $1.2 trillion by 2027, while JPMorgan cites rising cloud order backlogs and improving AI investment returns as key reasons for raising S&P targets. In other words, a Republican victory would not create the AI super-cycle but could push the existing AI CapEx super-cycle from a "fundamental bull market" to a "valuation record expansion plus short squeeze" bubble phase.

Strong earnings and outlook from key AI computing upstream players like ASML and TSMC, along with recent results from CoreWeave and Hon Hai, are sending a key signal to global stock markets: the AI computing chain is transitioning from a "training AI large model CapEx super-cycle" to a new phase of "exponential growth in inference computing demand driven by agent-based applications." These signals refute the recent "overcapacity" bearish narrative that led to the sell-off in AI stocks, especially semiconductors. Ed Yardeni, a veteran analyst known as "Wall Street's prophet" and founder of Yardeni Research, recently raised his S&P 500 year-end 2026 target from 8,250 to 8,400, while boosting his 2026/2027 EPS estimates from $330/$375 to $375/$415. He maintains a long-term view that the S&P 500 could reach 10,000 by late 2029, or even earlier, driven by AI-driven productivity gains and sustained EPS expansion. Conversely, if election shocks combine with runaway long-term bond yields or a sharp miss in AI CapEx returns, the same high-beta, crowded positioning could again act as a liquidation amplifier.