Taiwan Semiconductor Manufacturing Company (TSMC) delivered record second-quarter results that beat expectations, yet semiconductor stocks plummeted across the board because investors shifted focus from revenue growth to the massive capital costs required to stay competitive in artificial intelligence chip manufacturing. The company raised its full-year 2026 revenue guidance to slightly above 40% growth, but simultaneously increased capital expenditure guidance to $60 to $64 billion, up from a prior range of $52 to $56 billion, along with additional commitments to expand Arizona facilities.

On July 16, the S&P 500 fell approximately 0.5%, the Nasdaq-100 dropped roughly 1.6%, and the Dow Jones Industrial Average eased around 0.2%. The VanEck Semiconductor ETF (SMH) lost 3.70%, while major chip companies suffered significant declines. TSMC shares fell roughly 4% despite the record profit beat, Broadcom fell 4.6%, NXP Semiconductors fell 3.6%, and memory-focused companies like Micron and SK Hynix ADRs extended losses, with Micron down 6% and SK Hynix down 14%.

What Changed Between TSMC's Strong Results and the Market Selloff?

The disconnect between strong fundamentals and stock declines reveals a critical shift in how investors are evaluating semiconductor companies. TSMC confirmed that artificial intelligence-driven demand remains "extremely robust," and the company posted 68% operating margins, demonstrating exceptional profitability. However, the market's focus pivoted sharply to cash generation and the explicit cost of maintaining technological leadership.

Management guided third-quarter operating margins roughly 70 basis points below consensus and warned that overseas expansion and 2-nanometer ramp costs would dilute gross margins in the second half of the year. Every incremental dollar of TSMC's capital expenditure increase represents a potential drain on near-term free cash flow, compressing the yields needed to justify the semiconductor sector's lofty valuation multiples. This concern rippled across the entire chip industry, as investors recognized that scaling artificial intelligence manufacturing capacity will be exceptionally expensive.

"The chips aren't falling on demand, every fundamental print this month screamed the opposite. They're falling on positioning. The fundamentals just went five for five this month ASML, Meta, TSMC, the inflation prints, HBM sold out. The street went the other way," noted Nicholas Mugalli, institutional TMT research analyst.

Nicholas Mugalli, Institutional TMT Research Analyst

How Are Investors Now Evaluating Semiconductor Companies?

• Capital Intensity Focus: Rather than celebrating revenue growth and strong demand signals, investors now scrutinize how much capital companies must spend to maintain competitive manufacturing processes, with TSMC's $8 billion spending increase raising concerns about near-term profitability.
• Free Cash Flow Compression: The market shifted from valuing top-line artificial intelligence demand to evaluating actual cash generation, recognizing that massive capital expenditures reduce the cash available for shareholder returns and debt reduction.
• Margin Dilution Risk: TSMC's guidance that overseas expansion and advanced chip manufacturing would compress gross margins in the second half of 2026 signaled that profitability gains may not keep pace with revenue growth.
• Valuation Multiple De-Rating: Analysts warned that the semiconductor sector faces a "multiple de-rating," meaning investors may be willing to pay less per dollar of earnings if profit margins absorb the burden of rapid expansion.

Which Other Chip Companies Felt the Pressure?

The selloff extended well beyond TSMC. Broadcom, a major supplier of artificial intelligence infrastructure chips, fell 4.6% despite having recently secured a multi-year partnership with Apple worth over $30 billion through 2031. This deal, announced just eight days before the selloff, had previously boosted Broadcom shares by 5.2%, but the broader sector weakness overwhelmed that positive news. Broadcom is now trading at $375.36 per share, down 22.1% from its 52-week high of $481.57 set in June 2026.

Memory chip manufacturers faced particularly steep declines. Micron, which had recently inked deals with Qualcomm, Hyundai Mobis, and several automotive technology suppliers, fell 6% amid ongoing valuation and supply concerns in the artificial intelligence memory trade. SK Hynix ADRs dropped 14%, the steepest decline among major semiconductor players, reflecting heightened anxiety about memory chip valuations in the artificial intelligence infrastructure buildout.

Nvidia, the dominant artificial intelligence chip manufacturer, closed down 2%, pulled lower in sympathy with the broader semiconductor complex despite its dominant market position. Even Apple, which bucked the predominant trend and edged higher, benefited from investor perception that it has "more or less stayed out of the data center arms race and instead focused on measured artificial intelligence integration".

What Does This Mean for the Semiconductor Industry Going Forward?

The market's reaction signals that investors will now demand evidence that downstream software monetization can justify the massive capital costs flowing through the hardware supply chain. The semiconductor sector faces a critical test: can hyperscalers and artificial intelligence companies generate sufficient revenue from their artificial intelligence services to justify the enormous infrastructure investments required from chip manufacturers?

TSMC's capital expenditure increase reflects the reality that competing at the cutting edge of semiconductor manufacturing requires relentless investment. The company's $60 to $64 billion spending plan represents an unprecedented commitment to manufacturing capacity, driven by insatiable demand for artificial intelligence chips. However, this spending spree comes at a time when investors are increasingly focused on profitability and cash generation rather than pure growth metrics.

The semiconductor sector's valuation multiples may face compression as profit margins absorb the burden of rapid expansion, according to analyst commentary. This represents a fundamental shift in how the market evaluates chip companies, moving away from celebrating artificial intelligence demand tailwinds and toward scrutinizing the actual financial returns generated by massive capital investments. Investors will closely monitor upcoming earnings reports from major hyperscalers to assess whether the software and service revenue generated by artificial intelligence infrastructure can ultimately justify the hardware supply chain's extraordinary capital requirements.