Broadcom's second-quarter results were, by almost any standard measure, exceptional.
Revenue climbed 48% year-on-year to a record $22.2 billion (€19.1bn), AI chip sales more than doubled, and adjusted earnings per share rose 54% to $2.44, comfortably beating consensus estimates.
Yet none of it was enough.
Investors, conditioned to expect not just strong results but results that keep rewriting the ceiling, punished the stock sharply after the California-based semiconductor giant declined to raise its long-term AI revenue target and after its third-quarter AI chip outlook came in below the most aggressive forecasts circulating on Wall Street.
After closing Wednesday's session at an all-time high of $495 per share, Broadcom dropped more than 12% in after-hours trading, wiping out more than $300 billion (€258.3bn) in market value.
According to CEO Hock Tan, who called AI demand "simply insatiable", the trajectory remains firmly intact.
Tan guided third-quarter consolidated revenue to $29.4 billion (€25.3bn), representing 84% year-on-year growth, and raised the full-year AI semiconductor revenue forecast to $56 billion (€48.2bn), up roughly 180% from the last fiscal year.
Ripples across global markets
The impact of Broadcom's results extended well beyond the Nasdaq.
Asian markets opened sharply lower on Thursday, with the sell-off concentrated in semiconductor and technology stocks.
The MSCI index for Asian shares fell 1.6%, snapping a four-day rally that had pushed it to an all-time high, while Nasdaq 100 futures retreated 0.5%.
South Korea's KOSPI, widely regarded as a barometer for global AI investment sentiment given the prominence of its chip industry, fell 1.8%, making it one of the region's worst performers of the day.
Samsung Electronics and SK Hynix, two of the world's leading memory chipmakers, declined between 2% and 4%, while LG Electronics dropped nearly 14%.
In Taiwan, key AI server assemblers Hon Hai Precision and Wistron fell close to 4% and 8%, respectively.
Broadcom is a critical supplier of custom AI accelerators and high-speed networking chips to the world's largest hyperscalers, including Google, Meta and OpenAI, meaning its guidance carries significant weight for investors tracking the pace of AI development.
