NextFin News - The artificial intelligence trade, once a monolithic engine of market gains, has fractured into a landscape of high expectations and punishing corrections. On March 21, 2026, Nvidia shares closed another week of stagnant performance, marking a seven-month stretch where the world’s premier chipmaker has delivered zero percent in profits to its shareholders. Despite reporting record fiscal fourth-quarter revenues of $68.1 billion—a 73% year-over-year surge—the stock remains more than 10% below its October highs. This "beat-and-bleed" phenomenon, where stellar earnings fail to ignite rallies, suggests that the market has moved past the phase of speculative fervor and into a period of rigorous valuation scrutiny.
U.S. President Trump’s administration has maintained a focus on domestic semiconductor manufacturing, yet the macro environment for AI remains clouded by what some analysts call "indiscriminate selling." As Nvidia’s first-quarter sales outlook of $78 billion failed to satisfy a Wall Street hungry for even more aggressive growth, capital is beginning to rotate. Investors are no longer simply buying the "shovels" of the AI gold rush; they are hunting for undervalued "miners" and industrial players that are quietly integrating agentic AI into tangible business models. This shift has brought two specific names into focus: SoundHound AI and Rivian.
SoundHound AI has become a lightning rod for the broader correction in speculative tech. Since the peak of the AI frenzy last October, the company’s shares have plummeted 63%. For a small-cap growth stock, such a drawdown is often a death knell, but the underlying fundamentals tell a more nuanced story. SoundHound holds a portfolio of over 200 patents in voice AI, a sector increasingly vital as "agentic AI" moves from software screens to physical interfaces like drive-thru kiosks and automotive assistants. With its market capitalization compressed to $3 billion from a near-$10 billion peak, the risk-reward profile has shifted. While it faces stiff competition from Big Tech, its specialized focus on voice-to-action technology provides a niche that general-purpose LLMs have yet to fully commoditize.
Rivian represents a different breed of AI play—one hidden behind the sheet metal of an industrial manufacturer. The electric vehicle maker is currently trading at roughly 3 times sales, a stark contrast to Tesla’s 15 times sales valuation. This discrepancy exists because the market still views Rivian primarily through the lens of capital-intensive manufacturing. However, the company’s strategic pivot toward proprietary AI chips and end-to-end autonomous software suggests it is building a vertical stack similar to the one that fueled Tesla’s premium. Next month’s launch of the R2 SUV, priced under $50,000, serves as the critical test for this thesis. If Rivian can scale production while maintaining its AI-driven design efficiencies, it may finally shed its "industrial" discount.
The divergence between these companies and the semiconductor giants like Nvidia and Broadcom highlights a maturing market. Broadcom, for instance, saw its AI semiconductor revenues jump 74% recently, yet it too faces the pressure of a market that has already priced in perfection. The current sell-off in Nvidia is not necessarily a verdict on the future of AI, but rather a recalibration of the price investors are willing to pay for that future. For those looking beyond the immediate volatility of the chip sector, the wreckage of the October correction has left behind assets that are significantly cheaper, even as their technological relevance continues to grow.
Explore more exclusive insights at nextfin.ai.
