A car dealer once sold a vehicle without a powered tailgate because electronic components were unavailable during the pandemic. Buyers accepted fewer features simply because those parts could not be sourced. The shortage made the tailgate optional, not the car itself. This demonstrated a key investment principle: the most valuable parts of any industry are not always the most visible. They become essential when demand outstrips supply, and their suppliers gain pricing power—even if the components themselves are unremarkable.
Artificial intelligence is now facing a similar dynamic. While investors debate which AI model will lead or whether the sector is overvalued, a quieter development is unfolding in the background. Someone must provide the memory, storage, and optical connections that keep these systems operational. When those components grow scarce, their suppliers can charge premium prices—despite the parts themselves being ordinary.
A component doesn’t have to be glamorous to become valuable. It has to be difficult to replace when everybody needs it. The pandemic-era car shortage proved this point. Now, AI infrastructure is creating its own version of that bottleneck.
AI’s Unseen Supply Chain Bottlenecks
Semiconductors operate in cycles. Demand spikes, manufacturers expand production, and supply eventually catches up—turning shortages into oversupply. A stock trading at six times earnings may seem cheap until those earnings are halved. For AI-related businesses, the critical question is not just growth, but whether the conditions sustaining that growth remain intact.
Two investors specializing in institutional money flows argued that current data does not signal an imminent downturn. One highlighted rising earnings expectations among memory and storage suppliers. The other noted that manufacturers and their customers continue to describe a multiyear demand outlook. Neither observation eliminates cyclical risks, but both challenge the assumption that a reversal is inevitable.
Consider Micron Technology. Its expected earnings peak has repeatedly shifted further into the future as estimates adjust. A valuation based on an approaching collapse requires reassessment if the business keeps postponing that collapse. However, forecasts can change in either direction. The decisive factors will be whether customers reduce orders, earnings estimates decline, or new capacity overwhelms demand.
Photonics: The Light Speeding AI Growth
This leads to photonics, the technology using light to transmit data between chips. As AI systems grow more powerful, their processors require faster data exchange. Without adequate optical connections, even the most advanced hardware hits performance limits. Two companies dominate this space: Marvell, which supplies semiconductor technology for data infrastructure, including optical connectivity, and Corning, a leader in fiber and optical communications. They operate at different stages of the system, but both stand to benefit if demand for high-speed data transfer continues rising.
The investment question shifts from whether photonics is compelling to whether these companies can convert potential into revenue, profits, and lasting advantages. Shareholders still require tangible delivery, and purchase prices must allow for meaningful returns.
Optical stocks have traded sideways for months despite the technology’s importance. The explanation may lie in trading pressures rather than fundamentals. Forced selling, portfolio adjustments, or uncertainty about interest rates could depress prices, even if underlying demand remains strong. The challenge for investors is distinguishing temporary market noise from genuine warning signs.
Jason Bodner, ever the practical one, described himself as a surfer of money flows, watching the conditions in front of him. His investment process combines institutional trading activity with business fundamentals: sales, earnings, margins, and financial strength. That framework doesn’t settle the ultimate return on every AI investment. It does help distinguish a possible future problem from deterioration already showing up in the data.
For now, the AI infrastructure buildout continues. Suppliers of scarce components may enjoy stronger pricing power. However, if those components become more abundant, buyers could see their costs drop, reshaping winners and losers within the sector. This shift would not end AI investment opportunities but would alter which businesses benefit.
Beyond Tech: AI’s Margins for Main Street
Beyond technology, the question is whether non-tech companies will eventually improve margins through AI adoption. Retailers and other consumer businesses may take longer to translate experimentation into measurable results. This potential broadening is worth monitoring; but it must first appear in financial performance.
The original lesson from the car lot remains relevant. Trace the shortage to its supplier, then analyze that supplier’s economics. That is where the investment case materializes.
Investor discussions also emphasized how institutional funds are tracking AI-related spending shifts. Some have reduced exposure to speculative AI startups, while others focus on companies with clear demand signals. Jason Bodner, who monitors money flows, observed that institutional activity in memory and storage stocks has held up better than in some high-profile AI plays. This suggests investors prioritize businesses with visible economics over those relying on future promises. The distinction matters because even as AI hype fades, hardware demand may persist. Manufacturers continue describing a multiyear outlook for data center expansion, with optical and memory components playing a central role.
Corning, for instance, has seen steady demand for its fiber-optic cables as data centers scale up. The company’s optical communications revenue has grown alongside AI workloads, though its stock has not yet fully reflected this growth in price. The challenge for investors is assessing whether these companies can sustain margins as competition intensifies. Some suppliers may struggle if new capacity enters the market faster than expected, while others could strengthen positions through long-term contracts with cloud providers.
Leave a Reply