Musk, Cook Warn of Memory Chip Crisis as Demand From AI Grows

Speakers

Neil Kaplan — Bloomberg Senior Strategist

Summary

A severe global memory chip shortage is emerging, driven by AI demand; new AI chips (like H100s) consume 6-10x the memory of previous generations.
The supply crunch is structural and expected to last, as building new memory fabrication plants takes 3-5 years.
Rising component costs are threatening corporate profits and inflating prices for end-products ranging from laptops to cars.
AI efficiency narratives are backfiring for some legacy firms (e.g., Real Estate), where cost-cutting announcements are interpreted by the market as a sign of low moats and disruption risk rather than bullish efficiency.

Trade Ideas

Ticker	Direction	Speaker	Thesis	Time
SAMSUNG /AAPL /TSLA	SHORT	Neil Kaplan Bloomberg Senior Strategist	Musk and Cook warn that memory shortages are "beginning to hammer profits" and "inflate price tags on everything from laptops to cars." Kaplan notes Samsung is "impacted negatively" because, despite making chips, they are heavily exposed to manufacturing smartphones and PCs where costs are rising. The memory shortage acts as a tax on hardware OEMs. Higher Bill of Materials (BOM) costs squeeze margins. Unlike pure-play chip makers, these companies cannot fully pass costs to consumers without hurting demand. SHORT. The "crisis" narrative suggests margin compression for heavy hardware manufacturers. If these companies successfully pass costs to consumers or if memory prices stabilize faster than expected.	0:54
SK HYNIX /MEMORY SECTOR	LONG	Neil Kaplan Bloomberg Senior Strategist	"It takes between three and five years to build a new memory fabrication plant... creating this bottleneck." Hynix states the situation will "get worse before it gets better." Basic supply and demand. Demand is exploding (AI chips use 10x memory) while supply is inelastic (3-5 year lag). This creates significant pricing power for pure-play memory producers who sell the commodity rather than the finished device. LONG. Structural scarcity benefits the upstream commodity producer. Global recession reducing demand for end-products (smartphones/PCs), which could offset AI demand.	—
CBRE	SHORT	Neil Kaplan Bloomberg Senior Strategist	CBRE announced they could reduce research costs by 25% using AI, but the "stock actually got hit hard." Investors interpreted the efficiency claim negatively ("Second-Order Thinking"). If CBRE can use AI to cut costs, external competitors or clients can use the same models to bypass CBRE entirely or force fee compression. AI is viewed here as a deflationary disruptor to their business model. SHORT. The market is punishing service intermediaries who claim AI benefits, fearing they have no moat against the technology. If CBRE demonstrates that AI actually expands margins rather than cannibalizing revenue.	1:42