The Samsung Electronics and SK hynix era is over. All the money will flow “here” next | Seong Sang-hyeon’s Macro Beyond, Part 3-2
- What to look for
- You can explore the discussion of the possibility that memory capital expenditure could be reallocated toward subsequent industries such as equipment and materials, along with the four steps presented for identifying leading stocks.
- Why it matters
- This helps you learn the criteria for distinguishing structural deterioration involving downward EPS revisions from valuation corrections, and how to verify capital inflows through relative returns.
AI summary
Seong Sang-hyeon explains that AI investment capital may not remain in one sector indefinitely, but may be reallocated to the next industries and companies that resolve bottlenecks. Rather than conclusively declaring that “the era of Samsung Electronics and SK hynix is over,” the video discusses the possibility that increased capital expenditure by memory companies could accelerate earnings and share prices at smaller companies receiving that spending, such as semiconductor equipment and materials suppliers. In AI infrastructure, logic semiconductors such as GPUs, CPUs, and accelerators handle computation, while communications chips and network semiconductors handle data movement between chips and servers, so both areas become more important as AI models grow. However, even for communications chip companies, investors need to check the actual share of data center products, adoption by new customers, and forecasts for gross margins and EPS. When existing market leaders undergo a correction, distinguish between structural deterioration that also brings downward revisions to EPS estimates and a valuation correction in which earnings forecasts remain intact. The process for identifying leading stocks is to find structurally growing industries, identify companies with accelerating revenue, operating profit, and EPS, verify market capital inflows through relative returns over 1, 3, 6, and 12 months, and then narrow the selection to companies with strong market share, cash flow, and ROE. AI data centers, communications chips, and logic semiconductors, which show strength in both growth and share prices, are classified as existing leaders; uranium, whose share prices are weak relative to growth, as a potential latecomer; and cloud computing, AR/VR, and quantum computing as areas to watch. Although he envisions AI’s next expansion moving through computation, connectivity, on-device execution, and physical AI, he emphasizes that actual investment decisions must be made by continually updating company-specific EPS revisions and relative strength, rather than relying on a predicted sequence of industries.