Bottom Line
KOSPI 10,000 is not a dream but a 2027 base scenario defined as PER 10x × EPS 1,000p — Korea holds the highest beta to the memory/packaging/power triple shortage.
The Triple Shortage's Korean Supply-Chain Transposition Opens the First-Ever 300T KRW Operating Profit Era
KOSPI 10,000 is not a dream but a 2027 base scenario defined as PER 10x × EPS 1,000p — Korea holds the highest beta to the memory/packaging/power triple shortage.
Three earnings calls in two days — Intel Q1, GE Vernova Q1, SK Hynix 1Q26 — simultaneously proved the Agentic AI-driven triple shortage with a single quarter of hard data.
Beneficiaries — memory (SK Hynix, Samsung), back-end packaging (Hanmi Semiconductor, HPSP), power equipment (HD Hyundai Electric, Hyosung). Pressure — legacy DRAM, traditional manufacturing.
Quarterly SK Hynix and Samsung HBM guidance and 12M Fwd EPS trajectory — the speed of progression from 824.5p toward 1,000p is the key timeline indicator for KOSPI 10,000.
The three events that emerged in global markets over April 22-23, 2026 are not coincidence. Intel (Apr 23), GE Vernova (Apr 22), and SK Hynix (Apr 23) — three Q1 2026 earnings releases from different industries (integrated semis, power equipment, memory) — delivered a single message simultaneously: the Agentic AI-driven memory/packaging/power triple shortage has been quantitatively confirmed by quarterly results data.
Intel's Q1 2026 results beat consensus across every line. Foundry revenue +16% YoY at $5.4B; Data Center & AI (DCAI) +22% YoY at $5.1B. CFO David Zinsner explicitly stated on the call that "advanced packaging backlog grew during the quarter and has reached annual demand of billions of dollars," adding that "despite improving fab output, demand exceeds supply, particularly in the client CPU market." CEO Lip-Bu Tan said, "the essential role of CPUs in the AI era is expanding, driving surging demand for Intel CPUs, wafers, and advanced packaging," and raised Q2 revenue guidance to $13.8~14.8B.
Two structural takeaways follow. First, AI infrastructure investment has moved past GPU concentration into a phase that pulls in CPU demand as well. The "tens of millions of CPUs" specified in the AWS-OpenAI $38B compute deal means a single transaction accounts for a meaningful share of global annual server-CPU shipments. Second, when Intel says "advanced packaging backlog billions per year," it directly implies multi-year order visibility for Korea's packaging equipment and materials supply chain — Hanmi Semiconductor, HPSP, INTEKPLUS, and Isu Petasys.
GE Vernova's Q1 new orders grew +71% YoY to $18.3B and total backlog reached $163B. The crucial data point sits in gas turbines — the company's combined backlog and slot-reservation commitments for gas turbines jumped from 83GW to 100GW in a single quarter, and CEO Scott Strazik guided to 110GW by year-end. The more decisive signal came from Electrification: data-center transformer and power-equipment orders alone hit $2.4B in 1Q, exceeding the entirety of FY24 in a single quarter. New bid and order pricing rose 10~20% above 4Q backlog pricing, and the company formally stated that "only 10GW of gas-turbine capacity remains available before 2030; orders beyond that are slotted into 2031 and later."
The implication is clear. The world's #1 power-equipment vendor has locked five years of capacity through 2030, creating a sellers' market where pricing power expands +10~20% per quarter. This gap is being directly filled by Korea's Doosan Enerbility (gas turbines + SMR) and HD Hyundai Electric / Hyosung Heavy (transformer Big Three).
SK Hynix on the same April 23 reported revenue 52.6T KRW, operating profit 37.6T KRW, and operating margin 72%. Outpacing TSMC's 1Q operating margin of 58.1% for the second consecutive quarter, with the gap widening to 14ppt. This is the decisive signal that the memory industry has been re-positioned from a simple subcontractor to AI accelerators into the highest-margin segment of the value chain. Nomura on April 24 raised its SK Hynix price target from 1.93M KRW to 2.34M KRW (+21%, the highest among global brokers), upgrading 26F and 27F OP forecasts by +9% and +4% respectively, citing LTA durability of memory margins. Per TrendForce, 1Q26 DRAM contract pricing rose +90~95% QoQ — the largest quarterly jump on record — with +58~63% additional expected in 2Q and NAND +70~75%. Goldman, JPM, Gartner, and TrendForce 4-house consensus aligns on the supply shortage continuing through H1 2027.
How does this global macro thesis transmit into KOSPI's and PER? On EPS: Korea owns global share in memory (SK Hynix + Samsung >70% DRAM, >50% NAND), back-end packaging (Hanmi Semi effectively monopolizes TC bonders, HPSP monopolizes high-pressure anneal), and AI-accelerator PCBs (Isu Petasys global Big-3). When Intel and SK Hynix create price transmission, Korean EPS reacts with higher beta than the global average. On power: with >70% of US large transformers over 25 years old and GE Vernova's capacity shortage on top, Korea's transformer Big Three are effectively locking US capacity, holding multi-year backlogs (HD Hyundai Electric ~10T KRW, Hyosung 13.85T KRW). Doosan Enerbility uniquely owns both gas turbines and SMR globally, capturing the largest share of the gap GE Vernova and Siemens cannot fill.
Net result: the global triple shortage transmits into EPS explosion in Korea's supply chain — and that is the fundamental reason KOSPI's 12M forward EPS jumped +144% in six months.
Sign in with Google to unlock the full body of every free report instantly.
Sign in with GoogleThis site runs on ads — the tier system rewards community contributions.
This report is provided for informational purposes only and does not constitute a recommendation to buy or sell any financial instrument. Investment decisions should be made based on your own judgment and responsibility. The analysis and opinions contained herein are based on information available at the time of writing and are subject to change.