Midway Through June's Event Cluster: This Was a Rates, Oil, and AI Crowding Stress Test

Context This post connects the earlier notes on the CPI-BOJ-FOMC event cluster, the May CPI preview, the U.S. jobs shock and KOSPI 8,000 gate, foreign investors’ June 12 return, Jensen Huang’s HBM4 comments, and the SpaceX IPO read-through for Korea.

TL;DR

• The first half of June was not an easing confirmation. It was a stress test for rates, oil, and crowded AI positioning.
• U.S. jobs, CPI, and PPI all made it harder for the Fed to speak comfortably about cuts.
• Korea was hit harder than the U.S. Local DB data show KOSPI fell from 8,788 on June 1 to 7,484 on June 8, or -14.8%. KOSDAQ fell -13.2%. By comparison, the S&P 500 low drawdown was -4.4%, and Nasdaq 100 was -6.6%.
• From June 1 to June 12, foreigners sold KRW 17.99tn in Korean equities, and Samsung Electronics plus SK Hynix alone accounted for KRW 15.92tn of foreign net selling.
• June 12 mattered: foreigners bought KRW 1.62tn, institutions bought KRW 3.01tn, program buying was KRW 1.86tn, and Samsung plus SK Hynix drew KRW 2.17tn of foreign buying. But it is still a one-day signal.
• Jensen Huang’s Korea visit and the SpaceX IPO shifted the market from a simple macro-risk tape to a conditional risk-on re-ignition test for AI, space, and frontier technology.

1. What Has Been Confirmed

The BLS May employment report showed that U.S. payrolls were still expanding and unemployment was stable. CPI and PPI then showed that inflation pressure was not cleanly resolved. The message was simple: growth has not broken, but inflation is still uncomfortable. That is not the backdrop for a clean broad-beta risk-on.

2. Why Korea Fell More

Korea’s drawdown was not just macro. It was macro hitting a market already crowded into AI memory, substrates, power, robotics, and semiconductor beta.

Local flow data for June 1-12 show the structure clearly.

Foreign selling was not evenly distributed. It was concentrated in the two memory megacaps. That is why the index moved more violently than the U.S. benchmarks.

3. Why June 12 Matters, but Is Not Enough

June 12 was the first real absorption signal.

This was not just retail dip-buying. Foreigners, institutions, and program flows all came in. But as discussed in the prior note, the KOSPI foreign buy of KRW 2.72tn was only 3.6% of the prior 24-session KRW 75.57tn foreign selloff. We still need two to three consecutive sessions to call it a regime shift.

4. Jensen Huang and SpaceX Changed the Narrative

NVIDIA’s official blog said Jensen Huang visited Korea to deepen cooperation with SK, LG, Hyundai, NAVER, and Doosan across AI infrastructure, physical AI, and agentic AI. The important point is not the visit itself. It is that Korea is moving from memory supplier to AI factory partner and physical AI industrial base.

Business Insider reported that SpaceX began trading as SPCX on Nasdaq on June 12, raising $75bn at a roughly $1.77tn valuation and closing its first day around +19% above the IPO price. This is not a simple “buy all space stocks” signal. It is a liquidity and valuation signal for frontier technology.

5. Current Regime

The early-June regime was risk-off:

strong jobs + hot CPI/PPI + oil anxiety + foreign semiconductor selling
= risk-off

After Jensen’s Korea visit, SpaceX’s IPO, oil relief, and the June 12 foreign/institution/program rebound, the regime improves one notch:

macro remains uncomfortable
but AI/space/frontier tech narratives remain alive
oil relief allowed buyers to return to beaten-down growth
= conditional risk-on re-ignition candidate

This is not a market rising because macro is good. It is a market willing to buy structural growth again if macro does not deteriorate further.

6. Late-June Checklist

If these pass, the Korean playbook can broaden from Samsung and SK Hynix into AI components, NAVER/SK Telecom AI cloud optionality, and space/defense names. If oil, U.S. yields, and foreign selling return together, the June 12 bounce was only technical.

Final View

The first half of June did not solve the risks. It tested whether Korean leadership stocks could still find buyers after a rates, oil, and AI crowding shock. The answer is: some buyers returned, but confirmation is still pending.

The practical strategy is not to increase the number of positions. Keep exposure focused on the leadership names that passed the stress test: foreign-bought memory, high-liquidity AI infrastructure, and selective space/defense names with turnover. The decisive late-June question is whether foreigners keep buying Korean semiconductors after FOMC.

Evidence Ledger