A detailed read-through of Samsung Electronics' 2024-2026 FCF return policy, 2026 AI memory earnings scale, and the separate DS performance-bonus treasury-share purchase flow.
HBM, HBF and HBC are not the same type of memory. This post maps bandwidth, capacity and inference-cost as three distinct walls, compares maturity levels honestly, and explains why the three are complements rather than competitors.
This note reframes the KOSDAQ recovery question. The trigger is not a collapse in Samsung Electronics and SK hynix, but a phase where strong memory earnings remain intact while the two megacaps stop absorbing all market flow. The realistic window is mid-July to early August, after Micron's June 24 earnings, Samsung Electronics' Q2 preliminary result expectations, and SK hynix's Q2 expectations are absorbed. The first landing zone is likely KOSDAQ semiconductor test, equipment and metrology, not broad biotech or unprofitable growth.
In the MSCI annual market classification announcement due early on June 24, 2026, Korea's odds of an immediate promotion to Developed Markets are around 1%. The real event is whether Korea is re-added to the Developed Markets Watch List. Combining two analyses, the Watch List probability is 30-35%, the base case is positive monitoring (about 53%), and the most actionable beneficiary is not a chase of semiconductor names but a conditional re-rating in KB Financial.
A proprietary Korea Invest Insights screen combining USD/KRW, May and early-June export data, sector FX sensitivity, 2Q consensus and recent flow. The key is not generic exporter exposure, but dollar revenue, won costs, improving volume or pricing, and under-reflected consensus.
A deep dive on TechWing as a Korean back-end semiconductor test equipment company. The note connects memory test handlers, recurring C.O.K revenue, HBM Cube Prober, Samsung Electronics, SK hynix and Micron customer expansion, derivative losses, valuation and flow. The conclusion moves TechWing from watchlist to conditional buy candidate, but only after support or breakout confirmation.
We test whether the 2027 earnings consensus for Samsung Electronics, SK hynix, Micron, and Nvidia is mere sell-side optimism or a level demand can actually afford to pay, all within one frame. The verdict splits. Hyperscalers can pay conditionally, government and sovereign AI are supplementary demand, and PC and smartphone OEMs are already in the unpayable zone. The crux is not whether AI demand exists, but whether AI revenue and GPU utilization rise fast enough to justify CAPEX beyond 2027.
A reproduced KOSPI disparity framework showing why +28.6% above the 60-day moving average is better read as a partial rebalancing signal than an outright market-top call.