$SKHY, Samsung
속보)9월 1~10일 수출 350억달러 '역대 최대'… 반도체 270% 증가
9월 1~10일 수출액이 350억달러를 기록하면서 기간 역대 최대치를 달성했다. 반도체 수출이 270% 이상 증가세를 보이며 전체 실적을 견인했다.
관세청이 11일 발표한 '9월 1~10일 수출입 현황(잠정치)'에 따르면 지난 1~10일 수출은 1년 전보다 82.6% 증가한 350억달러로 집���됐다.
특히 반도체 수출의 증가세가 두드러졌다. 반도체 수출은 270.1% 증가한 165억달러로, 같은 기간 기준 역대 최대다. 전체 수출에서 반도체가 차지하는 비중은 47.1%로 전년 대비 23.9%포인트(p) 확대됐다. 이 밖에 석유제품(57.0%), 승용차(10%) 등이 증가했다.
국가별로는 중국(103.0%), 미국(137.5%), 대만(176.0%), 베트남(42.2%), 유럽연합(38.9%) 등으로의 증가 수출이 증가했다. 상위 3국인 중국·미국·대만의 수출 비중은 51.9%를 기록했다.
한편, 9월 1~10일 수입은 20.7% 증가한 246억달러였다. 반도체, 반도체 제조장비 등에서 수입이 크게 늘었다. 수출에서 수입을 뺀 무역수지는 104억 달러 흑자를 기록했다.
$MU $SKHY $DRAM Reports say Chinese HBMs yielded 25%.
Want the "Original Content" version from someone who has actually worked in this field?
3 defective out of 4 (if true) does not make the 1 that barely yielded good. That die has maybe a 10% shot at surviving reliability and stress testing, and it was almost certainly a partial test flow, not the full quals. Best case it bins into a lower tier, worst case it dies early in the field. Yielding a part is one thing. Holding to spec (passing reliability test) for years is a completely different bar.
TSV know-how is not built overnight. The 4 to 5 year gap is the conservative read, not the generous one. Meanwhile the memory trio is already onto HBM4E on advanced-node base dies and developing 3D DRAM.
Tune out the Chinese FUD.
These new memory revenue forecasts (see below) from Nomura are mind boggling.
SK Hynix currently has ~25% of global memory revenues... if these Nomura estimates are correct and SK Hynix maintains their current market share... then it means SK Hynix would have approx $918 billion (USD) of revenues in 2030.
SK Hynix is expected to do ~$260B revenues in 2026... in order for them to hit $918 billion in 2030 they'd need to grow revenues at a 38% CAGR over the next 4 years.
Even if net income margins declined by 200 bps per year... SK Hynix would still be doing more than $500B of net income in 2030.
In this scenario... from 2026 through 2030... SK Hynix would generate $1.8T of net income which is 80% bigger than their current market cap.
Even if net income margins did decline by 200 bps per year... with these growth rates and profits... SK Hynix could easily buyback 15-20% of their shares every year (not including special dividends).
Because of the margin compression... net income would only be growing at a 30-32% CAGR however with the 15-20% annual buybacks... EPS would be growing at a 55-60% CAGR through 2030.
Crazy stuff.