Publicar

Long-term bullish on AI, so why not dare to chase Hynix? Hynix just released its earnings report, with profits up 557%, breaking records, but the stock price first fell then rose because the market thought it "didn't rise sharply enough" — the expectation was 64 trillion, but the actual was only 60.5 trillion. Where's the gap? It sold a lot of HBM (high-end AI memory), but the main driver of this price increase was ordinary storage chips. HBM's proportion is too high, so it didn't fully benefit from the price hike, making the performance "good but not good enough." However, management quickly came out to boost confidence: AI investment hasn't slowed, HBM4 has already shipped, and long-term contracts are signed for 5 years. As soon as this was said, the stock price turned positive. This morning, South Korean stocks: Hynix rose 4%, Samsung rose 6%. But strangely, the day before, US AI hardware stocks collectively crashed, the Philadelphia Semiconductor Index fell 6%, SanDisk dropped 16%, and the Nasdaq 100 corrected by 10%. Only Seagate (which makes hard drives) rose against the trend because it said capacity is booked through 2028, and customers even see 2029. So now in this industry chain, on one side, record-breaking performance triggers sell-offs; on the other, capacity for the next three years is fully booked. Short-term valuations are high, expectations are high, and any slight miss causes a sell-off; long-term, AI demand is still there, so everyone rushes to lock in long-term orders. So it's not that people don't believe in AI, they're just afraid of "buying too high" in the short term, so they say long-term but don't dare to chase — that's the dilemma. In short: the industry boom hasn't stopped, but stock market sentiment is already very tense. $BTC $ETH $SNDK #海力士业绩创纪录但不及预期，存储股剧烈波动 #财报观察员：微软Meta亚马逊今夜交卷 #停火48小时告吹，美伊边打边谈