美股投资young

美股科技投资者 ai狂热者分享每日快讯. all in ai

303Following

764followers

Feed

Livestream

Feed

Storage is now priced as "peaking next year," but given current profitability, already signed long-term contracts, and ongoing buyback cancellations, this discount is too steep. First, let's see how much it has dropped. As of the close of US markets on 11/9, $MU closed at $733.15, about 42% below its 52-week high of 1255; $SNDK closed at $886.51, about 62% below its peak of 2354. On the Korean market close on 11/10, Samsung was at 198,900 KRW, SK Hynix at 1,212,000 KRW, about 47% and 59% below their 52-week highs respectively. Just in the past month, these four companies have dropped between 20% and 40%. There are three main pressures. The heaviest is the Democrats taking both houses of Congress after the midterm elections, plus winning governorships in major data center states like Texas and Georgia. The market fears that tighter electricity prices and approvals will delay new data centers, pushing back orders for HBM and server DRAM. The second is Nvidia's internal investigation scandal; Korean media say the short-term impact on HBM shipments is limited, but export compliance and subsequent share allocation have become more uncertain. Hynix, which supplies the highest proportion of HBM4 to Nvidia, is hit hardest. The third comes from The Information's report: TSMC is researching advanced storage, and Nvidia wants TSMC to take over the underlying dies of HBM. In the long term, storage manufacturers' influence may be partially diluted. This is far from mass production and mainly affects sentiment. How low have valuations fallen? Based on market data on 11/9, $MU's trailing P/E is about 9x, forward about 4x; $SNDK's trailing about 7x, forward about 4x. Samsung and Hynix, as of 11/6 close, have 12-month forward P/E of only 2.7x and 2.2x respectively; I couldn't find reliable trailing P/E figures on the same basis. For comparison, at the peak of the last supercycle in 2018, these two were still at 4 to 6 times. Low P/E ratios in the storage industry are often traps, with valuations looking cheapest when profits are fattest. What's different this time is that orders are signed far into the future. Micron recently disclosed that 26 long-term supply agreements cover over 35% of expected revenue through 2030; media reports say its HBM is sold out this year, backed by about $100 billion in firm contracts and $22 billion in customer prepayments. For Hynix, Korean media and KB Securities say over half of capacity for 2027 to 2028 is already under five-year contracts, and with ongoing negotiations, DRAM and HBM are basically sold out through 2029. Some brokers estimate Samsung's long-term contracts account for about 70% of shipments. $SNDK said in its August earnings report that it has signed 10 new commercial model agreements. Next quarter guidance remains firm: $MU guides revenue of $61.5 billion, gross margin about 86%, $SNDK guides $12.2 to $12.6 billion. Brokerage opinions diverge mainly on 2028. Morgan Stanley maintains an overweight on $MU with a $1200 target, and overweight on $SNDK with a pre-split target of $1750; UBS rates $MU a buy with a $1625 target. The same Morgan Stanley cuts Samsung's target to 309,000 KRW and Hynix to 1,600,000 KRW, citing new capacity releases starting end of 2027 and possible margin pressure in 2028. Goldman Sachs believes the two Korean companies have bottomed, while KB Securities maintains targets of 600,000 KRW for Samsung and 3,500,000 KRW for Hynix, calling for a W-shaped rebound in November to December. How to spend the money is the most undervalued part. Cloud providers are pouring massive capital expenditures into storage, and storage manufacturers are using the cash not only for capacity expansion but increasingly for buyback cancellations. Hynix approved a 40 trillion KRW buyback in August and canceled all shares, about 3.3% of total shares outstanding, raising its shareholder return target from "within 50%" of cumulative FCF from 2025 to 2027 to "over 50%," with total returns this year around 100 trillion KRW. Samsung is executing a policy to return 50% of FCF to shareholders from 2024 to 2026, with this year's scale at 90 to 110 trillion KRW, about 30 trillion of which is dividends, with the rest to be decided in January next year, including buybacks and cancellations. $MU's buyback authorization has reached $35.16 billion, $SNDK has $15.5 billion remaining, about 8.6% of the company at August prices. So the judgment is straightforward: given current profits, orders signed through 2030, and real cancellations, storage should not be discounted as a cyclical stock at its peak. But not being discounted and rising immediately are two different things. The next things to watch are: whether there is clear political guidance on data center approvals; whether DRAM and HBM contract prices reverse; whether $MU's 12/16 earnings report shows any renegotiation or delay of long-term contracts; and whether buybacks are executed with real cash quarterly. If buying, do so in batches, avoid leverage, don't chase after single-day spikes, and accept the sector's volatile temperament.

This week's $NVDA is like a chaotic weather forecast, what to do? young has a clever trick! Use young's strategy on OKX to carve out your own path in such a volatile market! 🌫 Fog As soon as the midterm elections ended, the market started worrying about the Democrats tightening data center approvals. On November 5, the Philadelphia Semiconductor Index dropped about 12% in one day, $NVDA fell about 9%. What’s being crushed isn’t the immediate orders, but the AI expansion pace for the next few years. ⚡ Thunder Reportedly, Jensen Huang admitted to being unfaithful to his wife, and the board has launched an independent investigation. The espionage allegations are not confirmed yet, and he still holds the CEO position. Oh, young is really unlucky, young didn’t cheat, so why is my account bleeding? But as long as the investigation isn’t clear, this cloud will hang over the stock price. 🌧 Rain After the US market closed on November 17, earnings were released. Consensus expected revenue around $108.8 billion, EPS about 2.47. At the end of October, it was still above 200, now it’s dropped near 178. How heavy this rain will be depends on whether the guidance can blow away the previous two clouds. Fog, thunder, and rain all squeezed into the same week. Going out empty-handed will get you soaked, rushing in bare-chested is even more dangerous. So this week I prepared three umbrellas on OKX 👇 ☂️ Long-handled umbrella: Buy 185 call Path: Options → Simple Options → xNVDA → Expiring 11/20 185 call, premium 6.67 Break-even 191.67 If it rises to 200 → more than $8 left per share Stuck below 185 → 6.67 all lost Betting on a decent rebound after earnings, maximum loss fixed at order time. 🌂 Folding umbrella: Hold xNVDA, then sell a 195 call Switch to “Sell” on the same page, select 195 call, can collect 2.74 premium now. Selling calls requires xNVDA as margin, so this is a “spot + options” two-legged position. If it doesn’t exceed 195 at expiration → collect 2.74 steadily, lowering cost basis to 175.32. If it breaks 195 → sell at 195, max profit about 19.68 per share, roughly 11%. If it really falls → 2.74 cushions the fall first. Suitable for those bullish on NVDA long-term but don’t believe one earnings report will push it back to 200. 🚶 No umbrella, wait for the rain to stop before going out: Stock perpetuals Don’t touch before earnings. Wait for guidance upgrades, next day’s open holds, volume follows. Then open a small position with perpetuals, can exit anytime if wrong. Among the three umbrellas, I like the folding umbrella best. NVDA is originally a stock to hold long-term; during earnings week, instead of guessing direction, better to turn waiting time into premium. If you really want to gamble, use the long-handled umbrella, max loss just over 6 bucks. Tokenized US stocks, simple options, stock perpetuals—all switchable within the same OKX account. Bullish or bearish, buy or sell, all done on one page. When US stocks open at night, the whole operation can be completed inside a crypto account. @misaENFP @star_okx @okx @okxchinese #OKX达人 #交易TradFi就来OKX

Today, the Asia-Pacific market is very likely to open higher, especially South Korea. Yesterday's rebound in the US stock market was not reflected at all during the day in South Korea. The time difference needs to be clarified. When South Korea closed on 11/9, the US stock market had not yet opened. The KOSPI closed down 0.81% at 5018.85, the lowest closing price since February. Samsung once rose 3.55% intraday, and Hynix once rose 5.39%, but both almost fully gave back their gains, closing up only 0.64% and 0.35% respectively. Foreign investors have been net sellers for seven consecutive trading days. Then the US market opened, with the Philadelphia Semiconductor Index surging 6.85%, $MU up 6.25%, $SNDK up 6.2%, and TSMC ADR up 3.5%. The US-listed Korean ETF $EWY rose 3.7%, and the Japanese ETF $EWJ rose 3.65%, effectively trading this rebound in advance on behalf of Japan and South Korea. Nikkei futures also closed at 60970, significantly higher than the Nikkei's 11/9 daytime close of 58775. So today's higher open is basically already priced in. What really matters is whether the market can hold after the high open. The issues with Hynix and Samsung are different. Hynix is the main supplier of Nvidia's HBM. Yesterday, the Philadelphia Semiconductor Index rose nearly 7%, but $NVDA actually fell 4.27%. Storage rose while Nvidia fell, putting Hynix right in the middle. It has dropped more than 30% from its one-month high, so theoretically it has the most room to rebound, but as long as Nvidia's investigation results are not out, it will be difficult for Hynix to rise decisively like Micron. Samsung faces less pressure; it does not rely entirely on Nvidia's HBM, and ordinary storage price increases affect it more directly. There was also a negative news last week: it was rumored that TSMC is researching 3D DRAM at Nvidia's request. Jefferies' view is that if TSMC really wants to enter storage, it might be as late as 2035, and the technology path and factories would have to start from scratch. This negative news is exaggerated in the short term. What I care more about is foreign investors. This round of sharp decline in South Korea was driven by foreign investors selling, retail investors selling, and only institutions buying. After today's high open, whether foreign investors continue to sell or stop is more telling than a few points of gains. If Samsung and Hynix weaken all the way after the high open and close below the opening price, it will be exactly the same pattern as yesterday: a pump to let foreign investors sell. If they can hold the opening price and foreign investors turn to net buying, then it can be said that the rebound in the US stock market has truly been connected. Japan is similar. The Nikkei has pulled back significantly from its October high. Today's high open is certain, but whether it can reclaim 60,000 points depends on whether semiconductor equipment stocks follow. This US stock market rebound is a gap fill, and the political obstacles have not yet been removed. Asia-Pacific is following the rebound with the same nature. Don't chase the high open; watch the close.