Post

峰哥的交易日记
峰哥的交易日记
Show original
从1255跌到698,美光到底了吗? 基本面炸裂到没朋友,营收暴增256%,净利润850亿,毛利率87%——但股价从高点腰斩40%。就在刚刚,698美元的多头拥挤到80%以上,全都等着反弹解套。这到底是黄金坑,还是万人坑? 先看表面:好公司,烂股价。 美光FY2026营收1332亿美元,同比暴增256%。Q4单季营收542亿,同比涨了379%,EPS 33.42美元,毛利率87%。下季度指引615亿,管理层说“这是全年低点”。 你见过哪家“低点”的指引比同行全年营收还高?这就是AI内存的疯狂。 但股价呢?从6月ATH 1255美元一路砸到662,现在698。一个月跌了33%,一周跌了18%。 基本面在天上,股价在地下,这种撕裂才是市场最残忍的地方。 第一件事:利空被放大了,但你不信。 民主党中期选举赢了,市场怕什么?怕数据中心审批变严,怕AI资本开支缩水,怕债务融资成本压垮数据中心建设。再加上中国供给担忧被媒体一放大,芯片股集体跪了。 但你冷静想想:1500亿美元的RPO(最低履约义务)摆在那里,长期客户协议大部分有价格地板。订单签了,价格锁了,你怕什么? 怕的是情绪。市场从“AI无限叙事”切换到“融资成本质疑”,这种切换一旦发生,好公司的股价也会被无差别砸。 市场恐慌的时候,从来不区分好公司和烂公司,它只区分“有没有人接盘”。 第二件事:估值便宜到离谱,但你不敢买。 基于已经实现的盈利,MU的P/E只有个位数到低双位数。前瞻更低。分析师共识Strong Buy,目标价1500+。 你花698买一个毛利率87%、净现金充裕、自由现金流爆炸的公司,贵吗?不贵。但你怕它再跌。 你不是不知道它便宜,你是怕买了之后别人比你更恐慌。 第三件事:多头太拥挤,这才是真正的雷。 OI约2亿USDT,账户多头占比80%以上,顶级交易员也是多头占优。但taker偏空,资金费率微正。 所有人都觉得“跌这么多了该反弹了”,全挤在多头一边。 这种结构最怕什么?最怕一根阴线,把所有人止损打掉,然后才真正见底。 多头拥挤不是因为看对了,是因为没人愿意认输。 多空对决,你自己看 一边是: FY2026营收+256%,净利850亿,毛利率87% Q1指引615亿,管理层称全年低点 1500亿RPO+价格地板,可见性极强 估值个位数P/E,分析师目标1500+ 12月回购限制到期,大规模回购预期 一边是: 从ATH跌40%,1个月跌33% 日线/周线空头结构,价格远低于均线 多头拥挤80%+,挤仓风险高 宏观通胀反复,利率压制成长股 中期选举后政策不确定性 关键位置 上方阻力:720-750(近期高点/枢轴)→ 800+ 下方支撑:650-670(近期低点)→ 620-600 操作策略 短线选手:698别追。等回落到670-680轻仓试多,止损650下方,目标710-730。放量站稳720再加仓看750。破650直接看620-600。 波段玩家:等日线企稳信号——站稳均线、RSI回升、成交量配合。目标800-900。如果宏观恶化再探600附近,那是更好的中长期布局区。12月回购启动是催化剂,提前埋伏但别满仓。 长线信仰者:600-650分批定投。基本面黄金坑,估值便宜到离谱。2027年目标看分析师共识1500+,但周期股要有耐心。记住:好公司不等于好股票,好价格才等于好机会。 MU现在就像2022年底的美光—— 99%的人觉得“芯片周期见顶了”,结果从50干到150。$BTC $MU $NVDA
峰哥的交易日记
峰哥的交易日记
NVDA at $176, do you dare to buy? A single-day plunge of 9%, market cap evaporated by 400 billion, CEO rumors flying everywhere, regulatory threats looming — everyone is shouting "The AI bubble has burst." Just now, the price dropped to around 175, RSI hit 25 in the oversold zone, and the Q3 earnings report on November 17 is just days away. Is this the end of the bull market or a golden buying opportunity? First, look at the surface: bad news lined up to crush the market, bulls are beaten down and can't lift their heads. From the October high of 243 down to 176, a drop of over 20% in one month. It broke below the 50-day and 200-day moving averages, the weekly rising wedge pattern broke down, and MACD shows a bearish alignment. The news is even harsher: post-midterm election regulatory concerns intensify, export controls may tighten again, CEO personal matters and board investigation rumors pressured pre-market. On November 5, a single-day drop of 9%, market cap evaporated over 400 billion USD. First thing: what's falling is the stock price, not computing power. Latest earnings: Q2 revenue 96.2 billion USD, +106% year-over-year; data center revenue 89 billion USD, +117% year-over-year; gross margin 75%. Q3 guidance 108 billion USD. FY2028 preliminary revenue growth expected around 70% — note, this is under "supply constrained" assumptions, actual demand is higher. The Vera Rubin platform has already started shipping, accounting for about 20% of data center revenue next quarter. Big customers like AWS are still increasing orders. In plain language: AI demand hasn't collapsed, orders haven't collapsed, gross margin hasn't collapsed. What collapsed is your leverage and sentiment. Regulation is a real risk, but it won't kill computing power demand. It will only concentrate the survivors. NVDA is the concentrated beneficiary. Second thing: the $176 price already factors in a lot of pessimistic expectations. Forward P/E has significantly declined, some analysts point to a range in the teens. Most Wall Street price targets remain at 250-300+. Do you know what this means? The market is not pricing in "NVDA is failing," but rather "NVDA can't skyrocket like before anymore." These are two different things. Customer self-developed chips, memory cost squeezing gross margin to 71-73%, China revenue basically zero — these negatives are already priced in. What you're afraid of now is what you should have feared three months ago. Third thing: technically oversold, earnings are the turning point. RSI daily has hit 25-30, the oversold zone. 175-170 is strong support, 164-165 near the 52-week low. Volume has increased confirming selling pressure, but after oversold, a technical rebound can come anytime. Key event: Q3 earnings on November 17. If guidance beats expectations and FY2028 growth is confirmed, a V-shaped rebound is not a dream. If it misses expectations or regulations tighten further, it may dip to 165 or even lower. Those who cut losses before earnings will likely regret it after, but those fully invested before earnings may also get buried. This is the current situation. Bull vs. bear showdown, judge for yourself: On one side: Revenue +106% YoY, data center +117%, gross margin 75% Vera Rubin shipments started, AWS continues to add orders Forward valuation significantly digested, Wall Street targets 250-300+ RSI oversold, 175-170 strong support Earnings could be a catalyst for reversal On the other side: Regulatory concerns intensify, export controls may tighten CEO-related rumors suppress sentiment Moving averages bearish alignment, weekly structure weakening Interest rates 3.75-4.00%, high rates suppress growth stocks Breaking below 175 may accelerate drop to 165 Key levels Resistance above: 181-185 → 190-195 → 200-205 (200-day MA) Support below: 175-170 (strong support) → 164-165 (52-week low) Trading strategy Short-term traders: Don't heavily short; oversold rebound can come anytime. Hold 175 with volume rebound, light long positions, target 181-185, stop loss 172. If 175 breaks with volume, wait or light short, target 170-165, stop loss above 178. Shorting here risks getting stuck at the bottom. Mid-term players: Wait for November 17 earnings clarity. Aggressive traders can build long positions in batches at 170-175 with low leverage, stop loss below 165, target back to 200+. Conservative traders wait to hold above 185-190 before adding. Long-term believers: If you are bullish on AI long-term, the value at 176 is already attractive. Dollar-cost average on dips or hold spot with low leverage, target 250+. But be prepared to endure a 10-15% further pullback. Long-term means not watching the chart obsessively, or watching but not panicking. NVDA now is like NVDA at the end of 2022 — 99% of people thought "The AI story is over," but it went from just over 100 to 240. $BTC $ETH $NVDA

Disclaimer: OKX Orbit content is provided for informational purposes only. Learn more

Replies

No comments yet. Be the first to reply!