𝑭𝒐𝒓𝒆𝒔𝒕𝒆𝒓

@BtcForester

TradFi & Crypto researcher | E/ACC practitioner 传统金融与加密研究者 & 风险管理实践者｜牛熊沉浮皆入局

Seattle

Joined May 2022

617 Following

167 Followers

2.2K Posts

Pinned Tweet

𝑭𝒐𝒓𝒆𝒔𝒕𝒆𝒓 @BtcForester

7 months ago

看着美股闪迪（SNDK）持续爆拉，这几天在尝试计算内存行业估值水平，查了些历史数据。先说总结：当前这轮更接近 AI 驱动的结构性行情，而不是简单补库存。HBM、DDR5、服务器 DRAM 的供需紧张短期很难缓解，即使厂商扩产，释放速度也有限。我个人判断高端内存价格高位至少还能维持到 2026 年 Q4，期间可能频繁震荡向上，但不太会快速回落。但 HBM 和服务器内存继续偏强。手机、PC 用的普通 DRAM 更可能在 2026 年进入平台期，市场份额持续萎缩。搜索到历史数据。第一轮，2016 到 2018 年。 DRAM 合约价从 2016 年下半年开始上涨，一直涨到 2018 年中后期，时间接近两年。期间服务器 DRAM 和移动内存涨幅最大，多数型号累计涨了 2 到 3 倍。真正见顶是在 2018 年下半年，前提是三星、海力士、美光同时扩产，客户库存明显偏高。第二轮，2020 年下半年到 2022 年初。疫情后需求反弹，叠加居家办公、云计算扩张，DRAM 价格从 2020 年三季度启动，上涨大约 12 到 15 个月。2021 年中是高点，之后因为 PC 和手机需求转弱，价格在 2022 年快速下行。第三轮，HBM 和 AI 相关内存。 2017 年第一次 HBM 放量，价格强势维持接近两年。 2023 年下半年开始的新一轮 HBM 行情，到现在仍在延续，核心原因是 AI 服务器需求远超预期，且产能爬坡慢，供给长期偏紧。从这些数据可以看到，只要是供需结构改变导致的上涨，至少持续一年以上，两年左右见顶。（以上照猫画虎主观预测）

297

𝑭𝒐𝒓𝒆𝒔𝒕𝒆𝒓 @BtcForester

1 day ago

@dapipi99 @shichuangang 老梁现在玩油管吧

𝑭𝒐𝒓𝒆𝒔𝒕𝒆𝒓 @BtcForester

2 days ago

@dapipi99 你没告诉他剪一个可以做男模的发型？

BtcForester retweeted

Mark Minervini

@markminervini

6 days ago

Don’t Fight the Fed. Don’t Fight the Tape. The Federal Reserve appears increasingly likely to be on a tightening course, with potentially one or two rate hikes ahead. The probability of a hike at the next meeting is currently around 55%, and much of that risk may already be discounted in the market. Still, the larger point shouldn’t be overlooked: the Fed is leaning toward tightening, not easing. There are two old market rules I still respect deeply: Don’t fight the Fed, and don’t fight the tape. Right now, those two signals are somewhat at odds. The Fed is becoming a potential headwind, while the tape remains relatively bullish. The major indexes are holding up well, and certain areas of the market continue to advance. But beneath the surface, participation is clearly thinning and the market is becoming increasingly fragmented. That distinction matters. Some of the leadership we had been looking to in the AI space—particularly companies tied to the data-center buildout—is beginning to encounter resistance. At the same time, a negative political narrative around data centers has been developing. Meanwhile, money continues to gravitate toward the largest, most established companies. $FNGS is rallying as investors seek the relative safety of mega-cap companies with proven earnings power. That can keep the indexes looking healthy even while conditions underneath become considerably more challenging. A narrowing market is not necessarily an immediately bearish market, but it is a more difficult market to trade. When participation contracts, fewer stocks respond favorably, breakouts become less dependable, and stock selection becomes increasingly important because the market is simply not lifting all boats. Add to that the seasonal backdrop. September has historically been the weakest month of the year. For now, the tape still gets the benefit of the doubt. But this is not an environment for complacency. I continue to hold select longs, most of which are profitable trades that I've owned but are still respecting stops. As far as a catalysts, I'm keeping a keen eye on the 10-year yield, which is at a very important level. Should the 10-year breakout to the upside (especially above 5%), that would likely put the market in a challenging position going into September. Why do I think the 10-year yield at an important level and a breakout to the upside would be bearish? Because right now most expect the Fed to tighten 25 basis points one or two times. However, it's not of the common opinion that the Fed is behind the curve. If interest rates keep rising, that will be an indication that the Fed is falling behind the curve and they may have to tighten more aggressively. So the 10-year right could be a key catalyst, in my opinion. https://t.co/JXzFFTmMtn

137

372

168K