· 个股论点

建议做多$XLU LEAPS,看好AI电力需求激增及降息利好。

涉及标的:

中文翻译

交易思路:做多虚值(OTM) $XLU LEAPS(2年期,2027年12月/2028年1月到期)。 受AI驱动,这感觉像是一次千载难逢的做多机会。 $XLU 集中持有 $VST / $CEG 等电力公司。 两个原因: 1. AI数据中心(DC)电力使用带来的范式转变。 2. 基于历史平均水平(自2000年代以来持平)的低隐含波动率(IV)(约14%)。 AI的电力消耗是天文数字。 这一点怎么强调都不为过。 历史上从未有过数据中心消耗如此多吉瓦(GW)电力的情况,尤其是当它们需要核反应堆的输出来训练大语言模型(LLMs)时。 这迫使 $META、$AMZN、$GOOGL 等公司签署多年协议以尽可能多地消耗电力。然而,电力仍然供不应求。 -> 因此,数万亿美元可能会涌入电网升级。 通常利率会损害该板块,但我们将进入更多降息周期,这使得该板块成为更好的做多标的。 OpenAI致国会的信恳请美国投资能源以与中国竞争。 因此,由于以下原因,这感觉像是一次十年一遇的做多机会: - AI消耗所有可用电力带来的范式转变 - 为与中国竞争而进行的数万亿美元电网升级 - 降息 以及基于历史平均水平的低隐含波动率定价。

英文原文

Trade Idea: Long OTM $XLU leaps (2 years, Dec 2027/Jan 2028). This feels like once-a-generation long due to AI. XLU has concentration in $VST / $CEG power companies. Two reasons: 1. Paradigm shift due to AI DC electricity usage. 2. Low option IV (~14%) based on historical averages (flat since 2000s). AI power usage is astronomical. This cannot be understated. Never before in history have DCs use up this much GWs in power, especially when they require outputs of nuclear reactors for training LLMs. This forces $META, $AMZN, $GOOGL, and others to sign multi-year agreements to consume as much power as possible. And yet they still don't have enough. -> So, trillions would likely be poured into grid upgrades. Usually interest rates hurt the sector but we're going into more rate cuts, so it makes the sector a much better long. OpenAI's letter to congress pleaded the US to invest in energy as well to compete vs. China. So, this feels like a once-a-decade type long due to: - paradigm shift eating up any available power from AI - trillions in grid upgrades to compete vs. China - rate cuts. And low IV pricing from historical averages.

在 X 上查看原推 ↗