🐦 推特JukanV3 · 实盘不做把握 5/10
北京时间·🕒 发布 06-28 14:43:45时区推断·⟳ 抓取 14:11:28回填·⚡ 决策 06-29 15:15:41·120.0 天前
Jukan转述IBM CEO播客金句力挺CSP资本支出可持续(企业AI第二年10倍回报飞轮)
信号本质是KOL转述IBM CEO的播客观点(企业AI第一年净亏、第二年10倍回报、CSP capex飞轮),是宏观叙事评论而非IBM个股催化,$255B巨头不会因一条播客金句异动。
📄 原文Jukan@jukan05↗ 原文
Many people point to CSPs already running negative cash flow and ask how on earth CAPEX can keep growing from here. It…
不做 IBM.US
纯宏观叙事评论(KOL二次转述CEO播客旧观点),非IBM个股催化;$255B巨头不会因一条播客金句异动;IBM近5日已涨9%(含周五+5.2%)系大盘/AI行情驱动而非此事件,无新信息、无edge,追多易追高。
展开详情(成交明细 / 研究档案 / 原文全文 / 决策记录链)
研究档案
🔎 新鲜度:扩散 high · price-in 风险 high ↗
关键理由
- 信号本质是KOL转述IBM CEO的播客观点(企业AI第一年净亏、第二年10倍回报、CSP capex飞轮),是宏观叙事评论而非IBM个股催化,$255B巨头不会因一条播客金句异动。
- Krishna这套'第一年亏/第二年拐点'是已被广泛传播的旧叙事,无新增、具体、可交易的信息,无edge。
- 若想交易其'CSP capex可持续'主题,推文未点名任何供应商、无任何新数据,顺此映射产业链等于无依据地凭空造一笔交易。
主要风险
- 把宏观叙事误当IBM催化去追多:IBM近5日已涨9%(含周五+5.2%)、13周轨迹91→124已回落至105,此时追高易套牢。
- AI capex宏观叙事方向偏多本身没错,但本信号无新增信息,误判为'可执行催化'是主要看错点。
原文全文
展开原文 · 1,023 字
Many people point to CSPs already running negative cash flow and ask how on earth CAPEX can keep growing from here.
It is a foolish question.
Look at what Arvind Krishna (IBM CEO) said in a podcast conversation.
“Year one of enterprise AI is a net loss. That is exactly why 80 percent of companies right now look like they have failed. It comes down to more engineers, higher token costs, and opportunity cost.
Year two is precisely when the 10x return kicks in. Most companies are not failing at AI. They are simply at the worst point on the curve, right before ROI compresses and improves.”
Ultimately, at this 10x return inflection point, the CSPs’ own revenue base will be expanded by AI, so even as CAPEX grows in absolute terms, the CAPEX to revenue ratio will decelerate, eventually setting off a virtuous cycle, a flywheel effect.
Why do people look to the future when judging CAPEX, but look only at the present when judging cash flow? Cash flow should also be judged by pulling forward the future cash flow.
It is a foolish question.
Look at what Arvind Krishna (IBM CEO) said in a podcast conversation.
“Year one of enterprise AI is a net loss. That is exactly why 80 percent of companies right now look like they have failed. It comes down to more engineers, higher token costs, and opportunity cost.
Year two is precisely when the 10x return kicks in. Most companies are not failing at AI. They are simply at the worst point on the curve, right before ROI compresses and improves.”
Ultimately, at this 10x return inflection point, the CSPs’ own revenue base will be expanded by AI, so even as CAPEX grows in absolute terms, the CAPEX to revenue ratio will decelerate, eventually setting off a virtuous cycle, a flywheel effect.
Why do people look to the future when judging CAPEX, but look only at the present when judging cash flow? Cash flow should also be judged by pulling forward the future cash flow.
决策记录链 · 只追加
每次领取产生一条记录,重新决策追加一条并以 supersedes 指向被替代的记录(ADR 0014)。
这条决策早于决策账本(2026-10-08 起逐条追加),没有追加记录。