The bottomless pit of AI infrastructure spending finally has an end in sight.
Oracle $ORCL laid out a “spending timeline” at its investor day on October 28 (U.S. time): capital expenditures will peak in fiscal 2027 and 2028, then decline starting in fiscal 2029; free cash flow will turn positive in the second half of fiscal 2028, with a target of $30 billion in fiscal 2029; and new financing through fiscal 2030 will be capped at $20 billion ($10 billion each in equity and debt; debt refinancing excluded).
The market voted with its feet: Oracle closed up about 10.7% on October 28 at around $144, after climbing more than 13% intraday. It recovered over $40 billion in market value in a single day, its biggest one-day gain in nearly five months.
The logic is clear: what’s been hammering AI stocks these past few days isn’t “weak growth”—it’s “bottomless spending.” Meta fell nearly 7% after its earnings, while Alphabet dropped about 10% after hours. The key question in both cases was when spending would turn into cash flow. Oracle’s timeline directly answers the market’s biggest fear: the spending has an endpoint. And there’s more to back it up than rosy promises: customer prepayments covered about 40% of data center spending last quarter; anchor customer OpenAI is already in the fold; and Nvidia’s next-generation systems and Google TPUs are set to go live through OCI starting in 2027. The orders are real.
But the risks are plain to see: ① These are forward-looking targets, not contracts, and delivery depends entirely on execution and AI demand; ② customer concentration: if a hyperscale customer cuts its orders, who will buy 10 gigawatts of computing capacity? ③ The stock is already down about 26% for the year, so this rally is a recovery in expectations. Until $30 billion in cash flow becomes reality, any delivery delays will be punished twice as hard by the market.
In a nutshell: Oracle has rewritten a “horror story” as a blueprint with a deadline. Before chasing the rally, remember what you’re betting on: cash flow in 2029, not today’s price.
Independent research; not investment advice.
#美股 #甲骨文 #AI基建 #资本开支 #EarningsSeason
Oracle $ORCL laid out a “spending timeline” at its investor day on October 28 (U.S. time): capital expenditures will peak in fiscal 2027 and 2028, then decline starting in fiscal 2029; free cash flow will turn positive in the second half of fiscal 2028, with a target of $30 billion in fiscal 2029; and new financing through fiscal 2030 will be capped at $20 billion ($10 billion each in equity and debt; debt refinancing excluded).
The market voted with its feet: Oracle closed up about 10.7% on October 28 at around $144, after climbing more than 13% intraday. It recovered over $40 billion in market value in a single day, its biggest one-day gain in nearly five months.
The logic is clear: what’s been hammering AI stocks these past few days isn’t “weak growth”—it’s “bottomless spending.” Meta fell nearly 7% after its earnings, while Alphabet dropped about 10% after hours. The key question in both cases was when spending would turn into cash flow. Oracle’s timeline directly answers the market’s biggest fear: the spending has an endpoint. And there’s more to back it up than rosy promises: customer prepayments covered about 40% of data center spending last quarter; anchor customer OpenAI is already in the fold; and Nvidia’s next-generation systems and Google TPUs are set to go live through OCI starting in 2027. The orders are real.
But the risks are plain to see: ① These are forward-looking targets, not contracts, and delivery depends entirely on execution and AI demand; ② customer concentration: if a hyperscale customer cuts its orders, who will buy 10 gigawatts of computing capacity? ③ The stock is already down about 26% for the year, so this rally is a recovery in expectations. Until $30 billion in cash flow becomes reality, any delivery delays will be punished twice as hard by the market.
In a nutshell: Oracle has rewritten a “horror story” as a blueprint with a deadline. Before chasing the rally, remember what you’re betting on: cash flow in 2029, not today’s price.
Independent research; not investment advice.
#美股 #甲骨文 #AI基建 #资本开支 #EarningsSeason
