Let's talk about the impact of Blue Owl's withdrawal from Oracle's $10 billion data center project financing in Michigan, which was also the main reason for the sharp decline in the US stock market last night. The market is concerned about whether high capital expenditures and debt can be recouped, as well as concerns about the sustainability of future performance for chip manufacturers without funding to support new capital expenditures. From a personal perspective, it is also a logical continuation of the backtracking discussed last week. 1. The tightening of private equity credit caused by weakened market sentiment Faced with Oracle's continuously rising AI spending commitments and expanding debt scale, lenders have begun to demand stricter lease and debt terms, and if the risk credit is confirmed, lenders will directly withdraw. New cloud providers with high debt scales such as Oracle, Coreweave, and Nubis will all face this problem. In Mag7, several major technology manufacturers have relatively small debt scales, low interest rates, and less interest expenses, with stronger cash reserves and cash flow creation capabilities. 2. Another issue is the uncertainty of data center construction. There will be various emergencies in the approval, construction and construction. Last week, Coreweave's data center fell sharply due to the delay of the rainstorm construction period. Delay for cloud providers means delayed revenue recognition and interest accumulation. For chips, there may be sales delays, and large models still face a shortage of computing power. 3. Going deeper, it's still market sentiment concerns rather than the real risk of default Oracle's latest debt is $108 billion, with an additional $25 billion added in the past year due to data center investments. The remaining debt is long-term debt issued between 15-23 years. Annual interest expenses exceed 4 billion US dollars, and debt with a maturity of over 10 billion US dollars over 26 years. Over the past four quarters, Oracle's operating cash flow has exceeded $20 billion, covering interest expenses. The default risk in the next two years is actually not significant Previously, here was https://(x.com)/qinbank/status/1992490737573724577? S=46&t=k6rimWSEbo2D2TXolYcM-A has discussed the debt issue of Coreweave: the total debt is $18.8 billion, with an average cycle of nearly 5 years. Before 28 years, there were almost no large debts due, which means that in recent years, the main interest payments were around $1.2 billion per year, and Coreweave's revenue in 25 years was $5 billion. I looked at his financial report, which showed a quarterly revenue of $1.3 billion, but the net cash flow from continuing operating activities was $1.68 billion, indicating a high proportion of prepayments from clients, not to mention CoreWeave's remaining unfulfilled contracts exceeding $50 billion in the third quarter. Basically, 28 years ago, there was no need to worry about Coreweave's debt crisis. 4. What is the impact on the market? Previously, we talked about the recent macro events and periods of significant volatility, which continue to undermine valuations by taking advantage of the negative issues of important companies in the industry. I don't think AI has reached a turning point now, but as before, https://(x.com)/qinbufark/status/1988826907534860695? S=46&t=k6rimWSEbo2D2TXolYcM-A There will be differentiation when we talk about it here: as we move forward, we also realize that not all companies on the AI mainline are winners (not everyone will eat the dividends), and there will definitely be those who fall behind. Like all major industrial transformations, after experiencing the Great Leap Forward and reaching the final stage of the decisive battle, the winner takes all and the loser loses everything. In the end, companies with strong revenue growth momentum, good capital structure, and high efficiency will have the last laugh. There may be reshuffling in various tracks of the AI sector, and funds will gather towards these ultimate winners. Just like Mag7 itself is a product of the evolution of the competitive landscape. In the short term, we have to wait until these major events of this week are resolved before watching