$Circle (CRCL.US)$ Successfully holding above 150, the outlook remains bullish…

$NVIDIA (NVDA.US)$ 191 low, 196 high...let see

When investors are willing to pay for $Dell Technologies (DELL.US)$ and $Oracle (ORCL.US)$, it is not because they are making money, but because they are spending it.
This is a typical characteristic of cycles: when an industry is considered strategically irreplaceable, capital first rewards spenders rather than producers. The resilience of AI investment cycles also stems from the temporary justification of this 'spending rationale'.
Over the past year, the flow of AI capital has taken a sharp turn.
The first phase was the software boom—large models, agents, and copilots emerged one after another, with capital focusing on models and applications.
The second phase involves hardware and cloud — NVIDIA’s H100 has sold out due to overwhelming demand. $Advanced Micro Devices (AMD.US)$   $Intel (INTC.US)$ Oracle reviving through AI cloud services.
Now we are entering the third phase: the capital loop closure—model demand → computing power investment → model performance improvement → further expansion of demand.
This is the prototype of what Altman referred to as the 'AI closed-loop economy.' The difference is that this cycle is heavier, more expensive, and faster.
Let us first explain the core logic of this cycle.The current AI investment cycle is like a self-reinforcing gear system: it starts with the massive infrastructure CapEx of technology companies, buying chips, building data centers, and training models; then these investments are transformed into products and services, stimulating demand growth; demand, in turn, drives up revenue and profits, attracting more capital influx...