Analyst Rick points out that Nvidia is playing a central bank-like role in the AI era—while it cannot issue currency or determine interest rates, it is influencing a more critical resource allocation method: capital, the scarcest element in the AI era. The entire AI industry is entering a financialization phase, with capital itself becoming an integral part of AI infrastructure, rather than merely an external financing tool. In this system, Nvidia is increasingly resembling a "quasi-central bank" of the AI era. It doesn't create currency, but it is a key driver of the financialization trend in the AI industry, leading GPU manufacturers, cloud service providers, NeoCloud, packaging manufacturers, HBM suppliers, banks, private lending institutions, infrastructure funds, insurance capital, and sovereign wealth funds to jointly build a new capital ecosystem.
The AI industry is gradually evolving from a simple semiconductor supply chain into an infrastructure asset system.
Financing to purchase GPUs → GPU leasing generating cash flow → cash flow supporting a new round of financing → new financing continuing to purchase GPUs—this forms a continuously expanding capital flywheel. GPUs are beginning to possess infrastructure asset attributes, no longer just electronic devices, but assets capable of generating long-term cash flow, being collateralized for financing, and securitizable. This also explains why Nvidia continues to lock in HBM, advanced packaging, and supply chain capacity. If capital becomes one of the future bottlenecks, then helping customers solve financing problems is the most effective way to expand GPU sales; helping suppliers solve financing problems is the most effective way to expand capacity. This trend has already spread throughout the industry. Microsoft, Google, Meta, and Amazon continue to expand capital expenditures, and the orders, prepayments, and cooperation agreements they offer to customers or suppliers also help upstream and downstream companies in the industry chain obtain funds more easily. NeoCloud uses debt financing extensively; HBM suppliers continue to expand production; advanced packaging companies sign long-term agreements; private credit funds, infrastructure funds, and sovereign wealth funds have begun to invest in AI data centers. The AI industry chain is shifting from traditional manufacturing financing models to infrastructure financing models. In the future, if hyperscale cloud service providers rely heavily on debt financing for their hundreds of billions of dollars in annual capital expenditures, they will continue to absorb long-term funds. New capital is flowing not only to GPUs but also to data centers, power, fiber optics, cooling systems, HBM, and advanced packaging.
The AI industry may become the world's largest capital-absorbing industry (if not the largest). The impact of increased capital demand on financial markets cannot be ignored. Limited long-term capital supply and increased demand for AI financing mean that long-term funding costs are likely to remain high. Even if short-term policy rates decline, long-term financing rates may still remain high, driven by capital demand.
Meanwhile, capital allocation will become stratified: financing costs for AI companies with stable cash flow and high growth certainty may continue to decline, while financing costs for traditional industries may rise relatively, potentially widening credit spreads.
