Hyperscalers including Meta are managing return on investment burdens rather than cutting AI infrastructure spending outright, according to DB Securities analyst Lee Jin-kyung. Meta decided to sell idle computing capacity externally and some datacenter projects faced delays or cancellations, prompting market concerns about AI investment cycle slowdown. The analyst stated that big tech firms are extending equipment useful lives, selling non-core assets, and reducing working capital to improve capital efficiency without signaling investment pullback. This approach transfers cost pressures to suppliers in the AI supply chain who must hold more inventory and wait longer for payments while hyperscalers optimize their own cash flows.
Hyperscalers Extend Asset Lives and Reduce Working Capital
AI infrastructure investment remains critical for competitive positioning among hyperscalers including Amazon, Microsoft, Google, and Meta, making spending cuts difficult to implement directly. Lee Jin-kyung of DB Securities stated that "reducing working capital is an effective means to improve capital efficiency without negative signals, unlike extending useful lives." The analyst explained that hyperscalers are responding by extending equipment useful lives, divesting non-core assets, and optimizing working capital rather than cutting investment volumes. These measures allow companies to maintain AI infrastructure buildout while improving financial metrics, as determining whether current investment levels constitute overinvestment requires waiting for actual demand and revenue confirmation after datacenter construction completion.
Supermicro Announces $7 Billion Fundraising Amid Inventory Buildup
DB Securities cited Supermicro as a representative case of supply chain burden transfer. The AI server manufacturer experienced revenue growth alongside rapid increases in inventory and accounts receivable, leading the company to announce a $7 billion fundraising plan. Market concerns centered not on dilution from equity issuance alone but on the structural inability of cash generation to keep pace with revenue growth. When hyperscalers reduce inventory and extend payment terms, their cash flows improve while server, component, and equipment suppliers must build inventory earlier, manufacture products in advance, and receive payment later. Lee stated that "AI CAPEX will continue not by easily slowing down but by transferring investment costs and financial burdens to companies with weak decision-making power."
DB Securities Forecasts Cash Flow Metrics Priority in AI Supply Chain
The analyst emphasized that evaluating AI beneficiary stocks now requires examining whether revenue translates into actual cash inflows, beyond sales growth alone. Investors should verify whether inventory is accumulating too rapidly, accounts receivable are building up, and operating cash flow is tracking profit growth. Lee Jin-kyung stated that "as the AI investment cycle matures, the market is likely to assign higher premiums to cash generation ability and capital efficiency rather than revenue or profit growth," adding that "in the AI supply chain going forward, 'who absorbs working capital burdens and who can transfer them' will become the core variable differentiating stock performance rather than revenue growth." The analyst noted that second-quarter earnings analysis should focus on working capital indicators rather than revenue figures alone.
FAQ
Why are hyperscalers reducing working capital instead of cutting AI investment directly?
DB Securities analyst Lee Jin-kyung explained that working capital reduction improves capital efficiency without sending negative market signals about investment pullback, unlike direct CAPEX cuts. Hyperscalers extend payment terms and reduce inventory while maintaining AI infrastructure spending levels.
What financial pressure did Supermicro face despite revenue growth?
Supermicro announced a $7 billion fundraising plan after inventory and accounts receivable increased rapidly alongside revenue growth. The company's cash generation ability failed to keep pace with sales expansion, illustrating supply chain burden transfer from hyperscalers to suppliers.
