Nvidia's Blockbuster Quarter Has One Ugly Loose End. $27 Billion of Free Cash Flow Vanished

Nvidia posted blockbuster fiscal Q2 FY27 revenue of $96.2B (+18% sequentially, +106% YoY), but free cash flow fell from $48.6B to $21.3B, a $27.2B sequential decline. The drop is largely tied to working-capital mechanics: accounts receivable rose $22B+ to $63.1B and inventory increased $5.8B to $31.6B, alongside an increase in DSO from 45 to 60 days due to extended payment terms for large investment-grade customers. Nvidia also disclosed $279B of supply commitments and $108.5B of guarantee obligations for AI cloud partners, raising concerns that a growing share of compute demand may be effectively vendor-financed even as operations remain strong.

Analysis

The market is still valuing NVDA like an operating-system monopoly, but the balance-sheet footprint is starting to look more like a structured-finance franchise. That matters because revenue growth alone is no longer the constraint; the incremental bottleneck is who funds the customer’s buy decision and how long the supplier waits to get paid. If this model persists, the true beneficiaries are not just NVDA’s top line but the capital providers around it: BLK, BX, KKR, GS, and BN get fee flow and deal flow, but they also inherit hidden correlation risk if AI capex slows and these structures are tested.

Near term, the stock can ignore this for days; the real catalyst window is the next 1-3 quarters when working-capital discipline is either normalized or confirmed as a financing habit. The key falsifiers are simple: DSO moving back toward prior levels, inventory growth cooling, or free cash flow rebounding faster than revenue. If that does not happen, the multiple should start to reflect cash conversion quality rather than just growth, which is where a 6-18 month de-rating can begin.

The contrarian read is that the market may be overreacting to a normal super-growth balance-sheet strain: extending terms and building inventory ahead of a product cycle is not the same as credit stress. But if the company is increasingly acting as the lender of first resort to its own buyers, then NVDA is no longer just a GPU story; it becomes a funding-availability story. That subtle shift is what consensus is missing.

Key Decisions for Investors

• Fade post-earnings strength in NVDA: initiate a 1-3 month bearish expression only on rallies, via put spreads or a small short, targeting a re-rating driven by cash conversion scrutiny. Risk/reward is attractive if the market keeps rewarding revenue while ignoring FCF quality; cover if next-quarter DSO and FCF normalize.
• Pair trade: short NVDA / long AMD for 1-3 months as a relative-quality hedge. Thesis: if investors start preferring cleaner balance sheets and less financing intensity, AMD can rerate on relative scarcity value. Falsifier: NVDA shows rapid cash conversion recovery or AMD fails to demonstrate share gains.
• Watch, don’t chase, BLK/BX/KKR/GS/BN on this theme. They may get a small fee-income tailwind from AI financing, but I would not underwrite them as primary beneficiaries unless disclosures show recurring, low-risk origination economics. Treat any rally there as a secondary trade, not the core expression.
• Set an alert for NVDA next-quarter DSO and inventory build: if DSO stays elevated and inventory rises again, add to the bearish NVDA view; if DSO falls back below ~50 days and FCF re-accelerates, exit the thesis.