Nvidia: Customers Wanted More Than It Could Build

NVIDIA reported Q2 FY27 revenue of $96.2B, above guidance, sending shares up 8.7%. The company guided to 70% FY28 revenue growth, though demand remains supply-constrained and gross margin is expected to trough at 71–72% before stabilizing. The outlook remains positive, but the investment case increasingly depends on easing supply limits, preserving margins and managing balance-sheet risk.

Analysis

The key underwriting change is from demand validation to conversion capacity: constrained supply makes near-term revenue visibility unusually high, but also limits upside relative to elevated expectations because incremental demand cannot be shipped. A low-70s gross-margin floor implies the mix is shifting toward more complex systems, networking content, and/or cost-intensive ramps; the market will tolerate this only if dollar gross profit and operating leverage continue to outpace margin dilution over the next 2-3 quarters.

The cleaner second-order beneficiaries may be bottleneck owners rather than direct GPU competitors. TSM benefits from high-end wafer and advanced-packaging utilization, while VRT and ETN capture the data-center power/cooling buildout that must occur before accelerator deployments can translate into revenue. AVGO is a more nuanced beneficiary: hyperscalers facing allocation constraints have greater incentive to fund custom silicon programs, creating a 6-18 month competitive pressure point even if it does not impair NVIDIA's current shipment cadence.

The underappreciated risk is balance-sheet quality rather than demand: investors should scrutinize receivables, inventory, purchase commitments, and any customer-financing arrangements. A rising share of revenue tied to concentrated hyperscalers can defer, rather than eliminate, cyclicality; a single capex digestion phase would simultaneously pressure utilization, mix, and the earnings multiple. Falsification of the constructive view would be a gross-margin outlook below 71%, sequentially rising inventory or DSO without corresponding backlog disclosure, or a major cloud customer reducing AI capex guidance.

Consensus likely treats margin compression as a temporary launch-cost issue. If margin stabilization arrives while supply remains constrained, the stock can rerate on earnings durability; if stabilization requires discounts, lower-priced configurations, or financing support, the multiple risk is material because expectations embed both scarcity and exceptional profitability.

Key Decisions for Investors

• Maintain a core long NVDA for the next 1-3 quarters, but add only on post-earnings consolidation rather than chase an 8%+ gap. Risk/reward remains favorable if gross margin holds at or above 71-72% and backlog conversion remains intact; reduce exposure on a sub-71% margin guide or evidence of receivables/inventory build.
• Express the infrastructure bottleneck thesis through long TSM and VRT over 6-12 months. This diversifies single-name platform risk while retaining exposure to accelerator shipment growth; invalidate on evidence that customer deployment schedules, rather than chip supply, are becoming the binding constraint.
• Initiate a relative-value watch: long AVGO versus NVDA only if hyperscaler disclosures show rising custom-ASIC commitments or NVIDIA allocation remains tight beyond two quarters. The missing data are customer-specific ASIC tape-out volumes and NVIDIA delivery lead times; absent confirmation, this is an alert rather than a funded pair trade.
• Buy 3-6 month NVDA downside protection via put spreads around the next earnings event if implied volatility is below the prior two earnings-cycle range. The catalyst is not a demand miss but a margin or working-capital surprise; structure should target a 10-15% downside move rather than pay for a tail scenario.