Nvidia Stock Has Become Deeply Undervalued, But Watch Out For Accounts Receivable

Nvidia was upgraded to Strong Buy, supported by strong earnings revisions, stock buybacks and favorable technical indicators while its forward P/E remains near 24x, roughly unchanged from four months earlier. The outlook calls for 70% revenue growth in FY2028, with operating-efficiency gains expected to sustain earnings growth despite anticipated gross-margin pressure in H2 FY2027. The report frames the unchanged valuation multiple following Q2 beats as an attractive entry point.

Analysis

The relevant question is not whether NVDA screens inexpensive on a static forward multiple, but whether consensus has properly modeled the transition from accelerator scarcity to a more normalized product-cycle mix. Buybacks are supportive only if free-cash-flow conversion remains ahead of incremental working-capital and customer-financing needs; they do not offset a de-rating if gross-margin pressure proves structural rather than a temporary ramp cost. The near-term setup is favorable for estimate momentum, but that is increasingly a consensus position rather than a differentiated catalyst.

Over the next 1-3 months, the key swing factor is evidence that hyperscaler capex remains broad-based beyond a handful of customers and that Blackwell/rack-scale deployments are converting to revenue without material acceptance or supply-chain delays. A sustained gross-margin guide below the market's expected recovery path would matter more than another revenue beat, because it challenges the operating-leverage thesis embedded in the multiple. Conversely, improving networking attach rates and software/support monetization could lift the earnings duration narrative and support multiple expansion despite lower hardware gross margin.

Second-order beneficiaries of sustained NVDA demand include TSM, AVGO, ANET, VRT and power/cooling infrastructure suppliers; however, each has different bottlenecks and valuation sensitivity. The contrarian view is that NVDA may be the cleaner hedge against AI-capex concentration: if customers rationalize accelerator spend, merchant silicon, networking and data-center infrastructure names can face sharper order volatility than NVDA, whose platform and supply allocation provide more demand visibility. The thesis is falsified by a material hyperscaler capex reset, a sequential deterioration in data-center backlog/conversion, or management guiding gross margin below roughly the mid-70% range without a credible offset from opex leverage.

Key Decisions for Investors

• Maintain or initiate a 3-6 month long NVDA position only on post-results volatility or a 8-12% pullback; target is upside from continued FY earnings revisions rather than multiple expansion. Size modestly because consensus ownership is crowded; reassess if next-quarter gross-margin guidance implies a slower-than-expected recovery.
• Express the more differentiated AI-capex view as long NVDA / short a basket of higher-beta infrastructure beneficiaries such as VRT and ANET over 1-3 months, with equal-dollar sizing. The pair protects against a broad AI-capex drawdown while retaining exposure to NVDA's platform economics; stop if NVDA underperforms the basket by 10% following an earnings update.
• Add TSM selectively on confirmation of sustained leading-edge packaging utilization, rather than chasing NVDA alone. TSM offers a broader AI supply-chain exposure, but requires monitoring of customer concentration and geopolitical-risk premium; a slowdown in advanced-packaging bookings would invalidate the spillover thesis.
• Do not treat the buyback as a standalone catalyst. Set an alert for disclosed repurchase pace versus quarterly free cash flow and for any increase in receivables or customer-financing terms; deterioration in either would reduce confidence in the quality of reported AI revenue.