Yes, Nvidia Is a Buy. Here's Why I'm Adding More Shares Right Now.

The article argues Nvidia could grow revenue 70% year over year in 2027 despite already being described as a $5 trillion company, reinforcing the strength of the AI infrastructure buildout. It cites Nvidia trading at 14.8x next-year earnings versus the S&P 500 at 21x forward earnings, and contends that a 30x trailing-earnings multiple after meeting estimates could support roughly 100% share-price upside over just more than a year. The view is strongly bullish on Nvidia relative to AMD and Broadcom, though it is opinion-based commentary rather than new company-reported results.

Analysis

The bullish setup is more fragile than the headline multiple arithmetic implies: NVDA’s upside requires both continued earnings delivery and sustained willingness to underwrite a premium terminal multiple despite an increasingly concentrated hyperscaler capex base. The relevant near-term variable is not aggregate AI spending but the mix of spend between merchant GPUs, internally designed accelerators, networking, memory, and power infrastructure. Any evidence that incremental workloads are shifting from training toward lower-cost inference will pressure GPU unit economics before it necessarily appears in headline cloud capex.

AVGO is the cleaner second-order beneficiary if hyperscalers seek bargaining power or workload-specific economics through custom ASICs; its custom silicon and networking exposure can gain even if total AI spend remains robust but becomes less NVDA-centric. AMD remains the highest-beta substitution candidate, but needs credible software adoption and recurring large deployments rather than benchmark-driven enthusiasm. A broad AI hardware de-rating would likely hurt AMD more than NVDA because AMD’s valuation support depends more heavily on share-gain expectations, while AVGO’s diversified cash flows offer relative downside protection.

For the next 1-3 months, the principal catalyst is customer commentary on 2027 capex, delivery timing, and utilization rather than another sell-side estimate increase. Over 6-18 months, depreciation expense and AI revenue monetization at hyperscalers become the binding constraint: if cloud providers cannot demonstrate rising AI revenue per deployed accelerator, procurement rationalization could arrive abruptly after existing commitments are fulfilled. The consensus risk is therefore not an immediate demand collapse, but a transition from supply-constrained ordering to return-on-investment discipline.

Contrarianly, extreme bullish retail framing is not itself a catalyst and may signal limited incremental buyer quality. NVDA can still compound materially, but a bullish thesis should be sized around execution durability rather than assuming multiple expansion; the more asymmetric expression is relative exposure to the AI compute stack rather than an outright chase.

Key Decisions for Investors

• Maintain or initiate a modest long NVDA / short AMD pair over a 1-3 month horizon: NVDA has the stronger software and installed-base moat, while AMD is more exposed if large deployments fail to convert into repeat orders. Target 10-15% relative outperformance; stop if AMD discloses a material multi-year hyperscaler deployment or NVDA cuts supply/demand commentary.
• Add AVGO on relative weakness versus NVDA as a 6-18 month custom-silicon hedge. The thesis is that hyperscaler procurement diversification can expand AVGO’s AI content even in a scenario where NVDA’s share of incremental accelerator spend moderates; reassess if custom ASIC program ramps are delayed or networking growth decelerates materially.
• Do not add outright NVDA exposure ahead of the next hyperscaler capex and monetization updates without confirming order visibility beyond committed deliveries. Use a post-results pullback or evidence of accelerating AI revenue at major cloud customers as the preferred entry trigger; invalidate on a meaningful reduction in customer capex guidance or adverse gross-margin commentary.
• Monitor the NVDA/AVGO relative-performance spread as an early signal of merchant-GPU versus ASIC preferences. A sustained AVGO breakout alongside unchanged aggregate cloud capex would support rotating part of AI exposure from NVDA into AVGO rather than reducing the sector gross outright.