Forget Waiting for an Nvidia Split: NVDA's Real Value Has Nothing to Do With Share Price

Nvidia reported Q2 FY2027 revenue of $96.2 billion, up 106% year over year, with a 75% gross margin and $89 billion in data center revenue. Its board added $150 billion to the share repurchase authorization, bringing the total to $235 billion, which the article views as a stronger shareholder-return signal than a stock split. The article argues that a split is unlikely in 2026 at a $239 share price and $5.8 trillion market cap, while noting Nvidia's dependence on continued AI spending as a key risk.

Analysis

The split question has little investment relevance: share denomination does not change economics, and any retail-access narrative is secondary to earnings revisions. The material debate is whether hyperscaler capex converts into durable accelerator demand or merely pulls forward purchases. Near term, continued infrastructure budgets support NVIDIA’s revenue visibility; over 6–18 months, Alphabet, Amazon, Meta, and Microsoft’s incentive to develop custom silicon could constrain NVIDIA’s share and pricing even if aggregate AI spending remains high. AMD may benefit from substitution, but a growing market does not guarantee it can capture attractive economics.

Treat the repurchase authorization as capacity, not a committed bid. Actual share-count reduction depends on execution and can be partly offset by equity compensation; at a large valuation, buybacks also compete with investment flexibility. Likewise, a low forward PEG is highly sensitive to the assumed earnings-growth path. If growth decelerates, multiple compression can outweigh EPS accretion.

The contrarian risk is not necessarily an abrupt collapse in AI spending: capex can remain elevated while returns on deployed infrastructure disappoint, prompting a slower purchase cadence and more in-house chips. The near-term catalyst is hyperscaler capex commentary and NVIDIA’s next data-center growth and gross-margin disclosures; the structural test is whether customer concentration and custom silicon erode growth or pricing. No split-related catalyst merits a trade.

Key Decisions for Investors

• Do not buy or wait based on a prospective stock split. For an existing NVIDIA position, keep exposure tied to earnings and risk limits rather than share price optics.
• For new exposure, scale in only if comfortable with AI-capex concentration; reassess after hyperscaler spending updates and NVIDIA’s next results rather than treating the buyback authorization as a price floor.
• Monitor reported diluted share count against repurchases, data-center growth, and gross margin. If share count fails to decline meaningfully or growth and margins weaken together, reduce the thesis rather than relying on authorization size.
• Avoid a reflexive short in AMD: custom silicon and AMD could gain share, but verify customer adoption and economics before positioning for substitution. A sustained slowdown in hyperscaler capex or clear deterioration in NVIDIA’s growth, margins, or guidance would falsify the bullish case.