Nvidia (NVDA)

Two forces put Nvidia back on the radar. First, the blowout Q2 FY2027 report (Aug 26): revenue of $96.2B, up 106% YoY and 18% QoQ, beating the ~$92.2B consensus, with Q3 guidance of $108B implying the growth hasn't peaked. Second, the daily financial summary headline — Nvidia teaming with Wall Street on a $500bn AI compute-financing deal (Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, KKR) — signals a structural shift: turning AI compute into a financeable asset class. Last covered July 29 (pre-earnings profile); this is the first post-earnings look.

Financial Snapshot

GAAP figures. Q2 FY27 gross margin hit 75.0% (GAAP & non-GAAP). Data Center revenue was $89.0B in Q2 FY27, +117% YoY. Q3 FY27 revenue guided to $108.0B ±2% (no China Data Center compute assumed).

Bull vs Bear

Year	Revenue	Rev Growth	Op Margin	Net Income
FY2024	$60.9B	+126%	54%	$29.8B
FY2025	$130.5B	+114%	62%	$72.9B
FY2026	$215.9B	+65%	60%	$120.1B
TTM (thru Q2 FY27)	$303B	+40%	65%	$193B

✅ Bull Case

Data Center revenue of $89.0B (+117% YoY) with Vera Rubin now in full production and demand "accelerating" across multiple frontier labs — the buildout has broadened well beyond a single hyperscaler.
The $500bn third-party financing platform converts lumpy customer capex into long-duration financed demand, extending the visible runway for AI factories and de-risking the biggest bear concern (a demand cliff).
Absurd profitability at scale: ~66% operating margin, $59.7B GAAP net income in a single quarter, $26.0B returned to shareholders in Q2, and ~$99B remaining on the buyback.

🔴 Bear Case

At a $5.25T market cap (~27x TTM earnings), near-perfect execution is priced in; any deceleration or guidance miss would be punished violently.
Circular-financing risk: Wall Street financing Nvidia's own customers raises systemic leverage questions — if AI revenue per token disappoints, the financing layer becomes a contagion vector rather than a backstop.
China structurally excluded (no Data Center compute revenue assumed in outlook) plus export-control overhang; Q3 gross margin guided down to 74.0%, and custom ASICs / AMD / China domestic chips are all gunning for inference share.

Verdict

The story has moved from "one customer, one product" to a financed, multi-lab AI-industrial buildout, and Q2's $96.2B print with a $108B guide shows demand is still compounding — not just growing. The main risk isn't the technology but the price and the novel financing architecture that is both a tailwind and, if AI economics sour, a systemic question mark. It's the most interesting name in the market, but at ~20x forward earnings for 40%+ growth it's now a "price-to-execution" bet rather than an obviously cheap one.

Why This Company Now

Financial Snapshot

Bull vs Bear

✅ Bull Case

🔴 Bear Case

Verdict