NVIDIA Corporation (NVDA)

The panel's take

Verdict: HOLD · Conviction: MEDIUM · Last price: USD 209.18 (as of 2026-06-08)

This is the rare panel that splits on the business itself, not just the price. The bulls see the dominant AI-infrastructure franchise — 85% revenue growth, ~75% gross margins, a PEG under 1 — trading below its own historical multiple; the bears see a $5 trillion company at the peak of a capex cycle whose spending runs roughly 10x the AI revenue it currently funds. With credible cases pulling hard in both directions and no clear edge at today's price, the synthesized read lands on HOLD.

Key levels

Key resistance runs from the ~$218 area (the reclaimed-50-day zone) up to historical resistance at the $232–240 all-time-high band; key support sits at $200–205, then the 200-day accumulation zone near $188–192, with the Supertrend level around $177 below that. The analyst consensus target spans roughly USD 215 to 500, averaging near USD 299.

What legendary investors think

We ran NVIDIA past a panel of six legendary investors' frameworks.

Where they agree — and where they clash

Common ground

• The financials are extraordinary and undisputed: FY2026 revenue $215.9B (+65%), Q1 FY2027 $81.6B (+85% YoY), ~75% gross margin, $96.7B free cash flow, near-zero debt.
• The CUDA software ecosystem is a real moat, and the Blackwell/Vera Rubin roadmap keeps NVIDIA ahead on raw performance.
• The recent ~12% pullback from the ~$237 high was largely Broadcom-led sector contagion, not an NVIDIA-specific stumble.

The real debate

• Is the demand durable or circular? The Disruptive-Innovation Seeker and GARP Operator see a multi-year secular buildout; the Forensic Skeptic sees ~$690B of capex chasing a fraction of that in real AI revenue — a cycle that compresses when the funding pauses.
• Are the margins permanent? The bull case needs ~75% gross / ~60%+ operating margins to hold for years; the Intrinsic-Value Modeler models them eroding as custom silicon (Google TPU, Amazon Trainium) and AMD scale, which collapses fair value toward $83.
• Own it now or wait for the entry? The Moat Compounder ($125–135) and Macro Opportunist ($188–195) want a lower price; the bulls argue a sub-1 PEG already is the discount.

The question it comes down to: Is NVIDIA's ~$5T valuation a reasonable price for the durable toll-keeper of the AI era — or is it pricing a peak-cycle, four-customer demand base as if it were permanent?

The numbers

Metric	Value
Price / Market cap	USD 209.18 / ~5.0T
P/E (TTM / fwd)	~32x / ~26x
ROE	76.3%
Operating margin	~65.6% (gross margin ~75%)
Dividend yield	~0.5%
Debt / equity	0.05
Free cash flow	USD 96.7B (FY2026, ~45% margin)

Figures as of FY2026 / Q1 FY2027 (April 2026) / June 2026; sourced from NVIDIA IR/SEC filings, GuruFocus, MacroTrends, StockAnalysis.

The bottom line

The disagreement here is unusually fundamental: not "great company, wrong price," but a genuine split over whether the earnings themselves are durable. On one side sit metrics few companies in history have matched — 85% growth, ~75% gross margins, $96.7B of free cash flow, a PEG under 1, and a software moat two decades in the making — which make the stock look cheap relative to its own history and its peers. On the other sits a $5 trillion valuation resting on a capex cycle that currently spends far more than the AI revenue it produces, on four customers who are also building the chips meant to replace NVIDIA, and on margins that most valuation models expect to compress. What would tip it bullish is durable evidence the AI buildout is self-funding — hyperscaler AI revenue scaling toward the capex — and the next architecture extending the lead; what would tip it bearish is a single large customer shifting meaningful workloads to custom silicon, gross margins breaking below 70%, or a capex-cycle pause. Until one side of that gets an answer, the case for and against roughly balance at $209.

The panel's take

Verdict: HOLD · Conviction: MEDIUM · Last price: USD 194.03 (as of 2026-06-29)

NVIDIA is the rare name where the bulls and bears are both standing on hard numbers. After a ~20% pullback from the early-2026 high near $240, the growth, momentum, and quality frameworks see an AI-infrastructure platform growing data-center revenue ~90% with a forward PEG near 0.4 and a CUDA moat that widens with every chip generation — structurally demanded, not cyclical. The valuation frameworks see something else: peak-cycle margins (~64% operating) on a revenue base where a handful of hyperscalers are simultaneously the largest customers and the best-funded competitors, and they place intrinsic value between roughly $80 and $165 — at or below today's $194. The chart is constructive (the long-term uptrend is intact and price is rebounding off support), which removes the timing objection, but it doesn't resolve the valuation one. With a strong bullish foundation offset by three frameworks seeing limited-to-negative margin of safety and a credible capex-digestion bear, the synthesis is HOLD — tilted constructive, but not a price the whole panel will defend.

Key levels

Key resistance starts at the ~$200 psychological mark, then the ~$215–220 breakdown zone from the pullback, up to the ~$240 all-time-high band. Key support sits at ~$186 (the 50-week trend and recent pullback low), then a ~$175 demand cluster, with a deeper historical accumulation zone near $153 (the long-term 200-week trend). The analyst consensus target spans roughly USD 215 to 500, averaging near USD 300 — notably, even the lowest Street target sits above the current price, while the panel's own intrinsic-value estimates run lower, which is the crux of the debate.

What legendary investors think

We ran NVIDIA past a panel of six legendary investors' frameworks.

Where they agree — and where they clash

Common ground

• The demand is real and currently accelerating: data-center revenue up ~90% YoY, a roughly $1T Blackwell + Rubin order backlog cited through 2027, and hyperscaler capex still rising.
• The financial profile is elite — 100%+ ROE, 60%+ operating margins, a fortress balance sheet — and the CUDA ecosystem is a genuine, durable switching cost.
• The risks are agreed even where their weight is not: customer concentration, custom-silicon substitution, China export policy, and the question of whether AI capex digests.
• After the pullback the chart is still in its long-term uptrend — nobody calls it technically broken above ~$186.

The real debate

• Structural demand or peak cycle? The growth and macro-bull lenses see AI infrastructure as a multi-year secular build; the valuation lenses see hyperscaler capex as inherently cyclical and today's margins as a peak to be normalized.
• Is the moat widening or being surrounded? The Disruptive-Innovation Seeker says each Rubin generation deepens CUDA lock-in; the Moat Compounder and Forensic Skeptic say the three biggest customers are spending billions specifically to escape it.
• Cheap or expensive? A 0.4 PEG and a sub-21x forward multiple read as cheap to the bulls; three intrinsic-value estimates ($80, $130–145, $165) read as fairly-to-richly valued to the bears — the same price, opposite conclusions.

The question it comes down to: Is NVIDIA's current earnings power a durable platform that justifies paying up even after a 20% pullback — or peak-cycle, concentration-heavy earnings that custom silicon and a capex digestion will normalize, leaving today's "cheap" forward multiple sitting on a shrinking base?

The numbers

Metric	Value
Price / Market cap	USD 194.03 / ~4.7T
P/E (TTM / fwd)	~29x / ~21x
ROE	~114%
Operating margin	~64%
Dividend yield	~0.02% (nominal)
Debt / equity	~0.07x (net cash)
Free cash flow	~USD 119B (FY2026); YoY growth turned negative on rising capex

Figures as of Q1 FY2027 / June 2026; sourced from NVIDIA IR/SEC 8-K, StockAnalysis, GuruFocus, MarketBeat. Margins and ROE reflect the current AI-capex peak; the valuation debate turns on how far operating margins normalize from ~64% as custom-silicon competition and capex cyclicality play out.

The bottom line

NVIDIA splits the panel not on whether it is a great business — everyone agrees it is — but on whether today's earnings are a durable base or a cyclical peak, and that single question pulls the fair-value estimates from $80 to $300. The bull case is grounded and currently being validated: data-center revenue compounding near 90%, a roughly $1T order backlog through 2027, a CUDA moat that deepens each generation, and a forward multiple that looks undemanding against that growth — with a chart still in its long-term uptrend after the pullback. The bear and valuation cases are equally concrete: roughly 40% of data-center revenue concentrated in a few hyperscalers who are funding custom silicon to escape NVIDIA's pricing, a circular-financing dynamic linking NVIDIA's investments to its own demand, free cash flow already growing more slowly than earnings as capex rises, and three independent intrinsic-value estimates that sit at or below the current price once margins are normalized. What would tip the synthesis bullish is evidence the capex cycle is durable rather than digesting — sustained hyperscaler spending into FY2028 and margins holding near 60% as custom silicon scales. What would tip it bearish is the first sign of digestion: a large customer guiding capex down, margin compression, or a weekly break of the ~$192 long-term trend. Until one resolves, the constructive tape and the unresolved valuation leave NVIDIA a hold rather than a clear call in either direction.

The panel's take

Verdict: HOLD · Conviction: MEDIUM · Last price: USD 210.96 (as of 2026-07-11)

NVIDIA is the panel's cleanest illustration of a split that runs straight down the middle of investing philosophy. The business is executing at a level with almost no precedent: Q1 FY27 revenue of $81.6B (+85% year over year), data-center revenue of $75.2B (+92%), a Q2 guide of $91B, ~75% gross margins, and ~$97B of free cash flow last fiscal year against a near-zero-debt balance sheet. And unusually, the stock is cheaper than it was — the trailing multiple has compressed from ~44x to ~31x and the shares are up only ~3% year to date despite record results, earning NVIDIA the "black sheep of the chip rally" label. That combination — accelerating fundamentals plus a de-rating to a sub-1 PEG — is why five lenses (technical, fundamental, sentiment, GARP, and disruptive-innovation) are constructive, with the fundamental read putting fair value at $210-270 and the disruptive lens carrying a five-year target of $450. But the valuation purists pull hard the other way: the intrinsic-value model lands at ~$128 base (−39%) once margin-fade and growth-deceleration are built in, the forensic lens sees ~$135-160 flagging depreciation games and circular AI-capex financing, and the value lens wants a 30-40% correction before there's any margin of safety. The macro lens sits on the fence, unwilling to fight a broken-but-recovering tape into an August 26 earnings print. With a quality franchise reasonably priced on forward numbers but carrying genuine two-sided risk — peak-cycle margins, customer concentration, and an AI-capex-bubble debate driving 5-8% weekly swings — the synthesis, as in the prior read, is HOLD.

Key levels

Key resistance starts at ~$220 (near-term supply from the recent consolidation), then ~$230 (a prior swing high), up to the ~$240 all-time high — a weekly close above ~$225-230 on volume is what the technical read treats as confirmation the ~20% correction is over. Key support sits at ~$195 (this week's low and the floor of the recent range), then ~$180 (a prior base), with the trend-defining Supertrend line near ~$176.53 — a weekly close below it would be the first genuine structural warning in this multi-year uptrend. The analyst consensus target runs from a low near $128 up to ~$350 around a ~$301 average — a ~43% gap to the price that reflects the same split the panel debates: the Street prices continued hyperscaler capex growth, while the low end of the range is roughly where the intrinsic-value model lands once the cycle normalizes.

What legendary investors think

We ran NVIDIA past a panel of six legendary investors' frameworks.

Where they agree — and where they clash

Common ground

• The business is executing at an extraordinary level: revenue +85%, data center +92%, a $91B Q2 guide, ~75% gross margins, and ~$97B of annual free cash flow with near-zero debt.
• The stock has de-rated even as fundamentals accelerated — the trailing multiple fell from ~44x to ~31x, and it is roughly flat year to date, the laggard of the chip complex.
• The moat (CUDA plus full-stack systems) is real but under well-funded attack from the same hyperscalers who are NVIDIA's largest customers — custom silicon (Broadcom ASICs, Google TPU, Amazon Trainium) is scaling toward >20GW by 2027.
• The bear case is multi-year, not imminent: margins have not yet cracked, and the near-term order book is still supply-constrained.

The real debate

• Cheap or peak? The growth lenses see a sub-1 PEG on 80%+ growth as an obvious value; the intrinsic-value and forensic lenses see a 64% operating margin on a $250B base as an unsustainable peak that DCF math discounts to $128-160.
• Is the demand organic or circular? Bulls point to Blackwell being sold out; the skeptic flags that NVIDIA is financing a slice of its own demand via equity stakes, and that frontier-lab revenue is ~6% of one year of hyperscaler capex — a gap that has to close.
• Does the depreciation matter? Whether hyperscaler useful-life extensions are prudent accounting or a smoothing of the true cost of the capex cycle that props up the demand curve.

The question it comes down to: Is NVIDIA a de-rated, still-accelerating AI-infrastructure monopoly whose sub-1 PEG makes today's price attractive — or a peak-margin, customer-concentrated, capex-circular cyclical whose normalized earnings power three DCF frameworks put 35-40% below the price, where the honest stance is to hold the quality and watch the cycle?

The numbers

Metric	Value
Price / Market cap	USD 210.96 / ~$5.1T
Revenue (Q1 FY27)	USD 81.6B (+85% YoY); data center $75.2B (+92%); Q2 guide $91B
Gross margin	~75% (recovering to mid-70s)
Free cash flow (FY26)	~$96.6B (~45% margin)
Balance sheet	~$54B net cash; D/E ~0.05; current ratio ~3.9
Capital return	$80B buyback; dividend $0.01 → $0.25
Valuation	P/E ~31x trailing (from ~44x) / ~20-24x forward; PEG <1
Market share	~80% of AI accelerators

Figures as of Q1 FY2027 (period ended Apr 26, 2026, reported May 20) / July 11, 2026; sourced from NVIDIA IR/SEC filings, CNBC, stockanalysis.com, MacroTrends. The stock de-rated on a trailing basis despite record results, as the market discounts the durability of hyperscaler capex (~$725B in 2026, guided toward $1T in 2027) and custom-silicon share risk. China is effectively modeled at zero after tightened export rules (H200 sales resumed May 2026; the proposed "AI OVERWATCH Act" could statutorily ban Blackwell exports); the Kyber rack for Rubin Ultra reportedly slipped to 2028. Insider selling is heavy but pre-scheduled (10b5-1), with no offsetting buying across the AI-hardware complex. Next print: Q2 FY2027 earnings, August 26, 2026.

The bottom line

NVIDIA is the panel's most philosophically split name, and — as in the prior read — the honest synthesis is HOLD. On the bull side sits a genuinely rare setup: a business compounding revenue 85% with ~75% gross margins and ~$97B of free cash flow, whose stock has actually de-rated to a sub-1 PEG and roughly flat year-to-date performance while the rest of the chip complex ran. That is why the GARP lens calls it cheap-for-growth, the disruptive lens carries a $450 target, and the fundamental read puts fair value at $210-270 with the price sitting at the low end. On the bear side sit three valuation frameworks that refuse to extrapolate the peak: the intrinsic-value model discounts margin-fade and growth-deceleration to ~$128, the forensic lens flags ~$176B of understated hyperscaler depreciation and a circular financing loop (NVIDIA funding the very customers who buy its chips) that put fair value at $135-160, and the value lens simply wants a 30-40% correction before there is any margin of safety. Both cases are internally coherent — which is precisely why this is a HOLD rather than a BUY or a SELL. It is not cheap enough for the value crowd, whose normalized math sees 35-40% downside; it is not confirmed enough on the tape for the momentum crowd, with the stock fighting its 200-day into an August 26 print and an AI-capex-bubble debate driving 5-8% weekly swings; and it is too dominant, too cash-generative, and too reasonably priced on forward numbers to sell. What tips it bullish is confirmation the cycle is durable — a reclaim of $225-230 on volume, another beat-and-raise, and evidence that margins hold above the mid-50s against custom silicon. What tips it bearish is the classic cyclical tell: any hyperscaler guiding capex digestion, or GPU share continuing to slide toward in-house chips. Until one side of that debate resolves, the quality franchise is a hold, not a trade.

The panel's take

Verdict: WAIT · Conviction: MEDIUM · Last price: USD 202.81 (as of 2026-07-18) · Changed from HOLD

What changed since 2026-07-11:

• Price: USD 210.96 → 202.81 (-3.9%) — a second straight week of range-bound chop: a probe to $213.8 was rejected, the stock flushed to $198.1, and it closed near the middle of the range.
• Panel: 2 of 6 → 1 of 6 now Bullish (The GARP Operator and The Intrinsic-Value Modeler moved to Neutral from Bullish and Bearish respectively, and The Forensic Skeptic moved to Neutral from Bearish) — the dissent narrowed to a question of timing rather than business quality.
• Trigger: The August 26 Q2 FY2027 print; until then the panel is watching for a volume-confirmed close above $225-230 or a break of the ~$176.5 Supertrend line.
• Verdict: HOLD → WAIT

NVIDIA's operating engine keeps compounding — a $75.2B data-center quarter (+92% year over year), a $91B Q2 guide, ~75% gross margins, ~$96.6B of free cash flow, and a fortress balance sheet with $54B of net cash and near-zero debt. Every dissenting lens is explicit that its objection is price and timing, not the business: the setup is intact and the PEG sits below 1 against 70-90% growth, but none are willing to chase a probe that has already failed twice near $213-214. Only the disruptive-innovation lens is buying today, treating the pullback as noise against a multi-year AI-compute build-out — and because the dissent is entirely about when rather than whether, nobody has a business-quality complaint even as the synthesis softens.

Key levels

Key resistance starts at ~$213.8, the level this week's probe was turned back at, then the $225-230 zone the panel treats as the volume-confirmed close needed to call the pullback over. Key support sits at ~$195-198 (this week's range floor), with the historical accumulation zone around the Supertrend structural line near ~$176.5 marking the deeper line that has held, untouched, for weeks; an analyst consensus target was not available from this week's panel data.

What legendary investors think

We ran NVIDIA past a panel of six legendary investors' frameworks.

Where they agree — and where they clash

Common ground

• Operating performance remains excellent: 92% data-center growth, a $91B Q2 guide, ~75% gross margins, ~$96.6B of free cash flow, and near-zero debt against $54B of net cash.
• The stock is stuck inside a well-worn $195-225 range with no fresh catalyst until the August 26 print resolves it.
• The Supertrend structural line near $176.5 hasn't moved in weeks and is treated by every lens as the trend-defining line, not yet tested.
• Nobody reads this week's chop as distribution or a warning sign — the pullback lacks the volume that would confirm a breakdown.

The real debate

• Wait for confirmation, or act now? The disruptive-innovation lens sees the pullback to $202.81 as noise on a multi-year S-curve and is a buyer today; the other five want either a volume-confirmed close above $225-230 or a deeper test of support before committing more capital.
• Are 62-64% operating margins sustainable? The GARP lens frames this as the market's live, unresolved question, while the intrinsic-value lens treats the sub-1 PEG as evidence the multiple hasn't caught up to the growth either way.
• How much information is in a failed probe? The forensic lens reads the twice-rejected $213-214 area as a market unable to commit in either direction; the macro lens treats the same probe-and-flush pattern as classic range behavior that erodes capital on both sides of the trade.

The question it comes down to: Is the pullback to $202.81 a short-lived pause in a still-accelerating AI-infrastructure compounder before it clears $225-230 for good — or is the range itself the market's honest signal that more information, starting with the August 26 print, is needed before the next leg resolves either way?

The numbers

Metric	Value
Price / Market cap	USD 202.81 / n/a
P/E (TTM / fwd)	29.8x / 19-23x
ROE	n/a
Operating margin	62-64%
Dividend yield	n/a
Debt / equity	Near-zero (net cash ~$54B)
Free cash flow	~$96.6B (~45% FCF margin)

Figures as of the week ended July 18, 2026 (price and range data) and NVIDIA's most recent quarterly results (fundamentals: data-center revenue +92% YoY, $75.2B quarter, $91B Q2 guide). PEG sits below 1 against 70-90% growth. Next print: Q2 FY2027 earnings, August 26, 2026.

The bottom line

If the range finally breaks, watch which side has the volume: a confirmed close above $225-230 would flip the three lenses that moved to Neutral this week (GARP, intrinsic-value, forensic) back toward Bullish, since none of them dispute the growth math — only the entry timing. A break of the ~$176.5 Supertrend line, untouched for weeks, would be the first read that turns a business-quality skeptic's argument, not just a timing one. Short of either, the August 26 Q2 FY2027 print is the next scheduled event with the power to move the panel off Neutral, given how tightly the current split centers on whether 62-64% operating margins hold or fade. A third straight week inside $195-225 without a volume signal would itself be informative — it would suggest the market, like the panel, is genuinely undecided rather than quietly leaning either way.