NVIDIA Corporation

NVDA · Technology · $225.07
✦ AI ANALYSIS — not investment advice
BUY Medium conviction 1-2 years
$301.46target
↗ 33.9% above the current $225.07

NVDA remains the core AI infrastructure winner, and with revenue guided to $78B next quarter, mid-70s gross margins, and a still-modest 28.31x P/E versus its growth rate, the earnings power is outrunning the stock even near the high.

Confidence low
◆Investment thesis

The market still underestimates how much of the AI value stack NVDA has already captured. This is no longer just an accelerator vendor selling scarce GPUs into a temporary shortage. The latest quarter showed $62B of data center revenue and $11B of networking revenue, with Grace Blackwell systems at roughly two-thirds of data center sales. That mix matters because it shows customers are buying integrated systems, fabrics, and architectures, not commodity compute. When the platform broadens, durability improves.

The second leg of the case is simple: earnings power is outrunning the stock. FY2026 revenue reached $215.9B, up 63.6%, while EPS rose 71.7% to $4.93. Yet the stock trades at 28.31x P/E, below the 40.1x peer average, even after a massive run. That does not make NVDA statistically cheap on every metric, because 17.97x P/S is still rich and the market cap at $5.45T leaves less room for multiple expansion. But for a business guiding $78B next quarter and saying revenue should grow sequentially through calendar 2026, the more relevant question is whether consensus earnings are still too low. The odds favor yes.

The case depends on three things. First, Blackwell and then Rubin must stay on schedule. So far, management has delivered against prior product-timing promises, and Vera Rubin samples have already shipped for 2H CY26 production. Second, customers must keep spending through the current build cycle. Supply commitments now extending into calendar 2027 help that argument. Third, margins must hold in the mid-70s gross range while operating expenses rise in the low 40s percent range. That is manageable if system-level revenue keeps taking share of the mix.

Conviction is Medium, not High, because the stock is only 4.8% off its high and because China remains shut out of the near-term model. This is not an undiscovered idea. It is the dominant franchise in the most important infrastructure market in tech, still posting growth and profitability that most mega-caps cannot touch. At $225.07, the setup still favors owning it, but the path will be less forgiving if execution slips even a little.

⋔Scenarios · 1-2 years
$180bear · 25%
$301base · 50%
$380bull · 25%
Bull

Inference demand broadens materially, Rubin ships cleanly in 2H CY26, and earnings power expands fast enough to support a premium multiple despite the large base.

Base

Blackwell and networking continue ramping, revenue follows management’s sequential-growth path through 2026, and valuation settles near the quant engine target.

Bear

AI infrastructure spending cools after the current build cycle, China stays shut, and margin slips toward the low end of mid-70s as mix broadens.

⊚Signal check
AIBUY
QuantBUY
AnalystsStrong Buy
aligned

Quant call aligns with analyst consensus (Strong Buy: 27 buy / 0 hold / 0 sell).

◑Quality & risk
78Quality
20Risk
Quality factors
Profitability100
Cash flow100
Financial health99
Growth95
Sentiment70
Shareholder62
Valuation56
Innovation41
Risk drivers · Low
Valuation55%
Market37%
Execution6%
Financial1%
futuregrowth
Never, ever invest in the present.
— Stanley Druckenmiller
→Forward outlook

Forward earnings power is still moving faster than the headline valuation suggests: trailing P/E is 28.31x on FY2026 EPS of $4.93, even as management guides the next quarter to $78B and says growth should continue sequentially through calendar 2026. The moat is stronger today than two years ago because NVLink, Spectrum-X, InfiniBand, Grace Blackwell, and the coming Vera Rubin platform deepen system-level lock-in beyond just GPUs, though 3-5 years out the main question is whether hyperscalers internalize enough silicon and software to compress returns. AI remains a secular tailwind, not a buzzword, because customers are now deploying profitable inference at scale, but export controls, sovereign build-outs, and architecture transitions will matter more than macro rates for the next leg of the story.

⚡Key drivers
Bull case
  • Scale is compounding, not slowing: FY2026 revenue was $215.9B, up 63.6% YoY, and the latest quarter accelerated to $68B, up 73% YoY.
  • Profitability remains extreme even while ramping new products: FY2026 operating margin was 62.4%, net margin 55.8%, and Q4 EBITDA margin reached 75.3%.
  • Execution keeps beating expectations: NVDA has beaten EPS estimates in 8 straight quarters, with recent beats of +6.2%, +6.3%, and +5.2%.
  • Demand visibility is unusually strong for semis: supply commitments extend into calendar 2027, Sovereign AI revenue exceeded $30B in FY26, and Grace Blackwell systems were roughly two-thirds of data center revenue in the quarter.
Bear case
  • China is still a real hole in the model: Q1 guidance assumes no China data center compute revenue, capping upside from a historically meaningful market.
  • Valuation is mixed, not outright cheap: 17.97x P/S is well above the 11.9x peer average, and the quant engine’s peer set shows wide dispersion of 4.4x on P/E comparables.
  • Gross margin is likely past peak mix benefits: Q1 non-GAAP gross margin is guided to 75.0% ±50 bps, below the prior 78.9% level seen in the 2025 Q1-era ramp.
  • The stock already reflects a lot of good news: shares are 12.8% above the 200-day average and just 4.8% below the 52-week high, which raises the bar for each print.
⌗Key metrics
MetricValueContext
Current price$225.07Reference price for the rating and scenario framework
Quant price target$301.46Low confidence, 33.9% upside vs current, basis: earnings_growth_blend
FY2026 revenue$215.9BUp 63.6% YoY
FY2026 EPS$4.93Up 71.7% YoY
Operating margin62.4%Flat versus prior FY, still elite at this scale
Net margin55.8%Shows unusual conversion of revenue into bottom-line earnings
P/E28.31xBelow peer average 40.1x, though peer dispersion is wide
P/S17.97xAbove peer average 11.9x, so sales valuation is not cheap
Balance sheet leverageDebt/EBITDA 0.27xVery low leverage, supported by 0.17 debt/equity and 4.59 current ratio
Free cash flow$21.4BSupports $19.7B buybacks, $6.0B dividends, and $2.7B capex
Next-quarter revenue guide$78B ±2%Management expects sequential growth throughout calendar 2026
52-week position-4.8% vs highShares are 12.8% above the 200-day average, so expectations are elevated
◎How the $301.46 target is built
P/E Multiple · 28%$420.81
PEG (growth-adjusted) · 3% (faded)$956.39
P/S Multiple · 20%$118.82
Quality-adjusted · 20%$275.49
DCF · 15%$224.46
Low confidence · earnings growth blend blend
⚑Risk & catalyst
Primary risk

The biggest risk is that AI compute demand normalizes before supply commitments convert into revenue, because the stock is priced for continued hypergrowth while Q1 guidance already assumes no China data center compute revenue and shares sit only 4.8% below the high.

Next catalyst

Next earnings on November 18, 2026, where the key issue is whether NVDA confirms the $78B quarter and keeps the promise of sequential growth through calendar 2026.

⇄Positioning & flows
Short interest1.3% of float · 2.1d to cover · -1.4%
Options GEXpositive regime · flip $216
put wall $200call wall $230

Ownership is crowded by design: the top institutions control 63.6%, insiders own 2.2%, and short interest is only 1.27% with 2.1 days-to-cover, so there is little classic short-squeeze fuel. Options positioning is supportive rather than explosive, with positive GEX and zero-gamma at $215.57, about 4.19% below spot, which can help damp volatility unless price breaks decisively lower. The practical read is that NVDA is a consensus long, so upside likely comes from earnings revisions, not positioning unwind.

▦Segments
Data Center$62B in the latest quarter, about 91% of Q4 revenue

Up 75% YoY and 22% QoQ, driven by Blackwell and Blackwell Ultra ramp.

Networking$11B in the latest quarter

Up more than 3.5x YoY, with demand led by NVLink, Spectrum-X, and InfiniBand.

Sovereign AIMore than **$30B** in FY26

More than tripled YoY, showing demand is widening beyond U.S. hyperscalers.

Gaming

Management expects supply constraints for a couple of quarters, so gaming is not the current earnings driver.

Geography

The important disclosed point is that Q1 assumes no China data center compute revenue.

▤Latest earnings call · Q4 2026

Q1 guide to $78B and management expects sequential growth through calendar 2026.

Call sentiment 88/100
“Management

Management has earned credibility. On the 2025 Q1 call, NVDA guided Q2 revenue to $28B ±2%, said gross margins would stay in the mid-70s, and promised Blackwell production shipments would start in Q2 and ramp in Q3. Since then, the company scaled from $26.0B of quarterly revenue in that period to $68.13B in the latest quarter, kept FY2026 operating margin at 62.4%, delivered 8 straight EPS beats, and now guides $78B ±2% for Q1 FY27 with gross margin again in the mid-70s. The specific promises on Blackwell also look largely delivered: Grace Blackwell systems are now roughly two-thirds of data center revenue, networking became an $11B business in the quarter, and Vera Rubin samples have already shipped with production still on track for 2H CY26. Direction has not changed materially, only the scale has.

Q1 revenue guide
So revenue is expected to be $78 billion, plus or minus 2%.
— Colette Kress, CFO · Q4 2026 earnings call
Gross margin outlook
GAAP and non-GAAP gross margins are expected to be 74.9% and 75%, respectively, plus or minus 50 basis points. For the full year, we continue to see gross margins in the mid-70s.
— Colette Kress, CFO · Q4 2026 earnings call
Quarter performance
Total revenue of $68 billion was up 73% year-over-year, accelerating from Q3.
— Colette Kress, CFO · Q4 2026 earnings call
Forward demand
We look ahead, we expect sequential revenue growth throughout calendar 2026, exceeding what was included in the $500 billion Blackwell and Rubin revenue opportunity we shared last year.
— Colette Kress, CFO · Q4 2026 earnings call
Demand inflection
I am certain at this point that we are at the inflection point, we've reached the inflection point and we're generating profitable tokens that are productive for customers and profitable for the cloud service providers.
— Jen-Hsun Huang, CEO · Q4 2026 earnings call
◈Moat & edge

NVDA’s moat is the full-stack AI system, not just the chip. CUDA, networking fabric, rack-scale architecture, and software-tooling lock customers into a platform that shortens deployment time and raises switching costs, which is why Grace Blackwell systems already represent roughly two-thirds of data center revenue. The moat looks durable over the next 2-3 years because customers still need integrated hardware, interconnect, and software at scale, but it is less permanent over 3-5 years if hyperscalers can absorb more of the stack internally.

⊞Capital allocation

Capital allocation has been aggressive but still sensible given the cash machine underneath it. FY2026 free cash flow was $21.4B, capex only $2.7B, buybacks $19.7B, and dividends $6.0B, while leverage stayed light at 0.27x debt/EBITDA and 0.17 debt/equity. That mix says management is funding capacity and returning capital without stretching the balance sheet, though the dividend level should be watched against FCF if working capital swings during heavy supply ramps.

∎The bottom line

NVDA rates a BUY with Medium conviction. The stock is no longer an easy valuation call, but the business is still compounding faster than the market is giving it credit for, with a $78B quarter ahead, mid-70s gross margins, and product cadence extending into Rubin. The main reason not to press to High conviction is setup, not quality: shares are near the high, China contributes zero in the next-quarter data center guide, and expectations are demanding. The call would weaken if sequential growth through calendar 2026 falters, if Rubin timing slips, or if gross margins break below the guided mid-70s range for more than a transition quarter.

≣ Full research note tap to expand

Executive Summary

BUY, Medium conviction. Rated BUY, Medium conviction. NVDA is no longer cheap on sales at 17.97x P/S, but the stock still looks mispriced against a business that grew FY2026 revenue 63.6% to $215.9B, EPS 71.7% to $4.93, and is guiding the next quarter to $78B with management expecting sequential growth through calendar 2026. The setup is less asymmetric than it was lower in the range, with shares only 4.8% below the 52-week high, but the company’s execution, margin structure, and AI demand visibility still support upside beyond the current $225.07.

Investment Thesis

The market still underestimates how much of the AI value stack NVDA has already captured. This is no longer just an accelerator vendor selling scarce GPUs into a temporary shortage. The latest quarter showed $62B of data center revenue and $11B of networking revenue, with Grace Blackwell systems at roughly two-thirds of data center sales. That mix matters because it shows customers are buying integrated systems, fabrics, and architectures, not commodity compute. When the platform broadens, durability improves.

The second leg of the case is simple: earnings power is outrunning the stock. FY2026 revenue reached $215.9B, up 63.6%, while EPS rose 71.7% to $4.93. Yet the stock trades at 28.31x P/E, below the 40.1x peer average, even after a massive run. That does not make NVDA statistically cheap on every metric, because 17.97x P/S is still rich and the market cap at $5.45T leaves less room for multiple expansion. But for a business guiding $78B next quarter and saying revenue should grow sequentially through calendar 2026, the more relevant question is whether consensus earnings are still too low. The odds favor yes.

The case depends on three things. First, Blackwell and then Rubin must stay on schedule. So far, management has delivered against prior product-timing promises, and Vera Rubin samples have already shipped for 2H CY26 production. Second, customers must keep spending through the current build cycle. Supply commitments now extending into calendar 2027 help that argument. Third, margins must hold in the mid-70s gross range while operating expenses rise in the low 40s percent range. That is manageable if system-level revenue keeps taking share of the mix.

Conviction is Medium, not High, because the stock is only 4.8% off its high and because China remains shut out of the near-term model. This is not an undiscovered idea. It is the dominant franchise in the most important infrastructure market in tech, still posting growth and profitability that most mega-caps cannot touch. At $225.07, the setup still favors owning it, but the path will be less forgiving if execution slips even a little.

Key Metrics

MetricValueContext
Current price$225.07Reference price for the rating and scenario framework
Quant price target$301.46Low confidence, 33.9% upside vs current, basis: earnings_growth_blend
FY2026 revenue$215.9BUp 63.6% YoY
FY2026 EPS$4.93Up 71.7% YoY
Operating margin62.4%Flat versus prior FY, still elite at this scale
Net margin55.8%Shows unusual conversion of revenue into bottom-line earnings
P/E28.31xBelow peer average 40.1x, though peer dispersion is wide
P/S17.97xAbove peer average 11.9x, so sales valuation is not cheap
Balance sheet leverageDebt/EBITDA 0.27xVery low leverage, supported by 0.17 debt/equity and 4.59 current ratio
Free cash flow$21.4BSupports $19.7B buybacks, $6.0B dividends, and $2.7B capex
Next-quarter revenue guide$78B ±2%Management expects sequential growth throughout calendar 2026
52-week position-4.8% vs highShares are 12.8% above the 200-day average, so expectations are elevated

Financial Strength

Financial strength is exceptional. Leverage is trivial at 0.27x debt/EBITDA, the current ratio is 4.59, and the quality framework scores 99 on financial health and 100 on cash flow quality. The more important point is that margins have stayed extraordinary while the company has scaled into a much larger revenue base, which usually breaks semiconductor models. Free cash flow of $21.4B and low capex intensity relative to revenue give NVDA room to fund capacity, absorb product transitions, and still return material cash to shareholders.

Competitive Position

NVDA’s moat is the full-stack AI system, not just the chip. CUDA, networking fabric, rack-scale architecture, and software-tooling lock customers into a platform that shortens deployment time and raises switching costs, which is why Grace Blackwell systems already represent roughly two-thirds of data center revenue. The moat looks durable over the next 2-3 years because customers still need integrated hardware, interconnect, and software at scale, but it is less permanent over 3-5 years if hyperscalers can absorb more of the stack internally.

Management & Guidance

Current guidance is aggressive but credible. CFO Colette Kress said, "So revenue is expected to be $78 billion, plus or minus 2%." She also said, "GAAP and non-GAAP gross margins are expected to be 74.9% and 75%, respectively, plus or minus 50 basis points. For the full year, we continue to see gross margins in the mid-70s." Credibility is supported by the beat record, 8 straight EPS beats, and by prior execution on Blackwell ramp promises, but investors should watch whether low-40s percent OpEx growth and ongoing supply tightness pressure incremental margins more than expected.

Segment Analysis

  • Data Center ($62B in the latest quarter, about 91% of Q4 revenue): Up 75% YoY and 22% QoQ, driven by Blackwell and Blackwell Ultra ramp.
  • Networking ($11B in the latest quarter): Up more than 3.5x YoY, with demand led by NVLink, Spectrum-X, and InfiniBand.
  • Sovereign AI (More than $30B in FY26): More than tripled YoY, showing demand is widening beyond U.S. hyperscalers.
  • Gaming (Not disclosed in the provided data): Management expects supply constraints for a couple of quarters, so gaming is not the current earnings driver.
  • Geography (Not disclosed in the provided data): The important disclosed point is that Q1 assumes no China data center compute revenue.

Capital Allocation

Capital allocation has been aggressive but still sensible given the cash machine underneath it. FY2026 free cash flow was $21.4B, capex only $2.7B, buybacks $19.7B, and dividends $6.0B, while leverage stayed light at 0.27x debt/EBITDA and 0.17 debt/equity. That mix says management is funding capacity and returning capital without stretching the balance sheet, though the dividend level should be watched against FCF if working capital swings during heavy supply ramps.

Management Execution & Track Record

Management has earned credibility. On the 2025 Q1 call, NVDA guided Q2 revenue to $28B ±2%, said gross margins would stay in the mid-70s, and promised Blackwell production shipments would start in Q2 and ramp in Q3. Since then, the company scaled from $26.0B of quarterly revenue in that period to $68.13B in the latest quarter, kept FY2026 operating margin at 62.4%, delivered 8 straight EPS beats, and now guides $78B ±2% for Q1 FY27 with gross margin again in the mid-70s. The specific promises on Blackwell also look largely delivered: Grace Blackwell systems are now roughly two-thirds of data center revenue, networking became an $11B business in the quarter, and Vera Rubin samples have already shipped with production still on track for 2H CY26. Direction has not changed materially, only the scale has.

Positioning & Flows

Ownership is crowded by design: the top institutions control 63.6%, insiders own 2.2%, and short interest is only 1.27% with 2.1 days-to-cover, so there is little classic short-squeeze fuel. Options positioning is supportive rather than explosive, with positive GEX and zero-gamma at $215.57, about 4.19% below spot, which can help damp volatility unless price breaks decisively lower. The practical read is that NVDA is a consensus long, so upside likely comes from earnings revisions, not positioning unwind.

Bull Case

  • Scale is compounding, not slowing: FY2026 revenue was $215.9B, up 63.6% YoY, and the latest quarter accelerated to $68B, up 73% YoY.
  • Profitability remains extreme even while ramping new products: FY2026 operating margin was 62.4%, net margin 55.8%, and Q4 EBITDA margin reached 75.3%.
  • Execution keeps beating expectations: NVDA has beaten EPS estimates in 8 straight quarters, with recent beats of +6.2%, +6.3%, and +5.2%.
  • Demand visibility is unusually strong for semis: supply commitments extend into calendar 2027, Sovereign AI revenue exceeded $30B in FY26, and Grace Blackwell systems were roughly two-thirds of data center revenue in the quarter.

Bear Case

  • China is still a real hole in the model: Q1 guidance assumes no China data center compute revenue, capping upside from a historically meaningful market.
  • Valuation is mixed, not outright cheap: 17.97x P/S is well above the 11.9x peer average, and the quant engine’s peer set shows wide dispersion of 4.4x on P/E comparables.
  • Gross margin is likely past peak mix benefits: Q1 non-GAAP gross margin is guided to 75.0% ±50 bps, below the prior 78.9% level seen in the 2025 Q1-era ramp.
  • The stock already reflects a lot of good news: shares are 12.8% above the 200-day average and just 4.8% below the 52-week high, which raises the bar for each print.

Valuation & Price Target

Engine price target $301.46 (+33.9% vs current), Low confidence.

  • P/E Multiple: $420.81 (28% weight)
  • PEG (growth-adjusted): $956.39 (3% weight)
  • P/S Multiple: $118.82 (20% weight)
  • Quality-adjusted: $275.49 (20% weight)
  • DCF: $224.46 (15% weight)

The rating aligns with the quant baseline of BUY, but with more emphasis on trajectory than on blended valuation outputs. The engine’s $301.46 target is reasonable as a base case, though confidence should stay low because the valuation methods range from $118.82 on P/S to $956.39 on PEG, and peer P/E dispersion is a very wide 4.4x. The reason to stay positive anyway is that this is not a typical peer-multiple story: FY2026 revenue grew 63.6%, EPS 71.7%, Q1 is guided to $78B, margins remain in the mid-70s gross range, and leverage is minimal. Sell-side consensus is also bullish, but that is context only, not the basis for the call.

Risk Assessment

Risk score 20/100 (Low). The biggest risk is that AI compute demand normalizes before supply commitments convert into revenue, because the stock is priced for continued hypergrowth while Q1 guidance already assumes no China data center compute revenue and shares sit only 4.8% below the high.

The Bottom Line

NVDA rates a BUY with Medium conviction. The stock is no longer an easy valuation call, but the business is still compounding faster than the market is giving it credit for, with a $78B quarter ahead, mid-70s gross margins, and product cadence extending into Rubin. The main reason not to press to High conviction is setup, not quality: shares are near the high, China contributes zero in the next-quarter data center guide, and expectations are demanding. The call would weaken if sequential growth through calendar 2026 falters, if Rubin timing slips, or if gross margins break below the guided mid-70s range for more than a transition quarter.

Go deeper in the app

Live charts, options analytics, the full earnings-call breakdown, price alerts, and every other stock — updated in real time.

Get Tickr