The sector is selling off. The business just posted $96.7 billion in free cash flow. These two facts are not as contradictory as they appear.
Nvidia is selling off alongside the broader chip sector, caught in a macro downdraft that has knocked nearly 300 S&P 500 names into negative territory for 2026. Market sentiment for the stock has collapsed, with news coverage dominated by sector fear and tariff-related macro concerns.
The financial picture tells a different story. In fiscal 2026, Nvidia generated $215.9 billion in revenue, $96.7 billion in free cash flow, and $120.1 billion in net income. Operating margins held above 60%. The company bought back $40.1 billion of its own stock while sitting on a net cash position.
When sentiment and fundamentals diverge this sharply, it is worth asking which one is more likely to be wrong.
The March 30, 2026 session saw chip stocks taking targeted selling pressure, with the semiconductor sector among the worst performers as the S&P 500 continued its roughly 7.5% year-to-date decline. News coverage on the day was dominated by negative headlines: sector selloffs deepening, optical names leading losses, macro headwinds intensifying.
The selling is not random. Semiconductors have been caught in a three-way squeeze: tariff uncertainty weighing on supply chain economics, broader equity market de-risking, and lingering questions about whether AI infrastructure spending can sustain the pace that drove the 2024 and 2025 earnings explosions.
Those are real concerns. They deserve serious treatment. But they should be evaluated against what the underlying business has actually produced, not what the tape says on a single trading day.
Nvidia's current market capitalization sits at approximately $4.07 trillion. At trailing free cash flow of $96.7 billion, the implied FCF yield is roughly 2.4%. That is an expensive stock by almost any traditional measure. The question is whether the business justifies it.
TickerXray Report
Get the complete Nvidia report with all 12 quantitative models, AI-generated investment thesis, and real-time data.
The revenue trajectory is the starting point. In fiscal 2024, Nvidia generated $60.9 billion in revenue. In fiscal 2025, $130.5 billion. In fiscal 2026, $215.9 billion. That is a compound annual growth rate above 85% over two years, and it is not happening through acquisitions or financial engineering. It is happening because every major cloud provider, hyperscaler, and enterprise AI deployment is running on Nvidia hardware.
The margin profile is as important as the revenue numbers. Operating margins expanded from 54.1% in fiscal 2024 to 62.4% in fiscal 2025 before settling at 60.4% in fiscal 2026. This is the signature of a genuine pricing monopoly: volume is growing rapidly and margins are not compressing. That combination almost never happens in hardware.
Free cash flow has tracked revenue closely, which matters. Fiscal 2024 produced $27.0 billion in FCF. Fiscal 2025 produced $60.9 billion. Fiscal 2026 produced $96.7 billion. Capital expenditures are minimal relative to the business size, just $6.0 billion on $215.9 billion in revenue. This is a fabless model, and the economics are extraordinary.
The balance sheet is clean. Net cash position after accounting for $7.5 billion in debt against $10.6 billion in cash. The company returned $40.1 billion to shareholders through buybacks in fiscal 2026 alone. Earnings beat consensus estimates in three of the last four reported quarters.
Nvidia's transformation from gaming GPU company to AI infrastructure monopoly happened because of a decade-long bet on CUDA, its parallel computing platform, that turned out to be the foundational software layer for the entire machine learning stack. Switching costs are embedded at the developer level, not just the hardware level, which is what makes the competitive position genuinely durable.
The question every serious investor is asking in 2026 is whether the AI capex cycle that drove this growth can continue, or whether hyperscalers will pull back once initial buildouts are complete. The honest answer is that no one knows with certainty. What the data shows is that Azure, AWS, and Google Cloud are all running at capacity constraints, and planned capex increases suggest the buildout is not finished.
Blackwell architecture deployment accelerated through 2025 and into 2026, with customer commentary suggesting demand continues to outpace supply. That demand environment, combined with the pricing power Nvidia has demonstrated, is why the operating margin held above 60% even as revenue doubled.
The buyback program is another signal worth noting. Companies buying back 25% of their market cap annually while sitting on net cash are usually expressing confidence in their near-term cash flows. Nvidia has been consistent on this, returning $9.5 billion in fiscal 2024, $33.7 billion in fiscal 2025, and $40.1 billion in fiscal 2026.
The bear case on Nvidia's competitive position rests on two arguments: that AMD will eventually close the performance gap, and that hyperscalers building their own chips (Google's TPUs, Amazon's Trainium, Microsoft's Maia) will reduce dependence on Nvidia over time. Both arguments are directionally correct. Neither has shown up in the financial statements yet.
AMD's MI300 series has made real progress, but market share data from the data center segment shows Nvidia's position has held or strengthened even as AMD shipped more units. The CUDA ecosystem is the moat. It is not just the hardware that customers are locked into, it is the software stack, the developer tools, the training frameworks, and the institutional knowledge built around the platform.
The internal chip programs at hyperscalers are more credible competition in the long term, but they are predominantly built for inference workloads on proprietary models, not general-purpose AI training. The use cases diverge enough that hyperscaler custom silicon and Nvidia hardware coexist rather than directly substitute.
None of this means the competitive position is permanent. It means the selloff is pricing in competitive threats that have not yet materialized in the financials, while ignoring the $96.7 billion in free cash flow that did.
The March 30 news coverage around Nvidia is dominated by two themes: the broader chip sector selloff, and tariff-related macro concerns. Both are real. Neither is a direct challenge to Nvidia's unit economics or pricing power.
Tariff risk is worth taking seriously. Nvidia's supply chain runs through TSMC fabs in Taiwan, and any escalation in US-China trade restrictions could complicate both manufacturing logistics and the significant China data center market. The company navigated the H100 export control restrictions by releasing the H20 variant for Chinese customers, but further tightening would compress that market.
What the current sentiment data is not capturing is the durability of the demand signal. The AI infrastructure build is not speculative at this point. It is actively deployed, actively generating revenue for the customers using it, and actively being expanded. The hyperscalers did not spend $100 billion on AI hardware to sit on it.
The risk for investors sitting on the sidelines waiting for macro clarity is that by the time sentiment recovers, the entry point will have moved significantly. Sentiment reversions in dominant technology franchises with strong FCF tend to be faster and more complete than the selloff that preceded them.
The concentration risk is real. Nvidia's top customers are a handful of hyperscalers who each represent a material percentage of revenue. Any slowdown in AI capex from one or two of these customers would show up quickly in Nvidia's numbers. The Q1 2026 earnings estimate currently sits at zero (pending report), which reflects uncertainty rather than a known result.
Export control escalation remains the highest-probability near-term risk. The H20 variant for China is a workaround, not a solution, and further restrictions could eliminate a market that represents somewhere between 10% and 20% of total revenue depending on the quarter. That is real dollar exposure.
Valuation compression is the silent risk. At a trailing PE above 34x on a business that has just tripled revenue, consensus estimates are embedding continued growth. If the growth trajectory decelerates to something more moderate, say 30% revenue growth instead of 60%, the multiple will not simply hold. It will reset, and the reset happens faster than the fundamental deterioration.
At $4.07 trillion in market cap and $96.7 billion in trailing free cash flow, Nvidia trades at roughly 42x FCF. The analyst community is split: 14 strong buy, 7 buy, 16 hold, 3 sell, 7 strong sell, with a consensus price target around $421.
The consensus target implies roughly 10% upside from where the stock was trading before the current selloff began. That is a fairly modest return expectation for a business growing at this velocity, which suggests analysts are either cautiously valuing the growth properly or are underweighting the durability of the demand cycle.
A reasonable framework: if Nvidia generates $100 billion in FCF in fiscal 2027 (roughly flat with fiscal 2026 growth rates decelerating) and the market values it at 30x FCF, that implies a $3 trillion valuation. At 40x FCF, it is $4 trillion. The current price implies the market thinks 40x is the right multiple for a business with this growth profile, which is a defensible position if the growth continues.
What the selloff is doing is temporarily pricing in a scenario where the multiple compresses without the FCF declining, which historically is where sharp recoveries originate.
Nvidia is not a cheap stock. It has not been a cheap stock for two years. The question has never been whether it is cheap; the question is whether the business quality and growth trajectory justify the premium.
Fiscal 2026 answered that question with $215.9 billion in revenue, $120.1 billion in net income, $96.7 billion in free cash flow, and operating margins above 60%. The chip sector selloff is a sentiment event layered on top of genuine macro uncertainty. It is not a fundamental development.
Investors who can distinguish between price and value, and who have a time horizon longer than the current macro news cycle, are looking at a business that is executing at a level that almost never happens in the history of public markets. Whether the entry point is right today depends on each investor's required return. The thesis itself is not broken by a bad tape.
Full forensic analysis of Nvidia
+ 6 more models included
150,000+ stocks covered
Global coverage across 60+ exchanges. Every report includes all 12 quantitative models and AI analysis.
View plansEvery report runs 12 quantitative models and generates an AI investment thesis. From Piotroski scores to manipulation detection -- get the full picture in seconds.
12 forensic models
Piotroski, Altman, Beneish, DuPont & more
AI investment thesis
Synthesized outlook on every stock
Manipulation detection
Spot red flags before they hit the news
150,000+ tickers
Global coverage across 60+ exchanges
Expected return
Forward return projections for every stock
Real-time data
Live prices, insider trades, news sentiment
Free accounts get 1 report per month. Pro gets unlimited.
Our earlier 'Five Trillion' narrative argued Nvidia was being underpriced for the AI capex cycle. Since publication, free cash flow has reached $96.7 billion, the multiple has held, and Wall Street's biggest AI bull just got louder. We update the thesis.
A $5.2 trillion market cap with revenue still compounding above 70% year over year is a profile the public market has never seen at this scale. The signals are telling a specific story.
NVDA's market cap has tripled in two years on FCF that grew from $3.8B to $96.7B. Most of the bear arguments are about the multiple, not the cash. Here's what the consensus is pricing wrong.