NVIDIA Corporation
NVIDIA dominates the AI accelerator market with an estimated 80%+ share in data center GPUs, anchored by a CUDA software moat that has compounded over 15+ years and currently translates into 70%+ gross margins and explosive revenue growth. However, the central question is whether current hyperscaler capex intensity represents a durable secular shift or a cyclical buildout peak, and whether competitive responses from AMD, custom ASICs (Google TPU, AWS Trainium), and customers' in-house silicon will erode pricing power before the addressable market expands enough to justify the valuation.
On June 22, 2026, NVIDIA's stock trades at $210.08, giving the company a market capitalization north of $5 trillion. That number is so large it almost stops being meaningful, which is precisely why I want to slow down and value the business rather than simply marvel at it. I have no position in NVIDIA, and I will not pretend to have special insight into the pace of AI adoption or the long-run economics of GPU clusters. What I do have is a framework, a story, and a willingness to be transparently wrong about both. My valuation of NVIDIA is not the valuation of NVIDIA; it is my valuation, built on assumptions you are free to reject, and the accompanying model lets you do exactly that.
One more thing before the numbers: the investment thesis I was asked to evaluate is largely right about the present and genuinely uncertain about the future. NVIDIA does dominate AI accelerators with an estimated 80%+ share. CUDA is a real moat. The margins are real. The central question, whether hyperscaler capex is a durable secular shift or a cyclical peak, and whether AMD, Google TPUs, AWS Trainium, and custom ASICs erode pricing power before the addressable market expands enough to justify the valuation, is also the right question. I will spend most of this report trying to answer it honestly.
Jensen Huang co-founded NVIDIA in 1993 to build graphics processors for gaming. For most of its first two decades, that is exactly what it did, and it did it well enough to survive brutal competition from ATI (later absorbed by AMD) and Intel. The pivot that changed everything was CUDA, introduced in 2006: a parallel-computing platform that let developers write general-purpose code for NVIDIA GPUs. At the time it looked like a clever engineering move to expand the addressable market. In retrospect it was one of the most consequential software decisions in the history of the semiconductor industry.
Today, NVIDIA is not a graphics chip company. Data Center revenue in Q1 FY2027 was $75.2 billion, up 92% year over year, and represents roughly 92% of total company revenue. The annualized run-rate is above $300 billion. The company's Compute & Networking segment generated $53.3 billion in operating income on $74.6 billion of revenue in that single quarter. Gross margin was 74.9% on a GAAP basis. These are not the numbers of a chip vendor; they are the numbers of a platform company that happens to manufacture the hardware its platform runs on.
The financial foundations of the business, as they stand today, are laid out below:
Source: NVIDIA Form 10-Q filed May 20, 2026; Form 10-K filed Feb. 25, 2026; my estimates.
As you can see, the profitability metrics are extraordinary by any benchmark: gross margins at 74.14%, operating margins at 65.6%, and a return on equity of 114.29% that reflects both the fabless capital model and the pricing power embedded in CUDA lock-in. The one number that demands attention is the beta of 2.20, a reminder that this is not a stable-cash-flow utility; it is a high-volatility growth platform whose fortunes are tied to the pace of AI infrastructure spending.
I capitalize NVIDIA's R&D in this valuation, treating it as a five-year amortizable asset rather than a period expense. The rationale is straightforward: CUDA, Blackwell, Rubin, Spectrum-X, NVLink, and AI Enterprise are platform assets, not one-period costs. NVIDIA spent $6.3 billion on R&D in Q1 FY2027 alone, an annualized rate above $25 billion, supporting 31,000 R&D employees, more than half of whom work on software. That is not an expense profile; it is an ecosystem-building program. Treating it as an expense understates the invested capital base and overstates returns, which matters when the terminal ROIC assumption is doing a lot of work.
To value a company today, I need the price of risk today: as of June 22, 2026, the 10-year U.S. Treasury yield stands at 4.495%, which I use as my riskfree rate, and the implied equity risk premium for the U.S. market is 4.46%. Those two numbers are the foundation of every discount rate I will use.
The AI infrastructure market is, by any measure, enormous and still accelerating. Hyperscaler capital expenditures from the five largest cloud providers are projected at $700 billion to $900 billion in 2026, with roughly 75% of that tied to AI infrastructure. AMD's Lisa Su has cited a $1 trillion AI accelerator TAM by 2030; institutional forecasts for total data center infrastructure run higher still. I am skeptical of the precision of any of these numbers, but the direction is not in doubt: AI compute is becoming a recurring production input for the largest companies in the world, and NVIDIA currently sits at the center of that supply chain.
That said, the TAM debate has a crucial nuance the thesis correctly identifies. Only 25% to 40% of hyperscaler AI capex flows directly to GPUs; the rest goes to land, power, cooling, data center shells, and networking. And there is a well-documented gap between infrastructure spend and monetized AI software revenue, with some analysts citing roughly $600 billion annually between what hyperscalers are spending and what downstream AI software is generating. That gap is not a reason to dismiss the AI buildout, but it is a reason to be careful about extrapolating current GPU demand as if it were fully validated by end-user economics. Power constraints are real: expanding grid capacity is a multi-year process, and data centers are already bumping against interconnection backlogs in multiple geographies.
The competitive landscape is more contested than it was two years ago, though not yet existentially so. AMD's CDNA 4 / MI350-class products are improving on memory capacity and cost-per-performance. Google is commercializing its TPU v7 (Ironwood) more aggressively, including a $3.2 billion guarantee backing Anthropic's deployment. AWS is reportedly exploring direct merchant sales of Trainium3 to third-party data centers. Broadcom's custom ASIC revenue is scaling rapidly, and custom silicon can be 30% to 50% cheaper per token for large-scale, stable inference workloads. JP Morgan has forecast that custom ASICs could reach 45% of the AI chip market by 2028. I do not know if that forecast is right, but I know it is not absurd.
The key competitive insight is this: NVIDIA's largest customers are also its most capable potential competitors. Hyperscalers represent approximately 50% of Data Center revenue (the thesis cited 40%; NVIDIA's Q1 FY2027 filing says 50%), and three direct customers accounted for 54% of total company revenue in the same quarter. Those customers have both the engineering talent and the financial incentive to reduce NVIDIA dependence. They will not do it overnight, because CUDA's nearly 20-year history, 7.5 million developers, and deep integration into PyTorch, TensorFlow, and every major AI framework make switching genuinely costly. But they will do it at the margin, and over a 10-year DCF horizon, the margin matters.
My story is that NVIDIA has become the toll road for AI factories. Data Center is now the company, with roughly 92% of revenue, a $300 billion-plus annualized run-rate, and an estimated 80%+ share of general-purpose AI accelerators. CUDA, NVLink, InfiniBand, Spectrum-X, Blackwell, Rubin, and full-stack AI factory reference architectures make NVIDIA a full-stack AI infrastructure operating system rather than a merchant chip vendor. That position converts directly into cash: gross margins near 75%, operating margins above 60%, and $50.3 billion in operating cash flow in a single quarter.
Over the next decade, my story is that NVIDIA becomes the dominant but less monopolistic backbone of AI compute. Hyperscaler spending moves from a panic buildout into a recurring replacement-and-expansion cycle as AI clusters age, model architectures advance, inference scales, and sovereign, enterprise, robotics, and physical AI workloads broaden demand. Growth decelerates from today's explosive pace, but it does not collapse into a one-time hardware cycle, because AI compute becomes a recurring production input rather than a discretionary experiment. Premium cash flows survive, but peak margins do not: AMD, Google TPUs, AWS Trainium, Broadcom-enabled ASICs, and customer in-house silicon take share in cost-sensitive inference workloads, and hyperscalers use those alternatives to pressure NVIDIA pricing. CUDA keeps switching costs high enough for NVIDIA to earn returns well above semiconductor averages, but not high enough to exempt it from normalization.
Is this story possible? Clearly yes; the AI infrastructure market is already hundreds of billions of dollars in annualized demand and NVIDIA has unmatched scale. Is it plausible? Yes, because CUDA, system integration, networking, supply-chain control, and the Blackwell/Rubin roadmap give NVIDIA advantages competitors cannot quickly replicate. Is it probable? I think so, for the base case of dominance with normalization. The improbable story is permanent 75% gross margins and 80% to 90% growth; the probable story is a dominant AI platform that matures into lower growth, lower margins, and still-exceptional returns.
The table below maps my narrative onto the specific value drivers before any model output appears:
Source: my estimates; NVIDIA SEC filings; Damodaran industry datasets.
As you can see, each driver connects the narrative claim to a named benchmark and a point estimate. Let me walk through the reasoning on each.
1. Year-1 revenue growth: 30.0%. NVIDIA's own-history benchmark is Q1 FY2027 revenue growth of 85% year over year and Data Center growth of 92%. I am assuming 30% for Year 1, which is a sharp deceleration from the current run-rate but still generous, reflecting Blackwell/Blackwell Ultra demand, Spectrum-X, NVLink, sovereign AI, and AI factory deployments that keep near-term demand supply- and power-constrained rather than demand-constrained. I am not extrapolating 85% growth; I am assuming the deceleration has already begun.
2. Years 2-5 revenue growth: 15.0%. This is the central operating bet of the valuation. The growth-to-mature life-cycle benchmark for a dominant platform is 8% to 25% annual growth, and 15% sits at the mid-to-upper range for a company already above a $300 billion Data Center run-rate. I am assuming hyperscaler capex shifts from panic buying toward multi-year Blackwell/Rubin replacement and expansion, while AMD, TPUs, Trainium, custom ASICs, power limits, and customer bargaining power bend the curve. I will be honest: this is the assumption I am least confident in, and I will return to it in the uncertainty section.
3. Terminal growth rate: 3.8%. The named cap for nominal stable growth is today's 4.495% riskfree rate, and 3.8% stays below it. The physical sanity check matters here: compounding from a $300 billion-plus Data Center run-rate through 30% Year-1 growth and 15% Years 2-5 growth would put Year-10 revenue in the neighborhood of the world's largest companies, comparable to and likely above Walmart's roughly $650 billion revenue scale. I am assuming 3.8% only after share and margins normalize, and even that is generous given the scale.
4. Year-1 pre-tax operating margin: 65.0%. NVIDIA's own-history benchmark is a 65.6% Q1 FY2027 operating margin. I start at 65.0%, allowing modest normalization from the current level while preserving the near-term pricing power embedded in AI accelerator scarcity and CUDA switching costs.
5. Target pre-tax operating margin: 42.5% over 10 years. This is where I reject the permanent-peak-margin assumption. The named benchmarks are NVIDIA's own historical R&D-adjusted margin of approximately 42.5% in 2020 and the sector anchor that top-decile tech margins run around 40%. Fifty percent is already pushing the limits of plausibility for a hardware-linked system business at massive scale. I assume the full 10-year glide to 42.5%, which is generous: CUDA lock-in, networking, system integration, and the Blackwell/Rubin roadmap slow competitive erosion, but hyperscaler ASICs and inference optimization mature over multiple architecture cycles. Put simply, 42.5% is a platform-level margin, not a permanent 65% peak-cycle margin.
6. Sales-to-capital: 1.25 in Years 1-5, 1.00 in Years 6-10. NVIDIA is fabless and software-rich, which structurally reduces capital intensity relative to integrated semiconductor peers. The semiconductor sector benchmark is NVIDIA's actual sales-to-capital of 0.65 versus the industry median of 1.15 and the tech median of 2.50. I assume 1.25 early (about $0.80 of reinvestment per $1 of added revenue), fading to 1.00 as growth becomes more physical-infrastructure constrained by power, cooling, and data center capacity. The $119 billion in manufacturing and supply commitments, $30 billion in cloud commitments, $32.4 billion in future lease obligations, and $27 billion in investment commitments all support a capital intensity assumption above the current fabless-only level.
7. Cost of capital: 7.72% initial, 7.0% terminal. The initial rate corresponds to the 75th percentile of the firm-weighted cost-of-capital cross-section, consistent with NVIDIA's growth-to-mature stage, real customer concentration, export-control risk, power-constraint risk, and capex-cycle exposure. The terminal rate glides to the cross-section median of 7.0% as risk normalizes. I am deliberately not making the cost of capital the receptacle of all my hopes and fears about AI; the risks are embedded in the cash flow assumptions, where they belong.
8. Terminal ROIC: 27.5%. The semiconductor industry average return on capital runs roughly 15% to 20%. CUDA, 7.5 million developers, full-stack AI factory design, NVLink, InfiniBand, Spectrum-X, and supply-chain scale keep NVIDIA above commodity semiconductor economics even after margins normalize. A terminal ROIC of 27.5% is within the 25% to 30% base-case moat range and implies terminal reinvestment of about 13.8% at 3.8% stable growth. It is generous, and I know it.
The full model on one page, with assumptions flowing into cash flows and then into value:
Source: my DCF model; NVIDIA SEC filings; Damodaran industry datasets.
As you can see, the model produces a base-case intrinsic value of $214.39 per share, against a market price of $210.08. The price is 98.0% of my estimated value, a 2.1% gap that is best read as a "fairly priced" signal rather than a large mispricing in either direction.
The loose ends: I use $10.6 billion in cash and $11.4 billion in debt from the balance sheet as of April 26, 2026. The June 18, 2026 $25 billion bond offering is material and shifts the net debt position, but the equity value is large enough that the per-share impact is modest. Share count is 24,221 million shares outstanding. The terminal value represents 75.0% of total enterprise value, which sits at the upper end of what I consider acceptable; it is not disqualifying, but it is a reminder that three-quarters of this valuation depends on what happens after the explicit 10-year forecast period.
The discounted cash flows by year are shown below:
Source: my DCF model; present values of projected free cash flows to the firm.
As you can see, the present value of each year's free cash flow to the firm builds steadily through the explicit forecast period, with the largest PV contributions in Years 5 through 9 as growth compounds on a large base and margins remain elevated before normalization. The cumulative PV of the 10-year cash flows is $1.30 trillion, with the terminal value (discounted) contributing $3.90 trillion. That ratio is the core tension in this valuation: the explicit-period cash flows are real and visible, but the terminal value is doing the heavy lifting.
The valuation bridge, showing how each component contributes to the total enterprise value, is below:
Source: my DCF model.
As you can see, the terminal value dominates the enterprise value calculation, which is the structural reality of valuing a company at this growth rate and this margin profile. If you do not like my assumptions, the Excel model that accompanies this report lets you change them and arrive at your own value.
A company valuation without a story to bind it together is just numbers on a spreadsheet, and a story that uses no numbers at all is a fairy tale. But the numbers I have produced are estimates, not facts, and the sensitivity of this valuation to a small number of inputs is extreme. Let me be direct about where I am most likely to be wrong.
The sensitivity analysis below shows the value range across the key scenarios I tested:
| Scenario | Value per Share | vs. Base ($214.39) |
|---|---|---|
| Sustained premium margin (operating margin target = 50%) | $248.29 | +15.8% |
| Near-term AI demand deceleration (Year-1 growth = 20%) | $197.92 | -7.7% |
| Higher early capital intensity (sales-to-capital Yrs 1-5 = 1.0) | $212.30 | -1.0% |
| Medium-term AI capex digestion (Yrs 2-5 growth = 10%) | $171.48 | -20.0% |
| Conservative mature AI infrastructure growth (terminal growth = 2.5%) | $168.81 | -21.3% |
The two-way sensitivity heatmap across discount rate and terminal growth assumptions:
Source: my DCF model; sensitivity computed at the base value of $214.39 per share.
As you can see, the valuation is most sensitive to the terminal growth rate and the Years 2-5 revenue growth assumption. A drop in terminal growth from 3.8% to 2.5% cuts fair value by about $45.58 per share; a drop in Years 2-5 growth from 15% to 10% cuts it by about $42.91. These are not tail scenarios; they are plausible outcomes if hyperscaler capex enters a digestion phase or if power and grid constraints slow deployments.
To justify today's price of $210.08, you would need to believe roughly my base-case assumptions: revenue grows about 30% in Year 1 and compounds around 15% annually in Years 2-5, while operating margin starts near 65% and gradually normalizes to about 42.5% over 10 years, with sales-to-capital of 1.25 early and 1.00 later, terminal ROIC of 27.5%, and a 3.8% perpetual growth rate at a 7.0% terminal cost of capital. The required belief is not fantasy, because the current price is already within about 2% of my base-case value, but it does require NVIDIA to remain one of the most profitable and strategically entrenched large technology companies in the world for many years.
The Monte Carlo distribution, varying the three most uncertain inputs simultaneously:
Source: my DCF model; 10,000 simulations; triangular distributions on revenue growth (Years 2-5), target operating margin, and cost of capital.
As you can see, the median simulated value is $212.09 per share, and the current price of $210.08 sits at approximately the 48th percentile of the value distribution. About 51.7% of simulated outcomes produce a value above the current price. That is not a comfortable margin of safety; it is a coin flip, which is exactly what "fairly priced" means. The 10th percentile value is $163.83 and the 90th percentile is $276.05, a range that reflects genuine uncertainty about the pace of AI infrastructure growth and the durability of NVIDIA's margin premium.
The Weak Links. I want to be honest about the four assumptions this valuation hinges on, because naming them is how you learn to trust the rest.
First, the Years 2-5 revenue growth rate of 15%. This is the central operating bet: NVIDIA must move from a panic AI buildout into a durable replacement-and-expansion cycle without a sharp capex digestion period. I would revise this downward if hyperscaler capex guidance turns flat, if lead times collapse, or if AMD and custom ASIC deployments visibly take incremental workloads. The $119 billion in manufacturing commitments creates real inventory risk if demand disappoints.
Second, the terminal growth rate of 3.8%. It is below the 4.495% riskfree-rate cap, so it passes the formal rule, but compounding from a $300 billion-plus Data Center run-rate through the growth path I have assumed would put Year-10 revenue in the neighborhood of the world's largest companies. I would reduce this if AI compute starts looking like a cyclical hardware replacement market rather than a permanent production input.
Third, the target operating margin of 42.5%. The weak link is not whether NVIDIA has a moat today; it is whether hyperscalers, custom ASICs, AMD, TPUs, Trainium, and procurement pressure can compress pricing faster than the 10-year glide assumes. The precedent from Q1 FY2026, when gross margin fell to 60.5% due to H20 export-related charges, shows how quickly margins can move when external shocks hit.
Fourth, the terminal ROIC of 27.5% at a revenue base approaching the largest companies in history. The bigger the revenue base, the harder it is to maintain both very high returns and high growth without inviting customer insourcing, regulatory constraints, or supply-chain bottlenecks. I would revisit this if NVIDIA has to make larger vendor financing or customer concessions to sustain growth.
Bankers don't value companies; they price them. And the pricing game for NVIDIA is genuinely interesting, because the multiples tell a more nuanced story than the headlines suggest.
The peer comparison table below shows where NVIDIA sits relative to its semiconductor and large-cap technology peers:
| Company | P/E | Fwd P/E | EV/EBITDA | EV/Sales | Gross Margin % | Op. Margin % | Net Margin % | ROE % | Rev. Growth % |
|---|---|---|---|---|---|---|---|---|---|
| NVDA | 32.2x | 16.5x | 30.6x | 20.0x | 74.1% | 65.6% | 63.0% | 114.3% | 85.2% |
| ADI | 65.4x | 29.8x | 35.3x | 17.0x | 64.5% | 38.1% | 26.0% | 9.6% | 37.2% |
| AMD | 181.2x | 41.6x | 116.8x | 23.2x | 53.1% | 14.4% | 13.4% | 8.1% | 37.8% |
| AVGO | 65.8x | 20.4x | 47.6x | 26.5x | 76.3% | 49.0% | 38.9% | 37.3% | 47.9% |
| INTC | n/m | 89.4x | 49.3x | 13.0x | 37.2% | 6.9% | -5.9% | -2.9% | 7.2% |
| MRVL | 103.5x | 48.9x | 100.7x | 31.3x | 51.5% | 14.5% | 29.0% | 16.0% | 27.6% |
| MSFT | 22.1x | 19.2x | 15.5x | 9.0x | 68.3% | 46.3% | 39.3% | 34.0% | 18.3% |
| MU | 55.5x | 9.9x | 34.6x | 21.9x | 58.4% | 67.6% | 41.5% | 39.8% | 196.3% |
| ON | 96.3x | 30.4x | 23.7x | 8.0x | 42.7% | 18.2% | 9.5% | 7.5% | 4.7% |
| QCOM | 24.7x | 21.5x | 18.8x | 5.5x | 54.8% | 22.1% | 22.3% | 36.1% | -3.5% |
| TXN | 56.7x | 35.0x | 34.9x | 16.4x | 57.3% | 37.8% | 29.1% | 32.4% | 18.6% |
| Peer Median | 65.4x | 30.1x | 35.1x | 16.7x | 56.1% | 29.9% | 27.5% | 24.2% | 23.1% |
Source: yfinance; market data as of June 22, 2026.
The pricing picture is counterintuitive and worth dwelling on. NVIDIA trades at a trailing P/E of 32.2x, well below the peer median of 65.4x, and a forward P/E of 16.5x, well below the peer median of 30.1x. On EV/EBITDA, NVIDIA at 30.6x sits between the peer lower quartile of 26.4x and the peer median of 35.1x. These are not the multiples of a stock the market is pricing for perfection on earnings. The market is paying for extraordinary current fundamentals, but it is also discounting the risk that today's earnings base is less durable than it looks.
The analyst consensus tells a different story. Sixty-three analysts cover NVIDIA, with 92.1% carrying bullish ratings (10 strong buy, 48 buy, 2 hold, 1 sell). Price targets range from $180 to $500, with a median of $286.50 and a mean of $298.07. The 107% target dispersion is the key signal: even the bulls disagree sharply about what multiple the market should pay for NVIDIA's forward earnings stream.
| Metric | Value |
|---|---|
| Low Target | $180.00 |
| Mean Target | $298.07 |
| Median Target | $286.50 |
| High Target | $500.00 |
| Rating | Count |
|---|---|
| Strong Buy | 10 |
| Buy | 48 |
| Hold | 2 |
| Sell | 1 |
| Strong Sell | 0 |
| No Opinion | 2 |
| Consensus | Buy |
Source: analyst consensus data as of June 22, 2026.
The Street's median target of $286.50 is about 34% above my DCF value of $214.39. That gap likely reflects greater Street confidence in near-term earnings durability and slower margin normalization than I am assuming. I am less bullish than the consensus central view, not because I think NVIDIA is a bad business (it is an extraordinary one), but because I think the consensus is pricing slower competitive normalization than the evidence supports. Note that "highly priced does not equal overpriced," and at 16.5x forward earnings with 85% revenue growth, NVIDIA is not obviously expensive on near-term numbers. The debate is about what happens after the current AI buildout cycle matures.
My DCF model values NVIDIA at $214.39 per share. The Monte Carlo analysis, varying revenue growth, target operating margin, and cost of capital simultaneously across 10,000 simulations, produces the following price-conditional verdict:
Below $184.89: NVIDIA is a bargain on my value distribution, sitting below the 25th percentile of simulated outcomes.
Between $184.89 and $201.34: Solid value, between the 25th and 40th percentile.
Between $201.34 and $224.01: Fairly priced, the range that brackets both my base-case value of $214.39 and the current market price of $210.08. Today's price falls squarely in this band.
Between $224.01 and $244.87: Richly priced, between the 60th and 75th percentile of simulated outcomes.
Above $244.87: Overvalued on my assumptions, in the upper quartile of the distribution.
At $210.08, NVIDIA is fairly priced by my model, with about 51.7% of simulated outcomes producing a value above the current price. That is not a margin of safety; it is a balanced bet on a dominant AI platform whose near-term fundamentals are extraordinary and whose long-term economics are genuinely uncertain. The margin for error is modest, and the valuation is highly sensitive to whether the AI buildout becomes a durable replacement-and-expansion cycle or a capex peak followed by digestion.
This is my valuation, not the valuation. The assumptions I have made are defensible, but they are not the only defensible assumptions. If you believe NVIDIA's margins will normalize faster, or that hyperscaler custom silicon will take more inference share than I assume, or that the terminal growth rate should be lower given the scale problem, the accompanying Excel model lets you change every input and arrive at your own number. Three things I would watch closely before revisiting my assumptions: (1) hyperscaler capex guidance in the next two earnings cycles, because any flattening would pressure the Years 2-5 growth assumption directly; (2) NVIDIA's gross margin trajectory over the next four quarters, because a sustained move below 70% would suggest competitive normalization is arriving faster than the 10-year glide assumes; and (3) the commercial scale of Google TPU and AWS Trainium deployments outside their own clouds, because merchant availability of those alternatives would materially alter the inference pricing dynamic.
Do your own research. I have shown you my work, and I am comfortable being transparently wrong about it. Time will tell!
DBOT is an automated AI research tool and does not provide investment advice. This report is generated for informational and educational purposes only and is not an offer, solicitation, or recommendation to buy or sell any security. AI-generated analysis may contain errors or omissions. You are solely responsible for your investment decisions — do your own research and consult a licensed financial advisor.
This is an example DBOT report. Run your own for any public company.
Get your own report