BUY at the dated $162.26 quote. Base present value is $237, with a conditional $67–$439 range. Exceptional operating cash generation supports the opportunity; customer funding, commitments and unresolved governance risks keep confidence measured.
| Metric | Value |
|---|---|
| Bear | 67.01 |
| Base | 237.25 |
| Bull | 438.68 |
| Scenario | Probability | Present value | vs $162.26 | What has to happen | The arithmetic |
|---|---|---|---|---|---|
| Bear | 25% | $67.01 | −59% | Demand and customer-credit stress; slower collections and weaker margins | FY2028 revenue $562bn; FY2038 gross margin 53% |
| Base | 50% | $237.25 | +46% | Strong platform growth with competitive fade and explicit support costs | FY2028 revenue $742bn; FY2038 gross margin 60% |
| Bull | 25% | $438.68 | +170% | Faster adoption, durable margins and better funding outcomes | FY2028 revenue $805bn; FY2038 gross margin 65% |
| Subjective 25/50/25 weights. Scenarios are not statistically calibrated probabilities or guaranteed bounds. | |||||
| Signpost | Now (Q3 FY2027) | Green if | Red if | Next check |
|---|---|---|---|---|
| Quarterly delivery | $130bn ±3% Q4 guidance | Delivery with sound collections | Revenue or acceptance delays | Q4 FY2027 |
| Margins | 71%–72% Q4 outlook | Rubin margins stabilize | Cost pressure or competitive loss | Q4 and FY2028 |
| Customer credit | 64-day Q3 DSO | Collections and financed deployment | More receivables and support dependence | Next quarterly disclosures |
| Support exposures | $118bn maximum guarantees | Conditions and recoveries remain sound | New defaults or material cash calls | FY2029 onward and earlier updates |
| Governance and filings | Investigation unresolved; 10-Q planned | Verifiable disclosures and oversight | New claims, restrictions or accounting concerns | Next filings and board updates |
Company guidance, reported balances and author assumptions remain distinct. A stronger quarter does not resolve the investigation or guarantee future funding.
Our call is BUY at the retained $162.2562 after-hours quote on November 17, 2026, at 6:03:45 p.m. Eastern. Base present fair value is $237.25 per share, with conditional bear and bull estimates of $67.01 and $438.68. The main question is whether NVIDIA’s extraordinary operating profitability can persist after funding the customers, supply chain and infrastructure on which its growth depends. The range is wide because commercial success and financing risk can move together. It is not a guarantee that prices remain inside those endpoints.
NVIDIA is already producing substantial cash. This is not a valuation of an unproven laboratory or a business that needs years of financing before its first commercial sale. The company’s integrated processors, networking, software and systems make its offering more valuable than an isolated chip specification. At the same time, the financial commitment to sustaining the ecosystem is expanding. Treating the business as a pure capital-light software royalty would miss the cash tied up in receivables, inventory, research infrastructure and customer support.
The main model forecasts customer-platform revenue, gross margins, operating expenses, taxes, owned equipment and operating capital. It then accounts for conditional capacity purchases, expected guarantee losses, financial investments and acquisition claims. We keep those categories separate so normal procurement is not also deducted as debt, and portfolio gains are not counted as both recurring operating earnings and a second asset benefit. Present value is the result of that model, not an average of several methods chosen to produce agreement.
Subjective scenario weights of 25% bear, 50% base and 25% bull produce $245.04 of weighted present value. These weights are judgments, not estimated market probabilities. The base uses a 10.5% business discount rate and a separate 12% investment-return hurdle. Confidence in the precise number is limited by the unresolved investigation, concentrated customers and incomplete current share detail. A 20% entry cushion below base value would be about $189.80; that is a purchase discipline, not another hidden reduction of fair value.
| Metric | Value |
|---|---|
| Quote | 162.26 |
| Base value | 237.25 |
| Weighted value | 245.04 |
Revenue for the quarter ended October 25, 2026, was $115.844 billion, up 103% year over year and approximately 20% sequentially. Gross profit was $86.676 billion and operating income was $77.520 billion. GAAP net income was $65.287 billion, with diluted earnings of $2.70. The operating result is economically important because it demonstrates that the demand surge is reaching profit. It does not, by itself, establish that every component of the quarter can be repeated indefinitely or converted to cash immediately.
Non-GAAP net income was $64.870 billion, slightly below the GAAP figure, and adjusted diluted earnings were $2.68. Equity-security gains and other nonoperating adjustments explain why the adjusted number need not exceed GAAP. Beginning in fiscal 2027, NVIDIA’s non-GAAP measures retain stock-based compensation, and the historical comparisons shown in the current package have been updated accordingly. Our analysis uses those current reconciliations. Applying an older assumption that adjusted profit removes all stock compensation would misstate the quality of this quarter’s earnings.
The written fourth-quarter outlook calls for $130 billion of revenue, plus or minus 3%, gross margin of 71% to 72% and approximately $11 billion of GAAP operating expense. The fiscal quarter has fourteen weeks, so part of the sequential comparison reflects an extra week. The outlook includes no Data Center compute revenue from China. Q3 contained $2.4 billion of licensed Hopper Data Center shipments to China, according to the CFO commentary. Geography based on customer headquarters is a different disclosure and should not be substituted for that product-specific amount.
Management’s earnings-call commentary raised its fiscal 2028 revenue-growth outlook to approximately 75%, from approximately 70%. We retain that statement through the identified transcript of management’s call and distinguish it from the written quarterly release. The base forecast of $742 billion for fiscal 2028 is an author assumption close to that ambition, not a separately disclosed company range. Supply, powered facilities, customer financing and final deployment all have to support the projected growth. A strong quarterly result does not remove those conditions.
| Metric | Value |
|---|---|
| Prior Q3 | 57.01 |
| Prior Q2 | 96.22 |
| Current Q3 | 115.84 |
| Metric | Value |
|---|---|
| Operating profit | 77.52 |
| Net income | 65.29 |
| Free cash flow | 43.78 |
NVIDIA reports two statutory segments, Compute & Networking and Graphics. Its market-platform discussion separately describes Data Center and Edge Computing, with Data Center split into Hyperscale and AI Clouds, Industrial & Enterprise. The model uses those market categories because they help distinguish customer demand and funding. They should not be presented as three audited operating segments with separately disclosed profit margins. Group margin assumptions are applied to the combined forecast, with the limitations of that approach stated explicitly.
Hyperscale revenue was $60.1 billion in Q3, growing 71% year over year. AI Clouds, Industrial & Enterprise contributed $47.5 billion, growing 195%, while Edge Computing contributed $8.2 billion, growing 42%. The first group includes large established cloud buyers; the second includes a broader range of customers with different credit profiles and access to capital. Those categories are not perfectly clean credit buckets, but their growth helps explain why a single aggregate revenue rate is insufficient for assessing cash and concentration risk.
The economic advantage is an integrated computing system that developers and customers can use together. Networking, software libraries, system design and a trained ecosystem can reduce time to deployment and improve utilization. That gives NVIDIA pricing power beyond the manufacturing cost of a chip. The principal substitutes include custom accelerators, competing GPUs and alternative architectures tailored to particular workloads. Customers do not need to replace every NVIDIA product for bargaining power to increase or for margins to normalize over time.
The base starts fiscal 2028 with $390 billion of hyperscale revenue, $310 billion from AI Clouds, Industrial & Enterprise and $42 billion from Edge Computing. Data-center growth then fades substantially over the explicit forecast. By fiscal 2038, total revenue is approximately $1.784 trillion. That remains a very large success case, even though growth slows. The model does not claim that today’s announced projects establish this long-range outcome. Customer monetization, capacity utilization and repeat purchasing are the evidence needed to support it.
| Metric | Value |
|---|---|
| Hyperscale | 60.10 |
| AI / industrial | 47.50 |
| Edge | 8.20 |
| Metric | Value |
|---|---|
| FY2028 | 742.00 |
| FY2032 | 1,339.27 |
| FY2038 | 1,784.20 |
Q3 operating cash flow was $48.253 billion. Purchases of property, equipment and intangible assets were $4.422 billion, with another $0.054 billion of related principal payments. NVIDIA’s free-cash-flow definition therefore produces $43.777 billion. That is substantial cash generation, but it precedes $15.696 billion of equity-security purchases and $2.718 billion of acquisition-related investing cash outflow. A label such as free cash flow describes a defined measure; it does not settle how much cash management can distribute after all strategic decisions.
Receivables were $81.033 billion at October 25, inventory was $39.724 billion and accounts payable were $21.074 billion. The CFO reported 64 days sales outstanding, up from 60, attributing the increase to launch timing and extended terms on certain large investment-grade customer arrangements. Those explanations may be reasonable while still representing a real funding requirement. Receivable collection and inventory turnover should be tracked alongside sales, particularly when customers themselves depend on financing to deploy and monetize computing capacity.
The starting core working-capital proxy is receivables plus inventory and $3.506 billion of prepayments, less accounts payable: $103.189 billion. We exclude mixed accruals and tax balances because the current supplemental disclosure does not supply a complete operating breakdown. Base forecasts use 65 receivable days, 115 inventory days and 65 payable days; prepayments are 0.8% of revenue. This deliberately inspectable approximation is preferable to claiming a complete cash conversion reconciliation that the available detail cannot support. Bear assumptions extend collection and inventory requirements.
The opening financial position is dated October 25, while valuation is November 17. We roll it forward with a full fourteen-week Q4 forecast, including an estimated 23 days already elapsed, and discount that cash to the fiscal quarter end. The elapsed portion contributes approximately $0.54 per share of base present value. It is a forecast, not an observed November cash balance. Removing that component would modestly lower the estimate. No unreported post-quarter buyback or share reduction is assumed to offset it.
| Model input | Base treatment |
|---|---|
| Opening selected core capital | $103.189bn |
| Receivable / inventory / payable days | 65 / 115 / 65 |
| FY2028 core capital increase | $57.024bn |
| Required cash floor | $15bn, then at least 2% of revenue |
| Elapsed Q4 estimate | Approximately $0.54/share present value |
Supply and capacity commitments increased from $279 billion to $390 billion, primarily because of multi-year memory procurement. The disclosed schedule includes $203 billion in fiscal 2028. These contracts can secure critical components and protect delivery capacity, but they also leave the company exposed if demand, architecture or prices change. We charge normal procurement through product costs and working capital. Deducting the entire commitment as debt would charge the business twice for inventory that is also assumed to generate future revenue.
Own-use cloud service commitments total $28 billion, and uncommenced data-center leases total $74 billion. These support engineering, research and model development. The leases can run for up to twenty years, with expected commencement between fiscal 2027 and fiscal 2033. The model retains cash rent and cloud spending in operating expenses, alongside owned equipment investment. It does not also subtract all operating lease liabilities from enterprise value. Potential assignment of leases to third parties is not assumed to eliminate the obligation without cost.
The separate $32 billion neocloud agreements depend on financing, deployment and acceptance conditions. In base, purchased capacity is assumed to recover 70% of cost, leaving $9.6 billion of net burden over the schedule. Bear recovery is 35%, and bull recovery is 90%. These are author estimates of economic recovery, not disclosed resale prices or guaranteed revenue shares. The long-dated $8 billion tail is spread across fiscal 2032 to 2035 for modeling. Capacity costs are distinct from default guarantees and are not counted a second time inside their loss severity.
Maximum gross guarantee exposure is $118 billion, including $105 billion associated with phased SB Energy facilities leased to OpenAI. The first phase is expected in fiscal 2029, with exposure declining as lease payments are made; NVIDIA reports no guarantee payments to date. Base expected losses are $5.9 billion, using 10% default probability and 50% loss severity. Bear and bull losses are $21.24 billion and $0.59 billion. These judgmental assumptions are not actuarial estimates, and actual maximum exposure is much larger. Losses are phased separately from ordinary capacity payments.
| Metric | Value |
|---|---|
| Supply / capacity | 390.00 |
| Own cloud / leases | 102.00 |
| Neocloud agreements | 32.00 |
| Metric | Value |
|---|---|
| Maximum gross | 118.00 |
| Bear loss | 21.24 |
| Base loss | 5.90 |
Cash of $29.089 billion plus marketable debt securities of $24.423 billion totals $53.512 billion. The supplemental disclosure identifies $3.471 billion of the debt-security balance as convertible into publicly held equity. We move that amount once into the investment portfolio, leaving $50.041 billion in the liquid category. This is a modeling classification, not a claim that every remaining security is identical to a bank deposit. At least $15 billion remains with operations, with incremental required cash funded as revenue grows.
Current public equity is $44.219 billion, and another $4.669 billion is classified in other assets, giving $48.888 billion. The nonmarketable balance is $65.528 billion, including $61.650 billion of private equity and other nonmarketable holdings. Those categories are separate from the public-equity convertible debt reclassified above. The company identifies $21 billion of investments subject to short-term lockups, while long-term public holdings are restricted through December 2027. Investment marks therefore cannot all be treated as immediate funding for distributions.
Base realizes 95% of public equity and convertible-security carrying values, and 75% of nonmarketable value, producing $98.887 billion of existing investment value. These combined realizability, tax and liquidity allowances are assumptions, not measured tax liabilities. Bear uses harsher factors; bull assumes better realization. Future equity commitments consume the full $20 billion of modeled cash before any later asset credit. Base recognizes $15 billion of resulting investment value at fiscal 2031, separately from the cash spent. It does not distribute those investment marks as if they were cash today.
Debt face value is reconstructed at $32.5 billion from the prior detailed schedule after the Q3 $1 billion repayment, versus current carrying value of $32.371 billion. The next major scheduled maturities are in calendar 2028. The pending Hugging Face purchase is approximately $11.9 billion, subject to adjustments and closing conditions, with a separate retention program of up to $1 billion. Full purchase funding is reserved; retention is future compensation. We also preserve the older $0.986 billion Groq accrued claim because no current detailed settlement reconciliation is supplied.
| Claim or allocation | Model treatment |
|---|---|
| Funded debt | $32.5bn face deducted once |
| Operating liquidity | $15bn minimum retained |
| Hugging Face | $11.9bn purchase funding reserve |
| Retention program | Up to $1bn inside compensation costs |
| Groq accrued claim | $0.986bn dated reserve |
| Existing portfolio, base | $98.887bn estimated realizable value |
The model uses 24.186 billion Q3 weighted-average diluted shares. That number is not a verified November spot fully diluted capitalization. Q3 weighted basic shares were 24.106 billion, while the August 21 cover-page common count was 24.1 billion. July’s equity-award disclosure reported 168 million outstanding restricted and performance-based awards. These figures describe different dates and concepts; combining them without a reconciliation could double-count some claims or imply a precision that the current supplemental package does not provide.
Future compensation remains an economic operating cost. We do not add stock compensation back to cash flow and then assume future engineering and employee services are free. We also avoid a second automatic future share increase for that same modeled compensation. Existing award dilution is approximated by the dated diluted denominator, with a sensitivity using alternative share counts. This is a practical proxy rather than a grant-by-grant valuation. The next current filing should improve the reconciliation of vesting, forfeitures, performance awards and tax withholding.
During Q3, repurchase cash outflow was $22.122 billion and dividends were $6.026 billion. The remaining $227.3 billion authorization is not completed repurchases or a guaranteed per-share benefit. Whether a buyback creates value depends on price and the alternative uses of cash. The return model keeps shares flat and distributes available owner cash, avoiding simultaneous credit for both a cash distribution and assumed buyback accretion. Actual capital allocation may differ substantially from that simplified distribution policy.
Sensitivity includes 24.1 billion, 24.186 billion and 24.354 billion shares. The last figure adds the full dated 168 million award balance to the diluted proxy as a deliberately conservative illustration, not as a verified upper bound. Some awards already overlap the accounting dilution and some service costs remain in expenses. The resulting per-share changes are modest relative to the commercial scenarios, but that does not make the limitation disappear. The packet preserves the dates so readers can distinguish reported capitalization from the author’s proxy.
| Metric | Value |
|---|---|
| 24.100bn | 238.09 |
| 24.186bn | 237.25 |
| 24.354bn | 235.61 |
The base starts with $742 billion of fiscal 2028 revenue, 72.5% gross margin and operating expense equal to 10% of sales. Gross margin then fades to 60% by fiscal 2038, while the expense ratio rises to 13%. This preserves strong platform economics but allows competition and customer bargaining to reduce profitability. The cash tax assumption is 20% on annual operating profit. Owned capital investment is 4% of sales and owned depreciation is 1.8%; operating capital and required cash increase with scale.
Fiscal 2028 base after-tax operating profit is $371 billion. Equipment investment is $29.680 billion, and the increase in core working capital is $57.024 billion. Full strategic equity purchases and expected capacity costs are also deducted before owner cash. The model’s definition is broader than the company’s reported free-cash-flow measure because it tracks the additional strategic uses of cash relevant to shareholders. Positive cash generation funds the modeled commitments without speculative new common issuance in the base, bear or bull cases.
Discounting the explicit cash flows and terminal franchise produces approximately $5.644 trillion of operating enterprise value. The net nonoperating bridge is approximately $94.2 billion after liquidity, investment values, debt, acquisition claims and investment funding are reconciled once. Dividing common-equity value by the 24.186 billion share proxy yields $237.25. A large operating franchise dominates this result, so small bridge adjustments do not determine the call. They still matter for avoiding systematic double counting and for understanding where the downside lives.
Terminal value contributes approximately 44.8% of base operating enterprise value. Terminal growth is 2.5%, with a 25% incremental return on capital; the formula requires reinvestment to support that growth. It replaces the explicit annual reinvestment rule at the terminal boundary instead of being added on top. The remaining value comes from the explicit forecast. Durability therefore matters materially, but the valuation is not created entirely by a distant terminal assumption. The bear uses lower growth and returns, higher discounting and more severe support losses.
| Valuation component | Base |
|---|---|
| Explicit present cash flows | $3,116.8bn |
| Terminal present value | $2,527.0bn |
| Terminal share of operating EV | 44.8% |
| Present fair value | $237.25/share |
| Discount / terminal growth / ROIC | 10.5% / 2.5% / 25% |
The independent near-term earnings check starts from reported nine-month adjusted net income of $164.374 billion, which retains stock compensation. It adds fourth-quarter operating earnings implied by the written revenue and margin midpoint, less $11 billion of operating expense and an assumed 17% tax rate. Quarterly net nonoperating income is assumed zero. The resulting normalized fiscal 2027 earnings estimate is approximately $9.61 per share. It is a guide-based estimate, not a reported full-year result or management EPS forecast.
Applying an author range of 18 to 24 times earnings produces approximately $173 to $231 per share. A retained TSMC market-data observation near 24 times trailing earnings provides independent market context. Primary TSMC results also show substantial semiconductor profitability, but its owned fabs, depreciation, Taiwan exposure and position in the value chain differ from NVIDIA. We do not automatically assign an identical multiple. The P/E check is already an equity value, so cash is not added again and debt is not subtracted a second time.
That check is lower than the main DCF because it capitalizes nearer-term earnings rather than explicitly valuing the full long-range growth path. The disagreement reduces confidence in treating the main estimate as a precise target. It does not justify averaging the two approaches until they appear to agree. A stock can look inexpensive on a near-term earnings basis and still disappoint if those earnings are near a cyclical peak, just as a premium multiple can be justified if durable reinvestment produces unusually high future returns.
The reverse calculation scales the data-center path while keeping edge revenue, later growth, margin fade, investment rules and discounting fixed. The dated quote is consistent with approximately $502 billion of fiscal 2028 revenue and $1.194 trillion by fiscal 2038 under those assumptions. That is not a literal forecast attributed to all investors. The same price could reflect lower margins, a shorter cycle or larger legal and financing losses. Operating sensitivity therefore leads with both revenue scale and gross margin, followed by discount-rate and terminal-growth context.
| Metric | Value |
|---|---|
| Check low | 172.95 |
| Check high | 230.61 |
| Main base | 237.25 |
| Scale | Margin shift | Value |
|---|---|---|
| 80% | −5% | $173.76 |
| 80% | +0% | $193.58 |
| 80% | +5% | $213.40 |
| 100% | −5% | $212.72 |
| 100% | +0% | $237.25 |
| 100% | +5% | $261.78 |
| 120% | −5% | $251.66 |
| 120% | +0% | $280.90 |
| 120% | +5% | $310.14 |
The bull case assumes profitable AI adoption continues across hyperscalers, enterprises, sovereign users and physical systems. Integrated hardware, networking and software preserve a premium while revenue scales faster than the base. Customer financing becomes more durable as deployed capacity earns cash, reducing the need for support. In that scenario, a larger business can generate more owner cash without requiring a permanent peak margin. The modeled bull is a conditional success path, not a claim that every announced project or partnership will deliver its headline potential.
The bear mechanism is a correlated demand and credit shock. Customers may commit to equipment ahead of monetization, encounter tighter capital markets and slow deployment. NVIDIA can then face lower sales, slower collections, excess supply, investment impairments and contingent support costs together. The scenario places each effect in its economic location: revenue, margins, working-capital days, portfolio factors and separate support-loss schedules. It does not apply the same loss repeatedly under different labels. An extreme outcome beyond these modeled assumptions remains possible.
Leadership and governance are unresolved risks. The filed November 12 disclosure confirms the November 11 resignation and appointment of Colette Kress as interim chief executive while she remains chief financial officer. A search for a permanent chief executive is ongoing. The new national-security and export-control oversight committee is a governance response, not proof that the board investigation or government and regulatory questions have been resolved. We make no prediction about the outcome and do not rely on unverified personal allegations to generate a valuation adjustment.
Current quarter financial information is unaudited and includes a supplemental package rather than a filed Q3 10-Q. The company says it plans to file the report in coming weeks. That statement does not establish that the filing is overdue, and the earnings release does not substitute for the eventual full notes. The next report should be checked for capital claims, commitments, customer concentration, accounting detail and governance disclosures. Confidence should improve only when actual evidence resolves uncertainties, not because a strong quarter makes them less comfortable to discuss.
| Metric | Value |
|---|---|
| Bear | 67.01 |
| Base | 237.25 |
| Bull | 438.68 |
The January 31, 2031 base price is approximately $280 per share. It is calculated from the remaining operating cash flows and terminal franchise at that future date, plus investment assets and the applicable cash and debt bridge. It is not today’s fair value mechanically compounded at a desired return. Available annual owner cash is assumed distributed after an estimated after-tax debt coupon. Strategic investment purchases consume cash before distributions, while their later asset value remains in the future equity bridge.
Including those timed modeled distributions produces approximately 21.8% annualized base return from the retained quote, versus the 12% required return. These are hypothetical capital distributions, not announced dividends. Shares remain flat, so the calculation does not also assume buyback accretion. The bear return is negative and the bull materially higher. A company choosing to retain cash, buy shares at different prices, complete other acquisitions or increase support commitments would produce a different pattern of shareholder receipts and future value.
The separate November 2027 base price is approximately $259.42. That price excludes the earlier fiscal Q4 distribution. The total-wealth output adds the earlier receipt separately, while the first full fiscal 2028 annual distribution falls after the one-year comparison date and remains in residual value. The $5 billion stub equity purchase is already deducted from cash before that earlier distribution; subtracting it again from the unchanged opening cash bridge would double-charge it. Price, distribution and total wealth are deliberately distinct concepts.
The next tests are concrete: delivery against the $130 billion Q4 revenue range, margins through the Rubin ramp, collection days relative to the current 64-day baseline, and the amount of support required to turn customer demand into financed deployments. Management’s fiscal 2028 ambition needs continuing operational evidence. The next filed financial statements and governance updates are equally important. We would weaken the call if customer dependence on NVIDIA funding rises faster than profitable deployment, even if reported revenue continues to set records.
| Dated outcome | Base assumption |
|---|---|
| January 2031 price | $279.92 |
| Annualized return with distributions | 21.8% |
| Required return | 12% |
| Distribution policy | Available owner cash; not declared dividends |
| Share policy | Flat proxy; no additional buyback accretion |
Reported financial inputs come from NVIDIA’s November 17 earnings release, CFO commentary and supplemental financial information, together with earlier filed notes where current detail is unavailable. The original GlobeNewswire release identifies 4:20 p.m. Eastern, or 3:20 p.m. Central, on November 17. CEO and acquisition terms come from the cited filed disclosures. The call-growth statement is attributed to management through the retained transcript rather than being presented as language in the written earnings release. Current security review does not imply that governance risks have been resolved.
The quote is a retained Yahoo Finance after-hours chart observation. Analyst context comes from a retained Stock Analysis page identifying S&P Global consensus, captured before this release; it is not claimed to reflect analyst revisions after the call. TSMC primary results and a dated secondary market multiple supply independent context. Prices and ratios move, and stale provider pages can differ from a current tape. The analysis fixes its quote and timestamps so readers can distinguish a valuation estimate from a live trading recommendation.
Financial tables use U.S. dollars in billions unless another unit is shown. Per-share values use billions of dollars divided by billions of shares. Fiscal-year labels refer to NVIDIA’s fiscal periods; debt maturity years and the expected Hugging Face closing window are calendar-year descriptions. Full-precision calculations underlie rounded displays, so small apparent addition differences may reflect rounding. Scenario probabilities, discount rates, realization factors, support losses, future distributions and long-range forecasts are author judgments, not company guidance or independently measured probabilities.
The tables expose the material driver paths, cash requirements, capitalization choices and sensitivities so the conclusion can be challenged using the cited evidence. The current diluted-share proxy, older detailed claims, full-quarter roll-forward and incomplete current filing are disclosed limitations. Neither a peer multiple nor a reverse calculation validates the entire DCF. This packet’s constructive conclusion rests on sustained profitable demand with controlled support costs; its uncertainty rests on how strongly those economic and governance conditions can diverge from the modeled base.
| Discount rate | Terminal growth | Value |
|---|---|---|
| 9.5% | 2.0% | $265.32 |
| 9.5% | 2.5% | $272.01 |
| 9.5% | 3.0% | $279.69 |
| 10.5% | 2.0% | $232.80 |
| 10.5% | 2.5% | $237.25 |
| 10.5% | 3.0% | $242.27 |
| 11.5% | 2.0% | $207.16 |
| 11.5% | 2.5% | $210.21 |
| 11.5% | 3.0% | $213.59 |
NVIDIA Q3 release · CFO commentary · Supplemental financial information · Prior Q2 filed notes · Hugging Face agreement disclosure · CEO transition filing · Oversight committee · Management call transcript · TSMC primary results · TSMC dated multiple · Dated analyst context · Dated quote chart