What does $190 imply?

Edit a number to recalculate. Monetary inputs and forecasts are USD billions; shares are billions. The default price is the case's July 29 close, not a live quote.

Revenue growth and operating margin

Years 7–10 growth fades linearly from the Year 6 rate to terminal growth. Margins fade from Year 5 to the Year 10 margin. The terminal year continues the mature assumptions.

Ten-year revenue and free cash flow

From operations to equity value

Sensitivity to mature margin and terminal growth

Each cell reruns the DCF, including the growth and margin fades. The outlined cell is the explorer. Invalid combinations are blank.

Bear: efficiency reduces spending and customers substitute custom silicon. Base: demand expands while inference mix and input costs reduce margins. Bull: usage grows faster than efficiency improves and the rack-scale platform protects margins.

These scenario drivers are independent of the explorer. Starting revenue, price, shares and asset balances are shared. Edit the three numeric columns; probabilities must total 100%.

The reverse DCF uses the explorer assumptions; the recommendation uses the weighted scenarios. They answer different questions and are labeled separately.

Model mechanics

Revenue grows annually. EBIT = revenue × operating margin. Cash tax = max(EBIT, 0) × tax rate. FCFF = EBIT − cash tax + D&A − capex − change in operating working capital. Change in NWC = the NWC absorption rate × change in revenue, so a contraction can release cash.

All ten future annual cash flows are discounted at year-end. The base is TTM revenue through April 26, 2026 (215.938 − 44.062 + 81.615 = $253.491B). Year 1 is a future annual modeling period from the July 30 case date, not reported FY2027. The roughly $393B Year 1 level is calibrated to the scale of the case's $393.6B FY2027 consensus, not its exact fiscal timing. A quarterly model would refine this approximation.

Terminal value = FCFF₁₁ / (WACC − g). Year 10 has already reached g and the mature margin, so FCFF₁₁ = FCFF₁₀ × (1 + g). Equity = operating EV + cash + investment values − debt − other claims. Divide by 24.22B shares by default. Financial debt and cash are not discounted again.

Why these starting assumptions?

The default 65% first-year operating margin is near Q1 FY2027's 65.6%. The base path fades to 52% as inference mix, custom silicon and memory costs affect pricing. The 17% tax rate is the midpoint of company guidance (16–18%). Capex at 3% is close to FY2026's 2.8%, D&A at 1.3% matches the exhibit, and working-capital absorption at 19.3% uses the FY2026 ratio rather than extrapolating Q1's release.

WACC at 11% reflects an approximately 1.49 beta using the case's 4.70% risk-free rate and 4.23% ERP. This is an equity-dominated approximation given very small debt funding. The 2.21 regression beta implies about 14.05% cost of equity; the bottom-up 1.35–1.75 range implies 10.41–12.10%. Test the disagreement. Growth, future margin paths, scenario probabilities and the 25% private-investment haircut are proposed judgments, not figures reported by NVIDIA. Terminal growth of 3% is nominal and deliberately much lower than near-term growth.

Accounting choices and limitations

GAAP operating margins already expense stock compensation; it is not added back. Share count stays constant: buybacks and dividends are financing distributions, not additional operating cash flow. R&D remains expensed. Investment gains are excluded from EBIT; the underlying investments enter the bridge once. Private investments and hardware demand can fall together, so haircuts rise in the bear scenario. No recurring future equity-investment spending or explicit customer-financing schedule is modeled. A forecast that relies on such funding would need additional reinvestment and credit-loss assumptions. The case's proposed large guarantees are not booked as debt. Operating leases remain operating costs; no separate lease-debt adjustment is made. Tax losses have no immediate refund and no modeled carryforward. No live quote feed is used.

Sources and verification

• Asaf Manela, NVIDIA Corporation: Valuing the Engine of the AI Revolution, July 30, 2026. Case sections 4–6 and Exhibits 1, 2, 4, 5, 6 and 8; supplied PDF and workbook. These are the controlling case inputs.
• NVIDIA Q1 FY2027 results and NVIDIA Investor Relations.
• Aswath Damodaran's valuation writing, with the break-even setup taken directly from Section 6.1 of the supplied case.
• Gradio documentation and Hugging Face Gradio Spaces.

The engine is checked against a hand-calculated perpetuity, an independent formula workbook, the case's $855B shortcut, and reverse-DCF repricing. This is a classroom valuation with substantial model uncertainty. The written reflection and final scenario choices require the student's review.