Is Marvell Technology, Inc. (MRVL) overvalued or undervalued?
Updated Nov 10, 2026 · hand-researched model · educational, not advice
At a recent price of $154.60, Marvell Technology, Inc. (MRVL) trades 59% above TickerWorth's blended fair value of $97.39 per share. For this company that blend leans on peer multiples: our discounted-cash-flow model holds margins flat and fades growth toward a long-run rate, which systematically understates a business growing this fast, so its figure (about $43.14) is shown as a reference beside the range rather than blended into it. Our estimate sits below the market — read that as a statement about how the market prices MRVL against comparable companies, not as a finding that it is overvalued on cash flow. This is an educational model, not a price target or a recommendation.
The valuation discounts Marvell Technology, Inc.'s projected free cash flows at a weighted-average cost of capital (WACC) of 14.6%, with a long-run terminal growth rate of 2.5%. About 66% of the enterprise value comes from the terminal value — the portion beyond the explicit forecast window — which is typical for a stable, cash-generative business and a reminder that small changes in the discount rate or terminal growth assumption swing the intrinsic value meaningfully.
Running the DCF in reverse — solving for the growth the current price already assumes — MRVL would need sustained revenue growth of more than 25%/yr to justify today's price. We don't quote a precise figure here, because past about 25% a year the reverse-DCF stops being informative: it solves for revenue growth alone, holding margins, capital intensity and the forecast horizon fixed, so a price well above our fair value has nowhere to go but growth. The honest reading is that the price implies growth this model can't reconcile — not that we forecast it.
Sensitivity matters more than any single number. Across a reasonable band of discount-rate assumptions, MRVL's modeled DCF value ranges from about $35.67 to $54.47 per share. If that range still sits below today's price, the market is optimistic relative to the model across the whole band; if it brackets the price, the stock is roughly fairly valued on our assumptions. Use the interactive slider to see exactly where your own cost-of-capital view lands.
How to read this page: the DCF fair value is what Marvell Technology, Inc. would be worth if its cash flows grow as modeled and are discounted at the stated WACC. The assumptions table shows every input and its source; the reverse-DCF shows what the market believes; the sensitivity band shows how fragile the answer is. None of this is investment advice — it is a transparent, auditable framework for forming your own view of MRVL's intrinsic value. Figures update as new filings and prices come in.
The assumptions behind MRVL's DCF
| Input | Value | Source |
|---|---|---|
| Base-year revenue | $8,194,600,000 | Marvell Q4/FY2026 earnings release, 8-K Ex-99.1 filed 2026-03-05; ties to FY2026 10-K (accession 0001835632-26-000011, filed 2026-03-11): Net revenue for the fiscal year ended 2026-01-31 (52 weeks), +42% on FY2025's $5,767.3M. End markets (10-K Note 3): Data center $6,100.3M (74%, +46%) and Communications and other $2,094.3M (26%, +31%). THE BASE CONTAINS A DIVESTED BUSINESS: the automotive ethernet business was sold to Infineon on 2025-08-14 for $2.5B cash (10-K MD&A), so FY2026 carries ~6.5 months of its revenue. The 10-K does not size that revenue, so I cannot strip it out; it slightly understates the organic growth rate implied by the forecast. UNCHANGED as the base; the FY2027 ramp belongs in the forecast. CROSS-CHECK: H1 FY2027 revenue was $5,157.1M, +32.2% on H1 FY2026's $3,901.4M (Q2 FY2027 release, 8-K Ex-99.1 filed 2026-08-27); Q2 alone $2,739.3M, +37%. |
| Base EBIT margin | 27.6% | Marvell Q4/FY2026 earnings release, 8-K Ex-99.1 filed 2026-03-05; FY2026 10-K (accession 0001835632-26-000011): THIS IS EBITA, NOT GAAP OPERATING INCOME — the AMD method in this tracker, not the IBM method, and it is the most important accounting call in the model. GAAP operating income $1,322.9M (16.14%) PLUS amortization of acquired intangible assets $942.0M (cash-flow statement; $639.0M in cost of goods sold + $303.0M in opex per the GAAP-to-non-GAAP reconciliation) = $2,264.9M / $8,194.6M = 27.64%. Everything ELSE stays expensed: stock-based compensation $590.8M (7.2% of revenue), restructuring $15.5M, acquisition costs. So this is NOT Marvell's non-GAAP operating margin (35.3% for FY2026), which also strips out SBC. WHY EBITA: the amortization is overwhelmingly Inphi (2021) and Cavium (2018), deals already paid for, largely in stock that is already in the share count, and it runs OFF fast on the schedule Marvell files: $385.3M for the rest of FY2027, $292.7M FY2028, $139.6M FY2029, $117.2M FY2030, $60.8M FY2031, $54.0M thereafter (Q2 FY2027 10-Q, Note 5). The engine holds D&A at a fixed % of revenue, so keeping GAAP EBIT would freeze an 11.5%-of-revenue charge that is ~1% by FY2029 into the terminal year forever. Celestial AI's $951M of acquired intangibles is all IPR&D, which is NOT amortized until the technology reaches production, so it is not yet in the schedule. CROSS-CHECK, same derivation: H1 FY2027 GAAP operating income $799.1M + amortization $440.1M = $1,239.2M / $5,157.1M = 24.03%; Q2 alone $459.7M + $214.9M = 24.63%. The drop from 27.6% is stock compensation: SBC jumped to $326.2M in Q2 (11.9% of revenue, from 7.7% a year earlier) on acquisition headcount and Celestial's assumed awards ($190.1M of post-combination expense to recognize). |
| Forecast revenue growth | 46% → 50% → 20% → 12% → 7% | Assumption — TickerWorth's 5-year explicit forecast |
| Risk-free rate | 5.2% | US 10Y Treasury constant maturity (FRED DGS10): 5.24% on 2026-10-01, the latest published observation. House input shared by every model in this tracker. |
| Beta | 1.90 | Assumption — normalized estimate; published Yahoo/yfinance 5Y-monthly beta 2.253 read 2026-10-03: Published 5Y-monthly beta is 2.253; I reproduced it from yfinance monthly prices vs the S&P 500 at 2.282 (60 returns to 2026-10). NORMALIZED DOWN to 1.90, which RAISES the value. Evidence the published figure is inflated by one event: dropping April 2026 alone (MRVL +66.8% against the S&P +10.4%, the month NVIDIA's $2B investment was announced) takes the regression from 2.28 to 1.91; the first 36 months of the same window, before the AI re-rating, give 1.90. But I do NOT go as low as the 1.60 this tracker uses for AMD and MU, because Marvell's beta was already ~2.1 before 2026 (window ending 2025-12: 2.12) — it is a structurally high-beta stock, with a 52-week range of $70.69-$329.88. Sensitivities: published 2.28 gives ~$36.40; the tracker's 1.60 gives ~$50.15. |
| Terminal growth | 2.5% | Assumption — house default: House default 2.5% (roughly long-run nominal GDP), kept for comparability with every other company in this tracker. For a company still guiding +50% two years out this is the structural reason the model lands far below the price: a 5-year explicit window that ends at +7% and then grows at 2.5% forever cannot represent a longer high-growth runway. 2.0% takes ~$1.31/share off, 3.0% adds ~$1.43. Well below the 14.55% WACC. |
| Diluted shares | 931M | Assumption — Marvell Q2 FY2027 earnings release, 8-K Ex-99.1 filed 2026-08-27; Q2 FY2027 10-Q Notes 3 and 15; Google warrant 8-K filed 2026-08-19 + estimate for warrants: REPORTED BASE + A STATED ADJUSTMENT. Q2 FY2027 weighted-average diluted shares were 921.2M, which already counts NVIDIA's $2B Series A preferred as converted (21.8M shares, conversion price ~$91.84) — the release labels the count 'common stock and preferred stock assuming conversion'; Q3 is guided at 921M. ADDED (+9.8M): (1) HALF of the Google warrant on the treasury-stock method — 58,970,907 shares at a $206.58 strike, issued 2026-08-18, vesting one tranche per $500M of Google custom-product revenue from Q3 FY2027 through FY2033. The forecast assumes Google revenue that vests part of it, so counting the revenue and ignoring its price would be one-sided; but full vesting needs ~$120B of purchases, far beyond this model's path. Half × (1 - 206.58/272.29) = +7.1M. (2) The unvested older customer warrants (3.0M at $87.77 and 1.0M at $87.00, 10-Q Note 3) on the same method = +2.7M. 921.2 + 9.8 = 931.0M. LOWERS value slightly: no warrants gives ~$43.60, the whole Google warrant ~$42.81. |
| Market price | $154.60 | Delayed quote · as of 2026-11-10 05:05 UTC |
Explore more
Most undervalued stocks (per our valuation) · Undervalued Technology stocks · All valued stocks · Interactive MRVL valuation & downloadable modelTickerWorth is an educational DCF tool, not a licensed investment adviser. Figures are estimates derived from public filings and market data and can be wrong.