W.003Marvell’s $115B offer for Lumentum is now worth $78B
Marvell pays a fixed 4.5 of its shares for each Lumentum share. Marvell has fallen 32% since the deal leaked, so the offer, $1,132 a share at the prices it was announced on, was worth $771 at Monday’s close: 16% below where Lumentum traded before the news. Lumentum closed at $684, 11.3% under even that, a 12.7% spread. It is the largest chip-industry deal ever agreed, at 30.7× trailing revenue, and it can wait on regulators, China among them, until August 2028.
A fixed ratio: Lumentum’s holders fell with Marvell
Daily closes, $ a share: Lumentum, and what 4.5 Marvell shares were worth. Bloomberg reported the talks after the close on October 28, the price the deal’s premium is measured from; the deal was announced before the open on November 2. The figure over each pair is the premium or, after the leak, the spread.
Data
| Close | Lumentum | Marvell | 4.5 Marvell shares | Premium or spread | Lumentum ÷ Marvell |
|---|---|---|---|---|---|
| Oct 28 | $913.95 | $251.65 | $1,132.43 | 23.9% | 3.632 |
| Oct 29 | $919.53 | $218.34 | $982.53 | 6.9% | 4.211 |
| Oct 30 | $840.70 | $196.40 | $883.80 | 5.1% | 4.281 |
| Nov 2 | $684.40 | $171.41 | $771.35 | 12.7% | 3.993 |
Why the offer shrank
Lumentum is paid in Marvell stock, at a fixed 4.5 shares; the filings describe no collar. So the offer moves with Marvell, and Marvell fell 13.2% the day after the leak, 10.0% the day after that and 12.7% on the announcement, −31.9% in all and, by Benzinga’s count, its worst three days since October 2000. Valued at the October 28 prices it was announced on, the offer was $1,132 a share; when the companies signed, on Sunday, November 1, it was already down to $883.80, and at Monday’s close it was worth $771. Measured the same way, the deal’s $115B of enterprise value was about $78B.
Lumentum fell with it, despite its own numbers. It reported the same day, three days earlier than planned: revenue of $1.26B, up 136.5% on a year earlier, a 42.3% non-GAAP operating margin, a second major near- or co-packaged optics (NPO/CPO) win backed by more than $5B of purchase commitments, and guidance of $1.575B–$1.625B for the next quarter, against a consensus near $1.48B. Its stock fell 18.6% on the day, and 25.1% from its unaffected close. With a fixed ratio, it now trades on Marvell’s price.
The market is marking down the buyer. The deal adds about 455M Marvell shares, 52% more than the 876.8M outstanding (BofA puts dilution near 50%), against $1B+ of yearly savings, mostly in cost of goods, with no timing given. BofA, downgrading Marvell to Neutral, put its earnings per share 4.9%, 10.5% and 14.7% below standalone in 2028, 2029 and 2030, with no revenue synergies. TD Cowen, also downgrading, worried the deal may come at or near the peak of an optical mini-cycle. Reuters put the price at about 33 times Lumentum’s projected core operating profit, against about 16 for its peers. Benzinga counts 25 of 27 analysts still rating Marvell a buy.
What the spread implies Lumentum is worth alone
If the deal closes, a Lumentum share becomes 4.5 Marvell shares, $771.35 at Monday’s prices; if it fails, it is worth whatever Lumentum is worth on its own. Today’s $684.40 is the probability-weighted mix of the two, so each closing probability implies a standalone value. Today’s money; it leaves out the return arbitrage funds demand for tying up capital, which would raise every value in the table.
| Chance the deal closes | Implied value of Lumentum alone | Against its unaffected $913.95 |
|---|---|---|
| 50% | $597 | −35% |
| 60% | $554 | −39% |
| 70% | $482 | −47% |
| 80% | $337 | −63% |
At any chance of closing, the implied standalone value is below Lumentum’s own price of $684.40, and above about 89% it falls to zero. On this arithmetic the market has marked down Lumentum on its own as well as the deal: its unaffected $913.95 isn’t a price it expects to see again soon. A semiconductor selloff (BofA counts the sector down 19% in a month) explains part of it. If the deal fails on regulatory grounds, Lumentum also collects the $3.2B fee, about $32–35 a share, so each implied value would be that much lower, and the zero point about 88%.
Lumentum’s revenue rose 3.7-fold in two years
Revenue by fiscal quarter (years end in June), and the non-GAAP operating margin over each bar. The outlined bar is the midpoint of guidance for the quarter to December.
The largest chip-industry deal ever agreed, though not the largest tech deal
Selected chip and tech deals by announced value, $B. Grey: closed. Light: never agreed or terminated. For Marvell–Lumentum, the solid bar is the deal’s value at the November 2 close; the outline runs to the $115B it was announced at. Deals between private companies, such as SpaceX’s $250B merger with xAI this year, are left out.
It is more than twice the $49B AMD–Xilinx was worth when it closed in 2022, the largest completed chip deal on that measure, and more than ten times Marvell’s purchase of Inphi in 2021. Calling it the largest merger of two public tech companies needs a qualifier: AOL and Time Warner, both public, agreed to an all-stock deal worth $164.75B in 2000, and Broadcom bid $117B for Qualcomm in 2018 before a presidential order blocked it.
The terms, and the clock
| Term | What it says | Detail |
|---|---|---|
| Exchange ratio | 4.500 Marvell shares per Lumentum share | Fixed; the filings describe no collar or adjustment; cash for fractions |
| Value as announced | $1,132.43 a share, about $115B of enterprise value | At October 28 unaffected prices: Marvell $251.65; 23.9% over Lumentum’s $913.95. Signed November 1 |
| Ownership | About 67% Marvell holders, 33% Lumentum holders | About 455M new Marvell shares (4.5 × Lumentum’s 101.0M diluted), 52% more than the 876.8M outstanding |
| Synergies | $1B+ a year, mostly cost of goods | No timing given |
| Termination fees | $5.8B from Marvell; $2.9B from Lumentum | If a board changes its recommendation or takes a better offer |
| Regulatory fee | $3.2B from Marvell, 2.8% of enterprise value | If approvals are denied or blocked, or the outside date passes, with everything else done |
| Outside date | November 1, 2027 | Extendable to February 1, May 1 and August 1, 2028 while approvals are pending |
| Approvals | Both sets of shareholders; HSR; “certain other required regulatory approvals” | The filings don’t name China; the target close is the first half of 2028 |
Big chip and tech deals took 6 to 21 months to close or fail
Months from announcement to close, or to termination. Broadcom–VMware, Synopsys–Ansys, Microsoft–Activision and Marvell–Inphi closed after China’s approval; Qualcomm–NXP ended without one. Marvell–Lumentum: 14 months to the start of its first-half 2028 target, 21 to the last outside date.
Why it takes so long
The filings don’t say, and they don’t name China. They list U.S. antitrust clearance and “certain other required regulatory approvals.” But both companies sell heavily into China: Marvell shipped 36% of its last fiscal year’s revenue there (41.2% over the trailing year), and Lumentum booked 26.7% of its fiscal 2026 revenue in China and Hong Kong. Both look comfortably above China’s filing test, RMB 800M of China turnover each, though those figures are by ship-to destination and Hong Kong is outside mainland merger control. Benzinga calls China’s approval “the biggest unknown”; Cantor calls the wait “deal purgatory.”
The fees put a price on failure. If regulators block the deal or the last outside date passes with everything else done, Marvell pays Lumentum $3.2B, 2.8% of the deal’s announced value and about 5.2% of Lumentum’s market value at the Monday close. China’s regulator cleared all seven earlier optical deals on Schulz Duggan’s list, four with conditions, after 5 to 18 months; Qualcomm walked away from NXP in 2018 when no Chinese decision came.
Until then, the spread pays for the wait. 12.7% if the deal closes, about 7.5% a year to the end of June 2028. That is on top of what cash earns: an arbitrage position, long Lumentum and short 4.5 Marvell shares, also collects interest on the short sale’s proceeds.
What this changes on 06.ai
- Tracker lead: the Marvell–Lumentum spread, daily, until the deal closes or breaks: the offer’s value, the spread, its yearly rate to each outside date, and the standalone value it implies. The same frame fits any large all-stock AI deal.
- Pack lead: optics in the AI buildout. Lumentum’s revenue rose 3.7-fold in two years and its guidance is above consensus again; the sell-side disagrees on the size of the optical market in 2027 by about two to one (TD Cowen’s $47B against about $100B implied by consensus). No 06.ai pack covers the optics market yet.
Sources and method
| Item | Figure | Source |
|---|---|---|
| Merger agreement | Exchange ratio 4.500; termination fees $5.8B, $2.9B and the $3.2B regulatory fee; outside date and extensions; approvals | Marvell 8-K, via StockTitan |
| Deal release and slides | About $1,132 a share on unaffected Oct 28 closes; 23.9% premium; about $115B enterprise value; 67/33 ownership; $1B+ synergies; first-half 2028 close | Lumentum 8-K, via StockTitan |
| Lumentum Q1 FY27 | Revenue $1,262.6M (Q1 FY26 $533.8M); non-GAAP gross margin 53.2%, operating margin 42.3%, EPS $4.57; Q2 guidance; a second NPO/CPO win backed by over $5B of purchase commitments | Lumentum, Nov 2 |
| Lumentum, earlier quarters | Q4 FY26 and FY26; Q1 FY27 reporting date first set for Nov 5 | Lumentum, Q4 FY26 release |
| Lumentum, earlier quarters | Q1 FY26 | Lumentum, via Business Wire |
| Lumentum, earlier quarters | Q1–Q2 FY25 | Lumentum, via Business Wire |
| Lumentum, earlier quarters | Q3–Q4 FY25 | Semiconductor Today |
| Lumentum, earlier quarters | Q3 FY26 | Lumentum |
| Lumentum, earlier quarters | Q2 FY26 | Lumentum, via Business Wire |
| Reporting date | Q1 FY27 results originally set for Thursday, Nov 5 | Lumentum |
| Prices | Lumentum daily closes, Oct 28–30 | StockAnalysis |
| Prices | Lumentum’s Nov 2 close, $684.40, down 18.59% | FinanceFeeds |
| Prices | Marvell daily closes, Oct 28–30 | StockAnalysis |
| Prices | Marvell’s Nov 2 close, $171.41 | The Motley Fool |
| Valuation | About 33 times projected core operating profit, against a peer median of about 16 (Reuters); Q2 guidance above the $1.48B analyst average | BERI |
| Analysts | BofA to Neutral ($400 to $225); TD Cowen to Hold ($350 to $175), worried the deal may come at or near the peak of an optical mini-cycle; TD Cowen’s $47B against about $100B; the sector down 19% in a month | Investing.com, via Yahoo Finance |
| Analysts | BofA: EPS 4.9%, 10.5% and 14.7% below standalone in 2028–30 with no revenue synergies; Cantor’s “deal purgatory” for at least 5–7 quarters; China’s approval “the biggest unknown”; Marvell’s worst three days since October 2000; 25 of 27 analysts rate it a buy | Benzinga, via TradingView |
| China | Lumentum FY26 revenue from China and Hong Kong, 26.7%; China’s filing test; SAMR cleared all seven earlier optical deals on its list, four with conditions, in 5–18 months | Schulz Duggan |
| Marvell | 876.8M shares outstanding; revenue shipped to China 36% of FY26, 41.2% over the trailing year | sec-api.io |
| Deal history | AOL–Time Warner $164.75B, all stock | CNN Money |
| Deal history | Broadcom’s $117B bid for Qualcomm, dropped after it was blocked | Light Reading |
| Deal history | Microsoft–Activision $68.7B, Jan 2022 to Oct 2023 | Wikipedia |
| Deal history | Dell–EMC $67B | Phys.org |
| Deal history | Broadcom–VMware $61B at announcement; closed after 18 months | Bloomberg Government |
| Deal history | Synopsys–Ansys $35B; closed July 2025 after China’s approval | Tom’s Hardware |
| Deal history | Nvidia–Arm, about $40B, terminated Feb 2022 | AI Business |
| Deal history | Qualcomm–NXP $44B, terminated July 2018 without China’s decision | Yahoo Finance |
| Deal history | Marvell–Inphi, about $10B, closed April 2021 | Optical Connections |
| Deal history | Avago–Broadcom $37B; Intel–Altera; AMD–Xilinx | Reuters factbox, via MarketScreener |
| Deal history | Largest deals by value: no agreed chip-industry deal above $50B before this one; JDS Uniphase–SDL $41.0B; AMD–Xilinx $35B at announcement | Wikipedia, largest M&A |
| Deal history | AMD–Xilinx worth $49B when it closed in February 2022 | M&A Community |
The offer. 4.5 times Marvell’s close. Its value as announced, $1,132.43, uses Marvell’s unaffected October 28 close of $251.65 and matches the companies’ “about $1,132”; the premium is over Lumentum’s unaffected $913.95. The spread is the offer’s value over Lumentum’s price, less one; its yearly rate compounds it over the 606 days from November 2 to June 30, 2028. The deal’s value at the Monday close scales the announced $115B by the offer’s fall; Lumentum’s net debt is small enough for that to hold. Marvell’s change in market value uses its 876.8M shares at August 1. Marvell’s November 2 close, $171.41, is The Motley Fool’s; Benzinga’s Tuesday quote, up 2.4% to $175.60, implies the same.
Implied standalone value. Lumentum’s price = p × the offer’s value + (1 − p) × its standalone value, solved for the standalone value at each probability p. It treats both outcomes in today’s money: a claim on 4.5 Marvell shares at closing is worth 4.5 shares today, less about $1.90 of Marvell dividends before then ($0.06 a quarter), so there is no time-value discount. It ignores the return arbitrage funds require and the cost of borrowing Marvell shares to hedge, both of which make the true probabilities higher than this arithmetic suggests.
Multiples divide the $115B enterprise value by trailing-twelve-month revenue (Q2 FY26 to Q1 FY27, $3.74B) and by four times the $1.60B guidance midpoint. Deal sizes are announced values as reported at the time, not adjusted for inflation; AOL–Time Warner is its equity value. Months to close run from announcement to completion or termination.
Wire reports are written within hours of the news, on a six-hour schedule. Each lists its sources and shows the arithmetic behind its figures. Not investment advice.