Celestica
A contract builder for the cloud giants: it assembles the network switches and AI server racks they order, and increasingly helps design them too.
The scores
Growth3/4 ↑
Fast and organic, but narrowing - the top ten customers went from 64% of sales to 79%, and one end market drives all of it.
Business quality2/4 ↑
Programs get re-bid and dual-sourced by its own account, and every design rival it names is larger. Its wider margin is a symptom.
Financial position3/4 →
Net cash and low leverage, but the plan was not funded from operations: it needed $3.45B of equity nine days after the filing said otherwise.
Execution record3/4 →
Eight straight quarters above guidance, but the July filing said operations would fund the plant spending and an equity raise followed nine days later.
Per share capture2/4 ↓
The count was flat for a year, then rose 9.7% in one week at a 16.5% discount, unrestated in the earnings guidance.
Potential2/4 ↑
It is somebody else's spending plan and nothing is contracted: the annual report never says backlog, and no agreement states a volume or term.
Risk profile2/4 ↓
Three customers are 63% of quarterly sales and Thailand makes 59% of what it sells; either concentration on its own would be the level above.
Valuation2/4 →
The price needs the plan and the acceleration beyond it. Remove the share-price swap and the trailing profit looks dearer still.
Evidence quality3/4 →
Every load-bearing figure is filed, but what the price is measured against is management's own guidance and a vendor's analyst panel, neither verified.
The Verdict
We rate Celestica unattractive. It builds network switches and AI server racks for the cloud giants, and increasingly helps design them. The business is mixed: sales are growing fast, but extra sales barely add cash, shareholders have just put in new money, and part of the profit is a gain tied to its own share price. The price is full, not cheap: buyers count this year's plan and next year's faster growth as done; any slip hurts. The view hinges on free cash rising with sales without more shares being sold. The growth is real; the profit the price rests on is flattered, and the growth costs ever more capital.
The long caseCloud giants are filling new AI data centres with switches and server racks Celestica builds; their spending must keep rising and its extra sales must start bringing in cash.
The evidence under this - 9 claims
- Reaching it needs fourth-quarter revenue of about $6.35B, 73.9% above Q4 2025 and 17.7% above the midpoint of the third-quarter guidance.
- The company has raised this outlook three times in 2026 and has beaten its own adjusted earnings guidance in each of the last eight quarters, so the record runs the other way.
- 2026-10-27editedthe September guide is now superseded by the 26 October outlook (C53)
- 2026-10-27editedre-struck at the October 26 close
- 2026-10-26editedQ3 release raised revenue to $21.0B and cut free cash flow to $460M
- The trailing figure is our roll-forward of the pack's, and the swap result in selling and administrative expense is not separated, so the multiple is approximate.
- The five competitors its own annual report names traded at a median near 25 times on trailing figures in September; that group was not re-struck after the fall, so the gap to it is not known today.
- 2026-10-27editedre-struck at the October 26 close of $246.65 after a 28.8% fall
- 2026-10-27editedthe September ranking is not re-struck after a 28.8% fall
The company
A cloud company that wants a switch built to its own design cannot buy one off the shelf, and does not want a factory. Celestica buys the chips, boards and power supplies, then assembles, tests and ships the finished system.
More and more it helps write the design too. That design-led work was 41% of sales in 2025, up from 21% in 2023. An older business, 18.9% of sales, builds electronics for aerospace, defence and industry.
The evidence under this - 3 claims
How it makes money
Celestica is paid for each system it builds. On a plain assembly job it earns a fee for the work; on a design it helped write, the price also carries its own design content. Margins are thin either way: the cloud business kept 8.7 cents of each sales dollar as profit in the June quarter, the highest in its history.
The evidence under this - 2 claims
- The chief executive's own description is that programmes are commonly dual-sourced and get offloaded to a second source as they mature, which is a moat that lasts one programme generation rather than one that holds a customer.
- Sanmina's own annual report says profitable revenue growth in this industry is often constrained by intense competition, which is a rival describing the same structure without a moat in it.
- Celestica's own annual report calls the cloud segment landscape very competitive, with aggressive bidding from contract manufacturers and increased competition from design manufacturers.
Where the money comes from, stream by stream
| Stream | Share of sales (Q2 2026) | How it earns it |
|---|---|---|
| Communications (CCS) | 56% | Network switches, routers and optical gear built, and increasingly co-designed, for hyperscalers and networking companies, paid per unit shipped. |
| Enterprise (CCS) | 25% | Servers, AI compute racks and storage systems built for cloud providers and computer makers, the fastest-growing line in 2026. |
| Advanced Technology Solutions | 19% | Contract manufacturing for aerospace and defence, industrial, health-technology and semiconductor-equipment customers, on longer, lower-volume programmes. |
Who runs it
Robert Mionis has run Celestica since 2015 and turned a low-margin assembler into a designer of cloud hardware. The finance chief of nine years moved to a sales role on 1 October; Todd Ankenmann, a finance executive there since 2018, replaced him. Insiders own about 1% of the company. The bonus pays on sales, profit margin and cash; most shares pay on earnings per share over three years, none on return on the money spent on new plants.
The evidence under this - 4 claims
The percentages predate the August offering of 11,129,031 shares (C13), which lowers every holder's share by about a tenth; and the chief executive has sold since under a plan (C38).
- Nothing pays on return on capital while plant spending runs to about 6% of revenue (C09), and free cash flow carries a fifth of the bonus only.
- The revenue component cannot pay above target unless target non-GAAP operating margin is met, so growth bought at a lower margin is not rewarded on its own.
- 2026-10-04editedthe interval was wrong: 1 October to the 27 October investor day is under four weeks, not six (six weeks ran from the 11 September announcement); the 8-K of 11 September 2026 (S20) confirms the 1 October date and is added as the primary source
- 2026-10-27editedthe investor day has come; the claim is a historical record now
Why it wins, and who could beat it
Why customers come back
Because Celestica co-designs the switch and has already ordered the parts, often more than a year ahead, a customer that moves a programme restarts two clocks: approving a new supplier's design, and rejoining a year-long queue for components. Management says customers now ask it to take back work they had given to second suppliers.
The limit is that this protects a programme already running, not the next one, which is bid again.
The evidence under this - 2 claims
- The chief executive's own description is that programmes are commonly dual-sourced and get offloaded to a second source as they mature, which is a moat that lasts one programme generation rather than one that holds a customer.
- Sanmina's own annual report says profitable revenue growth in this industry is often constrained by intense competition, which is a rival describing the same structure without a moat in it.
- Celestica's own annual report calls the cloud segment landscape very competitive, with aggressive bidding from contract manufacturers and increased competition from design manufacturers.
- Nothing outside the company measures this. Neither share nor design wins are disclosed by Celestica or by any of its five filed competitors, so the claim cannot be checked against a number.
- The same answer concedes that products are commonly dual-sourced at award and offloaded as they mature, which is share moving in both directions rather than a one-way gain.
The threat
The rivals it loses work to are Taiwanese builders such as Hon Hai and Quanta, often already approved as second supplier on the same programme. Jabil has built a whole division around the same buyers. And NVIDIA sells its own network inside the rack, so a buyer of NVIDIA systems needs no separate switch.
The evidence under this - 4 claims
- The chief executive's own description is that programmes are commonly dual-sourced and get offloaded to a second source as they mature, which is a moat that lasts one programme generation rather than one that holds a customer.
- Sanmina's own annual report says profitable revenue growth in this industry is often constrained by intense competition, which is a rival describing the same structure without a moat in it.
- Celestica's own annual report calls the cloud segment landscape very competitive, with aggressive bidding from contract manufacturers and increased competition from design manufacturers.
The rivals: NVDA, JBL, SANM, BHE, PLXS, FLEX, and 2 more that are private, in-house or listed abroad.
Who else does this, rival by rival
| Ticker | Rival | How far along | Priced against it |
|---|---|---|---|
| NVDA | Nvidia's NVLink and InfiniBand rack-scale interconnectsPriced against it: One segment only | On sale | One segment only |
| Hon Hai, Quanta, Wiwynn and AcctonPriced against it: Listed abroad | On sale | Listed abroad | |
| JBL | Jabil's Intelligent Infrastructure segmentPriced against it: Yes | On sale | Yes |
| In-house | The operator's own hardware teamPriced against it: In-house | On sale | In-house |
| SANM | SanminaPriced against it: Yes | On sale | Yes |
| BHE | Benchmark ElectronicsPriced against it: Yes | On sale | Yes |
| PLXS | PlexusPriced against it: Yes | On sale | Yes |
| FLEX | FlexPriced against it: Yes | On sale | Yes |
How big it could get
How big this could get
Four of its customers, Amazon, Microsoft, Alphabet and Meta, spent $407.8B on capital in their last filed years, and Celestica's guided $20.5B is about 5% of that. But much of that spending is land, buildings, power and chips, which Celestica does not sell, and no buyer publishes the split. The market it actually bids for is unmeasured.
The evidence under this - 3 claims
- The denominator is not an addressable market. Much of the four buyers' spending is land, buildings, power and the accelerator silicon, none of which Celestica sells, so the true ratio against what it can bid for is materially higher than 5%.
- The numerator is the whole company, including the 18.9% that sells to aerospace, defence, industrial and semiconductor-equipment customers and touches these four buyers not at all.
- The two sides are not the same period: Microsoft's figure is a year ended June 2026 and Celestica's is guidance for calendar 2026, so the ratio is indicative rather than struck.
- Reaching it needs fourth-quarter revenue of about $6.35B, 73.9% above Q4 2025 and 17.7% above the midpoint of the third-quarter guidance.
- The company has raised this outlook three times in 2026 and has beaten its own adjusted earnings guidance in each of the last eight quarters, so the record runs the other way.
- 2026-10-27editedthe September guide is now superseded by the 26 October outlook (C53)
The results
The numbers
Sales grew 62% over the twelve months to Sep 2026, faster than the 28% in the year to Dec 2025.
Operating margin is 8.3%, up from 4.2% in the year to Dec 2023.
Running costs take 3¢ of each sales dollar, down from 5¢ in the year to Dec 2022.
Each dollar put into the business earns 26% a year after tax, up from 13% in the year to Dec 2022.
It turned 3.0% of sales into free cash, $545M over the last twelve months.
It holds $3.96B of cash against $819M of debt, $3.14B more cash than it owes.
The share count fell 1.5% in 3.8 years, and pay in shares costs 0.5% of sales.
All the numbers
Filed figures from the company's 10-K and 10-Q, read 2026-10-26. Fiscal years end in December.
| Dec 2022 | Dec 2023 | Dec 2024 | Dec 2025 | 12 mo to Sep 2026 | |
|---|---|---|---|---|---|
| Size and growth | |||||
| Sales | $7.25B | $7.96B | $9.65B | $12.4B | $18.2B |
| Sales growth | - | +9.8% | +21% | +28% | +62% |
| Profit | |||||
| Gross margin | 9.0% | 9.5% | 11% | 12% | 11% |
| Operating margin | 4.0% | 4.2% | 6.2% | 8.4% | 8.3% |
| Net margin | 2.5% | 3.1% | 4.4% | 6.7% | 6.8% |
| What growth costs, per sales dollar | |||||
| Research and development | 0.6% | 0.8% | 0.8% | 1.0% | 0.9% |
| Selling, general and admin | 3.7% | 3.8% | 3.0% | 2.1% | 2.0% |
| Pay in shares | 0.7% | 0.7% | 0.6% | 0.6% | 0.5% |
| Capital spending | 1.5% | 1.6% | 1.8% | 1.6% | 4.7% |
| Cash | |||||
| Cash from operations | $211M | $326M | $474M | $660M | $1.41B |
| Free cash flow | $102M | $201M | $303M | $458M | $545M |
| Balance sheet | |||||
| Cash and short-term investments | - | $370M | $423M | $596M | $3.96B |
| Debt | - | $675M | $797M | $776M | $819M |
| Net cash | - | −$305M | −$373M | −$181M | $3.14B |
| Returns and shares | |||||
| Return on capital | 13% | 13% | 19% | 31% | 26% |
| Shares (diluted) | 124M | 120M | 119M | 116M | 122M |
What growth costs
Growth has made Celestica more profitable: overhead and share pay barely rose while sales soared, so margins widened. Part of that comes from the share-price bet booked as lower overhead. The real cost of growth is now new factories, which take cash rather than margin, and that bill is rising faster than sales.
The evidence under this - 3 claims
- Part of the fall in selling, general and administrative cost is the share-price swap: its gains are booked in that line, $64.8M in Q3 2025 and $33.7M in Q4 2025 (S01), so the overhead leverage is overstated while the shares rise (C04, C05).
- Capital spending is guided to about 6% of 2026 revenue (C09), so the cash cost of growth is rising even as the operating costs fall.
- 2026-10-26editedthe first 2027 capital spending guide arrived, above the $1.5B previously indicated at its top
The last report (27 Jul)
Quarter ended June 2026, filed figures.
The September quarter beat the top of the company's own forecast: sales of $5.83B, up 83%, and adjusted profit per share of $3.30. It raised the 2026 forecast to $21.0B of sales and gave a first 2027 forecast of $40.0B, up 90%, with a profit margin of 9.0%.
The cash side moved the other way: free cash for 2026 was cut from $600M to $460M and 2027 plant spending set at $1.5B to $2.0B. On the call, managers called 2027 a year of investment and expected free cash to turn positive in 2028.
What it did to the thesisThe growth doubt eased: sales and margin beat and the 2027 plan is far larger. The cash doubt did not: the call gave no 2027 cash figure and called 2027 a year of investment, so the view stays Unattractive with low conviction.
The quarter in detail
Sales growth from the previous quarter, filed figures.
Earnings per share $2.54 against $2.29 expected; $3.17 as filed, the gap being what the company adds back.
The evidence under this - 5 claims
The 2027 outlook rests on customer demand forecasts, not signed orders; three customers are about two thirds of revenue.
- 2026-10-26editedQ3 release raised revenue to $21.0B and cut free cash flow to $460M
- 2026-10-26editedthe first 2027 capital spending guide arrived, above the $1.5B previously indicated at its top
Management gave no figure for 2027 free cash flow or working capital, and 2026 free cash flow was cut to $460M (C53); the 2028 target is a statement of confidence, not a number.
What management did with the money
On 27 July its filing said operations would pay for this year's plant spending. Nine days later it sold $3.45B of new shares, 16.5% below the market price, adding 9.7% to the share count. The earnings forecast of $11.30 a share was never restated for them.
The evidence under this - 4 claims
The price, and where we disagree
In short. Full. The growth is real, but the price pays for next year's faster growth too, on a profit flattered by share-price gains and growth partly paid for by shareholders; it is cheaper than in September.
The shares have fallen 28.8% since our review began, and the 27 October call and investor day have now come. Management gave no 2027 free cash figure and called 2027 a year of investment, with positive free cash expected in 2028.
The next test is the fourth-quarter report in late January, which should bring the 2027 guidance and any free cash figure for next year.
The evidence under this - 3 claims
Management gave no figure for 2027 free cash flow or working capital, and 2026 free cash flow was cut to $460M (C53); the 2028 target is a statement of confidence, not a number.
- 2026-10-04editedthe interval was wrong: 1 October to the 27 October investor day is under four weeks, not six (six weeks ran from the 11 September announcement); the 8-K of 11 September 2026 (S20) confirms the 1 October date and is added as the primary source
- 2026-10-27editedthe investor day has come; the claim is a historical record now
How we read the multiple
Counting debt and new cash, the whole company cost about 32 times last year's operating profit in September, against about 25 for its rivals; those rivals were not re-checked after the fall.
Remove gains from a swap tied to its own share price and the filed profit makes it about 21.6 times at the 26 October close, down from 36.6. On this year's raised forecast, about 15.9. The price still asks for this year's plan and next year's faster growth, none of it contracted, but the premium is smaller.
Against its own history. Many times dearer for each dollar of sales than its own six-year normal. The business changed, but the price moved much further.
The evidence under this - 4 claims
- The trailing figure is our roll-forward of the pack's, and the swap result in selling and administrative expense is not separated, so the multiple is approximate.
- The five competitors its own annual report names traded at a median near 25 times on trailing figures in September; that group was not re-struck after the fall, so the gap to it is not known today.
- 2026-10-27editedre-struck at the October 26 close of $246.65 after a 28.8% fall
- 2026-10-27editedthe September ranking is not re-struck after a 28.8% fall
- No contract, order book or backlog figure supports it: the FY2025 10-K does not use the word backlog once in 611,829 characters.
- The chief financial officer of nine years, who framed these numbers, moves to a different role on 1 October, three weeks before the detailed 2027 outlook is due.
- 2026-10-26editedthe 2027 outlook was published on 26 October 2026
- Management says customers have placed non-cancellable, non-refundable orders for long-lead silicon and that some forecasts are binding 12 to 18 months out, which is a real commitment even though no document states its size.
- Component lead times above 52 weeks mean an operator that has committed the silicon has largely committed the programme, since restarting elsewhere costs a year.
- On the Q3 call management said it is awarded programmes whose silicon is not yet developed and that four of ten announced 1.6T wins are scale-up, but still gave no order or backlog figure.
- 2026-10-27editedthe call described the programmes again, still without a contract
Priced against its rivals
The competitors its own annual report names, each confirmed against the rival's filing. Measured on company value for each $1 of operating profit, cheapest first:
| Ticker | Company | company value for each $1 of operating profit |
|---|---|---|
| SANM | Sanmina | 18.7× |
| BHE | Benchmark Electronics | 23.2× |
| JBL | Jabil | 24.7× |
| PLXS | Plexus | 29.6× |
| FLEX | Flex | 30.5× |
| CLS | Celestica | 31.6× |
The middle of the other 5 is 24.7. Celestica Inc. is dearest of 6.
Rivals' figures come from a data provider, so read the spread of the group rather than any single pair.
The dearest of the companies its own annual report named in September, for each dollar of operating profit. The shares have fallen since, and that group was not re-checked.
Private or foreign rivals that cannot be priced: Hon Hai Precision Industry Co., Ltd., Quanta Computer Inc., Wiwynn Corporation, Accton Technology Corp.
Where we disagree with the market
An AI infrastructure winner whose faster growth and rising margins in 2027 justify paying far more than for a contract manufacturer.
The growth is real. The profit it is priced on is flattered by a share-price swap, and the growth now costs far more capital.
Most analysts rate the shares a buy, and their 2027 sales figure sits 75% above this year's forecast. What troubles us sits where headlines do not reach: the swap in a note to the accounts, and the cash the growth swallows.
We would be wrong if free cash turns positive on 2027 sales with plant spending at or below $2.0B, the opposite of what management said on the 27 October call.
What analysts say, and the evidence
The evidence under this - 5 claims
- 2026-10-26editedQ3 release raised revenue to $21.0B and cut free cash flow to $460M
- 2026-10-26editedthe first 2027 capital spending guide arrived, above the $1.5B previously indicated at its top
What could go wrong, and how we would know
The risks
- Three customers, 63% of sales. The top ten went from 64% of sales to 79% in two years, and the two new programmes deepen that rather than spread it.
- Nothing is contracted. The OpenAI and AMD programmes are described as multibillion-dollar 2027 opportunities, with no volume, term or signed agreement behind them.
- Growth that eats capital. Plant spending is about 6% of sales this year, against 1.5% to 2% historically, and was paid for with new shares.
- One country makes most of it. Thailand made 59% of 2025 sales, up from 46% in 2023.
The evidence under this - 5 claims
- Management says customers have placed non-cancellable, non-refundable orders for long-lead silicon and that some forecasts are binding 12 to 18 months out, which is a real commitment even though no document states its size.
- Component lead times above 52 weeks mean an operator that has committed the silicon has largely committed the programme, since restarting elsewhere costs a year.
- On the Q3 call management said it is awarded programmes whose silicon is not yet developed and that four of ten announced 1.6T wins are scale-up, but still gave no order or backlog figure.
- 2026-10-27editedthe call described the programmes again, still without a contract
- 2026-10-26editedthe first 2027 capital spending guide arrived, above the $1.5B previously indicated at its top
What we still do not know
- How much of the half-year's cash came from selling unpaid customer bills early: $879M went that way, against about $10M a year earlier.
- Whether the doubled bills owed to suppliers, which paid for this year's inventory build, stay or have to be paid down.
- What 2027 plant spending will be. About $1.5B is indicated, and the finance chief who would set it is leaving.
The evidence under this - 5 claims
- 2026-10-26editedthe first 2027 capital spending guide arrived, above the $1.5B previously indicated at its top
- 2026-10-04editedthe interval was wrong: 1 October to the 27 October investor day is under four weeks, not six (six weeks ran from the 11 September announcement); the 8-K of 11 September 2026 (S20) confirms the 1 October date and is added as the primary source
- 2026-10-27editedthe investor day has come; the claim is a historical record now
What would change our view
- Unattractive→AttractiveThe promised margin arrives, at a peer-like multiple
It needs all of:
- ○Adjusted operating margin tops 8.4% in a 2027 quarter, so the design work shows up in profit, not only in the story.checked at the next results, due by 27 Jan
- ○The company's value stays at or below about 25 times its filed operating profit, where it stood on 26 October.checked at the next results, due by 27 Jan
- Reopens the researchThe growth stops being funded by shareholders
It needs:
- ○Once 2027 sales have landed, free cash reaches at least a quarter of adjusted earnings, with no more shares sold.checked at the next results, due by 27 Jan
We would also be wrong if
- Plant spending settles above 6% of sales with no matching rise in operating margin.
- Any of the three biggest customers falls below 10% of sales for two quarters running.
- Celestica sells shares again before 2027 sales reach the guided level.
What is coming
- January 2027 Buyback programme in use after its renewal: Whether a company that sold $3.45B of stock at $310 buys shares back at market after renewing its programme for up to 2.5% of shares.
Every dated event ahead
- January 2027 Buyback programme in use after its renewal: Whether a company that sold $3.45B of stock at $310 buys shares back at market after renewing its programme for up to 2.5% of shares.
- January 2027 Q4 2026 results and the 2027 guidance: Whether $21.0B of 2026 revenue arrived with about $460M of free cash flow, and whether the 2027 growth claim survives.
Since we wrote this
- 27 OctEarningsShares fell 28.8%; the call gave no 2027 cash figureThe price premium behind Unattractive has largely gone and half the route to Attractive is met, but 2027 cash is unstated. View held, conviction cut to Low.
- 26 OctEarningsCelestica guided 2027 sales to $40B but cut its cash forecastSales and margin beat; free cash flow guidance fell and 2027 cash is unstated, so the Unattractive view held.
- 5 OctThesis reviewThe long case now sits under the verdictAfter the five-name pilot, the owner asked for the long case on every thesis.
- 4 OctThesis reviewOwner's checklist: revenue streams, cost of growth, management, catalystsThe owner's checklist found the method silent on who runs the company, what growth costs, where the money comes from and what is dated ahead; this pass fills those four blocks from the proxy and filings.
- 11 SepThesis writtenUnattractive, medium convictionDrafted as Neutral, it moved once the business was scored: a concentrated, uncontracted contract manufacturer, a quarter of whose profit is a share-price gain, priced at $2.80 per dollar of sales against $1.00 for its rivals.
Also checked 2 times with nothing new to act on.
Sources 24
Sources - 24
Every claim above cites one of these, or states the reasoning that derived it. The quote is frozen at the date it was read: a link can rot, and an agent fetch returns a reading of a page rather than the page. Backs counts the whole claim ledger, so it can name a claim this page does not surface.
- S01primarycompany filing2026-02-27segment and end-market split, customer concentration, the total return swap description and its gains, the competition list, manufacturing location mix, and the new capital-spending risk factorbacksC02C03C04C05C09C29C36C43C46C50C52
- S02primarycompany filing2026-07-27balance sheet, cash flow and the working-capital block line by line, the swap notional and strike, segment income reconciliation, customer concentration, receivables sold, and the statement that 2026 capital spending would be funded from operationsbacksC01C02C04C05C06C09C10C11C12C22C34C35C39C45C46C50
- S03primarycompany filing2026-07-27the 2026 annual outlook and its components, Q3 guidance, the free cash flow reconciliation, the diluted share count the guidance was struck on, and the 2027 acceleration statementbacksC06C07C08C16C18C19C20C37
- S04primarytranscript2026-07-28Celestica Q2 2026 earnings call transcriptadjusted gross margin, cash cycle days, net debt and liquidity; the 2027 capital-spending placeholder; the OpenAI and AMD programme detail; management's answers on supply, second-sourcing and customer funding; and eleven analysts' questions
we have a very sizable design engineering organization, which we've been vesting in for over a decade. That organization will be close to 2,000 engineers by the end of this year.
backsC03C06C09C19C25C34C39C41 - S05primarycompany filing2026-01-28the starting point of the 2026 guidance walk - revenue, adjusted earnings per share, operating margin and free cash flow as first framed for the yearbacksC08
- S06primarycompany filing2026-08-06the share count offered, the share count outstanding afterwards, net proceeds, the stated use of proceeds, and the closing price the offer was struck againstbacksC13C14
- S07secondarycompany release2026-08-12confirmation that the over-allotment option was taken up in full, which fixes the final share count
11,129,031 common shares, which includes the full exercise of the underwriters' option of 1,451,612 common shares, were sold by Celestica for aggregate gross proceeds, before underwriting discounts and Offering expenses, of approximately $3.45 billion.
backsC13C14 - S08primarycompany filing2026-08-14the transaction codes behind the aggregate insider figure: open-market sales under a Rule 10b5-1 plan adopted 11 March 2026, from two grantor retained annuity trustsbacksC38
- S09secondaryvendor data2026-09-11Research pack, frozen at publicationprice, the 252-day return and range, the analyst split, short interest, the measured news reactions, the filed-competitor cohort multiples and the company's own six-year multiple mediansbacksC15C20C22C23C37C38C44
- S10primarycompany filing2026-02-06a customer's own capital spending, filed: $128.3B of cash capital expenditures in 2025 against $77.7B in 2024backsC30C31
- S11primarycompany filing2026-07-29additions to property and equipment of $115,948M in the year to June 2026 against $64,551M and $44,477M in the two years beforebacksC30C31
- S12primarycompany filing2026-02-05$91.4B of 2025 capital expenditures and the statement that 2026 investment in servers, network equipment and data centres will increase significantlybacksC30
- S13primarycompany filing2026-01-29$72.22B of 2025 capital expenditures including principal payments on finance leasesbacksC30
- S14primarygovernment data2026-01-13the rate behind the driver, measured by a government agency rather than by anyone selling into it
EIA expects U.S. electricity use to grow by 1% this year and 3% in 2027. The driving factor behind this surge is increasing demand from large computing centers.
backsC32 - S15primarycompany filing2026-02-17the attack described from the other side of the table: customers developing switches for internal use, white-box vendors on open-source operating systems, and Nvidia's InfiniBand and NVLink competing with Ethernet for back-end AI clustersbacksC28
- S16primarycompany filing2025-10-17a named rival's own account of why buyers outsource, its dedicated per-customer business-unit structure, and an entire reportable segment pointed at AI infrastructure, cloud and data centrebacksC26C27
- S17primarycompany filing2025-11-13a rival's own statement that profitable revenue growth in this industry is often constrained by intense competition, and that customer payment terms and receivables factoring swing its quarterly cashread, but no claim rests on it
- S18secondarycompany release2026-09-11the chief financial officer of nine years moving to a new commercial role on 1 October 2026, under four weeks before the investor day that sets the 2027 numbers
Celestica is experiencing unprecedented growth, and as we look to the future, we are evolving our leadership structure to ensure we can scale effectively and continue delivering exceptional value to our customers and shareholders.
backsC42 - S19primarycompany filing2026-04-09beneficial ownership of directors, officers and 5% holders as of 27 March 2026; the executive biographies; the 2025 team incentive measures, weights and payout; the RSU and PSU mix and the PSU vesting measuresbacksC47C48C49
- S20primarycompany filing2026-09-11Mandeep Chawla moving from chief financial officer to Group President, Global Markets and Todd Ankenmann appointed chief financial officer, both effective 1 October 2026, with their backgrounds and new pay termsbacksC42C49
- S21primarycompany filing2026-10-26Q3 results, raised 2026 outlook, first 2027 outlook, cut free cash flow outlook, new buyback intention
We now expect revenue to reach $21.0 billion and adjusted EPS (non-GAAP)* of $11.55
backsC17C19C53 - S22primarycompany filing2026-10-26customer concentration, operating cash flow and plant spending
Three customers (all in our CCS segment) individually represented 10% or more of total revenue in Q3 2026
backsC54 - S23primarytranscript2026-10-27management's account of 2027 cash, plant spending, margins, funding and customers; the 2028 cash target; the renewed buyback
2027 is going to be a year of investment to support that incremental $20 billion of revenue. We did raise $3.5 billion of equity in August, and we feel this should be sufficient to cover our needs.
backsC52C55 - S24primarycompany filing2026-10-27nothing readable: the exhibit carries the slides as imagesread, but no claim rests on it