Filed 14 August 2026; year ended 30 June · Audited statements, notes 8/13–15/17/21–22, liquidity and risks. Annual-report PDF snapshot has equivalent financial statements. Primary financial source.
https://www.sec.gov/Archives/edgar/data/820318/000082031826000020/iivi-20260630.htmWatchlist · cash conversion and valuation discipline
The investment decision
At $221.50, the demanding base returns 16.7% annually. Its conditional 20% entry is $192.78; the separate cash-flow valuation is lower at $127.31. Neither is a guaranteed price.
Edition update. 27 October Sydney initiation. Reference: 26 October US regular close. Actual financials end in June; September product and share disclosures are dated separately. Company guidance and house assumptions are distinguished.
01. A strong photonics business is not automatically a cheap share
Research view: watchlist; wait for evidence of cash conversion or a larger valuation cushion. At the $221.50 reference, our demanding base returns 16.7% annually over five years. A 20% entry under that same 18x exit-multiple case is $192.78. However, a separate cash-flow valuation is only $127.31 per share. The higher entry figure is a conditional scenario price, not intrinsic fair value. [S1] [S9]
The business thesis is credible: faster AI chips need faster connections, and Coherent supplies the lasers, optical components and assemblies that move information between them. Broader integration can increase the content sold into each system. That opportunity is already widely recognised. NVIDIA’s $2bn equity investment and strong datacenter growth are evidence of strategic relevance, not proof that the current stock price offers a margin of safety. [S1] [S3] [S4]
The variant question is how much of that demand becomes profitable, collected cash after factory construction, inventory, employee awards and the capital structure. Fiscal 2026 revenue rose 22.5% and GAAP profit improved sharply, yet operating cash flow fell to $79.5m while cash equipment spending reached $1.103bn. Our base needs the production ramp to reverse that weak conversion rather than merely sustain impressive sales growth. [S1]
This is an absolute-return research assessment. We do not claim a verified mispricing, assign probabilities to the cases, or infer an attractive entry simply from a fall from a recent share-price peak. Improving cash conversion and successful product qualification would strengthen the ownership case; delays, inventory ageing or persistent reliance on external capital would weaken it.
02. Dates, definitions and the evidence boundary
Research date: 27 October 2026, Australia/Sydney. Reference price: USD 221.50 at the 26 October regular-session close, 4 pm EDT, equivalent to 7 am Sydney on 27 October. StockAnalysis and a second public quote provider corroborate that close. After-hours prices are excluded. The latest complete financial statements reviewed are the audited year ended 30 June 2026; September-quarter results are scheduled for 4 November. The newer share count is separately dated 21 September. [S1] [S6] [S8] [S9]
Coherent Corp. is NYSE: COHR, the former II-VI business renamed after its Coherent acquisition. It reports Datacenter & Communications and Industrial segments. Historical comparisons below use the company’s recast segment series. The June quarter means Q4 FY2026, not calendar Q2 as a separate reporting convention. Dollars are US dollars; model tables use billions unless noted. [S1]
The supplied September and October broker PDFs inform the questions and mechanisms, but their dated estimates and price targets are not current company guidance or an independently refreshed consensus. The public model is original Capital Thesis underwriting. Primary filings control historical financials and capital structure; company product announcements support stated plans, not guaranteed commercial outcomes.
03. What Coherent sells, and why more optics can mean more content
An optical link converts an electrical signal into light, carries that signal through fibre and turns it back into usable electrical information. Coherent participates across that chain through semiconductor materials, lasers, silicon photonics, precision optics, specialty fibre, detectors, transceivers, assembly and testing. It also sells industrial lasers, systems, optics and materials into manufacturing, semiconductor equipment, research and other markets. Industrial contributed $1.844bn, or about 26% of FY2026 sales; the company is not solely a transceiver supplier. [S1] [S4]
The economics depend on both demand and the content Coherent supplies per connection. More bandwidth, longer reach and power constraints can favour optical links over electrical alternatives. Supplying several qualified components or a finished assembly can increase revenue per system and reduce the customer’s integration burden. Revenue still depends on design selection, actual units shipped and pricing; a larger theoretical market does not flow directly into the income statement. [S1] [S4]
PhotonLink, introduced in September, combines technologies for co-packaged optics, near-packaged optics and emerging chip-to-chip connectivity. The company disclosed more than ten CPO engagements, more than ten NPO engagements and more than five chip-to-chip engagements, with anchor customers and long-term agreements for CPO and NPO. Management expected platform revenue to start ramping in calendar Q4 2026. Engagements are not recognised revenue, and the release does not provide enough contract economics to independently value each programme. [S4]
Architecture flexibility is useful but not costless. CPO can improve power and performance by bringing optics close to silicon; serviceability and packaging complexity can favour different layouts. Customers may buy selected components rather than the entire platform. Our model credits overall segment growth, not the sum of overlapping market-size headlines, each engagement as a separate sale, or every possible optical architecture at full penetration.
04. The manufacturing advantage must survive the ramp
Vertical integration gives Coherent control over materials, devices, assembly and testing. Scale, proprietary processes and customer qualification can create barriers to entry. The advantage is economic only if manufacturing yield, delivery reliability and unit cost remain competitive. Additional production space is not equivalent to qualified output: tools must be installed, staff trained and usable devices produced at sufficient yield. [S1] [S4]
The June Sherman announcement describes an expansion intended to double production space and quadruple wafer capacity at the six-inch indium-phosphide facility. The associated CHIPS support was a letter of intent for up to $50m, not a booked grant or unconditional cash receipt. Larger wafers and better utilisation can reduce unit cost, but we do not insert the full planned capacity increase into revenue automatically. [S7]
The supplied September conference note attributes stronger growth ambitions, price and mix improvement, a 24-month InP expansion plan and an 18-month payback target to management discussions. These are useful directional evidence, not formal full-year guidance or audited project returns. The base assumes meaningful operating leverage, while the bear explicitly models slower growth and heavier working capital. An apparent shortage today can later become excess capacity if customers revise deployment plans.
NVIDIA’s investment is separate from its commercial relationship. The March agreement combined an equity purchase with a non-exclusive strategic arrangement and multi-year purchasing/capacity provisions. Exact future volumes and unit margins are not publicly sufficient to reconstruct a guaranteed revenue stream. Coherent’s 10-K says the arrangement may require material equipment, labour and working-capital investment through 2030, and may affect concentration, margins and capex. [S1] [S3]
05. Growth and margins improved; GAAP and adjusted profit tell different stories
| Reported metric | FY2024 | FY2025 | FY2026 |
|---|---|---|---|
| Revenue, $bn | 4.708 | 5.810 | 7.118 |
| Datacenter & Communications, $bn | 2.631 | 3.755 | 5.275 |
| Industrial, $bn | 2.076 | 2.055 | 1.844 |
| Operating cash flow, $m | 545.7 | 633.6 | 79.5 |
| Cash PP&E purchases, $m | 346.8 | 440.8 | 1,102.9 |
Datacenter & Communications sales grew 40.5% in FY2026. Industrial sales declined 10.3%, materially affected by portfolio divestitures rather than representing a clean constant-portfolio demand comparison. In the June quarter, total revenue rose 33.8% to $2.046bn; GAAP gross margin improved from 35.7% to 38.5%. Management attributes annual margin improvement to volume leverage, manufacturing yields, pricing and other cost/mix effects. The revenue and margin evidence supports operating progress, but not every improvement is recurring or independent of the cycle. [S1] [S2]
FY2026 consolidated net income was $786.884m; parent-attributable income was $804.998m because $18.114m of subsidiary losses belonged to minority owners. Income available to common shareholders was $769.896m after preferred dividends and related accretion. Reported diluted EPS was $4.12; adjusted EPS was $5.61. These are distinct numerator and denominator conventions, not interchangeable measures of cash earned by today’s common shares. [S1] [S2]
GAAP profit benefited from a $124.133m business-sale gain and a low 7.2% effective tax rate; the cash-flow reconciliation also removes a $73.998m equity-investment gain. Acquisition amortisation, impairments and restructuring move the other way. The scenario model excludes acquired-intangible amortisation and exceptional gains/charges but retains a 3%-of-sales employee-award expense and normalises tax at 21%. Its economic operating profit is neither a GAAP forecast nor a reproduction of issuer non-GAAP earnings. [S1]
Formal Q1 FY2027 guidance is $2.2–2.4bn revenue, 39.5–41.5% adjusted gross margin, $400–420m adjusted operating expense and $1.85–2.05 adjusted EPS. There is no full five-year company forecast underlying our cases. Dividing an old broker target by current adjusted EPS would hide the ramp, investment and dilution assumptions that matter. [S2]
06. Why profit did not become cash
| FY2026 cash evidence | USD million | What it means |
|---|---|---|
| Consolidated net income | 786.884 | Accounting earnings before the cash reconciliation |
| Receivables cash use | (367.631) | Revenue collection lag absorbed cash |
| Inventory cash use | (1,183.000) | More materials and products tied up capital |
| Payables cash source | 748.533 | Supplier credit partly offset the build |
| Other operating net-asset cash use | (408.333) | Further operating balance-sheet absorption |
| Operating cash flow | 79.514 | After all noncash and working-capital adjustments |
| Cash PP&E purchases | (1,102.909) | Factories and equipment |
| CFO less cash PP&E | (1,023.395) | Analytical FCF before $7.174m intangible purchases |
The cash problem is not that every inventory dollar is lost. Some stock is necessary to fulfil a large production ramp, and receivables may subsequently be collected. The investment question is whether those balances turn into collected sales at the expected margin. If demand slows or technology changes, inventory can require discounts or write-downs; if supplier payment terms tighten, financing needs rise. The listed cash movements are selected explanations, not a complete bridge that should be summed to net income. [S1]
Receivables ended June at $1.343bn and inventory at $2.581bn. Accounts payable were $1.905bn, including noncash equipment additions; treating all of that as permanent free supplier financing would overstate conversion. The model removes $371.779m of equipment payables from its opening operating-working-capital measure and reserves the full amount as FY2027 cash capex. This is deliberately conservative: it does not assume the outstanding equipment bill vanishes. [S1]
Net cash from common-share issuance was approximately $1.999bn, primarily the NVIDIA financing. Business-sale proceeds of $436.992m and equity-investment sale proceeds of $89.384m also supported liquidity. They are financing or asset monetisation, not recurring operating cash. Similarly, the net $825m placed in short-term investments is a liquid-asset allocation rather than an extra $825m factory cost. [S1]
The forward model requires operating working capital to fall from roughly 36% of sales to 28% in base while revenue scales. That is an efficiency assumption, not an observed improvement. Capex remains substantial: FY2027 cash spending is $2.185bn including the opening equipment bill, and FY2031 spending is still $1.604bn. The base therefore does not manufacture free cash flow by immediately reducing investment to zero.
07. Count the cash, debt and shares once
| Capital item | Observed amount / date | Model treatment |
|---|---|---|
| Cash plus short-term investments | $1.987bn, 30 June | Opening liquid cash; includes retained NVIDIA proceeds |
| Restricted cash | $0.606bn, 30 June | Excluded from spendable cash and terminal surplus |
| Debt face / carrying value | $3.257bn / $3.222bn, 30 June | Use face for repayment; issuance-cost discount is not forgiven debt |
| Common shares | 196.931m, 21 September | Newer proxy count; not FY2026 weighted-average shares |
| Series B preferred | Zero outstanding, 30 June | Converted into about 30.1m common in December; already included |
| NVIDIA common issuance | 7.788m shares at $256.80 | Already in current shares and June cash |
| SiC outside ownership | 25% subsidiary minority stake | Terminal minority-equity deduction, not a second parent share issue |
Debt maturities total $7.916m in FY2027, $34.375m in FY2028, $93.629m in FY2029, $2.136bn in FY2030 and $985.937m in FY2031. The apparently back-loaded schedule reflects July 2029 term-loan and December 2029 note maturities falling in FY2030, and the September 2030 term-A maturity falling in FY2031. Base repays these from model cash; debt is not silently rolled beyond its stated maturity. [S1]
The June revolver had $664m available. A foreign subsidiary subsequently arranged approximately $945m of facilities, undrawn at the August filing, with a 36-month tenor after first use. These are borrowing resources, not cash in the bank, and access remains conditional. Bear draws about $291m in FY2030 to maintain a $500m cash floor, with a 2% financing fee and 7.5% interest assumption; successful new lending/refinancing is an assumption. Base and bull need no marketed equity financing in this path. [S1]
No common dividend or buyback is assumed. The former preferred claim is gone, but employee dilution remains: the filing lists unvested units/options, and August special performance awards have a $100m target value with share-price, relative-return and service conditions. Those awards are not immediate guaranteed cash compensation. Rather than adding the older June award pool again to the newer September count, the model uses 2% annual net share growth inclusive of existing and future awards. Economic shares reach 217.428m in all three cases. [S1] [S5] [S6]
The model’s operating profit expenses awards at 3% of revenue. Cash flow adds that noncash expense back and shareholders bear the explicit dilution; retained cash is then included only once in terminal equity. This resembles an earnings-and-cash forecast, not an owner-cash measure that subtracts the same award cost again. Operating lease costs remain inside margins, so lease liabilities are not also treated as extra financial debt.
08. What the base must achieve
| Capital Thesis base | FY2027 | FY2028 | FY2029 | FY2030 | FY2031 |
|---|---|---|---|---|---|
| Revenue, $bn | 10.07 | 12.77 | 15.54 | 18.05 | 20.05 |
| Economic EBIT margin | 21% | 23% | 25% | 26% | 27% |
| Economic EBIT, $bn | 2.12 | 2.94 | 3.89 | 4.69 | 5.41 |
| Cash capex, $bn | 2.19 | 1.92 | 1.87 | 1.62 | 1.60 |
| CFO less capex, $bn | -0.86 | 0.58 | 1.70 | 3.00 | 4.11 |
| End gross debt, $bn | 3.25 | 3.22 | 3.12 | 0.99 | 0.00 |
| End cash, $bn | 1.12 | 1.67 | 3.27 | 4.14 | 7.26 |
Base Datacenter & Communications growth is 55%, 32%, 25%, 18% and 12%; Industrial grows 3% then 4% annually. That produces $20.050bn revenue in FY2031, almost three times FY2026. The rationale is a strong near-term optical ramp, followed by deceleration as the comparison base expands. It still requires broad adoption, competitive content and capacity delivery. The formal Q1 revenue guide supports near-term acceleration but does not validate this full path.
Pre-award operating margin rises from 24% to 30%; subtracting the 3% award expense gives economic EBIT margins of 21–27%. Depreciation is 5% of sales and remains inside operating profit. Cash capex falls from 18% to 8% of sales as the buildout matures, with the opening equipment-payable reserve added in FY2027. Working capital moves from 36% to 28% of sales. These improvements are the heart of the thesis, not mathematical consequences of AI demand.
Normalised taxes are 21% on positive pretax earnings. Base interest is 5.5% of opening gross debt, with 2.5% interest on opening cash above a $500m reserve. Full-year interest on opening debt is conservative when principal is repaid during the year. Model cash remains above the floor and reaches $7.261bn after all face debt is repaid. The $606m restricted cash is never added as a second surplus. Minority owners receive no interim distribution in the model; their terminal operating-equity claim is valued separately at $400m.
Forecast years FY2027–FY2031 approximate the five years from October 2026 to October 2031. FY2031 earnings stand in for the exit run rate; the fiscal-year-end precedes the investment anniversary by about four months. The June starting cash balance is the latest filed balance, not a claimed October bank balance. No unreported quarter is invented to bridge the gap.
09. Five-year shareholder cases after investment and dilution
| House case | FY2031 sales / EBIT margin | Exit EV/EBIT | Net cash / NCI claim | Terminal share value | Annual return |
|---|---|---|---|---|---|
| Bear | $11.66bn / 16% | 12x | $0.74bn / $0.25bn | $105.19 | -13.8% |
| Base | $20.05bn / 27% | 18x | $7.26bn / $0.40bn | $479.71 | +16.7% |
| Bull | $25.62bn / 32% | 24x | $11.64bn / $0.60bn | $955.73 | +34.0% |
Bear five-year total-wealth CAGR: -13.8%. Base five-year total-wealth CAGR: 16.7%. Bull five-year total-wealth CAGR: 34.0%.
Bear combines slower D&C growth, flat-to-low Industrial growth, worsening margin to 16% after awards, heavier working capital and a 12x exit. It still assumes Coherent remains a functioning business and obtains needed debt; capital-market failure or technology displacement could be worse. Base needs the full $20bn revenue and 27% margin path at 18x. Bull requires faster platform adoption, $25.62bn sales, 32% economic margin and a 24x multiple. It is an upside illustration, not an analyst consensus forecast.
Base enterprise value = $5.413bn economic EBIT × 18 = $97.441bn. Add $7.261bn retained cash, subtract zero remaining face debt and a $0.400bn minority claim to obtain $104.302bn common equity. Divide by 217.428m shares for $479.71. Common dividends are zero, so five-year total wealth equals that terminal share value. CAGR = (terminal value / $221.50)^(1/5) − 1.
Base five-year total-wealth CAGR without rerating: 20.5%. This holds the reference economic enterprise value / FY2027 house EBIT multiple of 21.38x constant. Reference EV uses September common shares, June face debt and liquid cash plus the minority book claim, excluding restricted cash. This is a forward house-model benchmark, not a historical GAAP multiple or an observable market consensus. The main base deliberately exits at a lower 18x multiple.
Returns are nominal USD before investor tax, costs and foreign-exchange effects. Cash is retained without distributions and without an acquisition premium. The scenarios use correlated operating and valuation assumptions but no probability weights. None is a promise, and a 20% hurdle is an analytical preference rather than a market forecast.
10. The entry test is conditional; cash-flow value is lower
A 20% annual base-case return over five years requires an entry price of $192.78, holding the operating, financing, dilution and 18x exit-multiple assumptions fixed. The calculation is $479.71 / 1.20^5. Changing the entry changes the purchase price, not the company’s earnings, future funding or terminal share count. It does not establish that the shares are intrinsically worth $192.78 today.
The separate DCF values equity at $127.31 per share using a 12% discount rate and 3% perpetual growth. It starts from the same base operating path, continues capex at 8% of sales, keeps working capital at 28%, and charges the 3% award expense as a cash-equivalent recurring cost. It does not add that award expense back or assume new-share settlement. Current shares are used in this alternative capitalisation. This is a financing-neutral cross-check, not a second method that should produce exactly the same explicit-dilution outcome.
The gap is economically meaningful: an 18x terminal operating-profit multiple can capitalise a growing franchise more generously than a cash-flow model that requires continuing reinvestment. A terminal multiple must therefore be defended, not treated as an accounting identity. We give the lower DCF substantial weight in the watchlist decision and would not describe the $193 scenario entry as a low-risk buy level.
| FY2031 EBIT margin | 14x exit: CAGR / 20% entry | 18x exit: CAGR / 20% entry | 22x exit: CAGR / 20% entry |
|---|---|---|---|
| 23% | +8.2% / $132 | +13.3% / $166 | +17.6% / $200 |
| 27% | +11.4% / $153 | +16.7% / $193 | +21.2% / $233 |
| 31% | +14.3% / $174 | +19.8% / $219 | +24.4% / $265 |
Sensitivity changes only terminal margin and multiple while holding the base cash/debt/share path fixed. It is a valuation illustration, not a fully re-financed operating case; lower margins throughout the period would also reduce accumulated cash. A more attractive real-world setup requires both a sensible price and evidence that the assumptions are becoming more likely.
11. Where the thesis can fail
| Risk | Economic mechanism | Evidence that would change the view |
|---|---|---|
| Capacity and yield | Tools, staffing or qualification delays postpone shipments while fixed costs and cash capex continue. | Slower qualified output, rising unit cost or repeated delivery delays. |
| Architecture and pricing | A different optical design or competitive supplier can reduce content, price or share. | Engagements fail to convert; lower margin despite higher volume. |
| Working capital | Inventory and receivables absorb cash; supplier terms can reverse. | CFO remains weak after the ramp, inventory ageing or collection deterioration. |
| Customer concentration | A large platform customer can change deployment timing and negotiate capacity economics. | Revenue concentration rises without adequate margin and cash protections. |
| Funding and commitments | FY2027 material/supply commitments are $3.4bn, with $8.4bn thereafter; they reduce flexibility. | Liquidity falls, capex rises or new debt terms become materially worse. |
| Export controls | BIS inquiry concerns past Huawei sales; potential penalties and future restrictions. | Adverse regulatory findings or material cash costs. |
| Valuation and awards | The share price needs future growth, while stock compensation increases claims on it. | A lower supportable multiple or dilution above the model’s 2% annual pace. |
The purchase commitments are primarily materials and supplies. They are included conceptually in operating cost and working capital, not added again as $11.8bn of factory capex. Early-termination provisions and timing still create downside if demand weakens. The model does not assume customer prepayments will cover every commitment or that undrawn credit will always be available. [S1]
The 10-K says Coherent stopped shipments to Huawei and is cooperating with BIS while conducting an internal review. No determination or reasonably estimable loss range had been disclosed in the reviewed filing. This is an unresolved regulatory matter, not a concluded violation, a quantified fine, or a clean bill of health. We also do not treat the supplied broker podcast’s discussion of possible future optical restrictions as enacted law. [S1]
Governance deserves monitoring alongside technology. The August leadership award programme can reach 200% of target only if specified share-price, relative-performance and service requirements are satisfied. Its alignment does not remove its dilution cost. Semiconductor and optical exposure also means the stock is sensitive to industry spending and valuation cycles; no portfolio sizing recommendation is made without the investor’s broader context. [S5]
12. The next evidence should be operational and financial
| Window | What to monitor | Why it matters |
|---|---|---|
| 4 November 2026 results | September revenue against $2.2–2.4bn guide, gross margin and FY2027 outlook. | Tests whether demand and supply are translating into reported sales. |
| Calendar Q4 2026 | PhotonLink’s stated initial revenue ramp and qualification progress. | Separates engagement counts from delivered commercial output. |
| Next two to four quarters | Inventory, receivables, equipment spending and supplier payments. | Tests the transition from investment-heavy growth to cash generation. |
| FY2029–FY2031 maturities | Debt repayment progress, usable liquidity and covenant capacity. | Ensures per-share returns survive financing obligations. |
The earnings release is scheduled after the New York close on 4 November, followed by a 4:30 pm EST webcast. Because US daylight saving ends beforehand, that webcast is 8:30 am AEDT on 5 November in Sydney. This time conversion is based on the dated schedule, not the October EDT offset. [S8]
Upgrade evidence would include growing qualified shipments, durable margins after pricing normalises, cash conversion catching up with earnings and funding needs staying within internally generated resources. Re-underwrite if margin gains disappear, working capital keeps expanding faster than sales, material regulatory costs emerge, or required dilution exceeds the model. An attractive industry narrative alone does not resolve the cash-flow and entry-price tests.
Source limitations: no September-quarter financial statements are yet available; current bank cash, future customer volumes and precise PhotonLink programme economics are not independently disclosed. Supplied broker forecasts are dated opinions. Our model is transparent about the gaps rather than filling them with invented consensus or treating an ambitious capacity plan as guaranteed sales.
Financial exhibits
Exact observations accompany every chart. Actuals, management guidance and Capital Thesis assumptions are labelled separately; the chart shapes do not imply probabilities.
Datacenter growth drives the portfolio change
Industrial remains meaningful; historical mix is not exclusively AI.
| Observation | Datacenter & Communications | Industrial |
|---|---|---|
| FY2024 | 2.6314 | 2.0763 |
| FY2025 | 3.7552 | 2.055 |
| FY2026 | 5.2746 | 1.8436 |
Recast segment history; Industrial includes divestiture effects. Series sum to company sales. Sources: [S1]
The June quarter delivered both volume and margin progress
Reported GAAP gross margin rose to 38.5%.
| Observation | GAAP gross margin |
|---|---|
| Q4 FY2025 | 35.7 |
| Q3 FY2026 | 37.7 |
| Q4 FY2026 | 38.5 |
Q4 FY2026 sales were $2.046bn, +33.8% year-on-year; gross-margin percentages are not cash-flow margins. Sources: [S2]
Factory investment overtook operating cash
FY2026 CFO less cash PP&E was negative $1.023bn.
| Observation | Operating cash flow | Cash PP&E purchases |
|---|---|---|
| FY2024 | 545.731 | 346.816 |
| FY2025 | 633.6 | 440.836 |
| FY2026 | 79.514 | 1,102.909 |
Capex shown as positive spending for comparison. Analytical FCF subtracts spending; intangible purchases are additional. Sources: [S1]
Inventory and receivables absorbed cash before equipment spending
Supplier credit provided only a partial offset.
| Observation | Cash source / use |
|---|---|
| Receivables | -367.631 |
| Inventory | -1,183 |
| Payables | 748.533 |
| Other operating net assets | -408.333 |
Selected cash-flow movements, not balance-sheet amounts and not a complete CFO reconciliation. Sources: [S1]
Debt principal must be funded before the five-year exit
The filed maturity wall is concentrated in FY2030–FY2031.
| Observation | Required principal |
|---|---|
| FY2027 | 7.916 |
| FY2028 | 34.375 |
| FY2029 | 93.629 |
| FY2030 | 2,135.625 |
| FY2031 | 985.937 |
Face debt totals $3,257.482m. Carrying value is lower because of unamortised issuance costs. Base repays scheduled principal. Sources: [S1]
The base requires cash conversion to improve
Growth alone is insufficient; capital intensity and working capital must ease.
| Observation | Operating cash flow | Cash capex |
|---|---|---|
| FY2027 house | 1.3231 | 2.1852 |
| FY2028 house | 2.4982 | 1.915 |
| FY2029 house | 3.5634 | 1.8652 |
| FY2030 house | 4.6295 | 1.6249 |
| FY2031 house | 5.7119 | 1.604 |
House assumptions, not guidance. CFO includes noncash-award addback; 2% annual share dilution reflects award settlement. FY2027 capex includes opening equipment payables. Sources: [S1] [S2]
A successful base still falls short of a 20% hurdle at $221.50
Price, execution, investment and the exit multiple jointly determine returns.
| Observation | Annual shareholder return |
|---|---|
| Bear house | -13.8365 |
| Base house | 16.7135 |
| Bull house | 33.9652 |
No common distributions. Economic EBIT after award expense, explicit dilution, net cash and minority claims. These cases are not probability-weighted forecasts. Sources: [S1] [S6] [S9]
Source register and research basis
Evidence checked for this edition on 27 October 2026. Dates below identify source vintages. Company guidance and house scenarios are conditional; source links and the downloadable model allow the assumptions to be inspected.
12 August 2026 · Quarterly reported growth, GAAP/non-GAAP reconciliation and formal next-quarter guidance. Financial tables read alongside audited annual filing.
https://www.coherent.com/content/dam/coherent/site/en/documents/investors/financial-releases/2026/august-12/earnings-release-fy26-q4.pdf2 March 2026 · 7,788,161 common shares at $256.80; $2bn equity transaction distinct from the non-exclusive strategic purchase/capacity arrangement.
https://www.sec.gov/Archives/edgar/data/820318/000119312526084366/d42735d8k.htm21 September 2026 · Generate/Shape/Guide/Receive platform; over ten CPO, over ten NPO and over five chip-to-chip engagements; management expected calendar-Q4 revenue ramp.
https://ir.coherent.com/news-releases/news-release-details/coherent-launches-photonlinktm-integrated-optics-platform-ai31 August 2026; awards dated 27 August · $100m combined target value; absolute and relative return thresholds, service and holding requirements. Not guaranteed vesting.
https://www.sec.gov/Archives/edgar/data/820318/000119312526375462/d110649d8k.htm9 October 2026 · 196,931,166 common shares on 21 September record date. Do not substitute the slightly different 31 August ownership-table denominator.
https://www.sec.gov/Archives/edgar/data/820318/000110465926114847/tm264288-6_def14a.htm16 June 2026 · Planned double production space and quadruple wafer capacity. Up to $50m letter of intent, not unconditional awarded cash.
https://www.coherent.com/news/press-releases/a-chip-letter-of-intent-for-50m-to-expand-world-leading-manufacturing-facility-for-ai-infrastructure14 October 2026 · 4 November after-close results, 4:30 pm Eastern webcast; 5 November 8:30 am Sydney after US DST change.
https://ir.coherent.com/node/18371/pdf26 October 2026, 4 pm EDT regular close · $221.50; independently corroborated through FinancialContent. After-hours quote excluded. Price-provider share counts are not used in preference to the proxy.
https://stockanalysis.com/stocks/cohr/history/12 August 2026 · Segment, quarterly operating and guidance context. All forecast scenarios remain Capital Thesis assumptions.
https://www.coherent.com/content/dam/coherent/site/en/documents/investors/investor-presentations/2026/august-12/investor-presentation-20260812.pdf