HOLD at the dated $4.72 quote. Base present fair value is $5.03 per share, with a conditional $1.59–$11.48 range. Cash funds the opportunity, but the quote still requires a large manufacturing ramp. The base December 2030 annualized return is 11.8%, below our 15% hurdle.
| Metric | Value |
|---|---|
| Bear | 1.588 |
| Base | 5.026 |
| Bull | 11.478 |
| Scenario | Probability | Present value | vs $4.72 | What has to happen | The arithmetic |
|---|---|---|---|---|---|
| Bear | 30% | $1.59 | −66% | Four-year failed ramp then recovery | 2030 units 0.45m; finite recovery, no perpetuity. |
| Base | 50% | $5.03 | +6% | Commercial scaling with pricing erosion | 2036 units 9.5m at $76; product gross margin 46%. |
| Bull | 20% | $11.48 | +143% | Multiple wins and further capacity | 2036 units 25m at $90; product gross margin 53%. |
| Subjective 30/50/20 weights; not statistical probabilities or a downside floor. | |||||
| Signpost | Now (Q3 2026) | Green if | Red if | Next check |
|---|---|---|---|---|
| Results update | $1.026m Q3 revenue | Accepted shipments and product economics | Growth without conversion | March 1, 2027 |
| Backlog | 30,000-engine component | Ship current backlog | Another deadline slip | Q2 2027 |
| Manufacturing | >$45m program | Usable capacity and yield | Equipment without qualification | Mid-2027 |
| Assembly | About 1m annual base capacity | Roughly tenfold planned line | Utilization fails to follow capacity | End 2027 |
| New products | 1.6T / Blazar programs | Commercial production and repeat orders | Delayed qualifications | 2027 / early 2028 |
Management milestones are distinct from author forecasts. Watch volume, margins, investment and ownership together.
HOLD at the $4.72 reference quote, recorded November 16, 2026 at 3:15:43 p.m. Eastern. Our base present fair value is $5.03 per common share, with conditional bear and bull values of $1.59 and $11.48. The base is the result of one commercialization model; it is not an average of methods or analyst targets. The subjective probability-weighted value is $5.29. Neither the range nor the weighted value is a statistical confidence interval. Prices can fall below the modeled bear outcome.
POET has cleared a modest but meaningful revenue milestone: its first quarter above one million dollars. It also has unusually large liquid resources for a business at this commercial stage. Those facts improve its ability to pursue production, but they do not demonstrate that customers will buy enough engines at profitable prices. The decisive question is whether qualification, yield and supply can turn orders into a repeatable manufacturing business before spending and ownership claims erode the opportunity.
At $4.72, the gap to base present value is only 6.5%. A twenty percent purchase cushion below $5.03 points to roughly $4.02; this is an entry discipline, not a second haircut to fair value. Our separately modeled December 2030 base price is $7.48, equivalent to an 11.8% annualized return without distributions. That falls below our fifteen percent purchase hurdle. The business can therefore have promising technology and strong funding while the stock remains a hold at the dated quote.
This judgment could change through better operating evidence or a lower price. We do not need to dismiss the bull case to require a margin of safety. Equally, a large cash balance should not force a favorable rating when the future product economics, customer concentration and eventual ownership remain uncertain. The quote is an intraday observation, not the November 16 closing price.
| Metric | Value |
|---|---|
| Reference quote | 4.720 |
| Base value today | 5.026 |
| 20% entry discipline | 4.021 |
Third-quarter revenue was $1,026,087, versus $569,925 in the second quarter and $298,434 a year earlier. Those figures imply approximately 80% sequential growth and 244% year-over-year growth. This was the seventh consecutive sequential increase. The eight-quarter history shows an improving trajectory, but also how small the starting point remains. A large percentage increase from a small base is not evidence that the eventual addressable market has already been captured.
The revenue includes nonrecurring engineering work and product sales. The release does not provide a sufficiently detailed split or unit shipment reconciliation to calculate a reliable realized product selling price. Our forecast therefore labels engine-equivalent volumes and selling prices as analyst assumptions. We do not divide the quarter's combined revenue by an unrelated shipment target and present that result as an observed customer price.
The IFRS operating loss was $18.786m, compared with reported net loss of $7.561m. Noncash warrant gains and other income helped bridge those figures. The company’s reported loss per share of four cents should be understood alongside the expanded share count and those nonoperating effects. Lower loss per share does not independently demonstrate better factory economics. The financial statements are unaudited interim statements expressed in U.S. dollars.
Management expects fourth-quarter shipments to exceed those in the first nine months combined. That is a directional shipment expectation, not a numerical revenue or earnings guide. It does not reveal product mix, acceptance timing or the proportion recognized as revenue in the fourth quarter. We retain the statement as a checkpoint, without inventing a quarterly sales forecast in the company’s name. The distinction becomes especially important because the previously targeted 2026 engine backlog now extends into 2027.
| Metric | Value |
|---|---|
| Q4 2024 | 0.029 |
| Q1 2025 | 0.167 |
| Q2 2025 | 0.268 |
| Q3 2025 | 0.298 |
| Q4 2025 | 0.341 |
| Q1 2026 | 0.503 |
| Q2 2026 | 0.570 |
| Q3 2026 | 1.026 |
POET develops integrated optical engines, light sources and photonic components for data-center connectivity. Its interposer approach aims to simplify integration and assembly. Customers need reliable components that meet power, cost, performance and manufacturing requirements inside a finished optical module. A technical demonstration is necessary, but commercial qualification and repeatable yield determine whether the design becomes a durable source of revenue.
Management identifies nine active customers across sixteen projects, with seven customers expected to enter production during 2027. Seven refers to customers, not projects. Four customers have placed orders exceeding one million dollars, and a $2.4m repeat purchase order from an existing customer adds evidence beyond initial sampling. Still, a purchase order and recognized sales are different stages. Supply availability, customer acceptance and final delivery can delay conversion even after commercial interest is established.
Lumilens represents a larger potential relationship: an initial $50m purchase order within a framework potentially exceeding $500m over five years. That framework is not guaranteed recognized revenue. Customer warrants also transfer economic value when milestones are achieved. We explicitly allocate those claims in the model, rather than treating every future customer dollar as if it came without a commercial concession. The relationship increases opportunity and concentration at the same time.
Wafer throughput rose from fifty per month in 2025 to one hundred twenty in 2026. Committed equipment is intended to support 1,250 wafers monthly, or potentially 1,500 with productivity gains. Separately, the existing assembly reference is about one million optical engines annually, with a new line intended to expand that capability roughly tenfold by the end of 2027. These are different stages of production. Multiplying both expansion ratios together would create fictitious sales capacity. Neither planned capability nor equipment installation proves customer utilization, yield or profitable sell-through.
| Metric | Value |
|---|---|
| Initial Lumilens order | 50.000 |
| Potential five-year framework | 500.000 |
| Repeat customer order | 2.400 |
Reported IFRS research and development expense was $7.010m in the quarter. The issuer’s separate pro forma table shows $5.377m after removing research-related share compensation. These figures are not interchangeable. Total quarterly share compensation was $5.720m, including $4.087m in selling, general and administrative costs and $1.633m in research. It reduces ownership economics even when it does not immediately consume bank cash.
Our base forecast starts with a twenty percent product gross margin in 2027 and reaches forty-six percent in 2036. Nonrecurring engineering revenue has a separate fifty percent assumed gross margin. These are analyst estimates because the current disclosures do not establish mature product profitability. The base also allows selling prices to decline from $100 to $76 per engine equivalent, rather than assuming that today’s technical opportunity preserves pricing indefinitely.
Economic operating expenses rise from $60m to $150m over that period. They include ongoing research, selling costs, overhead, cash rentals and future new compensation. Existing awards are handled through the ownership allocation. Their historical accounting amortization is not charged again as a second claim. This distinction prevents both treating stock compensation as free and double-counting the same grant. It remains an approximation because future grant policy is unknown.
The key commercial test is the relationship among yield, throughput, product mix and cost. Low yields can consume wafers and labor without producing accepted engines. Higher shipment volumes can require additional support and quality spending. Competition from incumbent suppliers or different integration architectures can reduce price before costs have fallen. Our scenarios locate those risks in units, gross margins, operating costs and reinvestment; we do not add an unexplained extra discount after already modeling weaker economics.
| Metric | Value |
|---|---|
| IFRS R&D | 7.010 |
| Pro forma R&D | 5.377 |
| Total share compensation | 5.720 |
The $7.561m reported net loss benefited from a $6.378m noncash gain on warrant valuation. Other income of $5.055m, including interest, also helped offset the much larger operating deficit. The company’s warrant-excluded loss was $13.939m. That adjusted figure remains different from operating loss and cash flow: removing one fair-value movement does not remove every noncash item or every nonoperating source of income.
A warrant liability can fall in accounting value when the share price or valuation assumptions change. The resulting gain does not mean that customers paid more, that engines became cheaper to manufacture, or that operations generated cash. Conversely, future share-price appreciation can create an accounting charge without an equivalent cash payment. The valuation therefore starts with operating cash economics and allocates the outstanding option claims separately.
The September balance sheet includes $30.142m of deferred customer consideration. This is not cash and is not added as a free financial asset in our valuation bridge. It relates to customer warrants and future commercial accounting. Our sales forecast represents contractual economic sales before allocation of the associated ownership concession. It should not be mistaken for a precise forecast of future IFRS revenue presentation. Deducting the grant economically and then subtracting the same deferred consideration again would count the concession twice.
Interest earned on financial assets is also separated from enterprise cash flow. Current liquid assets are included in the equity bridge, so their income is not capitalized again as if it were operating profit. For the dated future-share-price calculation, retained cash does earn an assumed return. That is a different calculation at a future date, not an extra source of present franchise value. These separations explain why a headline earnings improvement can coexist with a cautious stock call.
| Metric | Value |
|---|---|
| Operating loss | 18.786 |
| Reported net loss | 7.561 |
| Noncash warrant gain | 6.378 |
Cash and cash equivalents of $276.018m plus short-term investments of $498.355m total $774.373m at September 30. The release rounds this balance to $774.3m; the exact statement total is used in the model. We subtract an estimated $12.5m of cash use from quarter-end through November 16, then model a separate $9m stub outflow through December 31. Neither estimate is a reported November bank balance. A $25m operating liquidity floor is reserved in the going-concern bridge.
Quarterly operating cash use was $12.7m. Nine-month operating cash use of $33.670m plus physical capex of $16.698m and patent spending of $0.064m produces a $50.432m outflow before strategic loans and investments. The company’s expansion commitment exceeds $45m, with installation expected through April 2027 and manufacturing readiness around midyear. Our base includes $38m of 2027 capex, followed by ongoing replacement and growth investment; it does not claim the full program is paid in that calendar year.
The $31.094m private loan receivable is distinct from cash. The underlying principal is $30m, carrying six percent compounded interest with a five-year or liquidity-event maturity and a conversion option. Base and bull retain carrying value; the failure scenario assumes no recovery. The $2.955m supplier investment receives the same scenario treatment. Neither asset is a quoted cash equivalent, and recoverability is a material limitation.
The SPX acquisition claim is $5.8m at September 30. Scheduled payments include $1m due October 2026, then $1.3m, $1.6m and $1.9m in successive Octobers. We cannot verify the October settlement and retain the full claim. If paid, cash and debt fall together; subtracting the payment from cash while retaining all debt would double count. All leases are normalized to cash rentals in operating costs, so the $2.467m lease liability is not deducted a second time. Forecast depreciation and capex concern owned assets.
| Metric | Value |
|---|---|
| Reported liquid assets | 774.373 |
| Private loan | 31.094 |
| SPX convertible claim | 5.800 |
| Metric | Value |
|---|---|
| Operating cash used | 33.670 |
| Physical capex | 16.698 |
| Patents | 0.064 |
The main model forecasts engine equivalents, prices, product margins, development revenue, operating costs, taxes, working capital and owned-equipment investment. Base annual units rise from 0.30m in 2027 to 3.50m in 2030 and 9.50m in 2036. At a $76 terminal-year selling price, product revenue reaches $722m. These are substantial assumptions against the still-delayed 30,000-engine milestone. The bull reaches 25m equivalents and explicitly spends for capacity beyond the currently proposed expansion.
Cash taxes are twenty percent of positive operating profit, without assuming an immediately usable tax-loss asset. Working capital consumes fifteen percent of incremental revenue. A 14.5% discount rate values the operating cash flows. Beyond 2036, growth of 2.5% requires reinvestment at an assumed eighteen percent return on incremental capital in the base, or twenty-five percent in the bull. Base terminal value is 79.1% of enterprise value. This signals dependence on distant competitive economics, even though current assets are a large part of total equity value.
Base operating enterprise value is $344.806m. Adding the $765.122m net asset bridge produces $1,109.928m before allocating common and award claims. The model uses 174.216m common shares, 6.170m assumed-vesting restricted units, employee options and the relevant warrant grants. It solves common price jointly with option values, using a ninety percent volatility and four percent risk-free assumption. Base option and warrant claims total about $203.343m; restricted units receive another $31.008m of value. Those claims explain why raw cash divided by basic shares is incomplete.
The ninety-percent volatility is an assumption, not a measured future outcome. Exercise cash is implicit in the call strikes, so it is not added again. The $15.606m accounting derivative liability is not separately deducted after valuing the same underlying claims. Existing awards are allocated here; future new compensation is charged in operating expense. This avoids duplicate penalties while acknowledging that the exact eventual ownership depends on vesting, exercise and price.
The independently motivated asset check assumes an orderly stop after one year. Liquid assets are reduced for interim use, debt, working liabilities, $60m of execution costs and $15m of shutdown costs. Intrinsic ownership claims are settled, with no recovery assumed for the loan, plant or supplier investment. It gives $3.67 at recovery, or $3.21 today at the same discount rate. This differs from the bear, which waits four years before winding down. Neither is a guaranteed legal liquidation value, and neither is averaged into the $5.03 base.
| Year | Units m | ASP $ | Gross margin | Capex $m | FCFF $m |
|---|---|---|---|---|---|
| 2027 | 0.30 | 100 | 20% | 38.0 | −88.8 |
| 2028 | 0.90 | 97 | 30% | 25.0 | −66.0 |
| 2029 | 2.00 | 94 | 36% | 20.0 | −33.1 |
| 2030 | 3.50 | 91 | 40% | 22.0 | 5.3 |
| 2031 | 5.00 | 88 | 43% | 25.0 | 48.0 |
| 2032 | 6.50 | 85 | 45% | 30.0 | 87.8 |
| 2033 | 7.50 | 82 | 46% | 32.0 | 115.6 |
| 2034 | 8.50 | 80 | 46% | 35.0 | 131.0 |
| 2035 | 9.00 | 78 | 46% | 37.0 | 138.5 |
| 2036 | 9.50 | 76 | 46% | 40.0 | 138.1 |
| Claim or value | USD millions |
|---|---|
| Enterprise value | 344.806 |
| Net asset bridge | 765.122 |
| Pre-award equity | 1109.928 |
| Option / warrant claims | 203.343 |
| RSU value | 31.008 |
| Common equity | 875.577 |
ChartMill displays a $11.81 average twelve-month analyst target, with a range of $8.08 to $15.49 and seven analysts in its November 16 snapshot. The displayed series is unchanged since October 4. Fintel separately records the same average on November 2 for a November 2027 projection. These are useful descriptions of an optimistic published expectation, but they are not verified post-earnings updates. Other aggregators use different analyst populations and stale individual targets; we do not blend them into a false unified consensus.
Our comparable November 2027 base is $5.44, calculated by interpolating the 2027 cash path after the year-end stub, retaining forecast cash generation and investment income, and revaluing the remaining business and claims. That is a future conditional estimate. The $5.03 headline is present fair value. Keeping those clocks separate prevents a misleading comparison between a future analyst target and an intrinsic value calculated today.
At $4.72, the reverse calculation needs 8.85m engine equivalents annually by 2036, compared with 9.50m in our base. It scales the unit path while holding selling prices, gross margins, operating costs, tax rates, discounting, terminal returns and the ownership method fixed. Higher incremental volume also carries additional equipment and depreciation assumptions; lower volume retains committed base spending. The solution is therefore an interpretation of price under a specified model, not a forecast attributed to the market or management.
Eight to nine million annual engines is far above the present backlog shipment milestone. Proposed capacity could make that scale physically possible, but does not demonstrate demand or acceptable yield. A lower selling price or weaker margin would require more units to support the same quote. Better economics could require fewer. The stock price consequently contains commercial optimism despite sitting near a large reported gross cash balance.
| Metric | Value |
|---|---|
| Street 12-month mean | 11.810 |
| Our Nov 2027 base | 5.440 |
| Our present base | 5.026 |
The company raised $550m gross in two 2026 placements: $150m in January and $400m in May. Those placements issued 39.737m common shares before option and warrant exercises. The May units included one common share and one $26.25 warrant each. Liquidity therefore reflects investor financing as well as operational activity. Evaluating per-share opportunity requires using the expanded ownership base, not a historical share count from before the raises.
The MD&A reports 174,215,937 common shares and 78,729,968 warrants at both September 30 and November 16. The total warrants already include 22,921,408 customer grants at an $8.25 strike. Only 2,292,140 customer warrants were vested at quarter-end; the other 20,629,268 depend on payment milestones. The base assumes half of those conditional awards eventually vest, the bull all, and the bear ten percent. This is a subjective scenario proxy, not a claim that actual contractual vesting is a linear function of revenue.
There are 5,216,495 employee options, including unvested awards. The September restricted-unit balance is 5,785,031. A November 2 release separately announces 384,615 new CFO units vesting in thirds over three anniversaries. The later MD&A repeats the old unit count as of November 16. We conservatively include both, giving 6,169,646 modeled units, while acknowledging the unresolved issuer reconciliation. This may modestly overstate awards if an offset is missing from the disclosures.
A further source inconsistency concerns the May warrants: generic financing prose describes a five-year term, while the explicit schedule gives May 18, 2029 and the MD&A supports a three-year term. We use the tabular expiry. No new financing, buyback or dividend is assumed in the model. The reviewed post-quarter SEC submissions reveal no new share placement, but that is not a guarantee that future capital allocation will remain unchanged.
| Ownership item | Millions | Qualification |
|---|---|---|
| Common shares | 174.215937 | Latest MD&A count |
| All warrants | 78.729968 | Includes customer grants |
| Customer warrants | 22.921408 | Subset; 2.292140 vested |
| Employee options | 5.216495 | Includes unvested |
| Reported RSUs | 5.785031 | MD&A repeats old count |
| November CFO grant | 0.384615 | Separate release |
| Modeled RSUs | 6.169646 | Conservative reconciliation |
Customer qualification can take longer than equipment installation. POET can complete a manufacturing line yet fail to win sufficient sockets at an acceptable cost. Incumbent optical suppliers and competing integration designs may improve during the delay. Those mechanisms affect unit volumes, selling prices and gross margins. The bear scenario assumes a prolonged but unsuccessful commercial effort rather than an immediate disappearance of all technology value.
The base cash balance remains above roughly $565m before financial returns, so the modeled ramp does not require another equity raise. That does not make execution risk small. Capital can be spent on excess capacity, acquisitions or projects with inadequate returns. A low probability of near-term insolvency is different from a high probability of attractive shareholder returns. The private loan also links part of asset recovery to another company’s credit and liquidity events.
Operating sensitivities show which assumptions change the call. Twenty-five percent fewer units at base margins gives $3.90 per share. Five percentage points less product gross margin at base volumes gives $4.49. Both changes together produce $3.49. These are alternative calculations, not deductions piled onto the bear. Higher-volume sensitivities include additional capex and depreciation; the lower-volume cases retain committed base equipment spending.
Discount-rate and terminal-growth sensitivities are secondary model-risk context. The base remains heavily dependent on cash flows beyond 2036. Architectural obsolescence could shorten the economic life, while unexpected manufacturing success could support better returns. The ninety-percent option-volatility assumption also affects common ownership allocation. Neither the model nor the evidence justifies treating a few cents of calculated fair value as precision. Changes in the cap table, loan recoverability or acquisition settlement should be incorporated before relying on future revisions.
| Unit scale | Margin −5 points | Base margin | Margin +5 points |
|---|---|---|---|
| 75% | $3.49 | $3.90 | $4.31 |
| 100% | $4.49 | $5.03 | $5.56 |
| 125% | $5.40 | $6.06 | $6.71 |
| Discount rate | 1.5% terminal growth | 2.5% terminal growth | 3.5% terminal growth |
|---|---|---|---|
| 13.0% | $5.42 | $5.47 | $5.54 |
| 14.5% | $4.99 | $5.03 | $5.06 |
| 16.0% | $4.67 | $4.69 | $4.71 |
Management has announced March 1, 2027 for its next results call. That update should help distinguish a strong fourth-quarter shipment comparison from meaningful revenue and cash conversion. The key question is not merely whether shipments exceed the first nine months, but whether customers accept repeat production and whether the economics improve as volume rises. A disclosed product-margin bridge would materially improve confidence in the forecast.
The second quarter of 2027 is the stated deadline for the current engine backlog, including the thirty thousand units previously targeted for 2026. Equipment installation through April and manufacturing readiness around midyear create a tight schedule. The roughly tenfold assembly expansion by year-end is a separate milestone. It should be evaluated through usable capacity, yield and customer qualification, rather than through floor space or installed machinery alone.
Other planned milestones include production of 1.6T optical engines in 2027 and early-2028 production of Blazar hybrid lasers after first-half-2027 sampling. These are management targets, not guaranteed revenue dates. Our model does not add separate speculative values for every product announcement; successful adoption contributes to the overall shipment and margin paths. Failure belongs in those same operating assumptions.
December 2030 conditional prices are $2.99 bear, $7.48 base and $19.08 bull. From the $4.72 quote, annualized returns are approximately minus 10.5%, plus 11.8% and plus 40.3%, with no distributions assumed. These are scenario-specific returns, not probability-weighted IRRs. The horizon calculation retains cash flows and investment income and revalues the remaining business. Expired warrants are represented through a simplified intrinsic net-settlement proxy; actual earlier exercise, cash proceeds and reinvestment timing can materially change ownership and value. The calculation does not claim expired contracts remain outstanding until 2030.
| Case | Future price | Total return | Annualized return |
|---|---|---|---|
| Bear | $2.99 | −36.6% | −10.5% |
| Base | $7.48 | 58.5% | 11.8% |
| Bull | $19.08 | 304.3% | 40.3% |
The financial scorecard uses the company’s November 16 release, unaudited September 30 financial statements and management discussion. The statements provide the exact cash, claims and expense figures; the release supplies operating milestones and management expectations. Where the release’s pro forma presentation differs from IFRS expense, both are labeled. Figures are in U.S. dollars, generally millions, unless the table specifies shares, physical units or per-share amounts.
The tables show the principal calculations behind our valuation. Product revenue is assumed units multiplied by selling price. Cash flow starts with economic operating profit, deducts cash taxes, adds owned-equipment depreciation, and deducts equipment investment and incremental working capital. Present enterprise value is the discounted explicit cash flows plus a reinvestment-funded terminal value. The separate asset bridge and ownership allocation then produce common-equity value. The failure case uses a finite recovery instead of a perpetual franchise.
Rounding is for readability. The model uses the exact September liquid-asset sum of $774.372772m and the stated common-share count. Present scenario values round to $1.59, $5.03 and $11.48; subjective weights of thirty, fifty and twenty percent produce $5.29. These weights express judgment about conditional commercial paths. They are not statistically calibrated probabilities and should not be interpreted as insurance against outcomes outside the range.
The strongest limitations are unproven product margins and yields, uncertain future customer mix, the discrepancy in restricted-unit disclosure, the warrant-expiry inconsistency and the absence of verified October acquisition settlement. The quote and Street snapshot are explicitly dated, and the latter is stale. The research does not infer a numerical company revenue guide, a verified liquidation floor, or a guaranteed ability to realize private assets at carrying value. Readers can use the cited primary documents and displayed driver tables to challenge the assumptions that matter.
Primary sources: POET Q3 2026 release; SEC financial statements and notes; SEC management discussion; November 2 CFO award release. Market context: Yahoo Finance POET quote (recorded November 16, 15:15:43 EST); ChartMill analyst target snapshot; Fintel dated target record. Forecasts are Charged Alpha author estimates.