CAPITAL THESIS

Independent equity research · AAOI · full edition

Applied OptoelectronicsThe AI optical ramp has a cash-and-dilution test.

Optical demand is accelerating. We separate qualified shipments from capacity ambitions, trace the $707m first-half cash requirement, and test funded five-year shareholder outcomes.

Research date: 23 October 2026 · USD common equity · Reference price: USD 68.07 · Observation: 2026-10-22

Watchlist · demand is stronger than valuation support

The investment decision

The $68.07 reference produces 13.6% annual return in an already demanding base case. A $51.66 entry clears a conditional 20% exit-multiple hurdle; the cash-flow valuation is only $25.07. The gap matters.

Q2 revenue growth+86.4%Reported year on year
H1 operating + investing−$707.5mCash use before financing
Base five-year return13.6% p.a.House case after funding/dilution
DCF cross-check$25.0714% discount / 3% terminal growth

Edition update. Initiation dated 23 October 2026 Sydney. Regular-close price frozen at 4 pm EDT on 22 October. Latest financial actuals end 30 June; Q3 results scheduled for 5 November.

01. The optical ramp is real. The shareholder payoff is conditional.

Research stance: watchlist; demand evidence is stronger than valuation support. At the frozen $68.07 regular close, the independent base case produces 13.6% a year over five years. It already requires revenue of $4.8bn and a 17% operating margin by FY2031. Its $51.66 entry for a 20% return is an exit-multiple scenario calculation, not intrinsic fair value. The cash-flow cross-check is materially lower at $25.07. We would not turn the former into an unconditional buy price.

The debate is whether a component supplier can convert AI infrastructure spending into qualified, profitable, cash-generating shipments before capital spending and dilution absorb the benefit. The potential upside comes from product qualification, manufacturing yield and scale. The counterargument is that investors fund the factories now while powerful customers control pricing, orders and qualification timing.

The key evidence to upgrade the view is growth in accepted 800G/1.6T shipments alongside improving gross margin and a credible fall in cash investment relative to revenue. A lower share price by itself does not establish better value if financing becomes more dilutive. Failure to translate the capacity ramp into margin and cash would weaken the thesis even if reported revenue continues growing.

Data cut-off: 23 October 2026 Sydney. Price basis: $68.07 at 4 pm EDT on 22 October, not the separate $66.80 after-hours observation. The regular close fell 12.0% on the day, but this report does not infer an unverified cause. The latest reported quarter ends 30 June; September results are scheduled for 5 November. No September cash balance or completed qualification is invented. [S2] [S5] [S11]

02. What AAOI sells, and why qualification matters

Applied Optoelectronics designs and manufactures optical networking products. Transceivers convert electrical signals into light and back again, allowing servers and switches to communicate over fibre. Its exposure extends beyond AI: the June quarter included $107.7m of data-centre revenue and $80.6m of cable-network revenue. These are end markets within one operating segment, not separately reported profit centres. [S1] [S2]

The commercial chain is design win → qualification → good manufacturing yield → accepted shipment → collection. Winning a design is valuable because changing an installed design can be costly and slow. It is not equivalent to a guaranteed order. The company states that customers generally buy on purchase orders, without commitments exceeding one year. This makes the timing and volume of capital spending difficult to underwrite. [S2]

Vertical integration can improve control over laser supply, engineering feedback and unit cost. The company manufactures its laser chips in Texas and performs assembly and manufacturing across the United States, Taiwan and China. This is a potential capability advantage, not proof of an enduring monopoly or current all-US production. The manufacturing-geography revenue table should not be misread as customer geography. [S2]

Competition turns this into a cost, power, reliability and delivery contest. Higher port speeds expand the opportunity, but price declines and rapid product transitions can erode the return on each production line. Co-packaged optics could change where value accrues; management also describes an external-laser opportunity. Neither automatic displacement nor a guaranteed new revenue stream is established. We exclude unqualified future products from near-term actuals. [S3]

03. Capacity is a ceiling, not an invoice

On the August call, management described roughly 200,000 monthly high-speed units of capacity, a year-end target above 650,000 and a 2027 year-end ambition above 930,000. It also discussed a $471m monthly data-centre revenue run-rate ambition by mid-2027. Those figures are management objectives; annualising the latter to $5.652bn does not create booked annual revenue or a binding backlog. [S3]

The gap between capacity and shareholder cash has four steps. First, the right customer must qualify each design. Second, the factory must turn installed tools into usable units at acceptable yields. Third, product mix and selling prices determine revenue and gross profit. Finally, inventory and customer invoices must convert to cash. A delay at any step can leave equipment and inventory financed well before the expected receipts arrive.

The June quarter produced only $12.8m of 800G revenue, according to the call, versus $48.4m from 400G. Management expected a steep 800G ramp and initial 1.6T shipments later in the third quarter; this report has not verified the subsequent qualification outcome. The July Pearland expansion announcement supports the investment programme, not realised sales. [S3] [S7]

Evidence typeRevenue / capacity statementHow it is used
Reported Q2 2026Revenue $191.922mHistorical base, not annual run rate
August Q3 guidanceRevenue $255–290m; adjusted gross margin 29–30.5%Forward management guidance, not actuals
August FY2026 commentaryRevenue around $1.1bnManagement expectation, not house FY2031 case
Capacity ambitionAbove 650,000 high-speed units/month at year endRequires qualification, yield and demand
Capital Thesis baseFY2031 revenue $4.8bnIndependent funded scenario; no claim of consensus

04. Growth has not yet established durable operating profit

Q2 revenue increased 86.4% year on year to $191.922m. GAAP gross margin fell from 30.3% to 27.7%; operating expenses were 40.6% of revenue. This produced a $24.727m operating loss and $22.781m net loss. The scaling opportunity is clear, but current volume has not yet covered the cost structure. [S1] [S2]

Adjusted net income of $5.475m is a different measure. It does not reverse the operating loss: adjusted operating income was also negative, at about $10.3m. Management attributed the better adjusted bottom line partly to foreign tax benefits and subsidy income. The reconciliation includes a substantial tax adjustment and other exclusions. A positive adjusted EPS headline is therefore weak evidence of a proven manufacturing earnings engine. [S1] [S3]

USD million, unless statedFY2024 actualFY2025 actualH1 2026 actual
Revenue249.365455.715343.066
Operating income−70.908−54.602−37.718
Net income−186.733−38.228−37.062
InterpretationNet loss includes a large debt-extinguishment chargeRevenue expansion, still operating lossSix months only; do not compare as a full year

Historical net income is not a clean trend in unit economics: the 2024 loss includes a large debt-extinguishment charge. Gross profit, operating cost absorption and cash investment are more informative at this stage. Trailing P/E and PEG are not meaningful because GAAP earnings are negative. A negative free-cash-flow yield is evidence of funding demand, not a positive income yield. [S4]

05. The $707m cash catch is a build-out, not a net loss

H1 operating cash flow was −$73.781m and investing cash flow was −$633.713m. Together, they consumed $707.494m before financing. This is a cash-flow sum, not an accounting loss, annual forecast or EBITDA adjustment. Financing supplied $980.279m; currency movements explain the remainder of the increase in reported cash and restricted cash. [S2]

H1 2026 cash flow, USD millionAmountInterpretation
Operating cash−73.781Working capital and operating performance
Property/equipment purchases−335.132Cash paid for capital assets
Equipment and other prepayments−289.744Cash already committed before asset delivery
Land, intangibles and net disposal proceeds−8.837Remaining net investing flows
Total investing cash−633.713Includes the prepayments above
Operating + investing−707.494Cash requirement before financing
Financing cash980.279Predominantly new common equity

A conventional operating-cash-minus-PPE calculation is −$408.913m. It excludes the large investing prepayments, making it incomplete for this particular funding question. We show the broader figure rather than calling the difference harmless. Prepaid equipment will later appear in assets and depreciation; the model must not charge that original cash a second time when delivery occurs.

Receivables and inventory absorbed $69.649m and $94.092m in operating cash respectively; payables provided $142.156m. Those are cash-flow movements, not the ending balances. More favourable supplier terms can ease the ramp, but relying on them indefinitely transfers risk to future payment dates. June receivables, inventory and payables imply trade working capital of $306.712m. Our $350m year-end starting point is explicitly a house assumption, not a September observation. [S2]

06. Cash raised is not free value per share

June unrestricted cash was $499.737m; restricted cash was a separate $9.021m. October’s completed ATM raised $600m gross and about $588m net through 5.695m shares at an average $105.36. The August prospectus separately reports $49.2m gross issued after June under the earlier programme. We estimate $48.216m net after a 2% fee; this is not a separately disclosed net figure. The $1.028bn H1 net equity proceeds are already inside June cash and are not added again. [S2] [S6] [S8]

The model cash bridge is $499.737m + $48.216m estimated net + $588.000m − a case-specific remaining-2026 cash-use allowance. The base allowance is $800m, leaving $335.953m at the start of FY2027. This is a forecast bridge, not a claim about October bank balances. Management said second-half capital-investment intensity would exceed first-half levels, so cash spending and additional financing are central monitoring items. [S3]

Starting financing claims include $58.915m bank borrowings, $33.940m bank acceptances and $124.925m face value of convertible notes. The notes mature on 15 January 2030 and carry a 2.75% coupon. Base and bull assume share conversion in 2030, adding about 2.884m shares; bear repays principal. Conversion and refinancing must actually be feasible then. The model omits the roughly $0.14m 2030 interest stub as immaterial, not a full-year interest bill. [S2]

Existing operating leases total $78.143m. Rent stays in operating margins and cash costs, so the lease liability is not subtracted a second time from a post-rent EBIT valuation. Available borrowing and covenant compliance at June do not guarantee funding capacity at the next ramp. The case model conservatively raises equity to protect a $150m year-end cash floor instead of assuming unlimited new borrowing. It does not capture all intra-quarter liquidity peaks. [S2]

07. Share count must follow the financing path

The starting common count is 84.906289m at 20 August plus 5.694845m shares from the later completed ATM, or 90.601134m. Earlier post-June issuance is already included in the August count. This is a dated pro-forma bridge, not a newly filed October count. August outstanding RSUs and PSUs add 1.719518m economic claims, before future financing or conversion. [S6] [S8]

The customer warrant can cover up to 7.945399m shares at $23.6956. Most vesting depends on cumulative purchases; the $4bn purchase threshold is not a guaranteed order. The base and bull assume full vesting by FY2031 and exercise only if economic. Exercise adds both shares and cash proceeds. Bear assumes only partial vesting and, at its terminal value, no exercise. Ignoring either the claims or the exercise cash would bias the outcome. [S2] [S8]

The base raises an additional $1058.3m gross across FY2027–31 at an assumed $55 per share, after 2% issuance costs. Terminal economic shares reach 122.391m. These are house financing assumptions, not a newly announced offering. The bull can finish with fewer shares despite a larger capital programme because issuance is assumed at $90; the bear assumes only $35. A business growing faster can still deliver disappointing per-share returns if financing terms deteriorate.

Operating margins retain ongoing employee compensation as a cash-equivalent cost. We do not add future stock compensation back as free cash and then ignore dilution; instead, its economic cost remains in margins. Existing awards are included separately. This is an economic scenario model, not a calculation of future GAAP diluted EPS.

08. The base path requires scale, yield and operating discipline

Capital Thesis base; USD millionFY2027FY2028FY2029FY2030FY2031
Revenue1,8002,7003,5004,2004,800
Gross margin32%34%36%37%38%
Operating margin4%9%13%16%17%
Cash investment700500400350330
Owner cash after investment-830.2-379.1-13.7310.5490.2
Ending cash150.0150.0150.0460.5950.7

The FY2027 $1.8bn base is above management’s $1.1bn FY2026 expectation but well below an annualisation of its more ambitious mid-2027 monthly target. Revenue then grows to $2.7bn, $3.5bn, $4.2bn and $4.8bn. This is still more than ten times FY2025 sales. We deliberately allow a slower ramp than the advertised capacity story, while acknowledging that even this lower path demands major commercial success.

An illustrative FY2031 product bridge is 12m high-speed units sold at $337.50 net blended revenue per unit, plus $750m of legacy, cable and other sales. Both are house assumptions, not disclosed orders, pricing or capacity. The implied 1m units sold per month requires qualified output beyond the stated 2027 capacity ambition. Faster product mix offsets price erosion in this illustration; it is a way to expose the sales hurdle, not a claim that unit forecasting is precise.

The terminal 38% gross margin and 17% operating margin imply operating expenses at 21% of sales, versus 40.6% in the latest quarter. Better factory utilisation and yield must overcome price erosion, depreciation and support costs. If those benefits fail, revenue alone cannot rescue the valuation. Cash investment includes new equipment and prepayment outlays; it does not repay equipment already prepaid in 2026. A 21% tax charge applies only to positive pretax income; loss tax benefits are not assumed.

Trade working capital falls from 35% to 27% of annual revenue as the base matures. Interest uses a 6% bank/acceptance assumption plus the note coupon until maturity. Surplus cash is retained; no common dividends or buybacks are modelled. A five-year fiscal path is mapped to a stylised 22 October 2026–22 October 2031 holding period; this is an economic comparison, not a quarter-by-quarter treasury forecast.

09. Funded scenarios: common-share outcomes, not revenue targets

House caseFY2031 salesOp marginExit EV/EBITEconomic sharesFuture priceAnnual return
Bear$1.8bn6%12x124.73m$11.64-29.8%
Base$4.8bn17%18x122.39m$128.56+13.6%
Bull$8.0bn23%22x117.87m$364.57+39.9%

Bear five-year total-wealth CAGR: -29.8%. Base five-year total-wealth CAGR: 13.6%. Bull five-year total-wealth CAGR: 39.9%.

Return = (terminal share value / $68.07)^(1/5) − 1. No common dividends, cash distributions, reinvestment receipts or buyback yield are added. Returns are nominal USD before investor taxes, fees and currency effects. There are no scenario probabilities. These cases test conditions, not likely share-price paths.

Bear models an incomplete ramp: $1.8bn terminal revenue, 6% operating margin and a 12x EV/EBIT exit, with weaker funding terms. Base reaches $4.8bn and 17%, capitalised at 18x. Bull reaches $8bn and 23%, valued at 22x. The latter needs both exceptional execution and continued willingness to pay a high multiple. Downside can exceed the bear case if liquidity, customer demand or technology change more severely.

Enterprise value is terminal EBIT times the selected multiple. We add retained cash and warrant exercise proceeds, subtract the remaining bank/acceptance debt, and divide by the funded economic share count. Convertible principal has already been repaid or converted by then. Unlike a P/E model, cash is an explicit enterprise-to-equity bridge; it is not added a second time elsewhere.

Base five-year total-wealth CAGR without rerating: 17.5%. The comparison freezes a 3.66x enterprise-value/house-FY2027-sales multiple and applies it to FY2031 sales, using the same terminal funding bridge. Starting enterprise value includes existing awards and contingent warrant/conversion premiums at the reference price, plus debt less model opening cash. It is model-relative because the forward sales and cash are assumptions; it is not an observed trailing P/E.

10. Cash valuation is materially below the scenario entry

Base DCF cross-check: $25.07 per share at a 14% discount rate and 3% long-term growth. It uses the same five-year operating path and cash investment, then funds continued growth through working capital and capital spending. Terminal annual unlevered cash flow is $588.0m. This is substantially below both the reference price and the $51.66 exit-multiple hurdle entry.

Discount rate / terminal growth2%3%4%
12%$31.18$34.49$38.63
14%$23.00$25.07$27.48
16%$16.90$18.32$19.96

In the perpetuity, cash investment is 7% of revenue, depreciation 6.25%, and incremental working capital remains funded. Enterprise cash flow excludes financing inflows. Starting bank debt and note face value are deducted; the note’s conversion premium and conditional warrant dilution are included using net share equivalents. This avoids subtracting note principal and then charging all conversion shares without the offset. Existing compensation claims remain explicit.

The DCF is lower because reinvestment comes before mature cash returns, the required return is high, and the perpetuity values cash rather than applying an earnings exit multiple. The terminal-value method is a major uncertainty, not a cosmetic choice. The scenario model additionally exposes future financing prices; the unlevered DCF captures funding needs in discounted cash flows instead of adding financing proceeds as value.

Base sensitivityFive-year annual returnConditional 20% entryAdditional gross equity
Base at 14x EV/EBIT8.4%$40.95$1058.3m
Base at 22x EV/EBIT17.9%$62.38$1058.3m
Extra $300m remaining-2026 cash use12.6%$49.42$1364.4m
New equity priced at $3511.6%$47.41$1058.3m

A lower exit multiple, extra 2026 cash use or cheaper equity issuance reduces per-share outcomes. These are one-at-a-time stresses; simultaneous misses can be much worse. The market does not have to wait until FY2031 to penalise a failed ramp.

11. A hurdle price is conditional on the whole path

The base-case conditional entry for a 20% five-year total-wealth CAGR is $51.66 per share, calculated as $128.56 terminal wealth divided by 1.20^5. The calculation uses unrounded outputs and zero common dividends. It requires the $4.8bn/17% base operating result, 18x exit EV/EBIT, stated investment and working-capital path, financing prices and economic shares. It is not a near-term price target or evidence that the stock is intrinsically worth that amount.

The entry explorer changes the investor’s purchase price while freezing the business and funding path. In practice a lower market price could also make future issuance more dilutive. Re-underwrite the cash bridge and share count before treating a price decline as an opportunity. The lower DCF cross-check argues for caution rather than treating the 20% hurdle as automatic margin of safety.

12. What would change the conclusion

Evidence windowInspectConsequence
5 November resultsQ3 shipments, gross margin, cash investment and fresh guidanceCheck whether the advertised capacity translated into accepted sales
Next filingReceivables, inventory, payables, prepayments and cashReplace the house 2026 bridge with reported data
Next 2–4 quarters800G/1.6T qualification, yields, price/mix and capacity utilisationTest whether scaling raises economic margins
Funding disclosuresNew shares, financing price, debt terms and warrant vestingRecalculate the denominator and liquidity runway
2030 maturity planningCash repayment or economically supported note conversionAvoid assuming dilution substitutes for refinancing automatically

Positive evidence would be repeated revenue growth with gross margin moving toward the low-to-mid 30s, narrowing operating losses and lower investment per dollar of sales. Warning signs are growing prepayments without shipment progress, persistent margins below 30%, another large funding need at lower prices, or customer concentration worsening. These are house monitoring thresholds, not management guidance.

Concentration is particularly important: the three largest customers accounted for approximately 92% of Q2 sales, based on disclosed rounded shares of 42%, 26% and 24%. A customer can slow deployments or negotiate harder even while the long-run AI market grows. Other risks include export/trade policy, supply-chain disruption, intellectual property, competing optical architectures and customer insourcing. [S2]

No current short-interest, options positioning, cost-to-borrow or institutional-holder analysis was verified. This report makes no short-squeeze prediction and no personalised allocation recommendation. The intended use is an absolute-return research watchlist with explicit operating and funding conditions.

13. Evidence boundaries and source conflicts

Two supplied AAOI broker notes date from February and March. Their targets, prices and forecasts are historical views with different horizons, not current consensus. The third supplied document concerns 5N Plus, not AAOI, and is excluded. We preserve all originals privately; no proprietary charts or lengthy text are reproduced. [S9] [S10]

The February note labels a roughly $74.9m Q4 figure as cable revenue; the company’s trend table identifies it as data-centre revenue, with cable at $54.0m. The March note contains inconsistent monthly run-rate totals and predates the later $471m management ambition. Primary company evidence takes precedence. A percentage row in the company’s trend sheet is also inconsistent with the underlying operating-expense/revenue calculation; the 10-Q’s 40.6% is used. [S2] [S4] [S9] [S10]

Market providers differed by a few cents on the close; the report consistently uses the dated Stock Analysis $68.07 observation rather than averaging prices. No post-August September-quarter balance sheet, comprehensive fresh consensus model or verified completed 1.6T qualification was available in this research set. The supplied broker material helps frame questions; company filings anchor actuals and capital terms. House scenarios and valuation judgements remain explicitly separate.

Financial exhibits

Exact observations accompany every chart. Actuals, management guidance and Capital Thesis assumptions are labelled separately; the chart shapes do not imply probabilities.

Chart 01 · USD million

Data-centre growth is the largest recent revenue driver

Cable remains material; the business is not yet only high-speed AI transceivers.

Data centreCable / CATV
Data-centre growth is the largest recent revenue driver (USD million). Exact figures in the table below.0255075100125Q3 2025Data centre: 43.935 USD million43.935Cable / CATV: 70.602 USD million70.602Q4 2025Data centre: 74.876 USD million74.876Cable / CATV: 54.002 USD million54.002Q1 2026Data centre: 81.404 USD million81.404Cable / CATV: 66.841 USD million66.841Q2 2026Data centre: 107.662 USD million107.662Cable / CATV: 80.578 USD million80.578
Exact values · USD million
ObservationData centreCable / CATV
Q3 202543.93570.602
Q4 202574.87654.002
Q1 202681.40466.841
Q2 2026107.66280.578

Reported actuals. Remaining telecom/other revenue excluded from this two-series exhibit, not treated as zero. Sources: [S2] [S4]

Chart 02 · percent of revenue

More sales have not yet produced a sustained margin uplift

Gross margin is before operating expenses and financing costs.

GAAP gross margin
More sales have not yet produced a sustained margin uplift (percent of revenue). Exact figures in the table below.010203040Q1 2025GAAP gross margin: 30.6 percent of revenue30.6Q2 2025GAAP gross margin: 30.3 percent of revenue30.3Q3 2025GAAP gross margin: 28 percent of revenue28Q4 2025GAAP gross margin: 31.2 percent of revenue31.2Q1 2026GAAP gross margin: 29.1 percent of revenue29.1Q2 2026GAAP gross margin: 27.7 percent of revenue27.7
Exact values · percent of revenue
ObservationGAAP gross margin
Q1 202530.6
Q2 202530.3
Q3 202528
Q4 202531.2
Q1 202629.1
Q2 202627.7

Reported quarters; do not mix with higher non-GAAP gross-margin guidance. Sources: [S2] [S4]

Chart 03 · USD million cash-flow contribution

H1 expansion depended on financing

Operating and investing cash together consumed $707.494m.

H1 2026 actual
H1 expansion depended on financing (USD million cash-flow contribution). Exact figures in the table below.-1,000-50005001,000OperatingH1 2026 actual: -73.781 USD million cash-flow contribution-73.781InvestingH1 2026 actual: -633.713 USD million cash-flow contribution-633.713FinancingH1 2026 actual: 980.279 USD million cash-flow contribution980.279Currency effectH1 2026 actual: 19.938 USD million cash-flow contribution19.938Net cash changeH1 2026 actual: 292.723 USD million cash-flow contribution292.723
Exact values · USD million cash-flow contribution
ObservationH1 2026 actual
Operating-73.781
Investing-633.713
Financing980.279
Currency effect19.938
Net cash change292.723

Cash and restricted cash reconciliation. The sum before financing is not net loss. Sources: [S2]

Chart 04 · USD million cash outflow

Equipment prepayments are a major cash commitment

CFO less ordinary PPE capex misses much of this build-out.

H1 2026 investing use
Equipment prepayments are a major cash commitment (USD million cash outflow). Exact figures in the table below.0100200300400PPE purchasesH1 2026 investing use: 335.132 USD million cash outflow335.132Equipment / other prepaymentsH1 2026 investing use: 289.744 USD million cash outflow289.744Land / intangibles, netH1 2026 investing use: 8.837 USD million cash outflow8.837
Exact values · USD million cash outflow
ObservationH1 2026 investing use
PPE purchases335.132
Equipment / other prepayments289.744
Land / intangibles, net8.837

Positive bars denote cash used; totals equal $633.713m investing outflow. Sources: [S2]

Chart 05 · USD million

Even the base requires a large successful ramp

Different outcomes require different execution, capital and funding conditions.

BearBaseBull
Even the base requires a large successful ramp (USD million). Exact figures in the table below.02,0004,0006,0008,000FY2027 houseBear: 1,200 USD million1,200Base: 1,800 USD million1,800Bull: 2,500 USD million2,500FY2028 houseBear: 1,500 USD million1,500Base: 2,700 USD million2,700Bull: 4,000 USD million4,000FY2029 houseBear: 1,700 USD million1,700Base: 3,500 USD million3,500Bull: 5,500 USD million5,500FY2030 houseBear: 1,800 USD million1,800Base: 4,200 USD million4,200Bull: 6,800 USD million6,800FY2031 houseBear: 1,800 USD million1,800Base: 4,800 USD million4,800Bull: 8,000 USD million8,000
Exact values · USD million
ObservationBearBaseBull
FY2027 house1,2001,8002,500
FY2028 house1,5002,7004,000
FY2029 house1,7003,5005,500
FY2030 house1,8004,2006,800
FY2031 house1,8004,8008,000

Independent Capital Thesis scenarios, not company guidance. No probability weights. Sources: [S2] [S3]

Chart 06 · percent annual total-wealth return

Per-share returns depend on the funded path

The base is below a 20% annual hurdle at the $68.07 reference.

Five-year CAGR
Per-share returns depend on the funded path (percent annual total-wealth return). Exact figures in the table below.-40-2002040Bear houseFive-year CAGR: -29.76 percent annual total-wealth return-29.76Base houseFive-year CAGR: 13.56 percent annual total-wealth return13.56Bull houseFive-year CAGR: 39.88 percent annual total-wealth return39.88
Exact values · percent annual total-wealth return
ObservationFive-year CAGR
Bear house-29.76
Base house13.56
Bull house39.88

House FY2027–31 model, explicit financing and dilution, zero common dividends; nominal USD before investor costs. Sources: [S2] [S5] [S6] [S8]

Source register and research basis

Evidence checked for this edition on 23 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.

S5 · Stock Analysis AAOI market snapshot

22 October 2026, 4 pm EDT regular close · Reference $68.07; separate after-hours $66.80 is not used. Snapshot is historical and the live page changes.

https://stockanalysis.com/stocks/aaoi/
S9 · User-supplied Needham AAOI results note

27 February 2026 · Private supplied document, pages 1–3, historical context only. Portal identifies provenance, not a publicly accessible copy.

https://app.godelterminal.com/
S10 · User-supplied Wolfe AAOI note

5 March 2026 · Private supplied document, pages 1–3, historical assumptions and run-rate discussion; not current consensus.

https://app.godelterminal.com/