Cirrus Logic Inc (NASDAQ: CRUS) — More Sockets, Same Customer, New Capacity Risk

Published: 2026-09-10 · Verdict: Accumulate · Entry price: $120 · Price target: $145 · Research confidence: High (88%)

Executive conclusion

Analyst Take

The investment conclusion is Accumulate below $120, with a $145 twelve-month base-case value and medium conviction. The verified September 10, 2026 closing price was $116.48. Using the company’s trailing diluted GAAP EPS of approximately $8.18, roughly 50.1 million basic shares outstanding, June-quarter cash, investments and lease liabilities, and trailing financial denominators reconciled to the filings, that price implies approximately 14.2 times trailing diluted GAAP earnings, 9.2 times adjusted enterprise value to EBITDA, and 10.4 times trailing free cash flow. The trailing equity free-cash-flow yield is about 9.6%. Those are undemanding multiples for a fabless mixed-signal company with a 52.8% trailing gross margin, approximately $572 million of trailing free cash flow, roughly $1.17 billion of cash and investments, no funded borrowings, and a period-end share count that fell more than 10% between FY2022 and FY2026. [S1] [S2] [S12]

The low multiple is not a free lunch. Apple represented 91% of FY2026 revenue and 90% of Q1 FY2027 revenue. Apple purchases through multiple contract manufacturers, but the underlying economic exposure remains one end customer. Cirrus’s sales arrangements generally do not provide recurring minimum-volume protection; customers can cancel, reduce, or reschedule orders, and Cirrus may commit R&D, inventory, equipment, and foundry capacity before final customer demand is known. On June 30, 2026, Cirrus entered a new GlobalFoundries agreement requiring defined quarterly wafer quantities for calendar 2027 and 2028, subject to shortfall provisions. Management estimates purchases of at least approximately $600 million over that period. That obligation is incremental context beyond the $354.2 million of unconditional commitments disclosed at FY2026 year-end and materially increases pre-revenue operating leverage. [S1] [S2]

There is also a material tax contingency absent from simplistic net-cash screens. The IRS’s final Revenue Agent’s Report asserted approximately $168.3 million of additional tax and $63.7 million of penalties, excluding interest, concerning transfer pricing between Cirrus’s US and UK affiliates for FY2017–FY2019. Cirrus disputes the positions, has entered the IRS appeals process, believes its existing reserves are adequate, and has not accrued the asserted additional liability. The assessment is therefore neither a reported payable nor something that should be deducted dollar-for-dollar in the base valuation. It is nevertheless a genuine contingent claim: the stated tax and penalty amounts alone equal roughly $4.63 per current basic share before interest. [S2]

The central thesis is narrower than management’s broad diversification narrative. Product diversification is real. Cirrus derived 42% of FY2026 revenue from high-performance mixed-signal products including camera controllers, haptics, sensing, and power components, and Q1 FY2027 HPMS revenue grew 26% year over year. Customer diversification is not yet real: Apple concentration rose from 87% in FY2024 to 89% in FY2025 and 91% in FY2026. Apple independently confirmed that it is working with Cirrus and GlobalFoundries on new process technology and mixed-signal solutions for several Apple applications, including advanced ICs that power Face ID systems. That is strong evidence of technical relevance; it is not evidence of production volume, Cirrus-specific economics, exclusivity, or minimum demand. [S1] [S2] [S9]

The strongest counter-case is that Cirrus owns renewable design wins, not a permanent franchise. Apple can redesign functions, internalize silicon, sponsor an alternative supplier, dual-source, or demand annual price reductions. The hoped-for independent growth leg also weakened. In May, management expected strong FY2027 PC growth despite market softness; in August it lowered expectations because of constrained platform supply, expensive and scarce memory and components, and delayed OEM models. Management characterized these issues principally as timing. IDC, however, expected material PC-market stress through 2027, while Gartner forecast substantial 2026 contractions in both PCs and smartphones. The mismatch does not prove lost sockets, but it means design engagement should not yet receive the valuation of recurring production revenue. [S6] [S7] [S10] [S11]

Financial quality supports the favorable side, with an important accounting qualification. Company Financials calculates trailing standardized ROIC at approximately 18.2%. Conventional ROIC understates cumulative engineering investment because R&D is expensed. A four-to-five-year straight-line capitalization, using a year-end cohort convention, produces an estimated research asset of roughly $1.1–$1.4 billion and reduces estimated economic ROIC to approximately 11–13%. The sensitivity remains above a reasonable cost of capital, but it is materially below the return suggested by screens that treat engineering as a cost with no enduring capital. FY2026 R&D of $434.0 million was also net of a $42.9 million UK research credit, so the reported line understates gross engineering expenditure. [S1] [S12]

The base case does not require every announced program to succeed. It requires existing smartphone audio and camera content to remain durable, gross margin to stay around 51–53%, PCs to recover partially, operating-expense growth to generate measurable new-product revenue, the foundry commitment to be substantially matched by production demand, and repurchases to remain net-accretive. The call would improve if non-Apple revenue grows faster than total revenue for several quarters, PC revenue establishes a credible path toward $100 million and then $200 million, and 2027–2028 wafer purchases are absorbed without abnormal inventory or shortfall charges. It would deteriorate if gross margin remains below 50%, normalized annual free cash flow falls below approximately $350 million, Apple-related Audio or HPMS sales contract beyond device-unit effects, inventory and purchase commitments rise without shipments, the tax dispute moves adversely beyond existing reserves, or diluted shares stop falling. Evidence quality is high for reported financials, liquidity, concentration, commitments, the tax dispute, and net share retirement; moderate for existing socket durability; and low for long-dated power, metrology, and fragmented general-market revenue.

Stock Price Action — Five-Year Event Map

Company Financials’ split-adjusted daily series from September 10, 2021 through September 10, 2026 shows a five-year closing low of $63.88 on October 12, 2022 and a closing high of $178.90 on June 3, 2026. The five-year intraday low was $61.94 on October 13, 2022. The 52-week intraday range was $106.26 on September 1, 2026 to $180.42 on May 27, 2026. The $116.48 September 10 close was 34.9% below the closing high and 9.6% above the 52-week intraday low, placing it near the bottom of the recent range. Prices are observed facts; causal explanations remain interpretations unless supported by contemporaneous disclosures. [S12]

• October 12–13, 2022 — trough region: the stock closed at its five-year low of $63.88 on October 12 and reached its five-year intraday low of $61.94 the next day. Semiconductor de-rating, higher interest rates, and smartphone-demand concerns were contemporaneous background conditions, but no Cirrus filing establishes a single cause. The defensible conclusion is that a concentrated mobile supplier can be valued as a cyclical component company when both end-market and discount-rate concerns rise.

• April 12, 2023 — down 12.3% to $89.28: the stock fell from $101.82 after reports that Apple had abandoned a planned solid-state-button architecture. Cirrus did not publicly confirm the program. The event cannot establish lost revenue, but it demonstrates the equity’s sensitivity to perceived content at one customer. [S19]

• February 7, 2024 — up 14.6% to $89.86: the move followed results that exceeded market expectations. The subsequent financial record supports an earnings recovery, but the daily price move itself remains a market response rather than proof of permanent design-win durability. [S20]

• May 8, 2024 — up 12.2% to $102.24: the move followed FY2024 results and an outlook supporting recovery from the $1.789 billion revenue trough. FY2025 and FY2026 revenue subsequently rose to $1.896 billion and $1.997 billion, validating the direction of the recovery without proving the exact cause of the one-day move. [S1] [S4]

• April 8–9, 2025 — sharp reversal: the shares closed at $77.43 on April 8 during a broader semiconductor and tariff-related selloff, then rose 16.5% to $90.23 on April 9. No corresponding Cirrus filing establishes either day as evidence of a design change. Treating the selloff or rebound as company-specific fundamental proof would be unwarranted. [S12]

• February 4, 2026 — up 8.3% to $134.91: the move followed Q3 FY2026 revenue of $580.6 million and record quarterly EPS, supported by stronger-than-expected smartphone demand and favorable device mix. The final FY2026 filing confirmed record annual revenue, operating profit, and EPS. [S1] [S5]

• May 27–June 3, 2026 — peak region: the 52-week intraday high of $180.42 occurred May 27; the highest close, $178.90, occurred June 3. Strong FY2026 results, Apple’s manufacturing announcement, and optimism around PCs and new HPMS programs were plausible contributors, but there was no single disclosure explaining the exact peak. The distinction matters because subsequent repurchases and insider sales occurred near elevated prices. [S5] [S9] [S15]

• August 6–September 1, 2026 — expectations reset: despite record first-quarter revenue and a $510–$570 million Q2 outlook, management reduced its PC expectations and guided to higher operating expense. The stock subsequently reached a $106.26 intraday low on September 1. The decline is consistent with lower value being assigned to future diversification, though that attribution remains inference. [S2] [S7] [S12]

• September 10, 2026 — up 5.2% to $116.48: the stock rose from $110.72 without an identified material company filing that day. A one-day rebound without new fundamental evidence does not confirm the thesis. [S12]

The event map exposes two common update errors. First, the 52-week low was the September 1 intraday price, not an August closing price. Second, valuation must be recomputed after the decline. Company Financials’ June-period valuation output used a $148.53 period-end price and therefore overstates the multiples applicable at $116.48. The price decline was much larger than the change in trailing earnings, so valuation is genuinely lower even though the forward PC narrative deteriorated. [S12]

Verdict: The range position and lower multiple create a better entry context, but price history confirms that customer-content expectations can dominate reported quarterly performance. A low price within its range is neither a catalyst nor proof that concentration risk is fully discounted.

Business Overview

Cirrus Logic is a fabless designer of low-power, high-precision mixed-signal integrated circuits and associated algorithms. Its products translate or control real-world analog phenomena—sound, current, voltage, power, motion, and sensing—inside digital devices. Cirrus retains product architecture, analog and digital circuit design, software, algorithms, systems engineering, qualification, customer engagement, and supply-chain management. It outsources wafer fabrication, assembly, and much testing. The result is low owned-factory intensity but substantial dependence on engineering talent, foundry processes, customer road maps, and external manufacturing execution. [S1]

The customer value proposition is system-level performance. A boosted amplifier can create louder sound from a smaller speaker without damaging it; a smart codec can combine conversion, signal processing, voice functions, and power management; a camera controller can improve stabilization or response; and a power device can improve efficiency, heat, size, or control accuracy. These gains can reduce the customer’s bill of materials, improve battery life, enable a visible feature, or lower software and integration burden. Cirrus earns attractive gross margins when its silicon and algorithms solve a sufficiently difficult system problem. It earns commodity-like economics when the function can be integrated, standardized, or sourced from lower-cost alternatives.

The company reports one operating segment because its chief operating decision maker receives consolidated profit and asset information rather than complete discrete financial information for separate product businesses. It nevertheless discloses two revenue categories. FY2026 Audio revenue was $1.160 billion, 58% of total sales, and grew 2%. Audio products include custom and general-market codecs, boosted amplifiers, smart codecs, DSP technologies, and algorithms. HPMS revenue was $837.4 million, 42% of sales, and grew 10.3%. HPMS includes camera controllers, haptic and sensing components, power-management and charging devices, and other precision mixed-signal products. In Q1 FY2027, Audio was $249.0 million, up 3.7%, while HPMS was $210.7 million, up 26.0%; HPMS provided most of the quarter’s incremental revenue. [S1] [S2]

These categories are not economic segments. Cirrus does not disclose product-line gross profit, operating income, capital employed, customer allocation, or R&D. HPMS growth therefore shows that non-audio functions are expanding, but it does not show whether they earn higher margins, reduce Apple dependence, or generate adequate returns on the Lion acquisition and subsequent research. The absence of segment profitability is especially important because management describes some general-market products as carrying margins above the corporate average, while the filings do not permit independent verification. [S7] [S8]

Revenue is transactional product revenue, not contractual recurring revenue. The performance obligation is generally satisfied upon delivery of goods, and most contracts have expected terms of one year or less. Customers and contract manufacturers place purchase orders against forecasts. Orders can be canceled, reduced, or rescheduled on short notice. Economic recurrence comes from winning positions in successive product generations and shipping through a device lifecycle, not from subscriptions or enforceable minimum purchases. This distinction explains how annual revenue can appear stable while the underlying equity exposure remains discontinuous. [S1] [S2]

A useful operating equation is device units multiplied by Cirrus content per device, product mix, and realized price. Gross profit then depends on product price, wafer and assembly cost, yield, freight, reserves, and mix. R&D determines the probability and timing of future sockets. Customer concentration determines bargaining power and the size of possible discontinuities. Working capital, test equipment, and wafer commitments determine how accounting earnings convert to cash. These mechanisms are intelligible, but investors cannot forecast them precisely because Cirrus does not disclose device units, dollars of content, customer-specific prices or margins, socket duration, or the economics of most individual programs.

Customer concentration is the defining characteristic. The ten largest end customers represented 96% of FY2026 sales. Apple represented 91%, up from 89% in FY2025 and 87% in FY2024. In Q1 FY2027, Apple represented 90%, up from 86% in the prior-year quarter. Contract manufacturers do not diversify that exposure: Cirrus identifies the end customer specifying the component, even when another entity buys and assembles it. The concentration is therefore economic, not an invoicing artifact. [S1] [S2]

Revenue stability is mixed. Annual revenue was $1.781 billion in FY2022, $1.898 billion in FY2023, $1.789 billion in FY2024, $1.896 billion in FY2025, and $1.997 billion in FY2026. That is a FY2022–FY2026 compound growth rate of approximately 2.9% and a relatively narrow annual range. Stability resulted from content, product mix, pricing, and execution offsetting device-unit volatility; it was not contractually protected. Quarterly revenue is materially seasonal around flagship consumer launches, and annual price reductions, redesign decisions, and device mix can produce changes disproportionate to end-market unit growth. [S1] [S4]

Cirrus’s reported international exposure requires interpretation. International sales were 99% of revenue in FY2024–FY2026, but the company explicitly includes sales to US-based end customers using overseas manufacturers. Ship-to geography therefore maps the electronics supply chain more than the residence of the ultimate consumer or location of product decisions. Sales are principally denominated in US dollars, limiting direct transaction-currency exposure relative to the apparent geographic percentage. [S1] [S2]

The fabless model transfers rather than eliminates capital commitments. GlobalFoundries and TSMC are important wafer suppliers, while third parties perform assembly and testing. Cirrus gains access to specialized processes without building multibillion-dollar fabs and can use different technologies for audio, power, or sensing. In exchange, it accepts supplier concentration, qualification risk, capacity reservation requirements, and limited direct control over yield and delivery. The June 30, 2026 agreement with GlobalFoundries creates defined quarterly wafer purchases estimated at at least $600 million for calendar 2027–2028. Publicly disclosed customer arrangements do not provide symmetrical minimum-volume protection. [S2]

Unrecognized economic assets include analog design libraries, software, algorithms, process-development knowledge, qualification history, customer-specific systems knowledge, and the accumulated skills of 1,668 employees. FY2026 R&D of $434.0 million was expensed and was net of a $42.9 million UK research credit. Immediate expensing is conservative for balance-sheet recognition, but it means reported book capital omits much of the investment supporting current revenue. These assets are perishable: employee turnover, lost sockets, process changes, or reduced customer relevance can impair them without a formal accounting charge. [S1]

Recognized goodwill was $435.9 million at June 2026, while remaining acquired intangibles were $19.2 million. Goodwill is principally associated with acquisitions such as Wolfson and Lion. Its continued balance does not prove those acquisitions earned their required return. Conversely, internally developed intellectual property is largely absent from the balance sheet. A sound return analysis therefore includes acquired consideration and capitalized research without treating either accounting goodwill or management’s pipeline descriptions as evidence of value.

Cirrus is a Delaware corporation whose ordinary common stock trades on the Nasdaq Global Select Market. It is not an ADR, partnership, MLP, pass-through security, or K-1 issuer. [S1] [S2]

Diligence conclusion — revenue stability: Annual revenue remained within approximately $1.78–$2.00 billion during FY2022–FY2026, but 90% customer concentration, cancellable orders, seasonality and redesign risk make the economic stream substantially less stable than the annual range implies. [S1] [S2]

Diligence conclusion — business understandability: The unit, content, price, margin, R&D and working-capital mechanisms are understandable, although undisclosed customer content and program economics prevent high-precision forecasts. [S1]

Diligence conclusion — unrecognized assets: The principal unrecognized assets are expensed mixed-signal research, engineering know-how, algorithms, qualification history and customer-development knowledge, all of which can lose value when design wins or pricing deteriorate. [S1]

Diligence conclusion — security status: CRUS is ordinary US common stock and is not an ADR, partnership, MLP or K-1 issuer. [S2]

Verdict: Cirrus has an attractive intellectual-property-led model with little owned manufacturing capital, but its apparent revenue recurrence is earned anew through design execution. The business is understandable at the mechanism level and unusually opaque at the socket, price, and customer-profit level.

Industry Dynamics

Cirrus participates in differentiated analog and mixed-signal semiconductors rather than one homogeneous market. Its serviceable opportunity consists of audio, camera-control, power, haptic, sensing, and precision-conversion sockets in smartphones, tablets, laptops, wearables, accessories, professional audio, automotive, and industrial applications. Total device-market value is a poor addressable-market proxy. What matters is the dollar content Cirrus can win in functions where power efficiency, heat, latency, noise, signal accuracy, software integration, size, or reliability justify differentiated silicon. [S1]

Profit pools are highly uneven. Commodity components compete mainly on price, availability, and manufacturing scale. Customized mixed-signal products can earn better margins when they eliminate board components, improve battery life, protect speakers, accelerate camera response, or enable a visible feature. Cirrus’s 52.8% trailing gross margin and 23.0% operating margin indicate participation in a differentiated pool. Those margins are well above current Skyworks levels but below Analog Devices’ 65.8% gross margin and Texas Instruments’ 58.3% gross margin. Diversified analog leaders distribute R&D, manufacturing, selling, and channel costs across thousands of products and customers; Cirrus’s concentrated customer possesses much greater negotiating leverage. [S1] [S12]

The filing identifies AKM, Analog Devices, Qualcomm, Realtek, Renesas, Awinic, Goodix, Skyworks, STMicroelectronics, Synaptics, and Texas Instruments among relevant competitors. This is not one stable league table. Qualcomm can integrate audio or power functionality into a broader platform. Realtek and Asian mixed-signal suppliers can compete aggressively on cost. ADI and TI possess broad catalogs, manufacturing assets, and channel reach. Apple itself is a potential competitor because it can internalize functions or sponsor alternatives. The most consequential competitive event may therefore be removal or integration of a merchant socket rather than a conventional share shift between component suppliers. [S1]

Barriers to entry are meaningful but permeable. Advanced analog engineering requires scarce talent, process-specific device models, intellectual property, system knowledge, software, packaging expertise, and repeated mass-production qualification. A supplier must demonstrate performance across voltage, temperature, noise, electromagnetic interference, physical variation, and yield. Smartphone launch schedules penalize delays heavily, increasing the value of reliable execution. Customer-specific knowledge and foundry co-development can shorten future programs. These barriers explain repeated design participation and margins above commodity levels. They cannot prevent a well-funded OEM or established semiconductor vendor from redesigning a function.

Switching costs have a clock. Within a qualified device generation, changing a codec, amplifier, or controller can require board changes, firmware integration, acoustic tuning, reliability work, supply-chain validation, and yield requalification. Switching cost is therefore real during production. At the next architecture cycle, however, the OEM can compare suppliers, integrate functions, or redesign the system. Cirrus’s moat is a repeated-game advantage: past execution improves the probability of the next award but does not turn the relationship into a permanent annuity.

The demand backdrop deteriorated during 2026. Gartner forecast global 2026 PC shipments down 10.4% and smartphone shipments down 8.4%, driven by higher memory costs, device-price increases, and longer replacement cycles. Gartner expected premium products to be less affected than entry-level devices, a partial offset because Cirrus is skewed toward premium systems. IDC forecast an 11.3% PC decline, worsening conditions through the fourth quarter and no meaningful memory relief before the end of 2027. These are independent estimates, not observed outcomes, but they contradict the idea of a rapid broad-based device recovery. [S10] [S11]

Cirrus can grow in a shrinking unit market if content per device rises faster than unit decline and pricing pressure. Adding a camera controller, extra amplifier, codec, or power component can raise dollars per device. This is the principal bull mechanism and helps explain why HPMS grew despite end-market concerns. The opposing mechanism is that OEM price pressure, function integration, model delays, or lower premium-device units absorb the content gain. Unit forecasts must therefore be paired with company shipment and content evidence.

The PC opportunity illustrates the difference between total market size and serviceable content. Global annual PC shipments are measured in hundreds of millions, but management described FY2026 Cirrus PC revenue only as being in the $40 million range, approximately 2% of company sales. Management has said a new business must plausibly become about 10% of company revenue to merit pursuit, implying roughly $200 million at current scale. Shipping to the top six laptop vendors and increasing SDCA adoption are meaningful indicators of engagement, but a large end market does not make a small supplier position economically material by itself. [S6]

Regulation affects Cirrus mainly through trade, export controls, taxation, environmental material restrictions, and semiconductor-manufacturing incentives rather than product-price regulation. Domestic manufacturing initiatives can support new process capacity and supply-chain resilience. They do not guarantee customer demand or supplier margins. The Apple–GlobalFoundries collaboration validates a technology and supply relationship, while the binding economic evidence for Cirrus is its wafer-purchase commitment, not the partnership label. [S2] [S9]

The foreign low-cost threat also requires precision. Cheap assembly labor is not the core threat because Cirrus already outsources assembly and fabrication globally. Lower-cost analog and mixed-signal designers willing to accept lower margins are more relevant, especially in Android devices. Customer-sponsored alternatives and system integration can be still more dangerous. Specialized engineering, software, and qualification slow replacement but do not eliminate price competition. Management has acknowledged severe pricing in parts of the Android market and has concentrated resources where differentiation appears more defensible. [S1] [S7]

The supply-side capital cycle creates both advantage and risk. Cirrus can pursue growth without constructing a leading-edge fab and can use mature or specialized process nodes where analog performance matters more than transistor density. However, foundries gain bargaining power during tight capacity, and a fabless supplier may reserve production before customer demand becomes firm. The new $600 million minimum estimated GlobalFoundries purchasing program equals roughly 30% of one year’s current revenue spread across two calendar years. It appears manageable relative to liquidity and projected cost of sales, but shortfall provisions create risk if customer programs slip, shrink, or are redesigned. [S2]

An additional capital-cycle issue is pricing rigidity. The 2027–2028 agreement fixes a framework for wafer pricing as well as quantities. If spot or market wafer prices fall materially below the contracted economics, Cirrus may be disadvantaged relative to competitors without comparable commitments. If the market tightens, the capacity reservation may instead protect delivery and cost. The agreement is therefore both insurance and an operating bet on future utilization. [S2]

Industry competition is likely becoming more demanding. OEM silicon internalization, aggressive Asian pricing, architecture integration, weak device units, and longer replacement cycles increase pressure. Offsetting those forces, system complexity and qualification requirements favor capable incumbents, and supply-chain localization may deepen existing technical partnerships. The net direction is not the disappearance of industry profit, but a widening gap between differentiated, repeatedly qualified sockets and functions that can be integrated or commoditized.

Diligence conclusion — market growth and geography: Cirrus serves large global smartphone and PC markets forecast to contract materially in 2026, while its 99% international ship-to mix mainly reflects manufacturing geography rather than ultimate consumer demand. [S1] [S10] [S11]

Diligence conclusion — competition direction: Competition is becoming more demanding through customer insourcing, architecture integration, aggressive Asian pricing and weaker device units, even as system complexity and qualification protect capable incumbents. [S1] [S2]

Diligence conclusion — industry profitability and barriers: Differentiated mixed-signal sockets can support attractive margins, but sustainable profit requires scarce analog talent, process IP, software, qualification, system co-design and repeated roadmap execution rather than patents alone. [S1] [S12]

Diligence conclusion — foreign low-cost threat: Low-cost assembly labor is not the main moat threat because Cirrus is already fabless; lower-margin Asian design competitors, integrated platforms and customer-sponsored alternatives are more consequential. [S1] [S7]

Verdict: The industry rewards differentiated engineering but does not provide permanent exclusivity. Cirrus benefits from a capital-light position in attractive sockets while accepting customer power, foundry commitments, and recurring redesign risk. Content growth can offset weak units, but only realized revenue and margins establish that it has done so.

Competitive Position

Cirrus’s brand has little direct consumer pull. A phone buyer rarely selects a device because it contains a Cirrus codec or controller. The brand matters inside OEM engineering, quality, procurement, and supply-chain organizations as a record of performance and execution. Its economic value must therefore appear indirectly—in repeated awards, qualification success, product longevity, pricing, margins, and customer breadth. [S1]

The competitive mechanism has five layers. First, precision analog design is difficult because circuits must operate across process variation, voltage, temperature, noise, and physical conditions. Second, Cirrus combines silicon, DSP, and algorithms, permitting system-level optimization rather than sale of an isolated component. Third, long custom programs build customer-specific knowledge. Fourth, qualification and software integration create switching friction inside a product generation. Fifth, continuing R&D funds the next road map. FY2026 R&D was approximately 22% of revenue before recognizing that the reported expense was net of a $42.9 million research credit. The moat is maintained through recurring expenditure, not inherited for free. [S1]

Financial outcomes support a genuine but bounded advantage. Gross margin rose from 50.4% in FY2023 to 52.8% in FY2026, while operating margin rose from 18.2% to 23.0%. Revenue recovered from its FY2024 trough without a proportional expense increase, and Company Financials calculates a trailing incremental operating margin near 30%. A vendor selling undifferentiated components under overwhelming customer power would be less likely to sustain those results. The qualification is that one customer still receives much of the created value through purchasing leverage, and aggregate margins cannot identify which products generated the improvement. [S1] [S12]

Apple’s March 2026 announcement is the strongest external validation. Apple identified Cirrus and GlobalFoundries as partners developing new semiconductor process technology at Malta, New York and said the work enables Cirrus to develop mixed-signal solutions for several Apple applications, including advanced ICs that power Face ID. This evidence is stronger than supplier-only pipeline commentary because it comes from the customer. It confirms relationship depth and relevance beyond legacy audio. It does not establish production timing, volume, price, exclusivity, minimum demand, or Cirrus’s share of Apple’s broader announced spending. [S9]

The relationship is simultaneously the moat’s best evidence and its greatest limitation. A long history of supplying a demanding OEM, combined with multiple functions and process co-development, makes arbitrary displacement less likely. Yet 91% concentration gives Apple extraordinary negotiating leverage. Apple can benchmark Cirrus against alternatives, require price reductions, dual-source, integrate functions, or develop components internally. More sockets at Apple can deepen technical integration while increasing the dollar amount exposed to one architecture decision.

Competition is socket-by-socket. In audio, vendors compete on sound quality, speaker protection, power consumption, latency, package size, integration, and software. Camera-control competition turns on precision, response, stabilization, system integration, and yield. Power components compete on efficiency, heat, size, voltage handling, and control accuracy. PC products compete on standards support, voice performance, reference-design inclusion, driver reliability, and OEM qualification. Price matters in every category but is rarely the only criterion in a flagship design.

The peer set should reflect different parts of this mechanism. Skyworks is a useful mobile-content and concentration analogue, although it is more exposed to RF and owned manufacturing. Its trailing gross margin was 40.7%, operating margin 8.7%, and ROIC approximately 4.1%, making its elevated GAAP P/E a consequence of depressed earnings rather than superior quality. Qualcomm competes through broader mobile platforms and integration; it had a 54.2% gross margin, 23.3% operating margin, and approximately 19.5% ROIC. Analog Devices and Texas Instruments are quality benchmarks rather than close replicas. Their catalog breadth and customer diversity support 65.8% and 58.3% gross margins and substantially higher valuation multiples. [S12]

Synaptics, Realtek, AKM, Awinic, Goodix, Renesas, and STMicroelectronics are relevant in individual functions, although simple consolidated comparisons can mislead. A broad analog supplier may accept a lower margin on one socket because it earns value elsewhere; a platform supplier may integrate the function; a regional competitor may prioritize share over margin. Cirrus’s above-Skyworks margin demonstrates differentiation but does not establish immunity from any of these strategies.

Switching costs are meaningful within production and weaker at redesign. A mid-generation supplier switch can impose board changes, software work, acoustic retuning, validation, and launch risk. At the next generation, those same activities become part of a planned redesign and shrink relative to the total program budget. Management’s claim that current custom codecs and amplifiers should have extended lifecycles is encouraging because it reduces near-term redesign frequency, but it remains a claim until sustained shipments and margins confirm it. [S5] [S6]

Cirrus’s PC position shows technical relevance without demonstrating scale. Management described approximately $40 million of FY2026 PC revenue, shipments to the top six laptop vendors, SDCA products near 60% of PC revenue, and an expectation that SDCA would approach 80% in FY2027. It also said more than half of expected FY2027 PC revenue would relate to mainstream products. These facts and claims establish wider engagement, but $40 million remains only about 2% of company revenue. The value of the franchise would improve materially if the PC position became large enough to affect customer concentration and contribution profit. [S6]

Brand relevance can therefore be tested rather than asserted. If the engineering reputation matters, Cirrus should retain multi-generation designs, protect gross margin, win adjacent functions, and broaden production customers. If it does not, revenue will weaken around redesigns and pricing will compress. If the brand were an impregnable moat, rising customer concentration and disclosed insourcing risk would be far less important.

Diligence conclusion — brand relevance: The Cirrus brand matters economically to OEM engineers, quality teams and procurement through execution credibility and qualification history, but it has little consumer pull and cannot override customer bargaining power. [S1]

Diligence conclusion — nature of competition: Cirrus competes on system performance, integration, power, latency, software, yield, reliability, roadmap execution and price at each device redesign. [S1]

Diligence conclusion — switching costs: Switching costs are meaningful within a qualified product generation but reset substantially at the next architecture cycle, making the moat renewable rather than permanent. [S1] [S2]

Verdict: Cirrus possesses a real engineering and qualification advantage supported by margins, repeated programs, and customer validation. Its moat is best described as a renewable design-win advantage. It raises the probability of future wins but supplies neither contractual protection nor independence from one customer.

Growth History and Forward Opportunities

The five-year record is stronger in earnings than revenue. FY2022–FY2026 revenue compounded at approximately 2.9%, while operating margin expanded and diluted shares declined. GAAP diluted EPS rose from $5.52 in FY2022 to $7.85 in FY2026, despite falling to $3.09 in FY2023. Content, product mix, tax, operating discipline, and repurchases contributed more than broad end-market unit growth. [S1] [S4]

The first growth pillar is the existing smartphone-audio franchise. Current boosted amplifiers and a 22-nanometer smart codec are in production, and management expects extended lifecycles. Audio grew only 2% in FY2026 and 3.7% in Q1 FY2027, consistent with a mature cash-generating business rather than an unlimited growth market. Content gains, new generations, and additional amplification can support modest growth, but annual pricing and architecture changes remain offsets. [S1] [S2] [S6]

The second pillar is HPMS content in smartphones and adjacent devices. Camera controllers are already material enough to influence results, and 26% Q1 HPMS growth confirms current shipment momentum. Power, haptic, sensing, and accessory products provide further possibilities. Apple’s Face ID reference increases confidence that Cirrus technology is relevant to a strategically important application. It does not establish a commercial production ramp. [S2] [S9]

Management described the HPMS pipeline as exceptionally strong. That is a claim, not reported revenue. Its reliability declines as the forecast horizon expands. Shipping camera controllers belong in the base case. A smart-power IC in design for a 3D-sensing application deserves option value because management indicated that introduction remained approximately two years away. Production awards, product-category revenue, and customer-specific economics would be stronger evidence than tape-outs or partnership language. [S6] [S7]

PCs are the most tangible route to customer diversification and the clearest recent disappointment. Management described FY2025 PC revenue as low tens of millions and FY2026 revenue as in the $40 million range. It reported shipments to the top six laptop vendors and growing SDCA penetration. Those are measurable advances. In May, however, management expected strong FY2027 growth even with market softness; by August it had reduced expectations because a key platform was supply constrained, memory and other components were expensive or scarce, and OEM models were delayed. [S6] [S7]

Management maintained that customer interest and design activity were healthy, including voice-codec projects and PCs based on an NVIDIA platform, and characterized the setback mainly as timing. IDC’s expectation of stress through 2027 is a credible contradiction: individual sockets can still ramp, but the external bottleneck may last longer than management’s initial framing implied. The clean test is production revenue, not engagement count. A credible path to $100 million would establish more than experimentation; approximately $200 million would begin to meet management’s conceptual relevance threshold. [S6] [S7] [S11]

General-market opportunities include professional-audio converters, automotive products, industrial imaging, timing devices, and a high-performance analog front end for smart-meter metrology. The metrology device had taped out and was expected to sample during the September 2026 quarter. Management discussed potential calendar-2028 deployment and extensions into data-center DC metrology, storage, vehicles, charging, and grid applications. It also acknowledged that it was too early to quantify the addressable market. Tape-out establishes engineering progress; sampling establishes product availability; neither establishes paid adoption or return on research capital. [S7] [S8]

These markets could improve durability because industrial and professional products often have longer lifecycles and broader customers. They also require channels, field support, application expertise, and qualification models different from Cirrus’s concentrated mobile strength. The company may develop those capabilities, but investors should not assume that success with one deeply integrated customer transfers automatically to fragmented markets.

The GlobalFoundries relationship may enable specialized high-voltage mixed-signal processes and geographic supply diversification. It also raises the cost of delay: defined wafer purchases arrive before public evidence of associated customer demand. Capacity should therefore be treated as an operating option purchased with a binding commitment, not a free catalyst. [S2] [S9]

The catalyst sequence should be staged. Near-term evidence is Q2 revenue, product mix, gross margin after the favorable wafer tranche, PC shipment commentary, inventory conversion, and repurchase activity. Medium-term evidence is non-Apple revenue growth, multiple quarters of PC production, qualification of smart-power devices, and disclosure that foundry quantities are being consumed. Long-dated metrology and industrial opportunities deserve increasing value only after sampling, customer qualification, production, and cash contribution occur in sequence.

Diligence conclusion — product outlook: Shipping smartphone audio and camera products support the near-term outlook; PCs offer a measurable but delayed diversification path; and Face ID power, metrology and other adjacencies remain staged options until production revenue is disclosed. [S2] [S6] [S7]

Verdict: The pipeline is technically credible and broader than legacy audio, but evidence quality declines sharply from current smartphone shipments to PC designs, power development, and metrology samples. The growth case should recognize milestones sequentially rather than capitalize the ultimate market at announcement.

Financial Quality

Cirrus combines modest top-line growth, strong margin expansion, low physical capital intensity, and excellent recent cash generation. The five-year filing-reconciled record is:

The income-statement and cash-flow values are reported facts. Free cash flow is operating cash flow less property, equipment, software, and technology investment, consistent with the company’s presentation. [S1] [S4] [S5]

FY2026 was a company earnings high, not a cyclical trough. Revenue rose 5.3%, gross margin reached 52.8%, operating income was $460.4 million, and EPS rose 30.7%. Q1 FY2027 remained strong: revenue rose 12.9%, operating income rose 16.6%, net income rose 26.6%, and diluted EPS rose 28.9%. Q1 operating margin was 18.4%, below the FY2026 average because of seasonality and higher R&D and SG&A. [S1] [S2]

The cycle position is unusual. Company earnings and content were near highs while external PC and smartphone forecasts weakened. A premium-device supplier can reach record earnings as the unit cycle turns down if content, mix, and pricing offset units. Normalization should therefore stress company-specific variables rather than assume either an automatic rebound or a conventional semiconductor peak collapse.

GAAP and non-GAAP results diverge materially. FY2026 non-GAAP operating income was $548.8 million, or 27.5% of revenue, compared with GAAP operating income of $460.4 million, or 23.0%. Non-GAAP EPS was $9.26 versus GAAP EPS of $7.85. The main exclusions were $81.8 million of recurring stock compensation and $6.6 million of acquired-intangible amortization. Acquisition amortization is non-cash in the current period, but the purchase price remains invested capital. Stock compensation is a recurring labor cost and potential claim on shareholders even when repurchases offset issuance. [S5]

Company Financials calculates trailing standardized ROIC at approximately 18.2%, ROA at 17.7%, and ROE at 20.9%. These are attractive returns, but standardized ROIC does not capture the accumulated research asset. A research-capitalization sensitivity using four- and five-year straight-line useful lives and a year-end cohort convention produces estimated unamortized R&D assets of roughly $1.1 billion and $1.4 billion. Adding the research asset to implied invested capital and adjusting operating profit for current research less estimated amortization reduces economic ROIC to approximately 11–13%. This is an analyst estimate dependent on useful life, timing convention, attrition, taxes, and treatment of the UK research credit—not a reported metric. [S1] [S12]

The exercise prevents the fabless model from making engineering investment appear free. A half-year timing convention would produce a smaller research asset and somewhat higher return; a longer useful life would increase the asset and generally lower the calculated return. The appropriate comparison with ADI, TXN, or Qualcomm must use a consistent research-capital convention. The result is not that Cirrus destroys value, but that its economic return is closer to a low-teens research-adjusted return than a costless-IP windfall.

Physical capital intensity is low. FY2026 property, software, and technology investment was less than 1% of revenue. Economic investment is much higher because R&D was approximately 22% of sales and gross expenditure before the UK credit was higher still. Q1 capital spending increased as Cirrus purchased test equipment to place at outsourced providers. Management argues this gives flexibility and better returns than owning fabrication. That claim should be tested through utilization, incremental margins, and cash returns rather than accepted from the asset-light label. [S1] [S2] [S8]

Cash conversion was excellent in FY2026 but included a working-capital release. Operating cash flow of $650.6 million was 1.57 times net income. Management attributed $103.4 million of favorable working-capital movement principally to lower inventory and prepaid wafer balances and higher accounts payable, partly offset elsewhere. Inventory declined from $299.1 million to $240.9 million. That release is real but not indefinitely repeatable. Through Q1 FY2027, trailing operating cash flow was approximately $599 million and trailing free cash flow approximately $572 million. [S1] [S2] [S12]

Q1 illustrates seasonal reversal. Net income was $76.9 million, while operating cash flow was $64.1 million compared with $116.1 million in the prior-year quarter. Accounts receivable increased by $33.8 million and inventory used $21.9 million as the company prepared for the September-quarter ramp. Accounts receivable ended at $254.0 million and inventory at $262.7 million. If balances convert during launch shipments, earnings quality remains intact. If inventory continues to rise, commitments are not matched by sales, or reserve charges increase, the interpretation becomes adverse. [S2]

Inventory is stated at the lower of cost or net realizable value and depends on management demand forecasts. Long lead times, proprietary products, and cancellable customer orders make obsolescence a real economic risk. Excess and obsolete inventory charges, including scrapped inventory, were $8.0 million in FY2026 versus $4.2 million in FY2025. That amount was not large relative to gross profit, but it demonstrates that proprietary inventory can lose value before the larger 2027–2028 foundry commitment begins. [S1]

Liquidity is excellent. At June 27, 2026, cash was $810.7 million, current marketable securities $80.6 million, and long-term securities $276.1 million, totaling $1.167 billion. Current and non-current lease liabilities totaled $130.5 million. There were no funded borrowings, and the amended $350 million revolving facility was undrawn. Company Financials categorizes leases as debt in its enterprise-value output; the report’s “no funded borrowings” description deliberately distinguishes leases from bank or bond debt. [S2] [S12]

Economic obligations are more substantial than the conventional debt figure. FY2026 unconditional purchase commitments were $354.2 million, including $322.2 million due in FY2027. The June 30 GlobalFoundries agreement subsequently created defined quarterly purchases estimated at at least $600 million across calendar 2027 and 2028, subject to shortfall provisions. The new amount should not simply be added to the March total because timing and overlap are not fully disclosed, but it makes the March commitment table an incomplete forward picture. Leases, software arrangements, owned test equipment, and the disputed IRS assessment are additional claims. [S1] [S2]

The IRS dispute requires balanced treatment. The asserted $168.3 million of additional tax and $63.7 million of penalties excludes interest and has not been accrued as an additional liability because Cirrus disputes the positions and believes its ASC 740 reserves are adequate. The company entered IRS Appeals in July 2024 and is prepared to litigate if necessary. The assertion is not an adjudicated obligation, but an adverse outcome could materially reduce liquidity and raise the normalized tax burden. [S2]

Accounting is conservative in expensing research and recognizing stock compensation under GAAP. It becomes less conservative when investors use non-GAAP earnings excluding recurring SBC, calculate returns without research capital, or treat working-capital release as permanent FCF. The $42.9 million research credit properly reduces accounting expense but should be identified when comparing engineering intensity. No material accounting-policy change was reported in Q1 FY2027; estimates for inventory, goodwill, useful lives, taxes, and research credits remain judgmental without being formal policy changes. [S1] [S2]

Diligence conclusion — earnings cycle: FY2026 company earnings were near a high rather than a trough even as external device forecasts weakened, so normalization must stress content, price, mix and R&D rather than assume an automatic cyclical recovery. [S1] [S10] [S11]

Diligence conclusion — business profitability: Standardized trailing ROIC was approximately 18.2%, while capitalizing four to five years of research reduces estimated economic ROIC to roughly 11–13%, still attractive but materially below unadjusted screens. [S1] [S12]

Diligence conclusion — off-balance-sheet obligations: Leases, software commitments and foundry purchases are the main non-funded-debt claims, including defined wafer purchases estimated at at least $600 million during calendar 2027–2028; the separately disputed IRS assessment adds a material unaccrued tax contingency. [S1] [S2]

Diligence conclusion — accounting conservatism: Immediate research expensing is conservative, while excluding recurring SBC, overlooking the research credit or omitting research capital and working-capital normalization would overstate owner economics. [S1] [S5]

Diligence conclusion — capital intensity: Cirrus is physically capital-light but economically research-intensive, with FY2026 capital spending below 1% of revenue and reported R&D near 22% before considering its $42.9 million research credit. [S1]

Diligence conclusion — income and cash divergence: FY2026 cash flow exceeded income partly through working-capital release, whereas Q1 FY2027 conversion weakened as receivables and inventory built ahead of the seasonal ramp. [S1] [S2]

Verdict: Financial quality is high but not effortless. Strong margins, net cash, genuine free cash flow, and double-digit research-adjusted returns support franchise value. Conservative analysis must count SBC, capitalize research, normalize FY2026 working capital, treat the foundry agreement as operating leverage, and retain the disputed tax assessment as a contingent rather than settled claim.

Capital Allocation

Management’s hierarchy is organic investment, selective acquisitions, and share repurchases; it does not plan a near-term dividend. The order is reasonable for a research-intensive company with volatile customer programs, provided R&D produces paid designs, acquisitions earn returns, and repurchases occur below intrinsic value. [S7] [S8]

Organic investment is the largest use. FY2026 reported R&D was $434.0 million and Q1 FY2027 R&D increased 11.8% to $115.0 million. The company is funding smartphone, PC, power, process-development, and general-market programs while device units are weak. Countercyclical investment can create value if competitors retrench, but pipeline breadth is not proof of return. The scorecard is future revenue, contribution margin, and customer diversification relative to cumulative research and capacity commitments. [S1] [S2]

The largest strategic capital decision now extends beyond R&D expense. The company has committed to defined quarterly GlobalFoundries wafer quantities through 2028, is purchasing test equipment for outsourced providers, and is advancing specialized process technology. These actions may protect supply and enable differentiated products, but they also make future free cash flow more sensitive to utilization. Management’s confidence is revealed through committed resources; shareholder value still depends on customer demand absorbing those resources. [S2] [S8]

The most important recent acquisition was Lion Semiconductor in 2021. The announcement described $335 million of cash consideration, while subsequent purchase accounting reported approximately $280.5 million of consideration transferred after acquired cash and accounting adjustments. Lion supplied switched-capacitor power technology that remains strategically relevant to the smart-power road map. Cirrus does not disclose Lion-derived revenue, profit, ongoing R&D, or invested capital, so a realized return cannot be calculated. [S14]

HPMS growth is not sufficient attribution. HPMS increased from $758.9 million in FY2025 to $837.4 million in FY2026 and grew 26% in Q1 FY2027, but that category also includes camera, haptics, sensing, and other products. Goodwill remains $435.9 million without impairment. The defensible conclusion is that Lion technology remains active, not that the acquisition has earned an attractive return.

Repurchases have produced real net share retirement. Period-end shares declined from 56.60 million in FY2022 to 50.60 million in FY2026, a 10.6% reduction, and shares outstanding were 50.12 million on August 3, 2026. Diluted weighted-average shares also declined. This is stronger evidence than an authorization announcement because it incorporates employee issuance and withholding. Repurchases were funded from internally generated cash while the balance sheet remained free of funded borrowings. [S1] [S2]

Price discipline is mixed. During FY2026, Cirrus repurchased approximately 2.5 million shares for $280.0 million at an average $113.91, slightly below the report-date price. In Q1 FY2027 it bought approximately 211,000 shares for $34.5 million at an average $163.43. After quarter-end it purchased approximately 359,000 shares for $50.5 million at $140.53 under a trading plan. The latter prices were about 40% and 21% above the September 10 close. Net retirement created per-share accretion, but recent execution does not demonstrate reliable market-timing skill. [S1] [S2]

Stock compensation remains economically material. FY2026 expense was $81.8 million, 4.1% of revenue. At March 28, approximately 2.83 million shares underlay outstanding awards at target, including roughly 2.40 million full-value awards. By May 7, the proxy’s overhang calculation used 3.02 million outstanding awards and 1.55 million shares available for future issuance. Stockholders subsequently approved an additional 3.5 million shares for the long-term incentive plan. The proxy estimated overhang would rise from 8% to 14% if the additional reserve were included. Authorization is not inevitable dilution, but the capacity is material. The disclosed three-year average burn rate was 1.73% annually. [S3] [S17]

Executive cash incentives use non-GAAP operating margin and a revenue-growth multiplier. The non-GAAP measure excludes stock compensation and acquired-intangible amortization. Long-term awards include restricted stock, market stock units tied to relative total shareholder return, and strategic-revenue PSUs. FY2026 strategic revenue of $186.5 million versus a $210 million target produced a 72.5% payout. Relative-return awards can align executives with shareholders but may pay when absolute returns are poor; non-GAAP margin can reward performance before the cost of equity compensation. [S3]

CEO John Forsyth received approximately $10.77 million of reported FY2026 compensation, and the CEO pay ratio was 61.4 times the median employee. The CEO ownership guideline is the lesser of three times base salary or 60,000 shares. That formulation creates a materially weaker requirement than a “greater of” test would. Non-employee directors receive cash retainers and annual equity awards subject to an aggregate annual cap. [S3]

Recent insider filings show grants, vesting, tax withholding, option exercise, and planned sales rather than open-market purchases. Forsyth sold 5,000 shares at a weighted-average $170.41 on May 8 under a Rule 10b5-1 plan adopted in August 2025. R&D executive Jeffrey Baumgartner exercised and sold 3,907 shares at $168.54 in June under a plan adopted in February 2026. Planned transactions carry less negative information than discretionary sales, but no recent code-P open-market purchase was identified in the reviewed filing set. [S15] [S16]

A limited management-role change occurred in March 2026: CFO Jeffrey Woolard also became principal accounting officer, while the prior principal accounting officer continued in a senior role; the filing disclosed no disagreement. This does not alter the operating thesis but should not be described as no leadership change. [S18]

Diligence conclusion — free cash flow generation and use: Cirrus generated $635.8 million of FY2026 free cash flow and used financial capacity principally for research, repurchases and liquidity, with no dividend and no funded-debt burden. [S1] [S5]

Diligence conclusion — acquisition record: Lion supplied strategically relevant power technology, but absent standalone revenue, profit and continuing-investment disclosure, its realized return cannot be verified. [S14]

Diligence conclusion — share repurchases: Buybacks reduced period-end shares by 10.6% from FY2022 to FY2026, although recent purchases at $163.43 and $140.53 were materially above the current market price. [S1] [S2]

Diligence conclusion — insider share issuance: Outstanding awards were material and the approved plan gained 3.5 million additional shares, although historical repurchases have more than offset actual issuance. [S3] [S17]

Diligence conclusion — compensation policy: Executive incentives combine non-GAAP margin and revenue growth with restricted stock, relative-return awards and strategic-revenue PSUs, while directors receive cash and annual equity. [S3]

Diligence conclusion — management motivations: Incentives emphasize growth, margin, strategic revenue and relative shareholder return, while high-priced repurchases, planned insider sales and the use of ex-SBC margin argue against assuming perfect alignment or price discipline. [S2] [S3] [S15]

Diligence conclusion — dividend policy: Cirrus pays no dividend and prioritizes organic research, selective acquisitions and repurchases; cash coverage is ample, so the absence is a policy choice rather than a financing constraint. [S7] [S12]

Verdict: Capital allocation has produced real per-share shrinkage and preserved exceptional liquidity, but it is not flawless. Research deserves priority only when it converts, Lion’s return remains unproven, the equity-plan reserve is material, and recent repurchase prices expose mechanical execution risk.

Changes and Headwinds — Last Two Years

The last two years transformed Cirrus from a recovering audio supplier into a record-earning company with a broader product road map, but not a broader customer base. Revenue recovered from $1.789 billion in FY2024 to $1.896 billion in FY2025 and $1.997 billion in FY2026. Gross margin rose to 52.8%, operating margin reached 23.0%, and period-end shares fell. Internal contributors included smartphone content, camera products, PC designs, cost discipline, and repurchases. [S1] [S4]

The largest strategic change was expansion of HPMS. Camera and power engagement deepened, and Apple publicly confirmed joint process work and Face ID power applications. That moved the evidence beyond Cirrus’s own pipeline narrative. Simultaneously, Apple concentration rose from 87% to 91%. Product breadth therefore strengthened current economics while increasing the amount of revenue correlated to one customer. [S1] [S9]

PC strategy progressed and then encountered a material headwind. Cirrus reached the top six laptop vendors and approximately $40 million of FY2026 revenue, but reduced FY2027 expectations between May and August. Some stated causes were external—memory inflation, component scarcity, platform constraints, and model delays. Internal design execution may remain intact, but management did not quantify the revenue removed or separate delayed from canceled programs. Realized shipments must resolve that distinction. [S6] [S7]

The broader device environment also weakened. Gartner and IDC forecast sharp 2026 contractions, higher device prices, and extended replacement cycles. Premium products may prove more resilient, but declining units can alter OEM launch timing, feature configurations, and price negotiations. External weakness is therefore not merely a multiple issue; it can interact directly with Cirrus content and inventory. [S10] [S11]

Supply strategy changed materially. Cirrus joined Apple’s manufacturing initiative, continued process development with GlobalFoundries, and entered the June 30 wafer agreement for 2027–2028. The arrangement may improve geographic resilience and specialized process access, but it introduces quarterly purchase and shortfall exposure before public evidence of matched customer commitments. Apple’s aggregate manufacturing-program announcement was not a commitment to spend that amount with Cirrus. [S2] [S9]

Facilities and capital deployment are evolving toward Cirrus-owned test equipment installed at outsourced assembly and test providers. This raises near-term capital spending but may improve flexibility. Cirrus remains fabless and has not announced construction of a wafer fab. Headcount was 1,668 at FY2026 year-end, with investment focused on engineering. [S1] [S8]

Senior leadership was broadly stable, although the CFO assumed the additional principal-accounting-officer role in March 2026. No disagreement was disclosed. The more consequential managerial choice is higher research spending during weak end markets and alongside future wafer commitments. That can be countercyclical investment or pre-revenue cost accumulation; subsequent shipments will distinguish them. [S2] [S18]

Tax treatment changed through 2025 legislation restoring immediate US research deductions beginning in FY2026. Together with jurisdictional mix, this contributed to a 16.6% FY2026 effective tax rate versus 25.5% in FY2025. Q1 FY2027’s effective rate was 18.4% versus 24.7% a year earlier. The lower tax rate improves cash and reported earnings but is not evidence of better product economics. Separately, the FY2017–FY2019 transfer-pricing dispute remains in IRS Appeals and can produce a material future cash claim despite the current lower tax rate. [S1] [S2]

No significant accounting-policy change was reported in Q1 FY2027. Core comparability remains intact, although inventory reserves, goodwill tests, useful lives, research credits, uncertain tax positions, and non-GAAP exclusions continue to require judgment. [S2]

Diligence conclusion — external and internal drivers: Recent gains came from internal content, mix, cost discipline and share retirement, while current PC pressure is substantially external; only production revenue can distinguish a temporary delay from lost demand. [S1] [S7] [S11]

Diligence conclusion — environment changes: Record company earnings now coexist with weaker device forecasts, memory inflation, delayed PC platforms, domestic foundry collaboration and materially larger forward wafer commitments. [S1] [S2] [S10]

Diligence conclusion — accounting policy changes: The latest quarterly filing reported no significant accounting-policy change, although estimates, research credits, uncertain tax positions and non-GAAP presentation still require scrutiny. [S2]

Diligence conclusion — markets, facilities and management: Device markets weakened, foundry and outsourced-test arrangements expanded, engineering investment increased, and the CFO added the principal-accounting-officer role while core operating leadership remained stable. [S1] [S8] [S18]

Verdict: Internal execution improved materially while the external environment and contractual supply exposure deteriorated. The key unresolved contradiction is strong design engagement versus reduced near-term PC shipments and binding future wafer purchases; the tax dispute adds a separate contingent claim without undermining current operating performance.

Risk Analysis

The central risk is customer concentration. A 10% reduction in Apple purchases would reduce total company revenue by approximately 9% before mitigation. On a $2.0 billion base and roughly 53% gross margin, that implies nearly $95 million of gross profit at risk before operating deleverage. The sensitivity is an analyst estimate; concentration and margin inputs are reported facts. [S1] [S2]

A catastrophic scenario would involve Apple internalizing or replacing several audio and HPMS functions across consecutive generations. Revenue could fall 30–50%, gross margin could compress, and R&D would be slow to resize because current projects support future products. Net cash would prevent an immediate liquidity crisis, but a solvent company could still suffer a majority equity-value loss. A literal total loss is remote and would probably require fraud, extraordinary liability, or prolonged destruction of the technical franchise.

Pricing and architecture are the next risks. Annual reductions are normal in consumer electronics, and both Audio and HPMS Q1 growth were partially offset by previously anticipated ASP reductions. Q2 FY2027 gross-margin guidance benefited temporarily from favorable wafer pricing expected largely to sell through during the quarter. If gross margin falls below 50%, each percentage point represents approximately $20 million of annual gross profit at current revenue. Architecture integration can be more damaging because it removes the merchant socket rather than merely lowering price. [S2] [S6] [S7]

End-market contraction is a high-likelihood near-term risk. Gartner and IDC forecast significant device declines and longer replacement cycles. Premium resilience and content gains may offset part of the pressure, but not necessarily all of it. Memory supply, pricing, and OEM schedule changes can delay otherwise successful Cirrus designs. [S10] [S11]

Diversification may be delayed or illusory. PC engagement can remain confined to delayed platforms or niche models. Camera and power growth can deepen Apple exposure instead of broadening the customer base. Industrial products can require unfamiliar channels and support. Product diversity is therefore not automatically portfolio diversity.

Pre-revenue investment has become more asymmetric. R&D is rising, test equipment is being purchased, and the GlobalFoundries agreement specifies quarterly wafer quantities estimated at at least $600 million. Customer orders lack matching public minimums. A canceled program can leave Cirrus with unused engineering, inventory, purchase shortfalls, or wafers requiring alternative demand. [S2]

Supply execution remains important. Process transfers can encounter yield, qualification, cost, and timing problems. Geographic diversification reduces dependence on one location but adds ramp complexity. Because products are proprietary and generally unavailable from second sources, disruptions can affect both Cirrus and its customer.

The IRS transfer-pricing dispute is a material but non-core risk. The asserted additional tax and penalties total $232.0 million before interest. Cirrus disputes the assessment, has not accrued the asserted additional liability, and believes existing reserves are adequate. Investors should neither treat the full assertion as settled debt nor ignore it. An adverse outcome would reduce excess cash, could raise effective tax rates, and might constrain repurchases, although it would not by itself threaten solvency. [S2]

Capital-allocation risks include buying stock above intrinsic value and pursuing acquisitions whose returns remain opaque. Recent repurchase prices materially exceeded the current quote. A large acquisition could consume the balance-sheet option that currently protects the downside. Lion shows why strategic fit cannot substitute for a disclosed return bridge.

SBC and governance are secondary but real risks. Existing awards, the expanded plan reserve, and ex-SBC incentive margins can transfer value even when basic shares decline. Relative-return awards can pay during weak absolute performance. Historical net retirement offsets this concern but does not eliminate the cost.

The structured risk matrix is:

Diligence conclusion — stock decline factors: Customer content loss, weaker device units, pricing, delayed diversification, rising research without shipments, foundry mismatches, tax liabilities, dilution and a higher technology risk premium can all reduce the equity value. [S1] [S2] [S10]

Diligence conclusion — catastrophic loss: The clearest catastrophic path is simultaneous design-out or insourcing of several Apple functions, producing a 30–50% revenue decline and severe operating deleverage before diversification can respond. [S1] [S2]

Diligence conclusion — total loss: A literal total loss is remote because Cirrus has approximately $1.17 billion of cash and investments, no funded borrowings, positive cash generation and valuable IP; fraud, extraordinary liability or prolonged franchise destruction would probably be required. [S2] [S12]

Verdict: Downside is driven primarily by discontinuous operating value, not conventional solvency risk. Net cash buys time but does not protect shareholders from a multi-socket design loss or a mismatch between cancellable customer demand and binding wafer purchases. The tax dispute is material to excess cash but not comparable in scale to the customer-franchise risk.

Valuation Discussion

At $116.48 and 50.12 million shares outstanding, basic equity value is approximately $5.84 billion. Adding $130.5 million of lease liabilities and subtracting $1.167 billion of cash and investments produces an analyst-adjusted enterprise value near $4.80 billion. Against trailing revenue of $2.050 billion, EBITDA of $524.7 million, diluted EPS of approximately $8.18, and free cash flow of approximately $571.9 million, the stock trades near 14.2 times trailing diluted GAAP earnings, 9.2 times adjusted EBITDA, 10.4 times trailing FCF, and 2.3 times revenue. [S2] [S12]

The calculation differs from Company Financials’ June-period output because that output uses the $148.53 June 30 price and subtracts cash but not all marketable securities. Recomputing with the September 10 price and filing liquidity is necessary. Cash and investments net of lease liabilities equal approximately $1.04 billion, or about $20.70 per basic share. Funded leverage is absent, although leases, software, purchase commitments, and the disputed tax assessment remain economically relevant.

The tax assertion should be scenario-adjusted rather than automatically treated as debt. Deducting the full $232.0 million stated tax and penalties, before interest and with no credit for existing reserves, would reduce value by approximately $4.63 per basic share. That is a useful stress but not a base-case liability because the company contests the positions and no adjudication has occurred. [S2]

Trailing FCF overstates a steady-state baseline because FY2026 released working capital. A normalized range of $450–$525 million is more prudent. At the current equity value, that implies a 7.7–9.0% normalized FCF yield before growth. The lower end assumes working-capital normalization, continued research investment, higher capital spending, and no repeat of the FY2026 release; the upper end assumes current margins and seasonal conversion remain durable.

Peer comparisons are diagnostic rather than mechanical:

Peer values are Company Financials trailing metrics at each company’s latest available period and can differ in lease, restructuring, and acquisition treatment. ADI and TXN deserve substantial premiums for customer diversity, catalog durability, and channel reach. Skyworks’ high P/E does not mean the market assigns it greater quality; depressed GAAP earnings inflate the ratio. CRUS’s adjusted enterprise multiple is more favorable than the group, but concentration prevents application of a diversified-analog multiple. [S12]

The company’s own June-quarter valuation was approximately 17.6 times trailing earnings and 12.8 times EBITDA, versus Company Financials period averages near 15.7 and 11.3 times. At $116.48, the recomputed multiples fall below those references. Historical averages are not intrinsic value and may be distorted by changing customer risk, but the post-peak decline clearly changed the valuation case. [S12]

Three scenarios make the assumptions explicit:

These are analyst estimates, not management guidance. The base case assigns value to existing smartphone content and partial PC recovery but does not require metrology to become material. The bear case assumes erosion without a franchise-breaking multi-socket loss. A catastrophic customer event could produce values below the stated bear range. An adverse tax outcome can reduce each scenario by several dollars per share depending on interest, reserves, settlement terms, and timing.

The current price appears to embed little durable growth and a permanent concentration discount. The market is correct that CRUS does not deserve ADI or TXN multiples, that FY2026 working-capital conversion should not be annualized, and that new products remain incompletely disclosed. The market may be too pessimistic if Apple content persists, HPMS maintains differentiated margins, and only one independent growth program scales.

The fragile bull assumptions are continuing Apple socket durability, PC conversion, productive capacity utilization, and margin preservation after the favorable wafer tranche ends. The fragile bear assumptions are that customer concentration implies imminent loss, that PC delays indicate weak competitive relevance, and that rising research creates no offsetting content. Each must be tested through revenue, margin, customer mix, inventory, and commitments rather than narrative.

The factor model dated September 9, 2026 estimates market exposure of 1.18, technology exposure of 0.49, financials exposure of negative 0.33, quality exposure of 0.14, small-size exposure of 0.11, and communication-services exposure of negative 0.04. Residual momentum was negative 0.021, residual Sharpe approximately negative 1.02, and residual volatility 0.299. Its R-squared was only 0.376. These are statistical return exposures and diagnostics, not business classifications, fundamental evidence, or causal explanations. Low explanatory power is consistent with company-specific customer and program news dominating the stock. [S13]

Verdict: Current valuation compensates for ordinary cyclicality, substantial concentration, and some execution risk, but not for a confirmed multi-socket Apple loss. The operating franchise is inexpensive if current content persists; it is a value trap if widening product breadth merely increases exposure to the same customer before binding capacity arrives.

Variant Perception

The observable market view is conflicted. The company produces high margins, free cash flow, and net cash, yet trades near 14 times trailing earnings near the bottom of its 52-week range. That combination suggests investors accept current financial quality but assign meaningful probability to peak earnings, Apple insourcing, pricing pressure, capacity mismatch, or failed diversification. [S1] [S12]

The strongest bull case is that the market confuses concentration with imminent displacement. Cirrus has a long execution record, multiple current smartphone functions, rising HPMS revenue, and independent Apple validation of future process and Face ID work. More than $1 billion of net liquidity and continuing share retirement can compound per-share value with only moderate revenue growth. If current audio and camera content persists, the valuation does not require every new program to succeed. [S2] [S9]

The strongest bear case is that relationship duration hides a changing merchant-silicon boundary. Product diversification has principally added Apple sockets, increasing revenue before reducing risk. PCs remain too small to offset a material Apple change, management reduced the PC outlook, research is rising, and $600 million of estimated wafer purchases must be absorbed during 2027–2028. Recent repurchases near the price peak reduce confidence in capital-allocation timing. [S2] [S7]

Five assumptions carry the thesis:

1. Existing audio and camera content persists. The favorable case is weakened by product-line contraction beyond device-unit effects, falling gross margin, or a lower Apple percentage caused by Apple revenue loss. The adverse case weakens if multi-generation revenue and margin remain stable.

2. PCs become economically relevant. The favorable case weakens if PC revenue remains below approximately $75 million through FY2028. The adverse case weakens if PC revenue exceeds approximately $200 million across several OEMs with positive contribution.

3. Power and metrology convert from engineering to production. Repeated slips, no production milestones, and rising research without revenue falsify favorable option value. Production awards and diversified paid revenue weaken the adverse case.

4. Margins remain differentiated. Sustained gross margin below 50% or operating margin below 18% outside a temporary transition indicates pricing or mix deterioration. Margins near 52–53% through weak device units would demonstrate stronger content power.

5. Capacity and capital allocation remain productive. Inventory, shortfall charges, or cash commitments without shipments weaken the thesis. Repurchases below value, continuing net share retirement, and matched wafer demand strengthen it.

Investor questions on the two latest calls focused on the boundary between Cirrus and customer-designed silicon, PC timing and ultimate scale, structural versus temporary gross-margin benefits, timing of smart power and metrology, R&D intensity, and repurchase priorities. Those questions identify the correct uncertainties. Management’s answers establish its beliefs and road map, not independent proof. [S6] [S7]

The factor model adds positioning context but cannot settle the debate. Market exposure above one and positive technology exposure are consistent with a volatile technology equity, while the negative financials exposure is merely a statistical coefficient. The 37.6% R-squared means most variation is not explained by the supplied factors, and negative residual risk-adjusted performance does not identify whether the cause was PC expectations, customer fears, valuation compression, or unrelated trading. [S13]

No prior dated CRUS report was available to score. Transferable ideas were re-tested rather than inherited. The repurchase rule is confirmed: gross dollars matter only after reconciling issuance, leverage, cash generation, and price; CRUS passes the net-share and leverage tests but recently failed the price-timing test. The price-gap rule is confirmed because current multiples differ materially from the June snapshot. The partnership rule is confirmed because Apple validates technical collaboration without guaranteeing demand. Unrelated biotechnology, banking, utility, retail, and other company-specific learnings do not transfer to this analysis.

The principal stale assumption is that product diversification automatically reduces risk. It has not: Apple concentration increased from 87% to 91%. A second contradiction is May PC optimism versus August’s reduced outlook. A third is the claim of a strong pipeline alongside limited program-level revenue and return disclosure. A fourth is the description of the balance sheet as simply debt-free without considering the new wafer commitment and disputed tax assessment. None proves the road map will fail, but all reduce justified confidence.

Diligence conclusion — investor questions: The most decision-useful investor questions concern customer insourcing, PC revenue conversion, gross-margin durability, power and metrology timing, foundry commitment coverage and whether repurchases occur below intrinsic value. [S6] [S7]

Verdict: The differentiated view is not that Apple risk disappears. It is that durable current content plus one successful independent growth leg could justify substantially more value. The thesis fails if diversification remains only a wider set of correlated sockets while price, units, and binding supply exposure move adversely.

Fact vs. Interpretation

Technical progress updates product-relevance confidence before it updates revenue or ROIC. Revenue growth updates adoption before it proves customer diversification or acquisition success. Repurchase dollars update capital return only after price, issuance, net shares, leverage, and cash generation are reconciled. [S1] [S3]

Open Questions

1. How much FY2027 PC revenue was removed from the May expectation, and what portion reflects delay rather than cancellation? [S6] [S7]

2. What annual price reductions and expected generation lives apply to the existing audio and camera portfolio?

3. When can PC revenue exceed $100 million and the approximately $200 million level representing 10% of current sales?

4. What production, qualification, and customer milestones remain before Face ID power products generate revenue? [S9]

5. What portion of the $600 million estimated GlobalFoundries purchases is matched by non-cancellable customer demand, and what are the shortfall economics? [S2]

6. How much HPMS revenue comes from camera, haptics, power, other products, and non-Apple customers?

7. What revenue, profit, and ongoing R&D can be attributed to Lion relative to acquisition consideration? [S14]

8. Will repurchase execution become more price-sensitive after purchases at $163.43 and $140.53?

9. How much of the expanded equity-plan reserve does the company expect to issue annually, and how will that affect diluted shares? [S3] [S17]

10. Can management disclose non-Apple revenue growth, production-program count, or customer count without compromising confidentiality?

11. What amount is already reserved for the IRS dispute, and what settlement range does the board use in capital-allocation planning? [S2]

These questions identify where public evidence is weakest. Missing disclosure does not prove poor economics, but it should limit conviction and the present value assigned to long-dated programs.

What Must Be True

Bull tests

• Existing Apple Audio and HPMS revenue must remain sufficiently durable that total annual revenue does not fall below approximately $1.9 billion during the device downturn. Monitor product-line revenue, Apple concentration, ASP commentary, and gross margin. [S1] [S2]

• Gross margin should remain above 50%, and GAAP operating margin should recover above 20% after seasonal investment. A sustained breach would indicate price, cost, utilization, or mix deterioration beyond ordinary launch noise. [S2] [S7]

• PC engagement must become production revenue. A credible intermediate path requires revenue approaching $100 million; approximately $200 million would begin to make the business independently material under management’s own conceptual threshold. [S6] [S7]

• Power and metrology programs must reach qualification and production milestones without repeated multi-year delay. Tape-outs, samples, and partnerships alone do not satisfy this test. [S7] [S9]

• Non-Apple revenue must grow faster than total revenue for several quarters. Otherwise product breadth will not become customer diversification. [S1] [S2]

• Normalized annual FCF should remain above approximately $450 million while research rises, and net diluted shares should continue declining without compromising liquidity. [S1] [S2]

• The 2027–2028 wafer commitment must be matched by product shipments without material shortfall charges, abnormal inventory, or sustained cash-conversion deterioration. [S2]

• The IRS dispute must resolve within existing reserves or at a cash cost that leaves the core capital-allocation program intact. A large new accrual or settlement materially above existing reserves would reduce excess-cash value. [S2]

Bear tests

• The bear case weakens if Apple concentration falls below approximately 80% because non-Apple revenue grows—not because Apple revenue collapses—while gross margin remains above 50%. [S1] [S2]

• PC revenue above approximately $200 million across several OEMs with positive contribution would establish a genuine independent growth leg and falsify the view that engagement remains immaterial. [S6]

• Stable multi-generation audio and camera revenue, plus independent production evidence for power applications, would weaken the view that customer insourcing is imminent. [S6] [S9]

• Sustained GAAP operating margin above 23% through weak device units would demonstrate stronger content and operating leverage than the bear case assumes. [S1] [S10]

• Four quarters of normalized FCF above $550 million, continuing net share retirement, and no material capacity charge would demonstrate that cash generation is not merely a working-capital peak. [S1] [S2]

Immediate monitoring sequence

The next two earnings reports should be scored against Q2 revenue guidance of $510–$570 million, Audio and HPMS growth, gross margin after the temporary favorable wafer tranche, PC-program timing, research growth, receivables and inventory, Apple concentration, repurchase price, the IRS dispute, and updated foundry obligations. The thesis should change when these measurable variables change—not when partnership language, design interest, or short-term price action merely sounds favorable. [S2] [S7]

Linked sources: FY2026 Form 10-K, Q1 FY2027 Form 10-Q, 2026 proxy, Q4 FY2026 shareholder letter, Apple manufacturing announcement, Gartner device outlook, and IDC PC outlook.

Public source appendix

• S1: Cirrus Logic FY2026 Form 10-K — primary filing; published 2026-05-21; Items 1, 1A, 7, and 8; product revenue, customer concentration, financial statements, R&D credit, leases, commitments, inventory reserves, stock compensation, and repurchases
• S2: Cirrus Logic Q1 FY2027 Form 10-Q — primary filing; published 2026-08-05; Financial statements; Notes 8–15; customer concentration, liquidity, cash flow, IRS dispute, repurchases, and subsequent GlobalFoundries commitment
• S3: Cirrus Logic 2026 Proxy Statement — primary filing; published 2026-06-03; Executive and director compensation, incentive definitions, ownership guidelines, awards, plan capacity, burn rate, and dilution
• S4: Cirrus Logic FY2024 Form 10-K — primary filing; published 2024-05-24; FY2022–FY2024 financial statements, business, competition, patents, and risk factors
• S5: Cirrus Logic Q4 and FY2026 Shareholder Letter — company primary release; published 2026-05-06; FY2026 reported and non-GAAP results, cash flow, repurchases, product commentary, and guidance
• S6: Cirrus Logic Q4 FY2026 Earnings-Call Transcript — management transcript; published 2026-05-06; Prepared remarks and Q&A on audio lifecycles, PCs, SDCA, smart power, R&D, and capital allocation
• S7: Cirrus Logic Q1 FY2027 Earnings-Call Transcript — management transcript; published 2026-08-05; Prepared remarks and Q&A on reduced PC expectations, gross margin, HPMS pipeline, metrology, wafer capacity, and repurchases
• S8: Cirrus Logic KeyBanc Technology Leadership Forum Transcript — management event transcript; published 2026-08-12; PC platform timing, SDCA claims, test-equipment strategy, general markets, and capital priorities
• S9: Apple Adds New American Manufacturing Program Partners — customer primary release; published 2026-03-26; Cirrus–GlobalFoundries process development, mixed-signal applications, Face ID reference, and aggregate multi-program spending
• S10: Gartner 2026 PC and Smartphone Shipment Outlook — independent industry estimate; published 2026-02-26; 2026 device-shipment, memory-cost, pricing, replacement-cycle, and premium-device forecasts
• S11: IDC Worldwide Personal Computing Device Outlook — independent industry estimate; published 2026-06-08; June 8, 2026 update; PC decline, memory constraints, pricing, and stabilization timing
• S12: Company Financials — CRUS and Peer Financial, Valuation, Transcript, and Price Data — third-party financial data; published 2026-09-10; NASDAQ:CRUS resolved primary profile; API v3 five-year statements, profitability, enterprise value, valuation, per-share data, earnings-call metadata, daily prices, and ADI, TXN, SWKS, and QCOM peer metrics; material values reconciled to filings
• S13: Factor model — CRUS Snapshot — internal quantitative diagnostic; published 2026-09-09; September 9, 2026 exposures, residual signals, and diagnostics
• S14: Cirrus Logic Lion Semiconductor Acquisition Disclosure — primary filing; published 2022-07-27; Acquisition note; consideration, acquired technology, goodwill, and strategic rationale
• S15: John Forsyth May 2026 Form 4 — primary insider filing; published 2026-05-11; 5,000-share Rule 10b5-1 sale at weighted-average $170.41 and plan-adoption footnote
• S16: Jeffrey Baumgartner June 2026 Form 4 — primary insider filing; published 2026-06-04; Option exercise, 3,907-share Rule 10b5-1 sale, and plan-adoption footnote
• S17: Cirrus Logic 2026 Annual Meeting Voting Results — primary filing; published 2026-08-03; Approval of amended and restated 2018 Long Term Incentive Plan and other voting results
• S18: Cirrus Logic Principal Accounting Officer Appointment — primary filing; published 2026-03-31; CFO Jeffrey Woolard’s additional appointment and predecessor transition
• S19: Contemporaneous Report on April 2023 CRUS Price Decline — contemporaneous news; published 2023-04-12; April 12, 2023 price move and attributed Apple architecture concern
• S20: Associated Press Report on February 2024 CRUS Results — contemporaneous news; published 2024-02-07; February 7, 2024 results and market response