Understanding Analog Devices (ADI) for the Long Term: The Strength of Analog Semiconductors That Hold the Key to “Digitizing the Physical World,” and the Quietly Emerging Risks

※ This report is prepared based on data as of 2026-02-19.

What is ADI? The middle-schooler-proof explanation

Analog Devices (ADI) designs and sells analog/mixed-signal semiconductors that convert small real-world signals—like “temperature, pressure, vibration, light, electricity, and radio waves”—into “data” that machines and computers can use. Put simply, ADI gives machines “senses,” plus the circuitry that turns those senses into usable digital information.

ADI doesn’t sell to consumers. Its customers are businesses (B2B) building equipment and infrastructure. It doesn’t manufacture end products (cars, factory equipment, base stations, etc.); it earns money by supplying the components embedded in the “systems that can’t stop.”

Who it serves (customers)

• Industrial equipment and factory automation (measurement, control, condition monitoring, predictive maintenance)
• Automotive (electrification, safety, in-vehicle networks, sensing)
• Communications infrastructure (RF-related components for base stations, etc.)
• Data centers / server adjacencies (power, monitoring, protection, stable operation)
• Medical devices, test & measurement equipment, aerospace & defense, etc.

How it makes money (revenue model)

At the surface, the model is straightforward: “sell semiconductor components at a per-unit price.” In practice, the real game is less about raw volume and more about winning the design slot.

• ADI’s components are selected during the customer’s product design phase (design-in)
• Once the customer ramps to mass production, the same components are typically purchased repeatedly for long stretches
• Many use cases demand high precision and high reliability, which makes swapping suppliers based on price alone difficult
• Beyond the chips themselves, ADI provides design support—development tools and reference designs—to lower adoption friction (even though the core profit engine remains component sales)

Today’s revenue pillars: what keeps the “field” running

1) Industrial (factories, infrastructure, measurement)

Industrial is the heart of ADI’s business. In factories, downtime is expensive, so “noise immunity, durability, and consistent long-term performance” are value propositions in their own right. ADI earns its place by delivering “on-site quality”—accurately measuring tiny signals, driving motors and equipment as intended, and spotting early warning signs through vibration, temperature, and other indicators.

2) Automotive (electrification, safety, new architectures)

As cars become more electronic, the need rises for components that handle physical signals with precision—“measuring and protecting batteries and power safely,” “reading many sensors accurately,” and “stabilizing in-vehicle power and communications.” ADI supplies the measurement, protection, and control building blocks behind the scenes of the “computer on wheels” shift.

3) Communications (base stations, etc.)

Radio equipment has to transmit and receive reliably at specific frequencies and power levels, where analog/RF fit is decisive. Communications also tends to be capex-cyclical, but demand can accelerate during standards transitions and technology upgrade cycles.

4) Power and protection (“the safety officer for electricity”)

Power is the foundation for every electronic system. In servers, industrial equipment, and vehicles, the value is in “not failing from a momentary anomaly” and “not going down.” ADI offers a broad lineup that distributes power efficiently, protects against faults, and monitors system health—and its relevance can rise as data centers push toward higher performance.

Future pillars: the next growth vectors for the AI era

ADI isn’t an “AI itself” company in the sense of making AI GPUs. Even so, it’s leaning into a clear message: as AI expands from cloud-based chat into “AI that senses and acts in the real world,” ADI’s domain—measurement, conversion, control, power, and synchronization—becomes more likely to show up as a constraint.

1) The foundation for Physical Intelligence (AI in the physical world)

As robots, factory equipment, and vehicles get smarter, “capturing real-world data correctly” becomes more important. If the input data is poor, AI decisions become unstable—making sensors and signal processing a quiet but decisive bottleneck. That’s where ADI’s core franchise sits, creating a setup where physical-world AI adoption can be a tailwind.

2) Enabling decision-making at the edge (on site)

When decisions can be made locally—without sending everything to the cloud—you can gain on speed, power, and availability. ADI emphasizes the “interface to the real world (the edge),” and the components and systems that support on-site decision-making can become a long-duration theme.

3) Upgrading power and protection in data centers

AI-era data centers are moving to higher power density, which raises the premium on stability and protection. ADI is positioned to benefit from demand for power, monitoring, and protection as “must-have adjacent components.”

Long-term fundamentals: seeing ADI’s “pattern” in the numbers

In Peter Lynch’s framework, you start by identifying “what kind of grower this is (its pattern)” using 5-year and 10-year trends. With ADI, the picture can look meaningfully different depending on the metric.

Growth rates: EPS is mid-to-low speed, while revenue and FCF can look strong in certain periods

• EPS CAGR: past 5 years +6.8%, past 10 years +7.6%
• Revenue CAGR: past 5 years +14.5%, past 10 years +12.4%
• FCF CAGR: past 5 years +18.3%, past 10 years +19.0%

Bottom line: on EPS alone, ADI doesn’t screen as “high growth” and reads more like mid-to-low growth. But revenue and free cash flow (FCF) have been stronger, so the growth narrative depends heavily on which yardstick you use.

Profitability (ROE): not consistently high, and currently modest

ROE in the latest fiscal year is 6.7%. Over 5 years it looks roughly flat to slightly improving, while over 10 years it looks more downward-tilted—both can be true depending on the window. This isn’t about FY vs TTM nuance; the key takeaway is simply that “the trend reads differently over 5 years versus 10 years.”

Cash generation: high FCF margin, without an unusually heavy capex load

On a recent TTM basis, revenue is $11.76bn and FCF is $4.56bn, implying an FCF margin of 38.8%. Capex is often benchmarked at about 8.0% of operating cash flow, which suggests this isn’t an especially capex-intensive model. In other words, “a business that tends to keep its profits as cash” is central to the long-term story.

ADI through the Lynch lens: a hybrid of “quality stock × cycle”

Rather than labeling ADI as a classic Fast Grower, it’s more accurate to view it as a hybrid: a maturity-leaning profile (mid-to-low EPS growth) plus quality-stock traits (strong cash generation), layered with the cyclicality that comes with being a semiconductor company.

• Rationale 1: Long-term EPS CAGR is past 5 years +6.8% and past 10 years +7.6%, which is not high-growth territory
• Rationale 2: Revenue (past 5 years +14.5%) and FCF (past 5 years +18.3%) are relatively strong, so the growth picture depends on the metric
• Rationale 3: As discussed later, the stock’s P/E is on the high side versus its own history, consistent with a profile where the market periodically tolerates a higher multiple

Short-term momentum (TTM + latest 8 quarters): is the pattern intact?

Once you’ve anchored on the long-term “pattern,” the next step is to check the current TTM (last 12 months) and the latest 8 quarters. This matters directly for timing and decision-making, and ADI’s recent results read as “a strong recovery/rebound.”

TTM growth: EPS, revenue, and FCF are all strong at the same time

• EPS (TTM): $5.51, TTM YoY +75.6%
• Revenue (TTM): $11.76bn, TTM YoY +25.9%
• FCF (TTM): $4.56bn, TTM YoY +43.2%
• FCF margin (TTM): 38.8%

On the most recent year alone, EPS is strong enough to break from the long-term “mid-to-low growth” impression. But when you pair that with the mid-to-low long-term CAGR (past 5–10 years), it’s reasonable to keep in mind that the latest period may “look strong due to a cyclical rebound.” That framing helps avoid treating short-term strength as permanent high growth.

Direction over the latest 8 quarters: a sustained uptrend, not a one-quarter spike

Across the last two years (roughly 8 quarters), the data points to an upward trend in EPS (correlation +0.63), revenue (+0.66), and FCF (+0.92). At a minimum, that suggests momentum has persisted rather than being “a one-off surge that faded.”

A profitability momentum marker: FCF margin pushes above its historical band

The 38.8% TTM FCF margin sits above the upper end of the typical range over the past 5 and 10 years. Unlike periods where revenue rises but underlying earning power deteriorates, this can be characterized—at least on a cash basis—as a strong phase.

Financial soundness (bankruptcy-risk framing): not over-levered, with room to service interest

Because semiconductors are cyclical, the balance-sheet question is simple: “Can it survive the trough?” ADI’s latest FY metrics suggest it isn’t operating under extreme strain.

• Debt-to-capital ratio (latest FY): 0.26
• Net Debt / EBITDA (latest FY): 1.00x
• Interest coverage (latest FY): 9.54x
• Cash ratio (latest FY): 1.13

From the standpoint of interest coverage and liquidity cushion, these figures don’t point to a structure that’s “likely to seize up quickly.” While you can’t declare bankruptcy risk away, the available evidence doesn’t support the idea that ADI is “forcing growth through excessive leverage.”

Dividends: a semiconductor where dividends matter, but earnings-linked pressure remains

ADI is a name where dividends are part of the story: 22 consecutive years of dividends and 21 consecutive years of dividend increases. The latest TTM dividend per share is $3.97094.

How to treat yield: latest TTM cannot be calculated

The latest TTM dividend yield can’t be calculated from this dataset. As reference points, the 5-year average is about 1.98% and the 10-year average is about 2.47%. Using a share price of $346.37 (report date), it could generally be a period that screens less attractive versus historical average yields, but this article does not make a definitive call because the latest yield is unavailable.

Dividend growth: roughly a ~10% annual pace

• DPS growth rate: past 5-year CAGR about 10.21%, past 10-year CAGR about 9.60%
• Most recent 1 year (TTM) dividend increase rate: about 8.46% (slightly below historical CAGR but not a major deviation)

Dividend safety: tight on accounting earnings, more comfortable on cash

• EPS-based payout ratio (TTM): about 72.13% (higher than the past 10-year average of about 69%, lower than the past 5-year average of about 76%)
• FCF-based payout ratio (TTM): about 42.81%
• Dividend coverage by FCF (TTM): about 2.34x

For the same dividend, it’s hard to argue the burden is light on the accounting earnings (EPS) side. On the cash side, though, it’s covered at more than 2x, which provides a cushion. That “tight on earnings, room on cash” setup supports framing ADI’s dividend as having “moderate safety.”

The key caution is that in semiconductor downcycles, when earnings fall, the EPS-based payout ratio tends to rise. With dividend increases often treated as a cultural commitment, flexibility in a trough remains an open question.

Where valuation stands today (own historical comparison only): high P/E and low FCF yield, but a calmer PEG

Here, without comparing ADI to the market or peers, we place the current level (share price $346.37) against ADI’s own history—primarily the past 5 years, and secondarily the past 10 years.

PEG: near the middle of its own range

• PEG (based on latest 1-year EPS growth rate): 0.83x
• Past 5-year median: 0.84x, past 10-year median: 0.84x

PEG sits within the historical range for both the past 5 and 10 years and is close to the median. Versus the median observed over the past 2 years (2.14x), the current level is positioned after a move lower.

P/E: above the typical band over both 5 and 10 years

• P/E (TTM): 62.91x
• Upper end of past 5-year normal range: 57.89x (current is above)
• Upper end of past 10-year normal range: 41.32x (current is well above)

On its own history, the P/E is clearly elevated. PEG looks more benign largely because “the latest 1-year EPS growth rate is unusually large,” and the contrast between P/E (TTM) and PEG (recent growth-based) comes down to different periods and denominators (TTM earnings vs recent growth).

FCF yield: below the 5- and 10-year ranges (= a higher-valuation posture)

• FCF yield (TTM): 2.69%
• Lower end of past 5-year normal range: 3.29% (current is below)
• Lower end of past 10-year normal range: 3.41% (current is below)

Within its own history, FCF yield is meaningfully low. A low yield (i.e., relatively little FCF versus market value) typically shows up during higher-valuation phases.

ROE: middle of the pack over 5 years, weaker over 10 years

• ROE (latest FY): 6.70%
• Past 5-year median: 6.70% (same level)
• Past 10-year median: 8.63% (below this, positioned lower within the 10-year range)

ROE screens as median over 5 years and modest over 10 years—not “exceptionally strong” on a recent basis.

FCF margin: above the normal range over both 5 and 10 years

• FCF margin (TTM): 38.79%
• Upper end of past 5-year normal range: 34.26% (above)
• Upper end of past 10-year normal range: 34.00% (above)

So while valuation metrics (P/E, FCF yield) point to a higher-valuation setup, the underlying cash-generation quality (FCF margin) is also above its own historical band. That combination supports a framing where “high valuation” and “a strong operating phase” coexist.

Net Debt / EBITDA: low within its own range (lower-multiple side)

• Net Debt / EBITDA (latest FY): 1.00x
• Past 5-year median: 1.00x (near the median, leaning toward the lower end of the range)
• Past 10-year median: 1.52x (below this, also low over 10 years)

This is an inverse-type indicator where a smaller value implies more balance-sheet flexibility. On that basis, 1.00x doesn’t suggest historically elevated leverage; it sits on the lower-multiple side of the range.

Cash flow tendencies (quality and direction): a business where FCF matters more than EPS

In ADI’s long-term data, there are stretches where FCF growth (past 5 years +18.3%) looks materially stronger than EPS growth (past 5 years +6.8%). Even in the latest TTM, the FCF margin is high at 38.8%, and capex does not appear unusually heavy (about 8% of operating cash flow). That makes “a model where profits tend to show up as cash” a persistent, central feature.

The dividend picture reinforces this. While the EPS payout ratio looks high at about 72%, the FCF payout ratio is about 43% and FCF covers the dividend about 2.34x—meaning the cushion is on the cash side. This isn’t just a generic “earnings vs cash” mismatch; it reflects ADI’s structurally strong FCF profile, and the dividend looks more moderate when evaluated through cash.

Why ADI has won (the core success story)

ADI’s intrinsic value is its ability to supply, reliably and over long periods, “components that convert real-world signals into forms machines can use,” enabling mission-critical environments—factories, vehicles, infrastructure, medical, and test & measurement—where downtime isn’t an option. Analog isn’t flashy, but the more an application demands precision, reliability, standards compliance, and long-term availability, the harder it is to displace a supplier through simple price comparisons.

And in markets with long design cycles, once a part is designed in, it often stays there for years. Switching typically requires re-evaluation, re-certification, and changes to designs and bills of materials. That “design-in → long tail” dynamic underpins ADI’s “industrial infrastructure-like” profitability.

What customers value (Top 3)

• Accuracy, noise immunity, repeatability: the ability to measure tiny signals and resist drift in real-world conditions
• Reliability and long-term supply: confidence in consistent quality and ongoing availability raises switching costs
• Ease of implementation via design support: reference designs, development tools, and know-how help shorten development cycles

What customers dislike (Top 3)

• Cost increases and procurement-planning difficulty tied to price revisions (a price revision notice from February 2026 has been reported)
• Lead-time and supply-constraint concerns, including mature-node products (risk of not being able to secure parts when needed)
• High design specialization makes replacements cumbersome (the flip side of barriers to entry, but it shows up as real friction in the field)

Story continuity: do recent moves still match the “winning playbook”?

ADI’s story, while exposed to near-term macro and inventory cycles, has been directionally consistent over time: “take on customer complexity and provide the foundation that links the physical and digital worlds.” Recent strategic messaging also fits that playbook, emphasizing “essential physical-side infrastructure” like Physical Intelligence (physical-world AI), an edge focus, and data center power.

Three areas where the narrative may be drifting

• How the recovery is described: in November 2025, the message was more of a “point” story—“automotive strong and industrial below expectations”—while by February 2026 it shifts toward a “surface” story of “YoY growth across all end markets, led particularly by industrial and communications”
• Pricing stance: when suppliers explicitly signal a push to “defend/raise prices,” it can change the tone around negotiations, redesigns, and multi-sourcing
• Shareholder returns: dividend increases (a long streak of consecutive raises) can be presented as evidence of “room,” but in cyclical troughs that same rigidity can become a pressure point

The key is not to label narrative changes as automatically “good” or “bad.” In cyclical industries, whether the recovery remains a “surface” or reverts to a “point” is itself a variable worth monitoring.

Invisible Fragility (hard-to-see fragility): five items to check, especially when things look strong

Without claiming anything is “already breaking,” this section lays out lagged risks that are structurally more likely to show up for ADI. For long-term investors, this is often where the biggest impact comes from.

• “Invisible concentration” in large customers: there may be periods when a small number of customers represent a high share of revenue or receivables, and shifts in their inventory/procurement policies can make demand appear more volatile
• Dual-sourcing/redesigns triggered by price revisions: helpful for near-term profitability, but it can also set up a lagged, medium-term “quiet defection” by customers
• Supply constraints in mature nodes: when supply tightens, customers may prioritize “parts they can actually get,” creating a path to lose sockets even if performance wins
• The gap between strong cash generation and capital efficiency (ROE): cash can look abundant in recoveries, but if long-term capital efficiency doesn’t improve, tensions can emerge across investment, M&A, and returns—“weakness that shows up late”
• Exogenous shocks from trade and regulation: as seen with China’s anti-dumping investigations, constraints can be imposed on procurement, design, and pricing for reasons unrelated to performance, potentially pushing customers to redesign to reduce future risk

Competitive landscape: who ADI competes with—and how it can lose

Analog/mixed-signal isn’t a market where winners are determined cleanly by benchmark-style performance the way CPUs or GPUs often are. Adoption decisions are frequently driven by application-specific fit—accuracy, noise, temperature behavior, reliability, functional safety, standards compliance, and long-term supply. Competition is a blend of “scale advantages” and “technology/application fit.”

Key competitors (core set, in descending order of overlap)

• Texas Instruments (TI): strong portfolio breadth, supply capability, and a direct-sales model; including reported price revisions, it can reset the competitive baseline
• STMicroelectronics: overlaps in industrial, automotive, power, and MCU contexts
• Infineon: strong in power semiconductors and automotive, with high overlap in electrification
• NXP: often enters designs alongside automotive/industrial control systems
• Renesas: a one-stop option with analog/power plus automotive/industrial control
• ON Semiconductor: partial overlap around automotive/industrial power and sensor adjacencies
• Microchip: overlap in industrial control and adjacent analog

As an additional note, in RF, Skyworks and Qorvo can be competitors, and on the power-device side, Wolfspeed and others can be competitors (though ADI’s main battlefield is less the devices themselves and more the surrounding components for measurement, control, and power management).

The key competitive reality: the “opponent” changes at the customer-system level

In automotive, for instance, companies anchored in power devices can have an edge, while in industrial measurement, measurement know-how can be the deciding factor. The competitive set shifts depending on “where in the customer’s design” a supplier is trying to win. That’s why ADI’s competition is structurally hard to capture through simple category labels.

Price, supply, and trade can reshape competitive behavior

Price revisions can lift profitability in the short run, but over time they can also motivate customers to dual-source or redesign to regain leverage. Trade and regulation (including anti-dumping investigations) can also affect adoption feasibility and procurement strategy on non-technical axes, potentially shifting the competitive map itself.

Moat (Moat): what is ADI’s moat, and how durable is it likely to be

ADI’s moat isn’t a network-effect story where value rises with user count. It’s driven more by “design-in switching costs” and the practical capabilities required in high-reliability applications.

• Design-in + re-certification costs: switching often requires re-evaluation, re-certification, and documentation updates
• “Measurement quality” such as accuracy, noise immunity, and long-term drift characteristics
• Reliability, long-term supply, and documentation readiness: the more mission-critical the application, the more these matter
• Application know-how and reference designs: the ability to make the full system work end-to-end

That said, moat durability can be challenged if customers institutionalize dual-sourcing after price revisions or supply constraints, or if trade/regulation tilts the playing field on “non-performance” dimensions. In other words, the moat is real—but “customer behavior” and “exogenous factors” can still reshape outcomes from outside the walls.

Structural position in the AI era: likely a tailwind, but not the “main character”

ADI isn’t an AI model company. It operates in the “physical layer”—sensors, conversion, control, power, synchronization, connectivity—that links the real world to compute. As AI expands from “perception → inference → action,” input-data fidelity and stable operation become more important, and ADI’s mission-critical role tends to rise.

AI-era strengths (structural)

• Because advantage comes from “measurement quality,” not “user data,” ADI is more likely to hold a bottleneck-like position in physical-world AI
• In factories, automotive, communications, and medical/test & measurement, failures translate directly into downtime or safety risk—raising the importance of power, protection, and synchronization
• Barriers to entry are reinforced by accuracy, reliability, long-term supply, and post-design-in re-certification costs, making them less likely to erode even in the AI era

New weaknesses created by AI (substitution/relativization)

The risk that generative AI directly disintermediates ADI’s revenue model appears relatively low. But as AI-driven design automation advances, parts of traditional design support may become more commoditized. This is less about “components becoming unnecessary” and more about “how adoption happens and where differentiation sits” shifting over time.

Management, culture, and governance: consistency is a strength—and can also become a pressure point

CEO Vincent Roche has described ADI’s strengths as rooted in a culture of “relentless innovation” and “deep customer engagement,” and he has been shown to discuss capital allocation with that priority in mind. Over the long term, he has also referenced a commitment to return all free cash flow, and the company has continued to raise dividends.

How leadership shows up in strategy

• Culture: focuses not just on component specs, but on translating performance into “customer system outcomes”
• Decision-making: tends to allocate resources toward high-reliability domains where design-in matters, and toward integration (analog + digital + software)
• Strategy: aligned with “essential physical-side infrastructure” themes such as Physical Intelligence, edge, and data center power

Where culture can get tested

In a cyclical downturn, a high EPS payout ratio (about 72% on a TTM basis) could, together with a dividend-raise culture, reduce flexibility. And if price revisions encourage customers to dual-source, the culture’s real edge—“earning trust over time and winning on TCO (total cost)”—is more likely to be put to the test.

Organizational changes and board strengthening

Executive changes tied to business-unit oversight were disclosed from late 2024 through 2025. In January 2026, ADI added a director with AI/robotics expertise, and another director is expected to retire at the March 2026 annual shareholders’ meeting. These moves can be read as consistent with an emphasis on integrating “physical × digital × AI,” but they are not, on their own, enough to conclude that the culture is changing.

How to think about “demand waves”: a name where structure and cycles coexist

ADI’s tailwinds include “factory automation and labor saving,” “automotive electrification/safety,” “communications infrastructure upgrades,” and “data center adjacencies (power, monitoring, high-reliability components).” At the same time, demand in these markets blends structural (long-term) and cyclical (short-term) forces.

Most recently, from November 2025 to February 2026, management commentary has described the end-market recovery as a broad “surface,” with industrial and communications leading. But the risk of momentum shifting due to customer inventories and capex cycles remains. That’s the essence of ADI’s hybrid profile: “quality-stock-like, with cyclical waves on top.”

The KPI tree investors should use: tracking ADI through cause-and-effect

ADI can be misread if you follow it only through headlines or the P/E multiple. Mapping the story into value causality (a KPI tree) helps clarify what to watch.

Outcomes

• Sustained expansion of earnings and free cash flow
• Improvement/maintenance of capital efficiency (ROE, etc.)
• Financial durability to keep investing and returning capital through demand waves
• Continuity of shareholder returns including dividends

Intermediate KPIs (value drivers)

• Revenue: accumulation of design-ins (new wins + continued shipments)
• Mix: share of higher value-added domains (applications where accuracy, reliability, standards, and long-term supply matter)
• Profitability: whether mission-critical value translates into margin resilience
• Cash conversion: how much profit turns into cash (FCF margin)
• Capex burden: whether FCF holds up without overinvestment
• Demand cycle: short-term optics can swing with inventory adjustments
• Switching costs: the weight of re-evaluation, re-certification, and documentation updates
• Supply: lead-time stability and long-term supply structure
• Price/contracts: how price increases affect near-term profitability and medium-term customer behavior

Bottleneck hypotheses (monitoring items)

• Whether the recovery continues as a “surface” or reverts to a “point” (fixed-point monitoring of how industrial/communications are discussed)
• Whether dual-sourcing and redesigns show signs of becoming normalized after price revisions
• Whether supply concerns, including mature nodes, are influencing customer design policies
• Whether trade/regulation developments are affecting adoption feasibility by region and application
• How the gap between strong cash generation (FCF) and modest capital efficiency (ROE) shows up in priorities across investment, returns, and operations
• Where AI-era tailwinds are showing up as revenue across industrial, automotive, and data center adjacencies
• As design automation advances, whether tool enhancements continue to function as “adoption-friction reducers”

Two-minute Drill (the core investment thesis in 2 minutes)

• ADI is an analog semiconductor company that converts real-world signals into accurate data and enables mission-critical sites that can’t stop, with design-in and long-term supply creating sticky revenue.
• Long-term EPS growth is mid-to-low speed (past 5 years +6.8%, past 10 years +7.6%), but there are periods where revenue and FCF look strong; the latest TTM shows a sharp rebound with EPS +75.6%, revenue +25.9%, and FCF +43.2%.
• The latest TTM FCF margin is 38.8%, above the upper end of its own historical range, supporting a view that current “earning power” is also strong.
• Financial leverage does not appear excessive, with Net Debt/EBITDA at 1.00x and interest coverage at 9.54x.
• At the same time, the P/E is high versus its own history (62.9x TTM), so investors should recognize that if short-term momentum fades, valuation headroom can compress.
• Hard-to-see risks include lagged customer dual-sourcing/redesigns triggered by price increases, supply, and trade, plus the long-running gap between strong cash generation and modest ROE as an ongoing quality question.
• The AI-era opportunity is not “being the main character of AI,” but being essential infrastructure for physical I/O plus power and synchronization that lets AI operate on site—and keeping that distinction straight is key to long-term understanding.