Monolithic Power Systems Deep Value Investment Analysis

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

• QHow large is its market ceiling? Is it expanding an existing pie, or creating an entirely new market?6/10

  A

  The ceiling is very high, but MPWR is mainly taking share in a fast-expanding existing pie, not creating a brand-new market from scratch. It makes high-performance power management and power electronics chips, designing them into server motherboards, automotive platforms, optical modules, switches, and industrial control boards. These "use cases" already exist. MPWR's role is to replace legacy discrete power solutions with higher integration, higher power density, and higher energy efficiency, while also fitting into the new power demands created by the AI era. This is fundamentally different from "creating a market that never existed before" such as smartphones or cloud computing in their early days: it is doing a better job at "powering every board" and capturing more share.

  The pie itself is indeed growing quickly, and both the report and external data support that. At the industry level, WSTS expects the global semiconductor market to reach about $975 billion in 2026, with AI infrastructure expansion as one of the core drivers. The hottest subsegment MPWR targets, AI data center power delivery, is growing even faster: in its announcement to acquire Empower Semiconductor, ADI noted that aggregate 2025 capital expenditure by the four major hyperscale cloud providers was about $413 billion and is expected to rise to $600 billion–$700 billion in 2026. The company's own execution is also direct: in 2026 Q1, Enterprise Data revenue was $262.8 million, up 97.7% year over year, and communications revenue was $111.5 million, up 33.1% mainly from optical modules and switch power. On this basis, the report also raises the company's addressable "capacity target" from the original $4 billion to nearly $6 billion. But it must be stressed that this is a management target, not revenue already realized.

  There are two hard pieces of evidence that this is "expanding an existing pie" rather than "creating a new market." First, the competitors named in the report are Analog Devices, Infineon, NXP, onsemi, Power Integrations, Renesas, ROHM, Semtech, ST, and Texas Instruments. This is a list of mature power and analog players, which means the pie has long had participants dividing it up. Second, AI power is attracting stronger capital into the field: ADI's roughly $1.5 billion cash acquisition of Empower Semiconductor, explicitly targeting AI power delivery, shows precisely that this is an "attractive existing profit pool," not "white space."

  For a Baillie Gifford LTGG lens, the implication is this: the "absolute height" of the ceiling is a positive. AI power demand is a real long runway with deep snow. But MPWR does not have the scarcity premium of "creating an entirely new market." Its upside comes more from "steadily raising share and value per unit in a market that is clearly expanding" than from "defining a new category others cannot yet see." This is a good business entering a good track, but the growth story has limited "0 to 1" imagination. It is more about "1 to N" execution.

  Jun 11, 2026
• QCan its revenue at least double over the next five years? Will growth be driven mainly by volume, price, or new businesses?6/10

  A

  There is a realistic path for revenue to double over the next five years to about $5.6 billion, and growth will be driven mainly by "volume," with help from mix upgrades in new use cases. It is almost impossible for it to rely purely on "price." Doubling from $2.79 billion to about $5.58 billion implies a five-year CAGR of about 14.9%. That sits squarely within the company's recent performance range and is not excessive. The reference point is the company's past trajectory: 2026 Q1 revenue was $804.2 million, up 26.1% year over year, with 2026 Q2 guidance of $890 million–$910 million. The report states that revenue rose from $1.821 billion to $2.790 billion from 2023 to 2025, about +53% over two years. Even if current growth slows materially, a five-year doubling remains "neutral to achievable," not something that requires optimism. The more important anchor is management's self-reported capacity ceiling, which has been raised from $4 billion to nearly $6 billion. In effect, the company itself has placed "doubling" within the visible design-pipeline framework, though this is a target rather than a realized fact.

  The driver mix is overwhelmingly led by "volume" for two reasons. First, the report clearly states that the company acknowledges in its 10-K that average selling prices (ASP) in the industry decline over product life cycles. This is the law of analog and power semiconductors, meaning growth through price increases would be swimming upstream. Second, the company was able to keep gross margin stable at 55%–56% from 2023 to 2025, and still maintained a 55.3% GAAP gross margin in 2026 Q1. That shows its capability is "maintaining gross margin through higher integration and higher-value products while ASP naturally declines," not expanding profit through price hikes. The real growth engines are shipments and higher content per platform. Price is more about "defense" than "contribution."

  The strongest "volume + new business" driver comes from structural ramp-up related to AI. In 2026 Q1, Enterprise Data revenue nearly doubled year over year (+97.7%) to $262.8 million, while communications revenue rose +33.1%. This is both "volume growth in the existing data center market" and an expansion of relatively new content such as optical modules, switch power, and server power delivery. The end-market mix disclosed in the report (Enterprise Data 32.7%, storage and computing 21.7%, automotive 18.9%, communications 13.9%) shows growth is increasingly concentrated on the AI/data center side. That is the main engine behind the doubling thesis. Automotive, especially high-power-density ADAS, is the second-order source of volume growth.

  The honest risk note: this is a semiconductor business, not a subscription business. The report clearly notes that backlog is described in the 10-K as subject to cancellation or delay and not a robust indicator of future revenue. Seasonality, channel inventory, and macro cycles can all disrupt quarterly cadence. So "doubling in five years" is realistic as a path, but realization depends heavily on the AI capex cycle not being interrupted and on the company continuing to raise share in Enterprise Data. If AI capex peaks or competition such as ADI+Empower erodes excess share faster, the doubling timetable will stretch. Conclusion: doubling is achievable and volume-led, but the certainty is not a "free win." It requires the twin engines of "cycle strength + share gain" to work together.

  Jun 11, 2026
• QAfter five years, what will take over as the next growth engine? Does this "second curve" exist today?5/10

  A

  Today the second curve is mainly a "new branch on the same trunk," not a new growth pole independent of power management. This is one of the weaker parts of MPWR's growth narrative. Based on the business mix and management commentary disclosed in the report, the candidates to take over after five years fall roughly into three categories: automotive (ADAS and high-power-density in-vehicle power), deeper sales of "complete silicon-based solutions" rather than individual chips (management explicitly described the 2026 Q1 transition as moving from "selling chips" to "selling complete silicon-based solutions"), and higher-value system-level power delivery inside AI data centers, moving from board-level power toward vertical power delivery and power modules closer to processors. The issue is that all three are essentially extensions and value upgrades of the main "power management" curve, not a true second engine orthogonal to the existing core business.

  Why say it "exists today but is not yet independent": the report's end-market mix shows that in 2026 Q1, Enterprise Data was 32.7%, storage and computing 21.7%, automotive 18.9%, communications 13.9%, consumer 6.8%, and industrial 6.0%. Growth is highly concentrated on the AI/data center side. Automotive and industrial are potential successors, but their scale and growth are not yet large enough to "take over independently once the main engine slows." Put differently, if the current main growth curve of AI data center power peaks five years from now, there is not yet another validated curve of comparable scale that can seamlessly replace it. That is a clear gap versus Baillie Gifford's preferred "compounder with built-in second-curve DNA," such as platform software or ecosystem companies.

  One structural positive is real and deserves credit: power is a hard constraint on AI compute, and generational upgrades in AI power demand, including rising GPU card power and changes in rack-level power architecture, will repeatedly create "new power delivery problems inside boards and racks." That gives MPWR's main curve a long self-extension runway. Industry capital is also confirming the value of this direction. ADI's acquisition of Empower Semiconductor targets precisely "bringing power closer to processors" through vertical power delivery, and Empower's CEO said the goal is to solve the "hardest problem" in AI power delivery. If MPWR continues winning designs in this "high-density vertical power" area, the ceiling of the main curve will keep being lifted. But that strengthens the argument that "the main curve is long," not that "a second curve has already emerged."

  The honest conclusion: MPWR's "second curve" today looks more like depth in the main trunk (higher value content and more system-level power solutions) plus horizontal migration across use cases (data center to automotive to industrial), not a business-orthogonal new engine. For a "good business," this is sound. For the Baillie Gifford search for a "great growth stock that can rise fivefold in ten years," it lacks the asymmetric imagination of "a hidden new S-curve outside the core business that no one has priced." Combined with a share price that has already pulled forward a great deal (see Q9/Q10), the weak second curve leaves a thin cushion for "renewal through a new engine if the main curve slows."

  Jun 11, 2026
• QWhat is its core competitive advantage? Will this moat widen or narrow over the next three to five years?6/10

  A

  The core competitive advantage is an engineering moat built from the combination of system-level power engineering capability, design-in stickiness, and proprietary process/packaging. Over the next three to five years it is likely to remain stable and slightly widen, but it is not widening dramatically, and new competitors are entering faster. The report breaks down the moat clearly: it does not come from a consumer brand or network effects, but from engineering capability, design-in stickiness, proprietary process/packaging, and the reputation built through long-term execution. Overall moat strength is rated 4/5. Item by item, the strengths and weaknesses are as follows: switching costs are moderately high (once a chip is designed into a server motherboard, automotive platform, or optical module, the customer must revalidate, redo engineering, and recertify before switching, so it will not do so lightly); patent/process barriers are medium to medium-high (the report says the company has 2,231 issued or pending patents, including 654 issued U.S. patents, with the latest expiring in 2045, and many processes are protected by trade secrets rather than patents); cost advantage is medium (fabless model + Asian supply chain + high integration); network effects are basically absent, and scale advantage is weaker than TI/ADI (medium to weak). This is a "real but bounded" moat: sticky inside B2B design chains, but without consumer-monopoly-style pricing dominance.

  Pricing power is medium-high, not absolute. The report identifies a key tension: the company admits in its 10-K that industry ASP declines over life cycles, yet it still kept gross margin stable at 55%–56% from 2023 to 2025 and maintained a 55.3% GAAP gross margin in 2026 Q1. This shows that its "pricing power" is essentially "protecting gross margin through higher performance, higher integration, and customer switching costs," not arbitrary price increases from brand monopoly. Holding 55% is already a strong capability, but it is different in kind from a top-tier moat with strong network effects and durable price increases.

  The next 3–5 years are best judged as "stable to slightly wider," for three reasons, all partly inferential. First, the company is moving from "selling chips" to "selling complete silicon-based solutions," a direction management explicitly stated in 2026 Q1, and solutions usually have higher stickiness. Second, the company is moving deeper into AI servers, optical modules, switches, ADAS, and other use cases with stricter requirements for power density and efficiency. Enterprise Data revenue rose +97.7% year over year and communications +33.1% in 2026 Q1, confirming continued design wins in these higher-barrier areas. Third, the capacity target was raised from $4 billion to nearly $6 billion, implying the company believes the design pipeline can keep converting for several more years. But the report itself emphasizes that this "widening" is more inference than fully time-tested fact.

  The bear-side signal deserves full weight: the largest moat threat is that "stronger capital is entering the same high-profit pie." ADI is acquiring Empower Semiconductor for roughly $1.5 billion in cash, explicitly targeting high-density AI power delivery, and larger players such as TI, Infineon, and NXP also have the resources to step up investment in AI power. The report therefore concludes that "industry demand is real, but the profit pool will not go uncontested." This means the moat's "width" is static and solid (design stickiness is hard to dislodge in the short term), but the "sustainability of excess profits" may be gradually eroded by competition. The moat itself is not narrow, but the water inside the moat, meaning excess returns, may be diverted. Conclusion: this is a real, stable, slightly widening engineering moat, but not a dominant moat that "keeps getting wider while competitors have no way in." Over 3–5 years, the more likely outcome is "share can be defended, but excess margins are slowly compressed by competition."

  Jun 11, 2026
• QIf its core business is disrupted, does it have the DNA to reinvent itself? How does it treat mistakes and bad news?5/10

  A

  MPWR has strong "engineering self-iteration" DNA, but it has not been tested by a life-or-death reinvention. Its recent record on "mistakes and bad news" is one of "honest disclosure, but slow correction," which is exactly why the report judges "whether management is trustworthy" as uncertain. Start with the source of the reinvention DNA: the company was founded by Michael Hsing in 1997, and he remains chairman, president, and CEO. The report emphasizes a culture that "stresses system-level problem solving and continuous innovation." The past few years do show "active evolution": from individual chips toward "complete silicon-based solutions," and from consumer/industrial power into higher-barrier areas such as AI servers, optical modules, and ADAS. Enterprise Data revenue rising +97.7% year over year in 2026 Q1 is the realization of that active migration. This shows the company acts when "the core business has not yet been disrupted but the next high ground is visible."

  But the hidden premise of this linked question, "can it reinvent itself if the core business is truly disrupted," is not something MPWR has really been tested on. Its moat is engineering capability and design stickiness, and it deals with "gradual technology substitution" (the report notes that analog power changes more slowly than digital chips). It has never faced a survival test in which "the main business is disrupted all at once and the business model must be completely remade." So the "reinvention DNA" can currently only be described as "the ability to keep iterating through engineering." Its resilience under a paradigm-level shock remains an unproven open question and should not be overestimated.

  How it treats mistakes and bad news is the most relevant current evidence, and the assessment is "mixed." The positive is honesty: the company disclosed the problem proactively and did not conceal it. The company explicitly stated that investors should no longer rely on its 2024 audited financial statements and 2025 quarterly results because of an unintentional tax accounting error related to a one-time foreign tax benefit; the issue was non-cash, affected only income tax accounting, and did not involve management misconduct. The negative is the speed of correction and the effectiveness of controls: the report states that improper design and execution of controls over deferred income tax accounting led to the 2024 financial statements being restated, and internal controls were deemed ineffective for both 2024 and 2025. As of 2026 Q1, the material weakness had still not been remediated and was expected to be completed only by the end of fiscal 2026. A problem that is essentially "tax accounting" stretching roughly two years from identification to promised remediation is a negative signal for "how fast problems get fixed."

  How to weigh it: I agree with the report's judgment: "trust its operating capability more than its financial control capability." In other words, at the business level the company shows solid active evolution DNA and reacts quickly to "business opportunities." At the governance level, its correction loop for "its own mistakes" is slow and exposes a systemic weakness in the control framework. It handles "bad news" honestly but not crisply enough. For Baillie Gifford's standard of finding a great company one can entrust capital to for ten years, founder continuity and an innovation culture are positives. But "unremediated material internal-control weakness + no life-or-death reinvention test" means this item can only be scored neutral-positive, not high. The key watchpoint: can the material weakness be fully remediated within fiscal 2026 with no new restatements? That will be the most direct test of "how it really treats mistakes."

  Jun 11, 2026
• QDoes management, especially the founder, have a long-term mindset and deep alignment with the company? Is it willing to sacrifice current profit for five to ten years out?5/10

  A

  The founder's long-term mindset and tenure are beyond dispute, but "deep alignment with the company" is only moderate. The low ownership percentage is the main deduction, while incentive design is a clear positive. On long-term mindset, the evidence is strongest: Michael Hsing founded the company in 1997 and still serves as chairman, president, and CEO nearly 30 years later. The report describes the culture as "emphasizing system-level problem solving and continuous innovation," and the company rarely makes large, risky acquisitions. Cash is mainly used for R&D, working capital, dividends, and moderate buybacks. This is a typical operating style "for the long term, not for the quarter." On "willingness to invest for the long term," the company's persistently high R&D spending (the report states that 2025 R&D expense was $382 million, about 13.7% of revenue), its active transition from single chips to "complete silicon-based solutions," and the increase in capacity target from $4 billion to nearly $6 billion all show a willingness to position for five to ten years out.

  But the "deep alignment of interests" is honestly not outstanding. The report states that as of the 2026 proxy statement record date, Michael Hsing held about 1.034 million shares, or about 2.1% of shares outstanding, while all directors and executives together held about 3.6%. The weight becomes clearer against the current market capitalization: at the June 10, 2026 closing price of about $1,473 and market capitalization of about $72.4 billion, the founder's 2.1% personal stake is worth about $1.5 billion, a large absolute amount, but the percentage of the company is low. Compared with the founder-heavy ownership model Baillie Gifford truly prefers, where the founder or controlling shareholder owns a double-digit stake and sits in the same boat as outside shareholders, MPWR's 2.1%/3.6% is "material personal interest, but far from deep alignment." After nearly 30 years since founding and ongoing dilution from equity incentives, insider ownership is no longer high.

  Incentive design is a clear positive and points in the same direction as "willingness to sacrifice the present for the long term." The report states that 2025 executive annual cash incentives were 100% tied to preset non-GAAP operating profit metrics. Long-term equity incentives were 100% PSU (performance stock units), with performance measured by "revenue growth relative to the industry" and "relative total shareholder return." The company also has explicit executive stock-ownership requirements and double-trigger change-of-control arrangements. The benefit of this design is that it links executive pay to "long-term growth that beats peers + long-term relative returns," not short-term share price or one-off metrics. Structurally, that encourages management to think across multiple years.

  On the direct question of "willingness to sacrifice current profit for five to ten years out," there is indirect evidence (high R&D intensity, upfront capacity investment, and no aggressive buybacks to beautify EPS; the report notes that in 2025 the company repurchased only about 8,000 shares for roughly $6.6 million, a restrained buyback), but no dramatic case of deliberately sacrificing current profit for the long term. The company is already highly profitable and does not need to make that kind of tradeoff. Another governance flaw deserves a deduction: the material internal-control weakness related to deferred income tax has still not been remediated (see the management-trust Q&A). This is not the same as "interest alignment," but it belongs in the same question of "whether this management team is worth entrusting for ten years." The report therefore labels "whether management is trustworthy" as uncertain. Overall: strong long-term mindset, good incentive design, slow correction, and medium-to-weak ownership alignment. This is a "trustworthy operating team," but not the kind of extreme Baillie Gifford alignment where "the founder's net worth is deeply in the same boat as yours."

  Jun 11, 2026
• QIf it disappeared tomorrow, how much would customers miss it? Is its growth model sustainable and not dependent on harming society or regulation?6/10

  A

  If MPWR disappeared tomorrow, its direct customers, including server, automotive, optical-module, and industrial-board manufacturers, would miss it considerably in the short to medium term because revalidating and recertifying a power chip already designed into a platform is costly and time-consuming. But it is not "irreplaceable"; alternative suppliers of similar high-performance power solutions exist. Its growth model is sustainable and basically does not depend on social harm or crossing regulatory red lines. Start with the "indispensability" half: the report notes that once a power management chip is designed into a server motherboard, automotive platform, optical module, or industrial control board, customers will not switch easily, because revalidation takes time, engineering resources, and certification costs. This is the source of MPWR's stickiness, and it means the short-term degree of being "missed" would be high, with real friction in switching. But the report also lays out the competitive landscape clearly: the 10-K names competitors including Analog Devices, Infineon, NXP, onsemi, Power Integrations, Renesas, ROHM, Semtech, ST, and Texas Instruments, and ADI has just acquired Empower Semiconductor for roughly $1.5 billion to enter the same AI power market. So the more precise wording is: it is temporarily indispensable for "specific design slots already locked in," but for "the overall power management supply base" it is one excellent supplier, not the only lifeline. That is a different order of magnitude from a utility or operating system where disappearance causes immediate paralysis.

  Now add one hidden premise of this linked question: "social/regulatory sustainability." MPWR's growth model is clean and sustainable on this dimension. It makes money by designing higher-efficiency, higher-power-density power chips into customer platforms. The business model itself does not depend on harvesting user data, regulatory arbitrage, or creating harmful externalities. On the contrary, higher-efficiency power solutions contribute positively to "reducing power consumption per unit of data center/AI compute." As AI electricity consumption becomes a growing social issue, this is a tailwind rather than a headwind. The report says the company's products emphasize higher integration, smaller size, and higher energy efficiency, which is aligned with ESG and energy-efficiency regulation trends.

  But one real fragility within "sustainability" must be stated honestly. It is not social or regulatory, but supply chain and geopolitical. The report clearly states that the company uses a fabless model, with manufacturing, packaging, and testing mainly distributed across China, Taiwan, South Korea, Singapore, and Malaysia, and many suppliers located in China. Tariffs, export controls, and geopolitics can affect cost and delivery. That means its "growth sustainability" depends heavily on Asian supply chains and the U.S.-China technology environment. This is not a "harming society" problem. It is a problem where "the external conditions growth depends on can be interrupted by geopolitical events." The report also notes that the company added $40 million of long-term refundable deposits in 2026 Q1 and already had $60 million of wafer-supply deposits at the end of 2025, showing that it is using working capital to buy supply stability. That is the cost of supply security.

  Conclusion: customers would miss it to a medium-high degree. Design stickiness is real, but substitutes exist. Its indispensability is "slot-level, not lifeline-level." Its growth model is sustainable on social and regulatory dimensions and even benefits from the energy-efficiency trend. There is no "growth by harming others" dynamic. The long-term item to track is not moral or compliance risk, but the effect of Asian supply chains and geopolitics on "delivery and cost." That is the most realistic external variable in its growth sustainability.

  Jun 11, 2026
• QHow are the unit economics of this business (gross margin, incremental returns)? Do they improve or deteriorate with scale? Where does the money it earns go?7/10

  A

  The unit economics are excellent: high gross margin, asset-light operations, high capital returns, and profits that convert solidly into cash. As scale increases, the business broadly "maintains a high level without visible deterioration," but there is limited room for marginal improvement in incremental returns. The money it earns is mainly spent on R&D, working capital for upfront growth, dividends, and restrained buybacks. Start with gross margin and profitability, MPWR's hard foundation: the report states that 2025 revenue was $2.790 billion, gross margin 55.2%, and operating margin 26.1%. 2026 Q1 GAAP gross margin still reached 55.3%, and non-GAAP diluted EPS was $5.10, up +26.2% year over year. On a full-year basis, 2025 operating profit was $730 million, up 35.1% year over year, while revenue grew 26.4%. Operating profit grew faster than revenue, showing positive operating leverage as scale expands. The fabless model keeps the capex burden light (the report states 2025 capex was only $172 million, about 6% of revenue), which is the structural source of strong unit economics.

  Incremental returns and capital returns are both high. Under the report's methodology, 2025 ROE was about 19.2% and ROA about 16.1%. If the large cash balance is treated as non-operating, rough operating ROIC is about 27%. A company that can earn about 27% operating ROIC without interest-bearing leverage or heavy capex is highly efficient at "earning money back on each additional dollar invested." That is exactly the "high incremental return" feature Baillie Gifford likes. But the honest point is this: gross margin has stayed in the 55%–56% range for three consecutive years from 2023 to 2025 and has not continued rising with scale (after peaking at 58.4% in 2022 and then falling back). So the "unit economics improve as scale grows" flywheel is, for MPWR, more about "maintaining excellence" than "continuous improvement." It is already on a plateau, with limited room to step up further, while the industry pattern of ASP declining over life cycles will keep applying pressure.

  The quality of profit conversion into cash is high, which is the key validation that the unit economics are "real." The report notes that excluding 2024, which was heavily distorted by a one-time tax benefit, the average free cash flow/net income conversion rate from 2019 to 2025 was around 100%. 2023 is especially illustrative: revenue grew only modestly, but operating cash flow reached $638 million and free cash flow $581 million. This shows MPWR's profit is "closer to real cash profit, not just accounting appearance."

  "Where does the money go?" Capital allocation is generally rational and conservative. The first destination is R&D ($382 million in 2025, 13.7% of revenue, the lifeline for maintaining design win rates). The second is working capital, invested upfront for growth. The report states that accounts receivable rose from $173 million at the end of 2024 to $302 million in 2026 Q1, inventory rose from $420 million to $619 million, and long-term refundable deposits were added to lock in capacity. The third is dividends: 2025 dividends were $6.24 per share, and the quarterly dividend was raised to $2.00 in 2026 Q1. The fourth is restrained buybacks: the report states that in 2025 the company repurchased only about 8,000 shares for roughly $6.6 million, meaning it did not aggressively repurchase at high prices to beautify EPS. The company makes almost no large, risky acquisitions. One item to watch: share count still rose from about 43.616 million at the end of 2019 to about 49.129 million in 2026 Q1, showing that equity-incentive dilution has not been fully offset by buybacks. That is a mild drag on "per-share value" outside the unit economics.

  Overall, unit economics are MPWR's strongest side: high gross margin, high ROIC, profits that convert into cash, and rational capital allocation. It fully deserves the label "good business." The two honest caveats are that scale expansion brings "holding a high level" rather than "continuing to move higher," and equity dilution mildly erodes per-share results. The money is spent prudently, not wastefully, but there is also no highly efficient use of excess cash that materially raises per-share intrinsic value through reinvestment or major buybacks. Capital allocation is "rational but not outstanding," as the report says.

  Jun 11, 2026
• QWhat conditions must be true for it to rise fivefold in ten years? Are those conditions realistic? What expectations are embedded in today's share price?2/10

  A

  For MPWR to rise "fivefold in ten years" from today's price, almost fully maxed-out optimistic conditions must all hold at the same time. A key fact must be stated first: the share price has doubled since the report was written, raising the fivefold hurdle to a level the report's author did not even face at the time. That materially reduces the realism of "fivefold in ten years." The report's valuation anchor was about $744.90 on May 22, 2026. But as of the June 10, 2026 close, MPWR was already about $1,473, with a market capitalization of about $72.4 billion and a 52-week range of $670–$1,714, almost doubling in less than three weeks. The intrinsic value ranges given in the report were: conservative $250–$320, reasonable $380–$460, and optimistic $560–$650 per share. On that basis, it judged $744.90 to be at a "15%–33% premium even to the optimistic range" and rated the stock "Watch." Today's $1,473 is about 2.3 times the optimistic upper bound of $650. In other words, today's price is not just far above the report's reasonable value; it is more than double even the most optimistic DCF scenario.

  First answer: "what conditions must hold for a fivefold return in ten years?" Starting from about $1,473, fivefold means about $7,365, requiring market capitalization to rise from about $72.4 billion to about $360 billion. That requires all of the following: (1) revenue and Owner Earnings sustain compound growth far above the report's neutral case (the neutral assumption is 12% for the first five years and 7% for the next five), roughly approaching the report's "optimistic+" path (about 15%+ in the first phase and about 9%+ later) without interruption; (2) gross margin remains around 55% despite intensifying competition; (3) valuation multiples do not compress over ten years. The third condition is the harshest: the company currently has a trailing PE of about 87–105 and a forward PE of about 52, far above the semiconductor industry's median forward PE of about 28.75. If valuation reverts toward the industry over ten years, the share price may not rise fivefold even if earnings triple or quadruple. For all three to hold at once is not very realistic. It requires "uninterrupted high growth + high gross margin not eroded by competition + high valuation not mean-reverting" all to be right.

  Now add the embedded-premise question: "what expectations are in today's share price?" At $744.90, the report already inferred that the market was betting on a near-max optimistic path (Owner Earnings growing about 18% for the first five years, then about 10%, with friendly terminal-value and discount-rate conditions) to barely support the price. Now that the price has doubled again, embedded expectations are even higher: the current forward PE is about 52, more than 75% above the industry median. The conservative Owner Earnings yield, using the report's TTM Owner Earnings of about $630 million, is only about 0.9% on today's $72.4 billion market capitalization, while the U.S. 10-year Treasury yield was about 4.57% on May 20, 2026 as cited by the report. In other words, today's buyer gets a "starting owner earnings yield" of less than one quarter of the risk-free rate. The price is packed with almost perfect expectations: a multi-year AI power supercycle, continued share gains, and no competitive pressure on margins.

  The honest Baillie Gifford conclusion: MPWR is a good business, and AI power demand is a genuine long runway. At the business level, "doubling revenue and growing profit significantly over ten years" is not fantasy (see Q2). But "the share price rising fivefold in ten years" is a different matter. From a starting point where the share price has already doubled and valuation is already extreme, fivefold requires the double stack of "maxed-out fundamentals + no valuation mean reversion." The margin of safety is negative. This is the continuation and amplification of the report's conclusion: "Watch, not in the buy zone, reasonable buy at $280–$340." If the report already considered $744.90 expensive, $1,473 only makes the implicit expectations for "fivefold in ten years" less realistic and hands more of the return to the least controllable assumption: "valuation must not compress." The business is worth following, but at the current price, betting on "fivefold in ten years" means paying for perfection with almost no room for error.

  Jun 11, 2026
• QWhy has the market not realized all this yet? Is it because it does not understand, does not respect, or cannot look far enough? What will become the "narrative inflection point"?2/10

  A

  The market has not failed to notice MPWR. Quite the opposite: the market has recognized its quality extremely fully, possibly excessively, and translated that recognition into near-maxed-out pricing. The honest answer to this question for MPWR is the opposite of most "undervalued" names. Its problem is not that "the market does not understand, does not respect, or cannot look far enough," causing underpricing. Its problem is that "the market understands too clearly, looks too far out, and is highly enthusiastic," causing overpricing. The evidence is the price itself: the report's anchor on May 22, 2026 was about $744.90, while as of the June 10, 2026 close it was about $1,473, with market capitalization of about $72.4 billion, year-to-date gains of about 67%, and a 52-week range of $670–$1,714. A near-doubling in less than three weeks is the classic price pattern of "full market recognition + emotional pursuit," not an "undiscovered gem."

  Break down the three classic causes of underappreciation, "does not understand / does not respect / cannot look far enough," and none applies to MPWR. First, it is not "not understood": sell-side coverage is dense, and the current rating consensus is Strong Buy/Buy. The market understands its AI power narrative very well. Second, it is not "not respected": it is already an institutional favorite and a star semiconductor stock, admired rather than dismissed. Third, it is not "unable to look far enough": the market is in fact looking "too far," linearly extrapolating boom-period signals such as Enterprise Data revenue up +97.7% year over year many years into the future. A forward PE of about 52, far above the semiconductor industry median of about 28.75, is the cost of "looking too far." So "why has the market not realized" is a false premise for MPWR. It has realized it, and may have over-realized it.

  One reverse signal is worth noting: the average sell-side target price is about $1,122–$1,193, already well below the current price of about $1,473. That means even sell-side analysts, who are usually optimistic, now have a central target price below the share price. This is usually a sign that "price has run ahead of fundamental consensus and is being driven by momentum/sentiment." It matches the report's judgment that "the market has priced it as an almost perfect company," and confirms the report's strongest bear-side logic: "the real danger for investors is buying the wrong price, not misunderstanding the company."

  Now add the hidden premise of this linked question: "what will become the narrative inflection point?" Since the current narrative is "AI power supercycle + MPWR's continued share gains + high valuation justified," the candidates that can reverse the narrative downward overlap strongly with the report's "signals that would trigger reassessment": (1) AI/data center capital expenditure peaks or growth stalls. This is the sharpest bull argument today, and if it weakens, high valuation loses support. (2) Enterprise Data growth slows visibly while inventory days keep rising. The report states that receivables rose from $173 million at the end of 2024 to $302 million in 2026 Q1, and inventory rose from $420 million to $619 million. If upfront stocking meets demand cooling, cash flow and valuation can be hit at the same time. (3) Competition materially erodes excess profits. ADI's acquisition of Empower Semiconductor to enter AI power and additional investment by TI/Infineon could shake faith in the 55% gross margin if margin pressure appears. (4) Governance black swan: failure to remediate the deferred-tax material internal-control weakness or another restatement (the report states the weakness is expected to be remediated only by the end of fiscal 2026) would damage the "high-quality management" narrative. (5) Spontaneous valuation multiple compression. Even if fundamentals remain fine, a move from about 52 times forward PE toward the industry median can itself consume a large part of returns.

  The honest conclusion: for MPWR, the real "narrative inflection point" is not "one day the market suddenly realizes how good it is." That has already happened and is already priced. It is "one day the market discovers it is not as perfect as the price implies." Any of the above signals could become the trigger for a shift from "perfect pricing" to "mean reversion." Baillie Gifford is looking for "great growth stocks the market has not yet recognized." MPWR is a great growth stock, but the market has already recognized it and paid a price with no margin of safety. That makes it look more like "the right company, the wrong entry point" than an "undiscovered opportunity."

  Jun 11, 2026