Equinix Long-Term Owner's Perspective Research

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 market, or creating an entirely new one?6/10

  A

  Conclusion: The ceiling is high, but Equinix is mainly expanding an already large market that is still structurally growing, rather than creating a brand-new market from scratch. Data centers and interconnection ecosystems are mature pools of demand. Equinix's role is to use neutral interconnection + global node density to capture the highest-quality part of that market: cross-cloud, cross-network, low-latency connection relationships. It does not create demand itself; it occupies a tollbooth position within an existing wave.

  Start with how large the market is and how long it can keep growing. This is not an industry that has peaked and stalled: the IEA Energy and AI report expects global data center electricity consumption to double by 2030 to about 945 TWh, with annual growth of about 15% from 2024–2030, more than four times the electricity-demand growth rate of all other sectors combined; the CBRE North America data center market report shows vacancy fell to a historic low of about 1.4% at the end of 2025, indicating that the bottleneck is not demand but power supply and delivery. The report's conclusion that long-term demand is strong while supply-side barriers are rising is consistent with primary data. So the TAM ceiling is not a meaningful constraint over the next decade. The market itself is still compounding at a double-digit rate.

  But a high ceiling does not mean Equinix can take the whole market. This is a market with strong leaders and a long tail, not a winner-take-all greenfield market: Synergy Research Group statistics show Equinix, Digital Realty, and NTT together account for about 30% of the global colocation market, while the rest is split among regional operators and hyperscalers' self-built capacity. The report's estimate that the top 6 hold about 37% comes from the same broad methodology and is of the same order of magnitude. Both point to the same conclusion: industry concentration is not high, and Equinix is competing for the most valuable customer layer in a fully competitive existing market.

  Equinix's own positioning also confirms that it is expanding an existing market, not opening a new one. The report states that 2025 total revenue was $9.217 billion, including $6.475 billion from colocation and $1.655 billion from interconnection, with recurring revenue at about 94.8% of the total. This is a business that sells power, space, and connectivity to existing enterprise IT, cloud, and network customers. What is truly scarce, and what lets Equinix capture a thicker slice of the market, is the neutral interconnection ecosystem repeatedly emphasized in the report: the Equinix investor relations page confirms the platform has more than 500,000 live interconnections, which it says is the most in the industry. This is quality upgrading and ecosystem positioning on top of existing demand, not the creation of a previously nonexistent demand category.

  One honest boundary point: the only incremental area with a new-market flavor is AI inference and neocloud deployments. The Equinix Q1 2026 results release says about 60% of the largest deals were AI-related. Even for AI, though, the essence is still larger-scale, higher-density compute colocation demand flowing into the existing data center track. It makes the market larger; it does not build a separate business from scratch. On Baillie Gifford's market-ceiling question, my judgment is: the ceiling is high enough and unconstrained over the next decade, but Equinix's growth story rests on using its moat to capture the highest-quality portion of a crowded large market, not on monopolizing a brand-new market. This business does not have the explosive repricing potential implied by the latter.

  Jun 10, 2026
• QCan its revenue at least double over the next five years? Will growth be driven mainly by volume, price, or new businesses?4/10

  A

  Conclusion: A revenue doubling over the next five years, requiring about 15% annualized growth, is highly unlikely. At the company's own pace, cumulative growth of about 50%–60% is more plausible. Growth will be driven mainly by volume: new capacity + customer deployment ramp. Price, through rate increases and power pass-through, and new businesses such as AI/neocloud, Fabric, and xScale are secondary but accelerating engines. This is a high-quality compounder, not a high-beta growth stock capable of doubling revenue in five years.

  Use primary data first to anchor the true growth slope. The Equinix Q1 2026 results release gives full-year 2026 revenue guidance of $10.144–10.244 billion. Against 2025 revenue of $9.217 billion, that implies about 10%–11% year-over-year growth; MRR grew 12% year over year, or 10% on a comparable-currency basis. The report also calculates a 2019–2025 revenue CAGR of about 8.8%. Extrapolating that 8%–12% compound slope for five years produces cumulative growth of about 47%–76%, implying 2030 revenue of roughly 1.35–1.6 times the current level. Can revenue at least double? A five-year doubling requires about 15% annualized revenue growth, materially above current guidance and the historical path. Unless AI demand drives capacity delivery far beyond today's run rate, a five-year revenue doubling is a low-probability outcome. The honest answer to this question is: it probably cannot double.

  In the growth breakdown, volume is the clear main driver. The report states that total 2026 capex is expected to be about $4.1 billion, including about $3.8 billion of non-recurring, expansionary capex. Almost all of that money goes into new data centers, incremental power capacity, and additional sellable cabinets. In other words, capital is exchanged for future sellable capacity. This model of heavy investment first and volume later means the revenue ceiling is determined by how fast capacity can be built and how much power can be connected. Industry supply is precisely the constraint: the CBRE report shows vacancy at only about 1.4%, which means newly delivered capacity can be leased, but also means delivery speed, not demand, is the real constraint on revenue growth.

  Price is a real but limited second driver. The report says the company has shown power pass-through and premium-node pricing ability in an inflationary environment. The Equinix Q1 2026 release also confirms record adjusted EBITDA margin of 51%, showing operating leverage and some pricing power. But the report's investment checklist is honest that its pricing power is not unlimited; it is closer to partial cost pass-through + premium-node uplift. Price will not be the main engine of a revenue doubling.

  New business is the most exciting part of the growth rate, but its share is still relatively small. Interconnection and Fabric are higher-quality increments: the same Q1 2026 release shows interconnection revenue rising from $393 million in 2025Q1 to $446 million in 2026Q1, up about 13% year over year and faster than total revenue. AI/neocloud demand is strong, with about 60% of the largest deals related to AI, and the Equinix investor relations page says 80% of leading AI model providers and 40% of leading neoclouds are expanding with it. The xScale joint ventures absorb hyperscale demand off balance sheet. But even if these businesses grow quickly, interconnection was only about 18% of total revenue in 2025, so they are not enough by themselves to double total revenue within five years.

  Overall, Equinix's revenue over the next five years is more likely to deliver high-quality, visible cumulative growth of about 50%–60%, underpinned by volume from capacity buildout and accelerated by price plus AI/interconnection new businesses. But the doubling threshold is high. This fits the report's broader conclusion: business quality is high and growth is real, but elasticity is not sufficient to support high-growth-stock valuation imagination. That is one of the root reasons it is rated Watch rather than Buy.

  Jun 10, 2026
• QFive years from now, what will take over as the next growth engine? Does this second curve exist today?4/10

  A

  Conclusion: Equinix does not have a brand-new second curve detached from its core business, but it does have real next-level engines inside the main curve that are already contributing revenue today: AI/neocloud capacity, interconnection and the Fabric software layer, and xScale hyperscale joint ventures. Strictly under Baillie Gifford's question of what takes over five years from now, these are more like extensions and upgrades of the main lane than new growth poles built from scratch. The handoff has high certainty, but limited imagination.

  Start with the next-level engines that already exist and can be quantified today. First is AI and neocloud deployment: the Equinix Q1 2026 results release says about 60% of the largest deals were AI-related, and the Equinix investor relations page says 80% of leading AI model providers and 40% of leading neoclouds are expanding with it. This is not a slide-deck vision; it is real demand already entering backlog. Second is interconnection and the Fabric software layer: the same Q1 2026 release shows interconnection revenue rising from $393 million in 2025Q1 to $446 million in 2026Q1, up about 13% year over year and faster than total revenue. The report also notes that Fabric and other software-defined connectivity products are stickier, higher-quality increments. Third is the xScale joint ventures: the report states that in 2024 the company created a new xScale JV with GIC and CPP Investments, with planned total investment capacity above $15 billion and more than 1.5GW of future U.S. capacity. This is a capacity second curve that absorbs hyperscale demand through an off-balance-sheet structure.

  But the honest answer to the hidden premise in this Baillie Gifford question, whether this second curve is an independent new growth pole separable from the core business, is no. All three grow from the same root: global nodes + neutral interconnection ecosystem. AI deployment is essentially higher-density colocation demand entering existing data centers. Interconnection/Fabric software-izes existing connection relationships and sells them again. xScale is the wholesale extension of the same site selection, construction, and operating capabilities. They make the main curve steeper and longer, but they do not create an independent second engine that can keep the company alive if the core business is disrupted. The risk section of the report actually exposes this point: if AI demand flows more toward hyperscale self-builds and large low-interconnection-density campuses, Equinix's interconnection premium could be bypassed. In other words, its next-level engines live and die with the moat of the core business; they do not diversify the risk like a true second curve.

  Put these engines into the financial picture to see the real strength of the handoff. The report calculates a 2019–2025 AFFO CAGR of about 11.8% and operating cash flow CAGR of about 11.9%. The Equinix Q1 2026 release gives 2026 AFFO/share guidance of $42.31–43.11, up 10%–12% year over year. So after the next-level engines take over, the most realistic result is to maintain overall growth in the high single digits to low double digits, not to reignite a period of explosive high growth. That is the nature of this business: the engine can be replaced, but the replacement is the same model with the same horsepower.

  One implicit premise needs to be added: the condition for this second curve to deliver. For the engine to truly take over, incremental capex must translate into sustained growth in AFFO/share and Owner Earnings per share. The report repeatedly flags the risk: 2026 total capex is about $4.1 billion, including about $3.8 billion of expansionary capex. If AI/xScale project returns fall below the cost of capital, the result will be larger scale without corresponding per-share value growth. In other words, the second curve does exist today and points in the right direction, but the quality of the handoff depends on capital-allocation discipline, not demand itself.

  Overall, the engines that can take over five years from now are visible and tangible today: AI capacity, the interconnection software layer, and xScale. This is where Equinix is stronger than purely cyclical assets. But it does not have an independent second curve that can pivot away if the core is disrupted; all increments attach to the same moat. For long-term owners, that means the handoff has relatively high certainty, but the ceiling remains tied to the core business. This is consistent with the report's overall judgment: excellent but not cheap, with real growth but limited elasticity.

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

  A

  Conclusion: Equinix's core competitive advantage is a compound moat combining neutral interconnection ecosystem + global node density + scarce resources in key metros, not a single brand or cost advantage. Over the next three to five years, this moat will most likely remain wide and slowly widen, rather than narrow. The report gives the moat a strength score of 4.5/5. I agree that this judgment holds up, but it is important to state that the widening is purchased through continuous heavy capital investment; it is not a naturally self-thickening moat.

  The true source of the moat is network effects, not the data centers themselves. The report says about two-thirds of recurring revenue comes from customers deployed in more than 10 IBX data centers. Once these customers migrate, they face a full rebuild of cabinets, networks, compliance, low-latency architecture, and business continuity, making switching costs extremely high. More important is interconnection density: the Equinix investor relations page confirms the platform has more than 500,000 live interconnections, the most in the industry; the report also says the company has close to 40% market share in private on-ramps to major cloud service providers. The more customers, the denser the interconnections, and the more cloud on-ramps, the more valuable the platform becomes for new customers. This is a classic network effect that reinforces itself with scale, and it is the core barrier separating Equinix from wholesale data centers.

  Why should the moat widen rather than narrow over the next three to five years? Three pieces of primary evidence support that direction. First, supply barriers are rising, which benefits incumbents that already hold the right locations: the CBRE report shows data center vacancy falling to a historic low of about 1.4%, and the IEA Energy and AI report expects data center electricity use to grow about 15% annually from 2024–2030 and double to about 945 TWh by 2030. The scarcer power and land become, the more valuable Equinix's locked-in key metro resources become, and the harder they are for new entrants to replicate. Second, the interconnection ecosystem is still densifying: the Equinix Q1 2026 results release shows interconnection revenue rising from $393 million in 2025Q1 to $446 million, up about 13% year over year and faster than overall revenue, indicating that the most valuable layer of the moat, connection relationships, continues to thicken. Third, pricing power and operating leverage are improving: the same release shows record adjusted EBITDA margin of 51% and 12% MRR growth, proving that in a strong demand environment the company can pass through costs and enjoy scale benefits.

  But two boundaries of the moat must be stated honestly to avoid overstating the wide-moat story. First, this is not a brand- or patent-based moat. It is built from capital + time + customer density. Replicating it requires tens of billions of dollars, multiple economic cycles, and years of operating history. That is why it is wide, but it also means its width requires continuous heavy reinvestment to maintain, rather than thickening while the company sits still. Second, the most realistic narrowing risk comes from customer mix migration: the report's risk section clearly warns that if more AI demand goes to hyperscale self-builds and large low-interconnection-density campuses, and if the customer mix tilts toward lower-margin hyperscale, interconnection share could stagnate and the most valuable layer of the moat could be bypassed. This is the only structural factor likely to narrow the moat at the margin over the next three to five years, and it should be tracked.

  A competitive comparison further locates the purity of the moat. The closest public comparable, Digital Realty, reported in 2026Q1 that Core FFO/share guidance was raised to $8.00–8.10 and net debt/adjusted EBITDA was 4.7x, with revenue of about $1.64 billion. It is very strong in large-scale wholesale/hyperscale capacity, but weaker than Equinix in retail neutral interconnection and ecosystem density. The report's view that Equinix has higher quality in interconnection ecosystem, customer density, and cross-cloud connection value holds up after verification: Equinix's moat has higher purity because it is built on the hardest-to-replicate connection relationships, not on the easiest-to-build capacity that capital can fund.

  Overall, the core advantage is a compound interconnection network effect. Over the next three to five years, the direction is slow widening, supported by supply scarcity + interconnection densification + operating leverage, and the 4.5/5 moat-strength score is justified. The one narrowing signal to watch is customer migration toward low-interconnection-density hyperscale. This moat is wide and durable enough to support the label of a good business. Its width, however, is maintained by sustained high capex, fully consistent with the report's judgment that the company must invest before it grows.

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

  A

  Conclusion: Equinix has some record of reinvention, evolving from simple colocation into a compound system of interconnection + Fabric + xScale + REIT structure. But its reinvention DNA is gradual product and capital-structure upgrading, not the kind of life-or-death pivot that rebuilds a company from the ruins after its core is disrupted. Its handling of mistakes and bad news is generally candid and governed by controls, but the 2024 impairments and restructuring show that it can still make capital-allocation mistakes. Under Baillie Gifford's added implicit premise, the ability to reinvent when the core business is disrupted, this is a relatively weak and not yet truly stress-tested dimension.

  Start with the positive evidence of reinvention: this is not a company standing still. The report describes how Equinix has evolved from selling power and space into a compound system of colocation + interconnection + Fabric + xScale JV + REIT tax structure. That reflects active product and capital-structure upgrades over more than the past decade. The most recent important structural evolution is using joint ventures to absorb hyperscale demand: the report states that in 2024 it created a new xScale JV with GIC and CPP Investments, with planned total investment capacity above $15 billion, placing the heavy-capital, low-interconnection-density hyperscale portion off balance sheet while preserving the higher-quality parent platform. Facing the AI wave, the company also quickly adjusted its customer mix to serve neocloud demand. The Equinix investor relations page says 80% of leading AI model providers and 40% of leading neoclouds are expanding with it. All of this shows an ability to reallocate resources in line with technology waves.

  But the honest answer to what Baillie Gifford is really asking is this: if the core is disrupted, can it rebuild from the rubble? Equinix has not truly been tested on that front, and its DNA appears relatively weak. There are 2 reasons. First, all of its adaptability grows from the same moat: interconnection ecosystem + node density. The report's risk section clearly warns that if AI demand shifts toward hyperscale self-builds and large low-interconnection-density campuses, bypassing the interconnection premium, the company would mostly be passively absorbing lower-margin hyperscale rather than opening a new path unrelated to interconnection. This is a machine that can be upgraded but is hard to transform. Second, it is a heavy-asset REIT: the report states that 2026 total capex is about $4.1 billion and net PP&E was $24.169 billion in 2026Q1. Such a heavy asset base means that once it bets in the wrong direction, the cost of turning is extremely high and the speed is slow. It does not have the rapid reinvention flexibility of an asset-light company.

  Now consider how it handles mistakes and bad news, which can be observed directly through governance and financial disclosure. The positive signal is that the company has institutionalized correction and accountability. According to the report's summary of the 2026 proxy statement, the company has clawback policies, hedging/pledging governance sections, and incentive metrics built around revenue, AFFO/share, and relative TSR, and it explicitly uses AFFO/share to drive non-dilutive growth. This design means mistakes, such as dilutive expansion, are penalized in compensation. It is a built-in correction mechanism.

  The more persuasive evidence is how it handled a real piece of bad news: the 2024 impairment and restructuring. The report states that 2024 GAAP operating margin fell to about 15.2%, mainly because of asset impairments, restructuring, and transaction costs, while 2025 quickly recovered to about 20.0% and adjusted EBITDA margin rose from 45% to 49%. The Equinix Q1 2026 release further shows 2026 guidance of about 51%. The report's interpretation is important and candid: 2024 was not a moat-collapse year, but more like a noisy GAAP year. The company chose to impair and clear problematic assets rather than cosmetically smooth the issue, which is a healthy way to handle bad news. Conversely, the impairment itself proves that the company can make capital-allocation mistakes. The report states plainly that when a global asset portfolio is large enough, capital-allocation errors do not show up immediately in revenue; they usually appear first in substandard asset returns, project adjustments, impairments, and restructuring.

  Overall, Equinix is an excellent gradual evolver and a candid corrector of mistakes, but not a reinvention machine that can survive core disruption and be reborn from fire. Its handling of mistakes includes clawbacks, timely impairments, and transparent disclosure, which deserves trust. But its heavy-asset, single-moat nature means a true disruptive shock would be its most fragile scenario. This is consistent with the report's broader judgment: the business is reassuring, but growth depends heavily on capital-allocation discipline. Reinvention DNA is not its strength; capital discipline is what must be monitored continuously.

  Jun 10, 2026
• QDoes management, especially the founder, have a long-term view and deep alignment with the company? Is it willing to sacrifice current profit for the next five to ten years?4/10

  A

  Conclusion: Equinix is run by an excellent professional management team, not a shareholder-founder team. Governance is sound, incentives are reasonably designed, and the time horizon is long, but insiders own less than 1% of the company, so their deep alignment with the business is clearly weaker than the owner-operator model Baillie Gifford prefers. The report gives management and capital allocation a 3.5/5 score. I agree with this middling assessment after verification: the company is willing to invest for the next five to ten years, and is doing so through heavy-capital expansion + xScale JV, but for Equinix, sacrificing current profit for the long term is more a business-model necessity than a founder-like act of self-sacrifice.

  Start with the alignment Baillie Gifford values most. This is precisely Equinix's weakest area. The report states that as of March 20, 2026, CEO Adaire Fox-Martin held about 20,630 shares, Keith Taylor (CFO) held about 26,587 shares, and all current directors and executives together held 265,917 shares, less than 1% of total shares outstanding. At the current share price of about $1,059.84 on June 9, 2026, the CEO's personal shareholding is worth about $20 million. That is not small for a large-company CEO, but relative to a company with a market cap of about $100 billion, it looks like an excellent salaried manager, not an owner who has most of their net worth in the stock. The report's wording is honest: this does not mean management lacks integrity, but it means this is not the owner-operator model in which management and shareholders have most of their net worth riding together in the stock. Under the Baillie Gifford framework, this is a material deduction.

  But long-term orientation and alignment through institutional constraints are solid strengths. The governance framework is standard and disciplined: according to the report's summary of the 2026 proxy statement, the company requires directors to hold at least 6 times their annual cash retainer, the CEO to hold 6 times salary, and other executives to hold 3 times salary. As of the end of 2025, covered executives were either compliant or within the allowed period to reach compliance. These mandatory ownership requirements partially offset low voluntary ownership. Incentive design also tilts toward per-share value rather than simple scale: according to the proxy summary in the report, annual and long-term incentives include revenue, AFFO/share, and total shareholder return, with 2025 bonus payout at 99% of target. AFFO/share is explicitly used to drive non-dilutive growth, and long-term incentives include relative TSR. One positive detail is pay restraint: 2025 base salaries for both the CEO and CFO were not increased.

  Is the company willing to sacrifice current profit for the next five to ten years? Yes, but the nature of that sacrifice matters. The evidence is hard: the report states that 2026 total capex is expected to be about $4.1 billion, including about $3.8 billion of non-recurring expansionary capex, and the Equinix Q1 2026 results release confirms that the company continues to invest heavily in capacity expansion. Reported free cash flow was therefore -$2.572 billion in 2025. Management is sacrificing current reported free cash flow to build the future recurring revenue base. That is a real long-term orientation. But for a heavy-asset REIT, continuous heavy investment is an inherent requirement of the business model. Without investment, there is no growth. So this sacrifice is not quite the same as a founder voluntarily suppressing short-term profits to invest in the long term.

  Capital allocation combines rationality with latent risk. The rational side: the report says the company mainly allocates cash to reinvestment, dividends, acquisitions, and debt management. It has raised its dividend for 11 consecutive years and lifted the 2026 quarterly dividend to $5.16, or about $20.64 annualized, and it uses the xScale JV, with planned total investment capacity above $15 billion, to place hyperscale demand off balance sheet and optimize the parent platform's balance sheet. This is rational structural thinking. The concern: the report states plainly that the company is naturally prone to loving a larger platform scale rather than expanding strictly around per-share intrinsic value, and it lists scale worship as a risk to monitor. Stock-based compensation is a real shareholder cost as well, about $498 million in 2025, diluting per-share value.

  Overall, management deserves trust, has a clear long-term view, incentives are built around per-share value, and dividend discipline is excellent. These are real strengths, and the 3.5/5 score is justified. But under Baillie Gifford's high standard of founder-level deep alignment, Equinix is an excellent professional-management case rather than an owner-led team. Insider ownership of <1% is a structural weakness. This fully matches the report's judgment: governance is sound and incentives are reasonable, but insider ownership is low. You are buying a good company run by excellent professional managers, not a founder-led company whose leaders share life-or-death exposure with you.

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

  A

  Conclusion: If Equinix disappeared tomorrow, its core customers would miss it badly, especially multi-site enterprises that have built cross-cloud and cross-network architecture on its interconnection ecosystem. Migration costs are extremely high and short-term substitution is almost impossible. Its growth model is also broadly sustainable and not built on harming society or fighting regulation. The only real social/regulatory tension comes from the industry-level issue of power consumption. Under Baillie Gifford's added dual premise, indispensability + social/regulatory sustainability, Equinix stands up on both dimensions, though power sustainability is a variable that must be watched over the long term.

  Start with hard evidence of indispensability. Equinix's customer stickiness does not come from a single service; it comes from an embedded network position. The report says about two-thirds of recurring revenue comes from customers deployed in more than 10 IBX data centers. These customers build their cabinets, networks, compliance, low-latency architecture, and business continuity around Equinix. If it disappeared, customers would not just be looking for a new landlord; they would be rebuilding an entire cross-region, cross-cloud connection topology. Interconnection scale makes the point even clearer: the Equinix investor relations page confirms the platform has more than 500,000 live interconnections, the most in the industry, and the report also says the company has close to 40% market share in private on-ramps to major cloud service providers. That means it is one of the physical chokepoints through which many enterprises reach AWS/Azure/GCP. This chokepoint position is exactly the kind of indispensability Baillie Gifford cares about.

  Operating resilience further confirms how needed it is. The report states that from 2020–2025, company revenue never declined year over year, and during the 2020 pandemic year revenue and AFFO still grew 8% and 13%, respectively. The Equinix Q1 2026 results release shows recurring revenue represented about 94.8% of 2025 revenue and MRR grew 12% year over year. Customers do not leave during macro weakness, and they add capacity during the AI wave, with about 60% of the largest deals AI-related. That is the clearest financial expression of indispensability. One boundary should be stated honestly: indispensability is strongest for retail customers with multi-site + high-interconnection-density needs. For hyperscale customers that only lease large blocks of wholesale capacity, Equinix is relatively more replaceable. This is the other side of the report's repeated warning about customer mix tilting toward low-interconnection-density hyperscale.

  Now address the other half of what Baillie Gifford is really asking: whether the growth model is sustainable and not dependent on harming society or fighting regulation. This is a relatively clean dimension for Equinix. Its business model does not rely on regulatory arbitrage or user harm. It is not gambling, abusive data extraction, or high-interest lending. It charges for infrastructure that supports the digital economy, and its social utility is positive. The report also shows sound governance and incentives built around per-share value and relative TSR, with no reliance on aggressive accounting or shareholder harm to grow. From a social-sustainability perspective, it is closer to a digital infrastructure toll road. Growth sources, including cloud migration, AI, and compliant localized deployment, are real long-term demands rather than extractive growth.

  The only real social/regulatory tension is concentrated in the industry-level issue of power, and it must be stated plainly. The IEA Energy and AI report expects global data center electricity use to grow about 15% annually from 2024–2030 and double to about 945 TWh by 2030, including about 240 TWh of incremental U.S. demand, or about 130% growth. This surge in electricity consumption is drawing public-policy attention around grid load, residential power prices, and carbon emissions, and could eventually translate into stricter energy-use permits, power allocations, or carbon regulation. This is not unique to Equinix; it is an industry-wide issue. But it is the variable most likely to tighten the social sustainability of its growth model. The report also lists power and construction constraints, and adverse changes in the energy regulatory environment, as risks and re-rating triggers.

  Overall, Equinix's indispensability is real and high, especially for high-interconnection-density customers. It is one of the few infrastructure platforms that would be strongly missed if it disappeared tomorrow. Its growth model does not harm society or fight regulation, and social sustainability is broadly good. The only long-term tension is power. That is both an industry moat, because supply scarcity helps incumbents, and a potential social/regulatory constraint if energy use is tightened. Balancing both sides, Equinix stands up on this Baillie Gifford dimension, consistent with the report's judgment that the industry is attractive and long-term demand is clear, but power supply is the core constraint.

  Jun 10, 2026
• QWhat are the unit economics of this business: gross margin and incremental returns? Do they improve or deteriorate with scale? Where does the money it earns go?5/10

  A

  Conclusion: Equinix's unit economics are excellent at the cash gross margin/EBITDA margin level and improve with scale, but only solid, not exceptional, at the capital-return or incremental-return level. This is a heavy-asset business with high margins and low capital turnover. The money it earns mainly goes into reinvestment for expansion, dividends, and debt management, rather than buybacks. In the investment checklist, the report honestly marks whether capital returns are excellent as uncertain. I agree with that restrained judgment after verification.

  Start with the genuinely excellent side of unit economics: margins and operating leverage. The Equinix Q1 2026 results release shows adjusted EBITDA of $1.245 billion and a record margin of 51%, with full-year guidance also around 51%. The report says adjusted EBITDA margin rose steadily from 45% in 2023 to 49% in 2025. This upward path is evidence that unit economics improve as scale grows: the fixed base of data centers, power, and operations is spread across more revenue, and the landing margin on incremental revenue is high. Interconnection is the highest-margin, highest-quality increment. It is almost a pure charge for connection relationships. The report says interconnection was about 18% of total revenue and about 18.9% of recurring revenue in 2025, and the same Q1 2026 release shows interconnection revenue up about 13% year over year, faster than total revenue. So under a gross margin/EBITDA lens, the conclusion that unit economics improve with scale is valid.

  But the key in this Baillie Gifford question is incremental returns: how much return does one more dollar of capital generate? That is precisely Equinix's weak spot. It is a very low-capital-turnover business. The report states that net PP&E was $24.169 billion in 2026Q1 and total assets were $40.898 billion, while 2025 revenue was only $9.217 billion. In other words, the company must first sink a huge asset base to generate relatively limited revenue. The report directly marks whether capital returns are excellent as uncertain and explains that project-level returns are attractive, but GAAP enterprise-level ROIC is not impressive. This is an honest and critical distinction: a single new data center project may produce decent returns, but after goodwill, historical acquisitions, and immature new assets are included, capital returns at the enterprise level are diluted. These are not the dream unit economics of a consumer company with light assets, high turnover, and high ROIC.

  Whether unit economics improve or deteriorate as scale grows needs a two-layer answer, not a single broad claim. At the margin level, they improve, as the record 51% EBITDA margin proves. At the capital-return level, they are roughly flat and may even deteriorate at the margin if the share of low-interconnection-density hyperscale rises. The report's risk section clearly warns that if the customer mix tilts toward lower-margin hyperscale, total capex keeps rising, and Owner Earnings per share fails to keep up, incremental returns will come under pressure. So unit economics do not improve monotonically. They combine upward margins with flat to pressured capital efficiency.

  Where does the money it earns go? This is where unit economics translate, or fail to translate, into shareholder value, and the evidence is clear. The report says capital mainly goes to 4 places: ① reinvestment for expansion is by far the largest, with 2026 total capex expected at about $4.1 billion, including about $3.8 billion of non-recurring expansionary capex; ② dividends, as the company has raised its dividend for 11 consecutive years and set its 2026 quarterly dividend at $5.16, about $20.64 annualized, or about a 1.95% dividend yield at the current share price of about $1,059.84; ③ acquisitions such as atNorth and xScale JV, with planned total investment capacity above $15 billion; and ④ debt management and maturity extension. One notable feature is that the company basically does not buy back stock and prioritizes using cash to widen the moat. The report also notes that this means shareholder returns depend heavily on whether these reinvestments truly translate into growth in AFFO/share and Owner Earnings, rather than being supported by buybacks or current cash distributions.

  Only by looking at true distributable cash can the quality of the unit economics be seen clearly. The report makes a conservative Owner Earnings estimate: 2025 AFFO to common shareholders was about $3.761 billion, and after deducting stock-based compensation of about $498 million, conservative Owner Earnings were about $3.1–3.3 billion. In other words, AFFO overstates truly distributable cash because it adds back SBC. This is a necessary deduction in the unit-economics analysis: a headline 51% EBITDA margin looks attractive, but after deducting real shareholder costs, per-share free cash is not as effortless as the numbers suggest.

  Overall, Equinix's unit economics fit the classic heavy-asset platform profile: high margins + scale dilution benefits + low capital turnover + only solid capital returns. The money it earns is almost entirely reinvested in expansion and dividends, with no buybacks, and shareholder value realization depends entirely on whether incremental capital can continue to produce returns above the cost of capital. This is fully consistent with the report's judgment: capital returns are uncertain, growth requires investment first, and AFFO/share must be monitored to see whether high capex and dilution are consuming value. Margin is the long board in the unit economics; capital efficiency is the short board that deserves continuous scrutiny.

  Jun 10, 2026
• QWhat conditions must all hold for it to rise fivefold over ten years? Are those conditions realistic? What expectations are embedded in today's share price?3/10

  A

  Conclusion: For Equinix to rise fivefold over ten years, about a 17.5% annualized share-price return, AFFO/share and Owner Earnings per share would need to compound far faster than current guidance, while valuation multiples do not contract and may even expand. Both conditions look unrealistic at the current price and with the current fundamentals. Today's share price of about $1,060 already embeds something close to the report's optimistic scenario, not an undervalued starting point that leaves room for a fivefold gain. Put plainly: a ten-year fivefold outcome is a low-probability case for Equinix. The report rates it Watch rather than Buy precisely because the price has already pulled forward a meaningful portion of the future.

  First quantify what a fivefold gain requires. A fivefold share-price gain over ten years is about a 17.5% annualized price return. Share-price return equals per-share cash-flow growth × multiple change + dividends. What is current AFFO/share growth? The Equinix Q1 2026 results release gives 2026 AFFO/share guidance of $42.31–43.11, up 10%–12% year over year, and the report calculates a 2019–2025 AFFO CAGR of about 11.8%. Even assuming optimistically that AFFO/share compounds at about 11% for the next ten years, AFFO/share would only increase about 2.8 times over the decade. To reach a fivefold total return, the valuation multiple would need to expand materially from its current level, on top of about 2% dividend contribution. In other words, a fivefold outcome requires two things to hold simultaneously for ten years: 11%+ per-share growth that does not decelerate, and a 25x forward AFFO multiple that does not fall and may even rise.

  Are these conditions realistic? Examined one by one, both are demanding. First, sustained per-share growth is hard. The report repeatedly warns that growth depends heavily on continuous heavy capital investment, with 2026 total capex of about $4.1 billion and expansionary capex of about $3.8 billion. If customers tilt toward low-interconnection-density hyperscale and xScale project returns fall below the cost of capital, AFFO/share growth could slow to the low single digits. Second, multiple expansion is almost a bet against gravity. The current share price implies about 25x forward AFFO, using the current price of about $1,060 divided by 2026E AFFO/share of about $42.7. That is already above the multiple implied by the closest comparable, Digital Realty's Core FFO/share guidance of about $8.00–8.10. The report judges that this ecosystem-quality premium has already been priced in. For the multiple to expand further from an already high level, market sentiment would need to remain euphoric, which is the opposite of an undervalued starting point that leaves room for a fivefold gain.

  Now answer the hidden core of this Baillie Gifford question: what expectations are embedded in today's share price? The report's own Owner Earnings discount work is the most persuasive evidence. It gives three intrinsic-value cases: conservative about $540, neutral about $808, and optimistic about $1,122, expressed as a fair range of $700–900 and an optimistic range of $950–1,150. The current share price of about $1,060 sits in the upper-middle part of the optimistic range, and the report clearly states that anchored to the neutral value of $808, the current price is about 34% higher. In other words, today's share price embeds expectations that the market is already pricing it close to the optimistic scenario: AI demand is realized smoothly, and high growth + high returns + high multiples all hold together for a long time. This is exactly the basis for the report's conclusion that the margin of safety is insufficient.

  Complete the picture with risks in both directions. The report calculates prospective ten-year annualized returns from the current price at 1%–3% in the conservative case, 4%–6% in the neutral case, and 7%–9% in the optimistic case. Even the company's optimistic case reaches only 9% annualized, far below the roughly 17.5% required for a fivefold gain. That means the ten-year fivefold outcome does not appear in any of the report's 3 scenarios. Achieving it would require an AI supercycle far beyond the report's optimistic assumptions plus sustained multiple expansion, a blue-sky tail case rather than a base expectation. Conversely, downside risk is real: the report warns that if valuation falls from the current high level to a more ordinary REIT range of 18–22x, combined with slower growth, long-term holders could face a 40%–50% permanent capital loss.

  Overall, a ten-year fivefold gain requires AFFO/share to compound far above guidance while the multiple does not contract and instead expands. Starting from about 25x forward AFFO, that is not realistic. Today's share price of about $1,060 embeds optimistic expectations that an excellent company is already fully, perhaps excessively, recognized rather than undervalued. Under Baillie Gifford LTGG's standard for a great growth stock capable of rising fivefold over ten years, Equinix is a good company worthy of the word great, but the current price leaves no room for a fivefold outcome. This is fully consistent with the report's core conclusion: excellent company, bad price.

  Jun 10, 2026
• QWhy has the market not realized all of this yet? Is it too hard to understand, too easy to dismiss, or too distant? What could become the narrative inflection point?3/10

  A

  Conclusion: For Equinix, this Baillie Gifford question has to be inverted. The market has not failed to realize how good it is; the market has already realized it fully, perhaps excessively. The problem is not that the company is misunderstood, dismissed, or underappreciated for the long term. It is understood too clearly and priced too fully. The real narrative inflection point is therefore two-sided, and a downward inflection, where the high multiple is punctured, is more likely than an upward inflection where the company is rediscovered. This matches the report's core view that an excellent company is already fully recognized and lacks valuation margin of safety. It is the honest answer when applying Baillie Gifford's market-not-yet-aware question to a crowded consensus stock.

  First, the evidence that the market has already realized it is strong. One, valuation already includes optimistic expectations: the report calculates that the current share price equals about 25.3x 2026E AFFO/share and is about 34% above the neutral intrinsic value anchor of $808. The Equinix Q1 2026 results release gives 2026 AFFO/share guidance of $42.31–43.11, and at the current share price of about $1,060, the forward multiple is about 25x. This is not an ignored or undervalued price. Two, coverage and institutional awareness are extensive: Equinix is a large blue-chip REIT with a market cap of roughly $10.2–10.4 billion at the current price of about $1,060, a benchmark data center REIT with dense sell-side coverage and a widely known AI narrative. There is no neglected awareness gap. Three, the AI narrative is already priced in: the Equinix Q1 2026 release says about 60% of the largest deals were AI-related and the company raised full-year guidance accordingly. The good news has occurred, and the market has reacted.

  Do Baillie Gifford's three mispricing categories, too hard to understand, too easy to dismiss, or too distant, still apply to Equinix? One by one, they do not. Too hard to understand: there is complexity, but it has been digested. The report acknowledges that REIT accounting, AFFO adding back SBC, and xScale off-balance-sheet structures make the company harder to understand than a utility, but that complexity is precisely what professional institutions have already studied in depth. Retail-level difficulty does not create systemic undervaluation. Too easy to dismiss: no. It is the industry leader, has raised dividends for 11 consecutive years, including a 2026 quarterly dividend of $5.16, and has record adjusted EBITDA margin of 51%. The market is not dismissing it. Too distant: in fact, the market is looking too far ahead. The current price pulls in high growth, strong returns, and high multiples from ten years out. The problem is not that the market cannot look far enough, but that it treats the optimistic long-term case as the base case.

  If there is one place where the market has not fully seen the truth, it is the only defensible perception gap, and its direction is negative. The report repeatedly points out a truth that can be obscured by AI excitement: AFFO adds back stock-based compensation, about $498 million in 2025. Under the more conservative Owner Earnings lens, about $3.1–3.3 billion and implying a higher true multiple, the company is not as cheap as headline AFFO suggests. Reported free cash flow has long been suppressed by expansionary capex, reaching -$2.572 billion in 2025. Under AI enthusiasm, what the market may not fully realize is not how good Equinix is, but that its truly distributable cash quality is weaker than headline AFFO and that growth must be bought through constant heavy reinvestment. That is a perception gap that should make investors more cautious, not more excited.

  Now address Baillie Gifford's added implicit premise: what could become the narrative inflection point? I think inflection points are two-sided, with the downside more likely. Downside inflection, the more likely case: if the report's re-rating triggers appear, the high-multiple narrative can be punctured. These include interconnection revenue and interconnection scale decelerating clearly for multiple quarters, recurring revenue share declining, 2026–2028 capex staying high while AFFO/share growth falls to low single digits, net debt/EBITDA rising with tighter liquidity, and the customer base concentrating noticeably in a few hyperscalers with compressed returns. If any of these occur, the market could move from a 25x forward AFFO multiple back to a more ordinary 18–22x REIT range. The report warns that this kind of multiple compression plus slowing growth could cause 40%–50% permanent capital loss. Upside inflection, a lower-probability case: the market would only revise upward further if AI/neocloud demand materializes far above guidance and xScale proves it can continuously lift per-share value rather than merely enlarge scale. But the current price already embeds a meaningful portion of that optimism, so upside is compressed by valuation.

  Overall, the most honest answer to Baillie Gifford's question of why the market has not realized this yet is that the market realized it long ago, and perhaps too fully. Equinix is a crowded-consensus high-quality blue chip, not a hidden perception-gap stock. The more likely narrative inflection is downward, where a high multiple is punctured by fundamental deceleration, rather than upward rediscovery. That is the essence of the report's Watch rating: a good company at a bad price. What investors should wait for is not the market finally understanding it, but market enthusiasm fading enough to create a thicker margin of safety.

  Jun 10, 2026