Digital Realty: An In-Depth Value Investing Analysis

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

• QHow high is its market ceiling? Is it enlarging an existing slice of the pie, or creating an entirely new market?6/10

  A

  The ceiling is very high, but in essence it is "enlarging an existing slice of the pie," not creating a new market from scratch. What DLR does is build long-term-leasable data centers at globally scarce power and network nodes and wholesale or colocate that capacity to cloud and large technology customers—this business form has existed for decades, and DLR is the leading player that has taken it global and to scale, not one that has defined an unprecedented category. The report labels it with restraint: "a good company in a good industry, but not a good industry where money comes easily," and "not an industry where 'mindless capacity expansion is guaranteed to earn high returns.'"

  The pie itself is indeed growing fast, which is DLR's firmest backdrop. AI training and inference push data centers from "ordinary cabinets" toward high-density, high-power-consumption workloads, and third-party industry estimates put the incremental space at rather astonishing levels: MarketsandMarkets projects the global AI data center market growing from roughly $236.4 billion in 2025 to roughly $933.8 billion in 2030 (a CAGR of about 31.6%); McKinsey goes further and estimates that by 2030 the world will need roughly $6.7 trillion of capital to build out data centers, of which roughly $5.2 trillion is earmarked specifically for AI-related facilities. The power side is likewise doubling: the report cites Reuters citing EPRI that by 2030 U.S. data center electricity use could rise to 4.6% to 9.1% of national electricity use.

  But we must honestly distinguish "industry ceiling" from "the share DLR can capture." DLR competes for supply share within the existing main channel, rather than opening a path no one has walked: in the report, of 2025 total revenue of $6.113 billion, "rental and other services" made up $5.969 billion, and the 2026 total-revenue guidance in the company's Q4 2025 results release is only $6.60 billion to $6.70 billion—growth is real, but it is "continuing to grow larger within a huge and rapidly expanding mature market," not "creating an entirely new market."

  From a Baillie Gifford perspective, this point is actually neutral-to-positive for DLR: the ceiling is high enough that volume growth will not hit the roof soon; but "enlarging an existing pie" also means it does not enjoy the near-monopolistic early dividend of a "new-category definer," and it must fight for share inch by inch with capital, power, and execution, so returns depend more on price than on narrative.

  Jun 11, 2026
• QOver the next five years, can its revenue at least double? Is growth driven mainly by volume, price, or new business?4/10

  A

  Doubling revenue over the next five years is "a stretch that is reachable, but not something that happens by default," and growth is driven mainly by "volume" (newly added leasable capacity), with price as a supporting factor and new business contributing very little. Doubling means going from $6.113 billion in 2025 to roughly $12 billion, corresponding to a five-year CAGR of about 14%—higher than the company's actual current pace. The 2026 total-revenue guidance in the company's Q4 2025 results release is $6.60 billion to $6.70 billion, up only about 8%-10% year over year versus 2025; extrapolating along this slope, the cumulative five-year figure is roughly +50% to +60%, leaving a gap to doubling. Closing that gap would require the AI cycle to keep the development pipeline maxed out for several years.

  On the composition of drivers, "volume" is the absolute mainstay. In the report, projects under construction reached 1,169MW (roughly 1.2GW on the company's Q1 2026 basis, up more than 50% from the prior quarter), of which 61% is pre-leased with an expected stabilized cash yield of roughly 11.4%—revenue growth comes mainly from these newly built capacities coming online one after another. Visibility comes from the backlog: the report discloses signed-but-not-yet-commenced leases of roughly $1.0 billion at DLR's share (on the company's Q1 2026 results basis, roughly $1.8 billion on a 100% basis), and newly signed leases carry an average commencement lag of roughly 19 months—meaning "volume" growth is real but time-lagged, not converted into revenue in the same quarter it is signed.

  "Price" is a secondary but real driver. In the report, Q1 2026 renewal leases of $193 million saw a blended cash-basis increase of roughly 5.0%, same-store cash NOI grew 7.9% year over year, and stabilized occupancy was 91.6%—supply tightness lets existing leases be renewed at moderate price increases, but this is a mid-single-digit assist that cannot support doubling. "New business" contributes the least: in the report, "fee income and other" was only $144 million in 2025, less than 2.4% of total revenue, and DLR remains at its core an asset operator centered on leasing and supporting services.

  Conclusion first: doubling in five years is a possibility under the optimistic scenario, not the base case. The report itself builds the optimistic-scenario intrinsic value ($190-210) on the simultaneous holding of three conditions—"AI demand persists for years, power scarcity continues to strengthen bargaining power, and the capital-recycling loop runs smoothly"—which are precisely the preconditions for doubling. Baillie Gifford wants "at least doubling in five years" high-certainty growth; what DLR offers is conditional growth—"there is a chance, but it requires betting fully on a long AI boom plus sustained large capital investment"—a notch lower in certainty.

  Jun 11, 2026
• QFive years out, what will take the baton as the next growth engine? Does this "second curve" exist today?4/10

  A

  Frankly, DLR has no "second curve" separate from its core business—its "second curve" is really an extension of the same main curve (AI/high-density capacity + the capital-recycling platform), not an independent new growth pole. This is where the company most deserves an honest label: nowhere in the report is there a narrative of "an entirely new business taking the baton five years out"; what can take the baton is still "building more, higher-density data centers at more scarce power nodes and leasing them out."

  The closest thing to a "new engine" is AI-driven high-density capacity shifting from incremental to mainstay. The report notes that AI brings "not ordinary cabinet demand, but high-density, high-power-consumption workloads with greater demands on power supply and cooling"; the company has 1,169MW under construction, 61% pre-leased, with an expected stabilized yield of roughly 11.4%—but this is the core business thickening naturally along the AI direction, not a second leg. Industry space is indeed feeding this extension line: McKinsey estimates that by 2030 the world's data centers will need roughly $6.7 trillion of capital, of which roughly $5.2 trillion goes to AI—the pie is big enough for DLR to keep feeding along the main curve for many more years.

  The second candidate is "capital recycling + the co-investment platform (JV/funds)" upgrading from a financing tool to a sustainable profit source. In the report, of 2025 GAAP net income to common of $1.268 billion, $996 million came from real-estate disposition gains; the company continually packages assets into funds/JV platforms, recognizes gains, and recycles capital for reinvestment. If this line becomes a stable "develop—deconsolidate—management fee" capital-light loop, it could in theory become a second curve; but the report is explicitly wary here: "if investors mistake such profits for 'recurring operating profit,' they will overestimate the true cash-generating ability," and the company in its 10-K also reminds readers that FFO does not reflect maintenance capex—this is more of a financing tool and does not yet constitute a stable, independent profit pole.

  The implicit premise of this Baillie Gifford question is "the core business will eventually decelerate, so who will hold up the firepower from years three to ten?" The honest answer for DLR is: today there is no second curve independent of data center leasing waiting in line to take the baton, and growth over the next decade heavily bets on "the same business + a long AI boom + sustained large reinvestment" not going out. Once the AI capex cycle recedes, or power/yield constraints tighten, DLR has no ready backup engine to hedge—this is a point where it is structurally more fragile than growth stocks that have already incubated independent new businesses.

  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 a four-piece set: "global scale + the ability to secure scarce power/network nodes + high-quality large-customer relationships + investment-grade financing capability"; over the next three to five years the moat broadly widens, but unevenly, and it is not consumer-goods-style strong pricing power. The report breaks it into five parts and states clearly that "the moat is still widening, but unevenly," giving a moat-strength score of 4/5.

  The two hardest pieces are scale and power/network density. In the report, the company spans 300+ data centers, 55+ metropolitan areas, and 6 continents, holds more than 5GW of future development capacity, and as of the end of Q1 2026 had 1,169MW under construction, 61% pre-leased. The report cites the 10-K emphasizing that certain campuses form "densely connected data communities" that management believes competitors find hard to replicate—in today's industry, what is truly scarce is "not the reinforced-concrete shell, but the ability to secure power, connect to networks, and form scalable campuses." The power constraint itself is reinforcing this moat: the report cites Reuters reporting that the pace of EMEA data center go-lives slowed in 2025, mainly because power supply is constrained rather than demand being weak—whoever already holds grid-connected power and mature campuses finds their existing assets more valuable.

  The third piece is customer stickiness and switching costs, but it is B2B engineering-grade stickiness, not brand emotion. In the report, the average remaining lease term is roughly 4 years, end-of-Q1-2026 stabilized occupancy is 91.6%, and renewal cash increases are 5.0%—once a customer deploys core workloads into a campus, migration involves networking, latency, cross-connects, compliance, and downtime risk, so they will not walk away as easily as switching a software subscription. The fourth piece is the financing moat: in the report, Q1 2026 net debt/adjusted EBITDA is 4.7x (below management's through-cycle target of roughly 5.5x), fixed or hedged-to-fixed-rate debt is 93.4%, the effective rate is roughly 2.85%, and ratings are BBB+/Baa2/BBB—investment-grade credit lets it support large-scale development at lower cost, something small and mid-size players cannot emulate.

  The directional judgment for the next three to five years: widening, but with a structural shortcoming. The reason for widening is that power/land/permitting are increasingly scarce, and first-movers' campuses and grid-connection allotments grow ever more valuable; the report judges that "for a competitor to replicate it would take many years and many billions of dollars of capital." But there are two shortcomings: first, in standardized, purely capacity-oriented wholesale hyperscale supply, the moat "depends more on execution and financing than on absolute pricing power"—this segment is more of a contest of cost and speed. Second, the strongest rival, Equinix, has stronger perception in interconnection/enterprise ecosystems; the report states plainly that "DLR is not a clearly superior choice to EQIX—just a different style," with Equinix's 2026 revenue guidance of roughly $10 billion and a market cap of roughly $106 billion, both larger in scale and interconnection premium. So on a Baillie Gifford basis this is a real, deepenable moat that is nonetheless "wide but not ultimate, strong but not monopolistic."

  Jun 11, 2026
• QIf its core business is disrupted, does it have the genes to reinvent itself? How does it treat mistakes and bad news?4/10

  A

  First the implicit premise, "does it have the genes to reinvent itself when its core business is disrupted": DLR's reinvention genes are on the weak side—it is an asset-heavy, long-cycle infrastructure platform that turns slowly and bets in a concentrated way, but fortunately the probability of its core business being "disruptively replaced" is itself low; its attitude toward mistakes and bad news is relatively candid, with restrained disclosure, which is a plus.

  Why the reinvention genes are weak: the asset form of this business makes it hard to pivot quickly. In the report, 2025 cash capex totaled $2.913 billion ($2.541 billion for development), and the 2026 plan rises to $3.25 billion to $3.75 billion—once hundreds of megawatts of concrete, power, and mechanical-electrical systems are invested, they lock in capital and location for many years to come, unlike a software company that can rewrite its business with a single version iteration. The report also notes that growth "depends heavily on sustained large capital expenditures, asset dispositions/JV capital recycling, and debt and equity financing"—the precondition for this model to work is that development yields persistently exceed the cost of capital, and it inherently lacks the flexibility to "start over after the core is disrupted."

  But to be fair: what DLR faces is not "overnight replacement by new technology" disruption, but "demand-slope/yield volatility" risk. Data centers are the physical foundation of AI and cloud; workloads merely migrate from one form (training) to another (inference), they do not vanish into thin air; the real threats are those the report repeatedly stresses—power/grid-connection constraints, development yields falling short, and changes in customer expansion pace—rather than "this DLR business being eliminated." So the shortcoming of "weak reinvention genes" does it less actual damage than it would to a pure technology company.

  On attitude toward mistakes and bad news, the report's evidence leans positive. The most persuasive point: the company in its 10-K plainly reminds readers that FFO "does not reflect the capital expenditures and capitalized leasing commissions needed to sustain operations, and so its usefulness as a performance metric is limited"—the report comments, "for a REIT, management's willingness to spell out this limitation is a plus." In its risk factors the company also directly admits that it "depends on third-party power supply and incremental power capacity; if incremental power is insufficient, it may be unable to meet customer demand or its growth plans"—putting the deadliest bottleneck on the table rather than glossing over it. The report also confirms that it "sees no obvious red flags of dishonesty," and on governance there are anti-hedging, anti-pledging, and clawback policies.

  This Baillie Gifford question asks for "regenerative ability in adversity + an honest culture." DLR's answer is: the honest culture passes (candid disclosure, not passing FFO off as real cash); but the regenerative genes are weak due to the asset-heavy nature, and its bets are highly concentrated on "the same business + a long AI boom." Once an industry-level headwind arrives, it is more likely to "tough it out + slow development + winter through on the balance sheet" than to "quickly reinvent itself into a different business."

  Jun 11, 2026
• QDoes management (especially the founder) have a long-term vision, with interests deeply aligned with the company? Are they willing to sacrifice current profit for five to ten years out?4/10

  A

  Management has a long-term vision and is a rational capital allocator, but the item "interests deeply aligned with the company" clearly falls short—DLR has no founder-level significant ownership, and the CEO's economic-interest alignment with shareholders can only be called moderate. The report gives management and capital allocation 3/5, the lowest tier across the whole report's moat/industry/business sections, a fair judgment.

  DLR does meet the two items "long-term vision" and "willing to sacrifice current profit for five to ten years out"—these are precisely two sides of the same capital-intensive model. In the report, 2025 total capex of $2.913 billion already exceeded operating cash flow of $2.412 billion, and 2026 rises to $3.25 billion to $3.75 billion; in other words, management is actively sacrificing current free cash flow in exchange for a decade of capacity and campuses. The compensation structure also points to the long term: the report discloses that in 2024 long-term incentives made up roughly 57% to 74% of named executives' total compensation, "indicating that incentives are not short-sighted"; the CEO is required to hold stock worth roughly 6x base salary, and the company has anti-hedging, anti-pledging, and clawback policies.

  But on the item "interests deeply aligned," the evidence is clearly "not enough." The report names CEO Andrew Power's beneficial ownership in the 2025 proxy at roughly 291,110 shares—at the current price of roughly $180.78, that works out to roughly $53 million, "a lot for an ordinary person, of course, but for a company with a market cap of roughly $64 billion it is not 'founder-level significant ownership.'" The report therefore states clearly that it "would not classify DLR as one of those 'management deeply aligned with shareholders.'" This is the most substantive gap between DLR and the "founder heavily invested, sharing the company's fate" growth stocks Baillie Gifford favors—it is a large REIT run by professional managers, not a growth machine helmed by a founder.

  The "rationality" of capital allocation earns a conditional pass. The report observes that in and after Q1 2026 the company issued roughly 7.3 million shares through the ATM, raising net proceeds of roughly $1.3 billion at an issue price of roughly $179-181 per share. The report's verdict is apt: "if the company can deploy the funds into projects with stabilized yields around 11%, then issuing equity when the stock is relatively rich in order to advance projects is rational; but if project yields decline in the future, this model turns into 'using share issuance to paper over insufficient free cash flow.'" In other words, whether this "issue high + JV capital recycling + redevelop" loop is rational hinges entirely on whether development yields can persistently outrun the cost of capital—a "rationality" that requires ongoing verification rather than one already set in stone.

  On a Baillie Gifford basis, conclusion first: management passes but is not outstanding. The long-term vision is there, and it is willing to sacrifice current profit for the future, but the hard indicator of "founder-level interest alignment" is absent, and the rationality of capital allocation depends heavily on external variables (development yields, the financing environment). This is precisely why the report holds it at 3/5 and makes management one of the important pieces in the "Watch" rather than "Buy" investment judgment.

  Jun 11, 2026
• QIf it disappeared tomorrow, how much would customers miss it? Is its growth model sustainable, without relying on harming society or regulation?5/10

  A

  Look at the dual premise separately: (1) indispensability—if DLR disappeared tomorrow, large customers would "miss it quite a bit," but not "irreplaceably fatally"; (2) social/regulatory sustainability—the growth model on the whole does not harm society, but power and energy consumption are becoming an increasingly real regulatory/social friction point, so sustainability "holds, but there are headwinds ahead."

  First, "how much they would miss it." The stickiness is real: in the report, the average remaining lease term is roughly 4 years, end-of-Q1-2026 stabilized occupancy is 91.6%, and renewal cash increases are 5.0%—after a customer deploys core workloads into a DLR campus, migration involves networking, latency, cross-connects, compliance, and downtime risk, so they cannot move out in the short term and would miss it a great deal. But the degree of "missing it" is diluted by two things: first, DLR is one of the "largest global providers" but not the only one—the strongest rival, Equinix, also spans six continents with 2026 revenue guidance of roughly $10 billion, plus a host of hyperscalers building their own data centers, so customers have alternative paths over the medium to long term; second, in the report, at the end of Q1 2026 the top twenty customers accounted for 51.9% of annualized recurring revenue, the largest customer for 11.4%, and Oracle for 9.8%—this level of concentration shows that the real "can't do without" tension lies more on the "DLR can't do without these large customers" side than on the "large customers can't do without DLR" side. So it is "an important supplier that would be missed," not "a chokehold with no replacement."

  Now social/regulatory sustainability—this is where this question most deserves digging for DLR. Its growth model does not rely on harming users, regulatory arbitrage, or data abuse; at its base it is "providing the physical foundation for cloud, AI, and enterprise digitization," a legitimate social value. But energy consumption is a real and rising social cost: the report cites Reuters citing EPRI that by 2030 U.S. data centers could account for 4.6% to 9.1% of national electricity use; McKinsey estimates that AI data centers will need roughly $5.2 trillion of investment by 2030—massive power consumption will inevitably draw regulatory attention around grid load, electricity-price livelihoods, water-cooling consumption, and community and environmental reviews. The report already treats this as a core bottleneck: the main reason the pace of EMEA go-lives slowed in 2025 was constrained power supply.

  This double-edged sword means "moat and regulatory risk share the same source" for DLR: power scarcity protects the pricing power of its existing campuses, but it also means new growth is increasingly constrained by grid-connection approvals, energy-consumption policy, and community acceptance—a structural headwind embedded within a structural tailwind. On a Baillie Gifford basis, conclusion first: DLR's indispensability is "upper-middle, not top-tier" (important but replaceable, with the bargaining balance tilted toward large customers), and the social/regulatory sustainability of its growth is "legitimate in direction, but the regulatory fuse of energy consumption is already lit and needs continuous tracking."

  Jun 11, 2026
• QWhat are the unit economics of this business (gross margin, incremental returns)? Do they improve or worsen as it scales? Where does the money it earns go?5/10

  A

  The unit economics are "pretty at the accounting level but actually heavy at the cash level": project-level development yields are decent (stabilized around 11.4%), but overall per-share free cash flow is not ample after deducting all growth capex—it was even negative in 2025; as the company scales, the unit economics show no obvious "bigger is better" software-style leverage, but rather "the more it grows, the more capital it consumes." Almost all of the money earned is plowed back into the development pipeline. The report gives financial quality 3/5, consistent with this judgment.

  First the positives: the returns on incremental projects are indeed attractive. In the report, 1,169MW is under construction, 61% pre-leased, with an expected stabilized cash yield of roughly 11.4%—the project-level return once a single new data center comes online is clearly above the company's roughly 2.85% cost of debt, which is the fundamental reason DLR is willing to keep investing heavily. The renewal side also has modest price leverage: in Q1 2026, $193 million of renewals saw a blended cash-basis increase of roughly 5.0%, and same-store cash NOI grew 7.9% year over year.

  But zoom out to "overall unit economics and per-share cash flow," and the story turns heavy. The key evidence is the report's conservative Owner Earnings calculation: operating cash flow $2.412 billion − recurring capex $344 million = $2.068 billion, or roughly $5.85 per share. But once all development-related capex is counted in, free cash flow no longer looks good—the report notes that "2025 total capex of $2.913 billion exceeded operating cash flow of $2.412 billion," and "after deducting all growth capex, free cash flow is not stable and was even negative in 2025." This is the direct embodiment of "unit economics not improving as the company scales": revenue grew from $3.904 billion in 2020 to $6.113 billion in 2025, and FFO reached $2.399 billion, or $6.96 per share, but capex rose in tandem, and the report characterizes it in one line: "DLR's growth is 'the more it grows, the more capital it needs,' not 'the more it grows, the more cash it spits out.'"

  We must also expose the "accounting profit" water content, which is critical for judging the true unit economics. In the report, of 2025 GAAP net income to common of $1.268 billion, $996 million came from real-estate disposition gains (2024 and 2023 also had $597 million and $908 million, respectively)—meaning a large chunk of the profit is "capital recycling/asset-revaluation realization" rather than recurring operating profit. The report therefore warns that if the market looks only at FFO, it "easily underestimates the capital intensity of this business."

  The answer to "where does the money it earns go" is very clear: the vast majority is reinvested into the development pipeline. Of 2025 cash capex of $2.913 billion, development projects took $2.541 billion; the 2026 planned capex rises to $3.25 billion to $3.75 billion; the gap is filled by ATM issuance (roughly $1.3 billion in and after Q1) and JV capital recycling. On a Baillie Gifford basis, conclusion first: DLR's unit economics are "asset-heavy infrastructure type"—project-level returns are respectable, but there is no marginal profit elasticity from scale amplification, and free cash flow is swallowed by growth itself. It is "a high-quality asset-heavy platform, not a capital-light compounding machine," which is precisely why "at what price to buy it" matters more than "whether it is good."

  Jun 11, 2026
• QFor it to rise fivefold in ten years, which conditions must hold simultaneously? Are these conditions realistic? What expectations does today's share price imply?3/10

  A

  First a list of hard constraints: for DLR to rise fivefold in ten years (roughly 17.5% annualized per-share value), five conditions must hold simultaneously—"AI/cloud demand not receding for years + the under-construction pipeline broadly realized at a stabilized yield of roughly 11.4% + net debt/EBITDA held within roughly 5.5x over the long term + external capital markets staying open at a reasonable cost + valuation multiples not contracting sharply." Each one alone is not outlandish, but for none of them to slip over ten years is fairly difficult—this belongs to the optimistic scenario, not the base case. Today's share price of roughly $180.78 already implies fairly optimistic expectations of "a long AI boom + a scarce-power premium + years of projects converting smoothly."

  First lay out the five conditions against the report: (1) demand side—the report explicitly builds the optimistic scenario on "AI demand persisting for years, power scarcity continuing to strengthen bargaining power"; (2) conversion side—the 1,169MW under construction, 61% pre-leased, expected stabilized yield of roughly 11.4% must actually land, and the report warns "if the expected 11.4% stabilized yield cannot be realized, the valuation will run into trouble"; (3) leverage side—the report lists "net debt/EBITDA over the long term not significantly above roughly 5.5x" as a key assumption, and the current 4.7x still has room; (4) financing side—the report stresses that growth "depends heavily on sustained large capital expenditures, asset dispositions/JV capital recycling, and debt and equity financing," requiring that "external capital markets stay open and can finance at a relatively reasonable cost"; (5) valuation side—this is the most hidden and most critical one, discussed separately below.

  What expectations does today's share price imply—this is the soul of this Baillie Gifford question. The report calculates it thoroughly: at the midpoint of 2026 Core FFO guidance, it currently trades at roughly 23.9x Core FFO (note: the report's snapshot price was $192.03; at the current roughly $180.78 it is about 22.3x), roughly 32.8x conservative Owner Earnings (about 30.9x at the current price), with a dividend yield of only about 2.5% and an Owner Earnings yield of only about 3.0%. Among the report's three valuation tiers—conservative intrinsic value $100-115, fair $135-155, optimistic $190-210—the current share price is "already close to the lower bound of the optimistic-scenario range and clearly above the conservative and neutral valuations," and "the market is already prepaying for AI, scarce power, and years of future project conversion." Cross-sectionally, DLR's current Owner Earnings yield of roughly 3.0% is still below the 10-year Treasury yield of roughly 4.55%—the return logic of buying it must rest on long-term 5%-8% per-share cash-flow growth rather than current cash returns.

  Are these conditions realistic? Item by item they are not harsh, but taken together they are tight. The report's long-term annualized return judgment for "buying near the current roughly $192" is: conservative 0%-3%, neutral 4%-7%, optimistic 8%-10%—note that even in the optimistic scenario, the annualized figure is only 8%-10%, still far from "fivefold in ten years (roughly 17.5% per year)." This shows that starting from today's price, "fivefold in ten years" requires not just all five fundamental conditions to hit but also valuation multiples to keep expanding (the market willing to grant a higher P/FFO), and the report judges the current multiple already on the rich side with limited upside. On a Baillie Gifford basis, conclusion first: the conditions DLR needs for a fivefold-in-ten-years return are "credible individually, but hard in aggregate," and today's price has already priced in the optimistic expectations, so the margin of safety is insufficient—this is precisely why the report's rating lands on "Watch" and compresses the ideal buy range to $135-155.

  Jun 11, 2026
• QWhy has the market not yet realized all this? Is it that it doesn't understand, doesn't respect, or doesn't see far enough? What would become the "narrative inflection point"?3/10

  A

  Frankly, the market has not "failed to realize" DLR's virtues—on the contrary, under the AI narrative the market already sees it quite clearly, even pricing it fully, so this Baillie Gifford question must be answered in reverse for DLR: it is not undervaluation caused by "not understanding/not respecting/not seeing far enough," but expensiveness caused by "seeing it very clearly, even a bit optimistically." The report's core judgment is that "the current market price has largely reflected optimistic expectations for AI, scarce power, and development conversion, and the margin of safety is not enough."

  Debunking the three standard options one by one: (1) not "doesn't understand"—the data center REIT is one of Wall Street's most heavily researched tracks, with 37 analysts giving an average price target of roughly $222 and extremely high AI-theme heat over the past year, so coverage is ample; (2) not "doesn't respect"—in the report it currently trades at roughly 22-24x Core FFO, EV/EBITDA of roughly 22x, and a dividend yield of only about 2.5%, with a market cap on the order of $64.6 billion—this is premium pricing, not a disdained discount; (3) not "doesn't see far enough"—the market has already capitalized years of future development pipeline, campus power scarcity, and JV platform capability into the share price; the report notes book total assets of roughly $49.410 billion versus an enterprise value of roughly $82 billion and a market cap of roughly $64 billion, and the difference is precisely the money the market has prepaid for "a future it sees very far into."

  So what the "market may be pricing insufficiently" is the bear side, not the bull side—that is, downside risk is masked by the AI narrative. The strongest bear-case view the report gives is: "the market is pricing it as an AI-infrastructure platform that almost never makes mistakes. If any one link—power, delivery, customer expansion, or cost of capital—goes wrong, the share price could revert from a 'high-valuation growth REIT' to an 'ordinary high-capex REIT.'" The underappreciated risk points include: the capital intensity of 2025 total capex of $2.913 billion already exceeding operating cash flow of $2.412 billion, the concentration of the top twenty customers accounting for 51.9% of annualized recurring revenue, and the "surface-profit" water content of $996 million of disposition gains within GAAP profit—these "not sexy enough" facts are easily overshadowed by the optimistic narrative.

  What would become the "narrative inflection point"—this is the implicit premise, and both sides must be given: the downward inflection (more likely to arrive first) = the reassessment-trigger signals the report lists, namely the backlog shrinking markedly for consecutive periods, stabilized occupancy falling below roughly 90% with slow recovery, net debt/EBITDA staying persistently and significantly above 5.5x, expected development-project yields declining markedly, continued issuance while Core FFO per share and Owner Earnings per share do not grow, or capital-market rates rising to make financing more expensive—if any one is confirmed, the market could downgrade the valuation from a "growth REIT" to an "ordinary REIT," and the report judges a long-term drawdown of 35%-50% "not unimaginable" in that scenario. The upward inflection = AI inference demand pushing data centers from "training-driven" toward larger-scale "inference-driven," power scarcity further strengthening existing-campus bargaining, or the capital-recycling/JV loop producing a sustainable capital-light-profit second curve—but most of these are already partly priced in, leaving limited room for marginal surprise.

  On a Baillie Gifford basis, conclusion first: DLR is not a "dust-covered growth stock the market has yet to discover," but a high-quality asset-heavy platform that "the market fully recognizes and prices to optimistic expectations." Its excess return does not come from "correcting the market's misunderstanding" but from "waiting for a cheaper price"—the report therefore sets the action as "patiently wait for a better price," with an ideal buy range of $135-155, while the current roughly $180.78 remains on the rich side.

  Jun 11, 2026