EQIX Equinix, Inc. : Bullish and Bearish Analyst Opinions

Bullish on Equinix due to data center demand.

"We've been the biggest investors in the world in digital infrastructure data centers... that foundation needs to be put in place first... building data centers in Europe is very hard." The speaker identifies data centers as the foundational, physical prerequisite for the AI revolution. Scarcity of supply, especially in constrained markets like Europe, increases the value and pricing power of existing, well-located assets owned by major public REITs. As the dominant private investor (BX) faces hurdles building new supply, the value of existing public data center operators (like DLR, EQIX) is underscored. This is a LONG on the sector leaders. Rapid overbuilding could eventually ease supply constraints. Technological shifts could change data center requirements.

"A lot of the infrastructure needs are currently being driven by some of our technology innovations... whether it's more power that we need for data centers, right, data center construction." "Infrastructure" is now a derivative trade on AI. To support LLMs, you need physical Data Centers (EQIX, DLR) and massive amounts of electricity/power generation (VST, CEG). These "Real Assets" have inflation-linked contracts and secular demand growth. Long Data Center REITs and Power Producers/Utilities. Regulatory pushback on power consumption or a slowdown in AI capex spending.

Taylor distinguishes between "training" data centers (high churn risk) and "interconnection/carrier hotels" (network hubs). He notes that interconnection assets are "prime real estate" and highly durable. Equinix (EQIX) is the primary "carrier hotel" REIT. As AI models move from training to inference, the value shifts toward connectivity and edge distribution, favoring interconnection hubs over generic server farms. Long Interconnection Data Centers. Power constraints preventing expansion; valuation concerns.

Ares Management is underweight Office but heavily invested in Data Centers, stating "we can't get enough" compute infrastructure. The AI boom requires massive physical infrastructure. While commercial office space is "bifurcated" and struggling, the demand for data centers is secular and disconnected from the broader real estate malaise. LONG. Real Estate exposure should be concentrated in the physical backbone of the internet. Power supply constraints or overbuilding in the sector.

Regarding AI, McVey warns against the "if you build it, they will come" speculation. Instead, he explicitly advises finding "contracted cash flow" within the Data Center/AI theme. Data Center REITs (Real Estate Investment Trusts) operate on long-term leases (contracted cash flows) with hyperscalers (Microsoft, Google, etc.). They represent the safe, rent-collecting side of the AI boom rather than the speculative hardware side. Long Data Center REITs as the "contracted cash flow" play on AI. Overbuilding in the data center space leading to lower rental rates; higher interest rates hurting REIT valuations.

She mentions a specific recent trend: "issuance for tracked to fund data center growth... Those are large chunky issuances." Companies do not issue large amounts of debt to build infrastructure unless there is massive, immediate demand. This confirms the "Growth CapEx" cycle is active specifically for data center operators (REITs). LONG Data Center REITs as the primary recipients of this capital expenditure boom. Over-leverage if interest rates spike; oversupply in the long run.

Chamath cites data showing 40% of data center projects face local opposition ("Bananas" movement), causing the industry to lose ~5GW of capacity and potentially $130B in revenue over two years. The demand for compute is inelastic, but supply is artificially constrained by regulation/NIMBYism. Trump's administration is pushing for "behind the meter" solutions where tech companies build their own power plants. This favors companies that can secure land/power or provide independent power generation. LONG. Scarcity of power + high demand = pricing power for energy infrastructure owners. Continued regulatory gridlock or successful "Bananas" lawsuits halting construction entirely.

Software CapEx has been slow, while Hardware CapEx is accelerating. Partners Group sold a data center portfolio but is reinvesting because they see continued growth. The AI trade is shifting from "AI Software" (which is easily disrupted) to "Hard Assets" (Data Centers, Power, Chips). You cannot disrupt a physical piece of equipment with a line of code. LONG. Capital flows are concentrating on the physical infrastructure required to run AI models. Overbuilding capacity leading to a glut in 2-3 years.

"A stabilized data center is going to throw off a margin in the mid 20s... Every dollar that we're putting to work today is ensuring that we are going to earn dollars over the next five years." Critics argue AI infrastructure is a money pit. Intrator provides the counter-metric: mid-20% margins on stabilized assets. This suggests the business model is durable, not just a cash burn. This benefits the entire value chain of data center construction, management, and software layers that optimize compute (CoreWeave's specific value add). LONG. The sector is transitioning from "speculative build" to "stabilized cash flow" generation. Margin compression if electricity costs spike or if hyperscalers (AMZN/GOOG/MSFT) aggressively undercut pricing to gain market share.

Equinix is acquiring atNorth for $4B to expand Nordic capacity; Dell surged on AI server demand. The "AI Infrastructure" trade is widening beyond chips (NVDA) to the physical layer: Data Centers (EQIX) and Server Hardware (DELL). The Equinix deal highlights the premium on power-efficient, sovereign-compliant data capacity. LONG Data Center Infrastructure and Hardware. Overbuilding capacity; regulatory hurdles in M&A.

"We're very, very early on in AI. It's like the dial up generation when we're just using a chatbot right now... If we don't build data centers to support AI advancement, we will be falling behind on other countries." The CEO frames the current AI boom as merely the "dial-up" phase, implying a multi-year secular growth cycle for physical infrastructure. The "national security" argument for AI development suggests the US government will ultimately support data center build-outs despite short-term grid concerns. LONG. Continued demand for power and rack space supports the broader data center theme. Local community pushback ("Occupy Silicon Valley" sentiment) and grid capacity bottlenecks.

Flatt states, "We can't build them fast enough because the demand is so large." He compares AI data centers to the "toll roads and pipelines" of the modern economy. While software valuations fluctuate, the physical infrastructure (power, cooling, real estate) required to run AI models is contractually secured by creditworthy counterparties (hyperscalers/nations). This creates a defensive growth moat. LONG Brookfield and related infrastructure plays. If hyperscaler demand evaporates suddenly (unlikely per Flatt), contracts could be renegotiated.

"You're going to still see [hyperscalers] to compete with each other. That does mean putting the metal in the ground... If you're taking Industrials you're taking Utilities. You're buying into the whole let's call it an adjunct theme of data centers." While direct AI tech stocks have valuation and monetization risks, the physical infrastructure required to support them (power, construction, machinery) is a certainty due to competition. Industrials and Utilities are the "picks and shovels" way to play this capex cycle without taking on tech valuation risk. Long Industrials and Utilities as the safer, valuation-sensitive AI play. A sudden halt in hyperscaler capex due to lack of AI ROI would hurt these sectors.

Goolsbee notes that while AI productivity is a long-term positive, in the short run, "data center investment demand [is] using up all of the HVAC... electrical equipment... computer chips." The Fed President is confirming a physical supply shock. The build-out phase of AI is inflationary for the hardware supply chain. While the Fed worries about the macro inflation this causes, it is a direct revenue boom for the companies supplying the physical constraints (chips, cooling, power). Long the "Pick and Shovel" providers for the AI build-out. Fed tightening specifically to combat this sector-led inflation.

Ram points to Blue Owl (OWL) stock struggling and notes that private credit firms are making loans to tech/software companies and data centers against assets (GPUs) that rapidly depreciate. This resembles "venture lending" disguised as safe private credit. If the AI hype cycle cools or software valuations (currently ~60x earnings for some) correct, the collateral backing these loans evaporates. Avoid Private Credit firms with heavy exposure to Data Center/AI lending. The AI boom continues unabated, sustaining the creditworthiness of these borrowers.

Rowan identifies massive long-term capital needs: "This is manufacturing. This is energy. This is AI and data. This is infrastructure... measured in trillions." The "re-industrialization" of Japan and the build-out of AI capabilities require physical infrastructure. The grid, power generation, and data centers are the "picks and shovels" required before the digital economy can function. LONG Infrastructure and Energy sectors supporting the AI/Data buildout. Regulatory hurdles in energy policy or supply chain bottlenecks.

"Demand does continue to be very strong... we're going to see very strong data center spending at least this year." She also notes "memory prices going up dramatically." The fundamental capex cycle for AI infrastructure is intact and accelerating. However, the trade is crowded. The direction of the entire Nasdaq depends on whether these companies can beat elevated expectations in upcoming earnings (specifically NVDA). WATCH. While fundamentals are bullish (strong spending), the speaker explicitly refuses to "make a call" until seeing if earnings justify the current price levels. Political hurdles or financing issues slowing down data center build-outs.

CRH is currently active on over 100 data center projects in the US. Construction accounts for ~50% of the total cost of a data center, and CRH provides the essential concrete/aggregates first. While software AI stocks face valuation questions, the physical build-out is non-negotiable. You cannot have AI without the concrete shell. CRH has visibility into 2026/2027 backlogs, making it a tangible beneficiary of the "AI Capex" spend regardless of software adoption rates. LONG CRH and materials suppliers to Data Centers. Federal funding delays or a sudden halt in hyperscaler capex.

"The shortage in memory is not going to be resolved in 26 and maybe not even in 27... We own a lot of Western Digital. We like it a lot." The AI build-out requires massive amounts of raw infrastructure (cooling, equipment, memory). Since the supply shortage for memory is structural and multi-year, pricing power and demand for companies like Western Digital and Micron will remain elevated longer than the market expects. LONG. Specific conviction on WDC and the broader memory/infrastructure theme. Unexpected rapid increase in memory supply or a sudden halt in hyperscaler CapEx.

Ads are being introduced to "offset some of the massive data center spending." The explicit mention of "massive" spending confirms that the capex cycle for AI infrastructure is nowhere near slowing down; companies are simply finding new revenue streams (ads) to sustain the hardware purchasing. Long the infrastructure providers (NVDA) receiving this spend. If ad revenue fails to materialize, capex might eventually be cut.

Green explicitly states we are in a "Giant AI CapEx Bubble" and that the amount of money being spent on infrastructure is "mind-boggling" relative to current revenue. Overbuilding is rampant. Similar to the telecom fiber bubble, capacity is being built that may not be utilized immediately, leading to massive depreciation cycles that will hurt the owners of this hardware/infrastructure. AVOID. The risk/reward for pure-play infrastructure build-out is skewed to the downside if utilization lags. If AI adoption accelerates exponentially (AGI), the demand for compute could outstrip even this massive build-out.

"There seems to be just tons of demand for it... we think that the next five years there's not an over supply situation." Despite fears of AI efficiency reducing space needs, the physical infrastructure required for compute (Hyperscalers) continues to outpace supply. Long Data Center REITs and infrastructure providers. Exit risk/valuation concerns if liquidity dries up; long-term AI efficiency reducing physical footprint.

Traditional Data Center REITs (Equinix, Digital Realty) trade at significantly higher multiples than Bitcoin miners. These serve as the valuation benchmark. If miners can prove they are "Data Center REITs" in disguise, they will trade up to these levels. However, miners are currently building faster/cheaper, potentially disrupting the lower end of this market. WATCH. Use these as the relative value comp to gauge the upside for miners. Miners failing to execute validates the premium on EQIX/DLR.

"There's obviously a huge boom in construction going on in data centers... but I don't have any sign... that the number of car plants in this country is increasing." While general manufacturing (factories) is in secular decline, the build-out of AI/Cloud infrastructure is the singular bright spot in US capital investment. This concentrates demand on construction machinery, power management, and HVAC specific to data centers. LONG. This is the only sub-sector of "manufacturing/construction" with actual momentum. Regulatory pauses on power consumption or AI capex spending cuts.

"Everyone thought when this whole emergence came that the traditional data center companies would just chew up all that demand and that's not happening." The consensus trade is that legacy data centers (Digital Realty, Equinix) are the sole beneficiaries of the AI boom. The speaker argues they are losing the "density" war to miners. If traditional data centers cannot handle the heat/power requirements of next-gen AI chips as efficiently as miners, their growth assumptions are overpriced. Avoid or underweight legacy data centers relative to power-dense mining infrastructure. Traditional data centers may acquire miners or retrofit faster than expected.

"The traditional data center world is saying well this is the way we usually do it." Legacy data centers are moving too slowly to capture the explosive initial demand for AI training capacity. Their adherence to standard procedures makes them uncompetitive against agile miners willing to cut corners or build off-grid. AVOID. They are losing the "land grab" phase of the AI boom. Enterprise clients may prefer the reliability and compliance of legacy centers over the "wild west" approach of miners.

While general manufacturing is struggling, Rattner states there is "obviously a huge boom in construction going on in data centers." He also cites the CHIPS Act as a success with factories building in Arizona. Capital expenditure is decisively shifting from traditional industrial plants to digital infrastructure. Companies providing the physical infrastructure (power/cooling like Vertiv) and the chips (Semiconductors) are the sole beneficiaries of this "manufacturing" spend. LONG the picks and shovels of the data center buildout. Overbuild/capacity glut or energy supply constraints.

"You're seeing a lot of jobs being created in construction, especially non-residential construction. Has to do with the data centers, the AI boom that's going on." The $600B capex spend requires physical infrastructure. This directly benefits engineering, construction, and industrial firms that build the shells and power systems for AI data centers. Long Industrial and Construction sectors exposed to data center build-outs. Regulatory halts on power consumption or a slowdown in AI scaling laws.

"What we hear from the HYPERSCALERS is they're looking for unique data... that unique data resides in space... We think that's one of the killer apps [AI]." The AI trade is evolving from "chip manufacturing" to "data acquisition." As terrestrial data becomes commoditized or exhausted, the premium shifts to companies that can harvest and process unique physical-world data from orbit. Long the AI value chain that extends into physical infrastructure and data acquisition. Overvaluation in the AI sector; the speaker notes the industry recently went through an "asset bubble" in 2021/2022 that has since burst and is recovering.