𝗔𝗺𝗲𝗿𝗶𝗰𝗮'𝘀 𝗔𝗜 𝗯𝗼𝗼𝗺 𝗰𝗼𝗺𝗲𝘀 𝘄𝗶𝘁𝗵 𝗮 $𝟵 𝘁𝗿𝗶𝗹𝗹𝗶𝗼𝗻 𝗯𝗶𝗹𝗹. 𝗘𝘃𝗲𝗻 𝗕𝗶𝗴 𝗧𝗲𝗰𝗵 𝗻𝗲𝗲𝗱𝘀 𝗵𝗲𝗹𝗽 𝗽𝗮𝘆𝗶𝗻𝗴 𝗶𝘁. 🏗️ 𝗧𝗵𝗲 𝘀𝗰𝗮𝗹𝗲: about 𝟯.𝟮% 𝗼𝗳 𝗚𝗗𝗣 a year through 2032. Relative to the economy, that's bigger than the railroad, highway or 90s telecom booms. 💸 𝗧𝗵𝗲 𝘁𝘂𝗿𝗻𝗶𝗻𝗴 𝗽𝗼𝗶𝗻𝘁: Big Tech's in

𝗔𝗺𝗲𝗿𝗶𝗰𝗮'𝘀 𝗔𝗜 𝗯𝗼𝗼𝗺 𝗰𝗼𝗺𝗲𝘀 𝘄𝗶𝘁𝗵 𝗮 $𝟵 𝘁𝗿𝗶𝗹𝗹𝗶𝗼𝗻 𝗯𝗶𝗹𝗹. 𝗘𝘃𝗲𝗻 𝗕𝗶𝗴 𝗧𝗲𝗰𝗵 𝗻𝗲𝗲𝗱𝘀 𝗵𝗲𝗹𝗽 𝗽𝗮𝘆𝗶𝗻𝗴 𝗶𝘁. 🏗️ 𝗧𝗵𝗲 𝘀𝗰𝗮𝗹𝗲: about 𝟯.𝟮% 𝗼𝗳 𝗚𝗗𝗣 a year through 2032. Relative to the economy, that's bigger than the railroad, highway or 90s telecom booms. 💸 𝗧𝗵𝗲 𝘁𝘂𝗿𝗻𝗶𝗻𝗴 𝗽𝗼𝗶𝗻𝘁: Big Tech's infrastructure spending is set to hit about $𝟴𝟬𝟬𝗕 in 2026, more than the cash its businesses bring in. 𝘛𝘩𝘢𝘵'𝘴 𝘢 𝘧𝘪𝘳𝘴𝘵. 🏦 𝗧𝗵𝗲 𝗻𝗲𝘄 𝗽𝗹𝗮𝘆𝗯𝗼𝗼𝗸: leases, private lenders and separate companies that hold the debt. Meta's Hyperion data center was funded by a record $𝟮𝟳𝗕 bond, held outside Meta's balance sheet. It even costs more than Meta borrowing directly. 🧩 𝗪𝗵𝗮𝘁 𝘀𝘁𝗼𝗼𝗱 𝗼𝘂𝘁 𝘁𝗼 𝗺𝗲: the risk is still there, it's just 𝘩𝘢𝘳𝘥𝘦𝘳 𝘵𝘰 𝘴𝘦𝘦. Count leases as debt and Big Tech's debt ratio jumps from 𝟮𝟬% 𝘁𝗼 𝟰𝟴%. 📄 𝗙𝗶𝗻𝗮𝗻𝗰𝗶𝗻𝗴 𝘁𝗵𝗲 𝗔𝗜 𝗕𝘂𝗶𝗹𝗱𝗼𝘂𝘁, a new NBER paper by Stijn Van Nieuwerburgh (Columbia Business School). He sizes the bill using about 3,400 US data center projects, then follows the money. 🔗 🔎 One caveat: the $9T is a scenario, not a forecast, and it rests on assumptions about which announced projects get built. #AI #ArtificialIntelligence #DataCenters #AIInfrastructure #BigTech #Hyperscalers #Finance #Economics #Investing #CapitalMarkets #PrivateCredit #ProjectFinance #Credit #FixedIncome #Infrastructure #RiskManagement #FinancialStability #Meta #NBER #ColumbiaBusinessSchool I started a monthly digest that cuts out all the noise in