US AI Boom Saps Treasury Demand as Foreign Funds Chase Big Tech

Bloomberg Originals

Circular capital flows into US artificial intelligence are pushing global imbalances to 2008 peaks, raising borrowing costs on the $40 trillion national debt while exposing Asian suppliers to a double shock if valuations collapse.

AI-assisted summary of Why AI Is Sucking Up the World’s Wealth — original source. Check the original for full context.

Source published: 2026-10-10

Key ideas

The artificial intelligence boom has concentrated global capital into a narrow cluster of US tech firms, with foreign investors now owning 20% of all US equities. (0:00)
Global current account imbalances widened to roughly $2.5 trillion in 2025, reaching extreme divergence levels unseen since the run-up to the 2008 financial crisis. (2:03)
Unlike the 2000s when foreign surpluses funded US Treasuries, modern capital flows prioritize high-risk equities, putting upward pressure on yields across the $40 trillion US national debt pile. (3:02)
Geopolitical fragmentation and AI concentration are starving developing markets of vital financing needed for fundamental infrastructure like factories, roads, and power grids. (5:38)

Detailed notes

## Background Published by Bloomberg Originals on YouTube on October 9, 2026, this video explores how the artificial intelligence boom has concentrated global wealth into a handful of American tech giants. The accompanying overview highlights how this dynamic drains capital from emerging markets and traditional sectors, driving global trade and current account imbalances to levels unseen since the 2008 financial crisis.

## TL;DR The video outlines a circular macroeconomic loop in which US tech firms buy high-end AI components from East Asian manufacturers, whose accumulated trade surpluses are promptly reinvested back into American equities and corporate debt. Unlike the pre-2008 era when foreign central banks channeled reserves into safe US Treasuries, modern asset managers seek riskier yields in tech stocks, exerting upward pressure on borrowing costs across America's $40 trillion national debt. Meanwhile, geopolitical alignment increasingly dictates international capital flows, leaving developing economies starved of essential funding for foundational infrastructure like roads and power grids. If tech valuations collapse, this concentration exposes Asian manufacturing hubs to severe dual shocks in both portfolio devaluations and cratering export demand.

## Atmosphere & style The presentation carries the sleek, sober gravitas of institutional financial journalism, blending lucid macroeconomic diagrams with an undercurrent of systemic unease. Its measured, data-driven narration methodically dissects global flows without sensationalism, yet deliberately evokes the haunting structural conditions that preceded the 2008 financial crash. Watching it feels like attending an urgent executive briefing that methodically punctures Silicon Valley hype to reveal fragile geopolitical fault lines beneath.

## Bulletin - [00:00](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=0s) The artificial intelligence boom has concentrated global capital into a narrow cluster of US tech firms, with foreign investors now owning 20% of all US equities. - [02:03](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=123s) Global current account imbalances widened to roughly $2.5 trillion in 2025, reaching extreme divergence levels unseen since the run-up to the 2008 financial crisis. - [03:02](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=182s) Unlike the 2000s when foreign surpluses funded US Treasuries, modern capital flows prioritize high-risk equities, putting upward pressure on yields across the $40 trillion US national debt pile. - [05:38](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=338s) Geopolitical fragmentation and AI concentration are starving developing markets of vital financing needed for fundamental infrastructure like factories, roads, and power grids.

## Key takeaways - [00:00](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=0s) Global capital has disproportionately concentrated into American technology giants, pushing foreign ownership of US equities to 20%. - [00:58](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=58s) A circular macro loop has emerged where US data center spending funds East Asian hardware producers, whose resulting trade surpluses are reinvested directly back into US tech assets. - [02:03](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=123s) Total global current account surpluses and matching deficits reached approximately $2.5 trillion in 2025, mirroring the conditions preceding the 2008 crash. - [03:02](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=182s) Capital allocation has structurally shifted from conservative central banks buying risk-free debt to return-seeking asset managers and sovereign wealth funds targeting equities. - [04:22](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=262s) The AI capital drain saps foreign demand for US Treasuries, forcing Washington to pay higher auction yields to service its more than $40 trillion national debt. - [05:38](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=338s) The US administration's aggressive promotion of AI dominance directly conflicts with its stated objective of eliminating American trade deficits. - [05:38](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=338s) Global cross-border capital flows plateaued after peaking at 141% of global GDP in 2008, increasingly fragmenting along geopolitical lines and UN voting alignments. - [06:52](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=412s) Emerging economies in the global South face an acute infrastructure financing vacuum as capital is diverted toward Western AI data centers. - [07:25](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=445s) Due to widespread, opaque exposure across sovereign wealth funds, private debt, and retail portfolios, the systemic contagion risk of an AI market correction is difficult to assess. - [07:45](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=465s) Asian manufacturing hubs face a double risk if AI multiples compress: heavy losses on US equity holdings and an immediate collapse in hardware export demand.

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=0s) The Circular Flow of AI Capital

Global capital flows—essentially "a nerdy way of saying the money that flows around the world"—naturally migrate toward wherever investors anticipate the greatest returns. In the current global economic landscape, that appetite points toward a single dominant destination: artificial intelligence. An unprecedented volume of worldwide capital has funneled into a tiny concentration of powerful, predominantly American technology companies. Foreign investors now account for 20% of all US equity holdings, drawn by the explosive scale of the AI revolution. However, economists warn that capital drawn toward AI leaves far less available for other sectors and regions, raising deep concerns over mounting balance-of-payments distortions. If Big Tech giants fail to deliver on these lofty return expectations, that roaring investment pipeline could abruptly turn into a catastrophic drain.

In tech industry parlance, circular deals often describe transactional loops between enterprise software vendors and their suppliers. From a macroeconomic perspective, the circularity operates on a far grander international scale. Asian manufacturing hubs—chiefly South Korea, Japan, and Taiwan—produce the critical building blocks of AI infrastructure, including semiconductors, advanced GPUs, and memory chips demanded by the data center build-out in the United States. In supplying this equipment, these economies amass vast trade surpluses and domestic savings. Seeking attractive yields, those foreign savings find their way right back across the Pacific into US equities and corporate debt, directly bankrolling the very American AI boom that created those profits in the first place.

```mermaid flowchart LR accTitle: The Circular Flow of AI Capital accDescr: Diagram illustrating how Asian chip suppliers receive capital from US tech firms, then reinvest savings back into US markets.

subgraph US["United States AI Hub"] A["US Tech Giants & Data Centers"] B["US Equities & Corporate Bonds"] end

subgraph Asia["East Asian Chip Producers"] C["Taiwan, South Korea & Japan"] D["Foreign Savings & Surpluses"] end

A -->|"Buys semiconductors & GPUs"| C C -->|"Generates wealth"| D D -->|"Reinvests for higher yield"| B B -->|"Finances infrastructure"| A ```

While this circular loop accelerates the speed of AI deployment, it also constricts the geographical distribution of global capital and fuels historic trade divergences. By 2025, economies running current account surpluses accumulated roughly $2.5 trillion among them, matched by an identical deficit across shortfall nations. Historical precedence shows that when massive pools of foreign savings accumulate abroad, investors inevitably push outward into higher-yielding, higher-risk assets—an exact mirror of the conditions leading up to the 2008 global financial crisis. Expanding the data back to 2000 reveals that the gulf between current account surplus and deficit nations has not reached this magnitude since 2008.

- Foreign investors currently hold 20% of all US equity assets. - In 2025, combined current account surpluses reached roughly $2.5 trillion, matched by equivalent deficits globally. - The gap between current account surplus and deficit economies has returned to extremes last recorded right before the 2008 crash.

---

[03:02](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=182s) From Safe Treasuries to High-Risk Equities

Although the current account imbalances evoke memories of the mid-2000s, the destination of cross-border capital has radically shifted. In the early 2000s, massive foreign surpluses pouring in from China, East Asia, and Middle Eastern petrostates were channeled into ultrasafe, fixed-income vehicles. Foreign reserves back then were predominantly steered by conservative central banks seeking security in US Treasuries and government-sponsored enterprise bonds issued by Fannie Mae and Freddie Mac. In today's market, capital allocation is managed by return-hungry asset managers with significantly higher risk tolerance: sovereign wealth funds, hedge funds, and private pension systems.

Rather than stockpiling risk-free sovereign debt, modern global capital is targeting equities and corporate obligations. While foreign capital has not entirely abandoned the Treasury market, its overwhelming preference has migrated to tech equities. This structural shift creates a severe financing headache for Washington. Analysts observe a literal "huge sucking sound" pulling capital toward the AI sector, sapping demand for freshly issued US sovereign bonds.

When global investors scale back their purchases of American sovereign debt, the US government is forced to offer higher yields to clear its bond auctions. These elevated borrowing rates directly increase the debt service burden on the US Treasury, Treasury Secretary Scott Bessent, and domestic taxpayers, who must now finance the compounding interest on more than $40 trillion in total US debt.

- In the early 2000s, petrostates and Asian central banks parked their surpluses in Treasuries and agency debt from Fannie Mae and Freddie Mac. - Today, risk-seeking institutions—such as hedge funds, sovereign wealth funds, and pension funds—prioritize tech stocks and corporate debt over Treasuries. - Diverted capital places upward pressure on Treasury yields, raising interest service costs on the more than $40 trillion US national debt pile.

---

[05:38](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=338s) Geopolitics, Fragmentation, and Emerging Market Squeeze

The capital drain is further complicated by internal policy contradictions within Washington. Donald Trump has long framed narrowing the US trade deficit as an economic priority, asserting that American manufacturing was decimated by overseas trade. Concurrently, Trump has championed an "all-in" posture on AI, viewing the technology as an existential theater of geopolitical competition with China. However, this aggressive backing of the AI complex directly fuels the very global imbalances and trade deficits the administration rails against, all while the US itself introduces fresh macroeconomic volatility through tariffs and protectionism.

Beyond technology, geopolitics is actively splintering international capital allocation. Cross-border capital flows have decoupled from purely commercial logic; recent investment data reveals that countries with aligned United Nations voting records are significantly more prone to direct bilateral capital toward one another. Historically, global cross-border investments expanded from 27% of global GDP in 1980 to a peak of 141% in 2008. Following the financial crisis, that integration ground to a halt, plateauing despite an unprecedented abundance of global liquidity.

This geopolitical fracturing and tech concentration starves the broader global South. In an uninhibited global marketplace, capital theoretically seeks out productive opportunities wherever efficiency dictates. In reality, developing and emerging markets that desperately need financing for foundational economic infrastructure—factories, roads, ports, and power grids—are sidelined as resources are redirected toward American artificial intelligence and geopolitically aligned allies.

- Cross-border investments surged from 27% of global GDP in 1980 to 141% in 2008, before plateauing in the post-crisis era. - Modern capital allocation mirrors diplomatic blocs, with foreign investment tightly tracking aligned voting records at the United Nations. - Developing economies face an infrastructure financing vacuum as capital is diverted from physical grids and factories toward Western AI centers.

---

[07:25](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=445s) The Trillion-Dollar Question and Double Risk for Asia

The concentration of wealth in the AI ecosystem leads to a fundamental dilemma: is the global economy standing at the threshold of a genuine technological epoch, or is it financing the false promise of one? The central danger lies in the opacity of modern financial exposure. Because capital has circulated so broadly through sovereign funds, private debt, and passive retail portfolios into AI, determining the true contagion risk should valuations plummet remains extraordinarily difficult.

If tech multiples undergo a sharp contraction, Asian manufacturing economies face a devastating double blow: 1. **Asset Value Destruction:** Their sovereign and institutional investments placed into US technology stocks would suffer severe devaluations. 2. **Economic and Trade Contraction:** The sudden collapse in American data center spending would simultaneously depress their export orders, trade balances, and broader domestic industrial health.

As global trade imbalances push toward historic extremes alongside soaring valuations across global risk assets, economists caution that markets have built a combustible and historically volatile combination.

## Notable quotes > [00:00](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=0s) a nerdy way of saying the money that flows around the world

> [03:02](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=182s) huge sucking sound

> [05:38](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=338s) all-in

## Mentioned Bloomberg Originals, Fannie Mae, Freddie Mac, US Treasury, Scott Bessent, Donald Trump, United Nations

## Video description Bloomberg Originals unpacks how the artificial intelligence boom has resulted in a concentration of immense wealth. By drawing investment away from emerging markets and traditional industries, this massive capital drain is pushing global trade imbalances to levels unseen since the 2008 financial crisis.

CHAPTERS 00:00 Introduction 00:58 The Circular Flow of AI Capital 01:09 Title Card 02:03 Rising Imbalances 03:02 Shift from Safe Assets to Riskier Investments 04:22 US Debt and Treasury Market Pressures 05:38 Geopolitics and Fragmenting Investment Patterns 06:52 Capital Flow Stagnation and Emerging Market Needs 07:25 AI's Future Valuation 07:45 Double Risks for Asian Economies 08:10 Conclusion

#AI #Artificial #Intelligence

-------- Like this video? Subscribe: http://www.youtube.com/Bloomberg?sub_confirmation=1

Get unlimited access to Bloomberg.com for just $1.99 your first month: https://www.bloomberg.com/subscriptions?in_source=YoutubeOriginals Bloomberg Originals offers bold takes for curious minds on today’s biggest topics. Hosted by experts covering stories you haven’t seen and viewpoints you haven’t heard, you’ll discover cinematic, data-led shows that investigate the intersection of business and culture. Exploring every angle of climate change, technology, finance, sports and beyond, Bloomberg Originals is business as you’ve never seen it.

Subscribe for business news, but not as you've known it: exclusive interviews, fascinating profiles, data-driven analysis, and the latest in tech innovation from around the world.

Visit our partner channel Bloomberg News for global news and insight in an instant.

[Source](https://www.youtube.com/watch?v=xZPPWP0ZKCI)

中文简报

外资追逐大型科技股，美国AI热潮削弱美债需求流入美国人工智能领域的循环资本流正将全球失衡推向2008年峰值，推高40万亿美元国债的借贷成本，且若估值崩塌，亚洲供应商将面临双重冲击。该视频概述了一种宏观经济的循环回路：美国科技公司从东亚制造商处采购高端AI零部件，后者积累的贸易顺差随后又迅速回流并投资于美国股票和企业债。与2008年之前外国央行将储备资金注入安全的美国国债不同，现代资产管理机构在科技股中追逐更高风险的收益率，从而对美国40万亿美元国债的借贷成本构成上行压力。与此同时，地缘政治结盟日益主导国际资本流动，导致发展中经济体极度缺乏用于公路和电网等基础建设的关键资金。如果科技股估值崩塌，这种资本集中将使亚洲制造枢纽遭受投资组合贬值与出口需求断崖式下跌的双重严重冲击。人工智能热潮已将全球资本聚集到少数美国科技巨头手中，外国投资者目前持有全美股票总量的20%。 (0:00) 2025年全球经常账户失衡扩大至约2.5万亿美元，达到了2008年金融危机前夕以来未曾见过的极端分化水平。 (2:03) 与2000年代外国盈余资金资助美债不同，如今的资本流动优先流向高风险股票，推高了40万亿美元美国国债的收益率上行压力。 (3:02) 地缘政治碎片化与AI资本集中，正导致发展中市场极度匮乏建设工厂、道路和电网等基础设施所需的核心资金。 (5:38)

Reading summary

Global capital is increasingly concentrating into a narrow cluster of American artificial intelligence companies, propelled by a circular financing loop with East Asian hardware suppliers and a structural shift toward high-risk equity investments. Economists warn that this dynamic has inflated cross-border balance-of-payments imbalances to levels unseen since 2008, draining capital away from sovereign debt markets and developing economies. The Transpacific Circular Loop of AI Capital (0:00) Capital is concentrating in the US artificial intelligence ecosystem, with foreign investors now holding 20% of all US equity assets. This dynamic is reinforced by an international macroeconomic loop: American tech giants purchase advanced semiconductors, GPUs, and memory hardware from East Asian manufacturing hubs in Taiwan, South Korea, and Japan to construct data center infrastructure. The massive trade surpluses and domestic savings generated by these Asian suppliers are then recycled directly back across the Pacific into US equities and corporate debt to capture higher returns, effectively bankrolling the continued expansion of the US AI boom. From Sovereign Debt to High-Risk Tech Equities (3:02) While current account imbalances mirror conditions preceding the 2008 financial crisis—with global surplus and deficit nations each reaching approximately $2.5 trillion by 2025—the composition of foreign capital flows has fundamentally shifted. In the early 2000s, Asian central banks and Middle Eastern petrostates directed their reserves toward conservative, ultrasafe assets such as US Treasuries and agency debt from Fannie Mae and Freddie Mac. Today, asset allocation is dominated by yield-seeking institutions, including sovereign wealth funds, hedge funds, and private pension systems, that heavily favor tech stocks and corporate credit over risk-free government debt. This migration has created a significant financing challenge for the US government. Analysts describe a capital drain pulling funds toward AI equities and away from newly issued sovereign debt. To successfully clear bond auctions, the US Treasury must offer higher yields, directly escalating the cost of servicing a national debt load that exceeds $40 trillion. Geopolitical Alignment and the Emerging Market Deficit (5:38) Cross-border capital deployment is increasingly splintering along geopolitical allegiances rather than following purely market-driven logic. Investment data shows that cross-border funding flows correlate strongly with aligned United Nations voting records. This alignment coincides with a broader stagnation in international financial integration: cross-border investments surged from 27% of global GDP in 1980 to a peak of 141% in 2008, but have plateaued ever since despite substantial worldwide liquidity. This geopolitical clustering and concentration in Western AI creates a pronounced financing vacuum across the developing world. Emerging markets that require capital for essential physical infrastructure—such as power grids, factories, ports, and transportation networks—are sidelined as global resources are diverted toward American technology giants and diplomatically aligned partners. Meanwhile, US policy exhibits internal tensions: the Trump administration maintains an aggressive push for AI dominance that actively expands the very trade deficits and international imbalances it seeks to curtail through tariffs and protectionism. Dual Contagion Risks for Asian Suppliers (7:25) The scale and opacity of modern exposure to the AI sector pose severe systemic questions if technological returns fail to justify current valuations. Because capital has permeated sovereign entities, private debt markets, and passive retail portfolios, the exact contours of contagion risk remain difficult to map. If market multiples contract sharply, Asian manufacturing economies face a compounding double exposure: their sovereign and institutional asset holdings in US tech stocks would suffer steep devaluations, while an abrupt slowdown in US data center spending would simultaneously collapse their industrial export volumes and domestic trade balances. The global economy faces an unresolved dilemma over whether massive AI expenditures will inaugurate a genuine technological era or culminate in overextended asset valuations. Economists caution that the combination of historic current account divergences matching 2008 peaks alongside concentrated exposure to risk assets has established a combustible macroeconomic environment.

Detailed notes — zh

## Background 本视频由 Bloomberg Originals 于 2026 年 10 月 10 日发布在 YouTube 平台。页面简介指出，节目探讨了人工智能（AI）繁荣如何引发巨大的财富集中，并指出这种资本虹吸效应抽离了流向新兴市场与传统行业的投资，正将全球贸易失衡推升至 2008 年金融危机以来未曾见过的危险水平。 ## TL;DR 节目揭示了人工智能（AI）热潮在全球掀起的新型资本重构：东亚半导体产业链向美国出口硬件积累巨额顺差，随后又将资金反向注入美股与企业债，形成了自我强化的“循环流动”。这种结构使全球经常账户失衡重回 2008 年金融危机前夕的极端水平，且海外资本的风险偏好发生了从安全美债向高风险科技股权的根本性转变。巨额资金被 AI 资产吸干不仅推高了美国国债的融资成本与财政压力，也与缩减贸易逆差的施政目标形成悖论。在地缘政治阵营化导致跨国投资停滞的背景下，发展中经济体面临基建失血困境，而一旦 AI 超高估值发生破裂，亚洲等深度绑定的经济体恐将承受资产暴跌与外贸受挫的双重打击。 ## Atmosphere & style 全片延续了彭博深度报道一贯的冷静、克制与高度专业感，语调严谨且充满宏观层面的警示意味。叙事层层拆解全球贸易与金融齿轮的内在传动，将公众眼中华丽的科技热潮置于冷酷的宏观失衡图景中，营造出一种山雨欲来的压迫感。观看过程宛如一场穿透市场躁动的硬核财务解剖，观众在条理清晰的数据对比与图表推演中，能深切感受到全球金融系统悄然走向失衡临界点的紧张与深邃。 ## Bulletin - [00:00](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=0s) 全球资本流动正高度向以美国为主的 AI 巨头集中，外国投资者在美股总持有量中的占比已攀升至 20%，导致传统产业和新兴经济体面临严重的资金挤压。 - [00:58](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=58s) 东亚半导体出口与美国 AI 建设形成了资金闭环：韩国、日本与中国台湾凭借芯片出口积累的贸易顺差与储蓄，又反向回流至美股和企业债市场，进一步推高 AI 估值。 - [02:03](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=123s) 2025 年全球经常账户顺差与逆差规模均达约 2.5 万亿美元，失衡程度已重回 2008 年金融危机前夕水平，且资金偏好从避险的美国国债转向高风险的科技股与企业债。 - [04:22](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=262s) AI 资产如抽水机般吸走安全资金，导致美国财政部发债承接力减弱、融资成本攀升，而美国总债务规模现已突破 40 万亿美元。 - [07:45](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=465s) 若 AI 超高估值发生破裂，东亚等经济体将面临既承受在美资产减值、又面临出口与贸易顺差骤降的“双重打击”风险。 ## Key takeaways - [00:00](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=0s) 全球资本正以前所未有的速度汇入极少数以美国为主的 AI 科技巨头，使新兴市场与传统制造业面临资金严重失血的困境。 - [00:00](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=0s) 外国投资者在美股持有量中的占比达到 20%，表明海外资金对美国资产尤其是科技股权资产的依赖度极高。 - [00:58](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=58s) 东亚经济体（韩国、日本、中国台湾）通过出口先进芯片与关键硬件获取巨额顺差，这些储蓄随后又回流美国股市和企业债市场，形成自强化的资本闭环。 - [02:03](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=123s) 2025 年全球经常账户顺差与逆差规模双双达到约 2.5 万亿美元，再现 2008 年金融危机前夕的巨大宏观失衡裂痕。 - [03:02](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=182s) 相较于 2000 年代主要由各国央行主导、重仓美国国债的避险模式，本轮周期的资金主要由追求高回报的对冲基金、主权财富基金和养老金驱动，偏好高风险股票与企业债。 - [04:22](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=262s) 海外资本对科技巨头股权的追逐分流了原本借给美国政府的安全资金，迫使财政部提高国债收益率以吸引买家，加重了纳税人的付息负担。 - [04:22](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=262s) 在全美债务总额超过 40 万亿美元的背景下，AI 投资对美债流动性的抽离令新任财政政策与利息支出承受巨大压力。 - [04:22](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=262s) 特朗普政府既试图缩减贸易逆差、保护制造业，又全方位押注 AI 争取技术霸权，这种政策内在冲突反而进一步加剧了全球资本与贸易失衡。 - [05:38](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=338s) 地缘政治阵营化导致跨境投资根据联合国投票倾向等政治立场发生割裂，阻碍了资本在全球范围内的自由流动与高效配置。 - [06:52](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=412s) 全球跨境投资占 GDP 比例在 2008 年触及 141% 后陷入平台期，急需资金建设电网、交通等实体基建的发展中国家遭遇更严重的资本获取瓶颈。 - [07:25](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=445s) 市场正面临 AI 究竟是开启新时代还是虚假承诺的估值考验，一旦预期无法兑现，超高估值可能引发系统性流动性危机。 - [07:45](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=465s) 亚洲经济体面临双重宏观风险：既要在 AI 估值回调中承担巨大的海外投资减值损失，又要在全球科技需求下滑时承受本国实体出口萎缩的冲击。 ## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=0s) 全球资本流向与 AI 的“循环流动” “全球资本流动”听起来是一个偏学术的术语，本质上就是资金在世界各地的流向。资本往往会流向投资者预期回报率最高的地方，而当下这个目的地几乎只有一个：人工智能（AI）。市场上充斥着对 AI 的狂热追求，巨量资本集中流入了极少数实力雄厚、以美国为主的技术巨头手中。目前，外国投资者在美国股票总持有量中的占比已达到 20%。这场 AI 革命耗资极为巨大，而资金加速流向 AI 领域的代价，就是流入其他传统行业与新兴经济体的资金被大幅挤压。经济学家们深感忧虑的正是这种资本流动的极度失衡以及系统潜在的脆弱性。如果这些科技巨头无法兑现预期的超高回报，原本支撑行业扩张的庞大资金管道可能会瞬间转变为吸干流动性的深渊。从更宏观的经济学视角来看，全球范围内正在形成一个前所未有的“循环流动”（Circular Flow）： ```mermaid flowchart LR accTitle: AI全球资本循环流动图 accDescr: 展示东亚半导体出口与美国AI科技巨头投资之间的资金闭环 subgraph Asia["东亚经济体 (韩国 / 日本 / 中国台湾)"] A1["出口先进芯片与组件<br/>(GPU、高带宽内存等)"] A2["积累巨额贸易顺差与储蓄"] end subgraph US["美国科技与金融市场"] B1["建设大规模 AI 数据中心"] B2["美股与企业债市场<br/>(AI 科技巨头)"] end A1 -->|"供应硬件支撑"| B1 B1 -->|"支付采购资金"| A2 A2 -->|"资金寻求高额回报<br/>回流美国市场"| B2 B2 -->|"提供融资，推高估值"| B1 ``` 韩国、日本以及中国台湾地区是制造人工智能数据中心关键硬件（如 GPU、存储芯片等半导体组件）的主力供应商。在向美国数据中心建设出口大量高精尖硬件的过程中，这些经济体积累了数额极其庞大的外汇顺差与国民储蓄。为了让这些庞大资金继续获取高收益，它们又流回了美国股票市场和企业债券市场，反过来为这场造就了巨额储蓄的 AI 狂潮提供进一步的融资支持。这种自我强化的闭环既推动了 AI 产业以难以想象的速度爆炸式扩张，也使得全球资本的流向变得高度狭窄。 --- ### [02:03](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=123s) 失衡加剧与风险偏好的历史性转变这种循环正在制造令人担忧的全球宏观失衡。2025 年，全球经常账户顺差国合计录得约 2.5 万亿美元的顺差，而逆差国合计的赤字规模也基本相当。如果将数据回溯至 2000 年，可以清晰地看出一条历史规律：上一次全球经常账户顺差国与逆差国之间出现如此巨大的鸿沟，恰恰是在 2008 年全球金融危机前夕。每当海外积累了海量的超额储蓄，资本总会自发寻找高回报率但伴随高风险的资产。尽管历史规律显示出相似的迹象，但本轮周期与 2000 年代初相比存在一个至关重要的区别——**资金的去向和风险偏好发生了根本性转变**： | 历史时期 | 资金来源主体 | 管理机构及特征 | 主要流向资产 | |:--- |:--- |:--- |:--- | | **2000 年代初至 2008 年** | 中国、东亚其他经济体、海湾石油富国 | 各国央行主导；风格高度保守、极度厌恶风险 | 美国国债（Treasuries）、房利美（Fannie Mae）与房地美（Freddie Mac）等政府支持机构债券 | | **当下（AI 繁荣期）** | 追求高回报的全球机构资本 | 对冲基金、主权财富基金、养老基金；风险胃纳极高 | 美国股票（尤其是科技巨头股票）、高收益企业债券 | 在低利率环境推动下，股票和企业债券投资在金融危机前曾短暂激增，而今海外资金对美股的青睐全面复苏。虽然海外投资者并未完全抛弃美国国债，但他们对股权资产的狂热显著分流了原本用于支持政府借贷的安全资金。 --- ### [04:22](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=262s) 美债市场压力与特朗普政策的内在冲突全球资本被 AI 资产疯狂抽离的现象，为美国政府带来了巨大的现实难题。AI 产业像一台马力全开的抽水机一样吸食着原本可能流入安全资产的资金，这种“巨大的吸吮声”（Huge sucking sound）直接冲击了美国财政部发行国债的承接力。 - **美债融资成本飙升**：当借钱给美国的全球买家变少时，财政部就必须开出更优厚的条件（即支付更高的国债收益率与利息），这直接加重了美国纳税人以及财政部长斯科特·贝森特（Scott Bessent）的利息负担。目前美国债务总额已突破 40 万亿美元大关，沉重的利息偿还规模正演变为严峻的财政压力。 - **政策目标的自相矛盾**：自 2016 年以来，唐纳德·特朗普始终将缩减贸易逆差作为经济施政核心，认为海外竞争掏空了美国制造业；但与此同时，特朗普又对 AI 产业进行了全方位押注，认定美国正与中国在这一关键核心技术的主导权上展开一场“生死存亡级的竞争”。大力押注 AI 在客观上又反向推高了全球贸易与资本流动的失衡，导致美国反倒成了全球宏观波动的主要发源地。 --- ### [05:38](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=338s) 地缘政治割裂与新兴市场的资本饥渴即使撇开 AI 这头房间里的大象，地缘政治的深度介入也在深刻改写资本流动的版图。最近的跨境投资数据显示，地缘政治阵营正在加速固化——在联合国投票倾向一致的国家之间，彼此互相投资的意愿明显更高，资本早已不再如过去全球化时代那样无国界自由配置。数据显示，全球跨境投资占全球 GDP 的比例在 1980 年仅为 27%，到 2008 年一路攀升至 141%，但金融危机之后这一扩张势头戛然而止并长期处于平台停滞期。虽然如今全球范围内飞舞的“热钱”总量远超以往，资本却不再遵循纯粹的经济效率逻辑流入最需要它的地方： - **新兴市场基础设施严重失血**：广大发展中经济体急需巨量资金用于修建厂房、交通道路、电力网络等关键实体基础设施，然而地缘政治对抗、美国日趋严厉的保护主义政策，加上 AI 对资金的无情吸纳，使得发展中国家的资本获取难度剧增。 - **估值暴跌的系统性外溢**：资本市场目前面临一个价值万亿美元的终极拷问：“我们究竟是正处于一个新时代的黎明，还是在为一个新时代的虚假承诺买单？”一旦 AI 的超高估值发生剧烈回调，全球投资者才会惊觉自己的风险敞口早已深不见底。亚洲经济体在这种剧烈震荡中极易遭受**双重打击**（Double Risks）：一方面，其先前大规模注入美国科技公司的顺差资金面临资产大幅减值与估值缩水；另一方面，伴随全球科技需求骤降与外部环境恶化，其自身的贸易顺差与实体经济也将遭到全面重创。全球贸易失衡加剧与全球风险资产泡沫化的叠加，正构成一个极度动荡且危险的宏观组合。 ## Notable quotes > [00:00](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=0s) 外国投资者在美国股票总持有量中的占比已达到 20%。 > [00:00](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=0s) 如果这些科技巨头无法兑现预期的超高回报，原本支撑行业扩张的庞大资金管道可能会瞬间转变为吸干流动性的深渊。 > [02:03](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=123s) 上一次全球经常账户顺差国与逆差国之间出现如此巨大的鸿沟，恰恰是在 2008 年全球金融危机前夕。 > [02:03](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=123s) 每当海外积累了海量的超额储蓄，资本总会自发寻找高回报率但伴随高风险的资产。 > [04:22](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=262s) AI 产业像一台马力全开的抽水机一样吸食着原本可能流入安全资产的资金，这种“巨大的吸吮声”直接冲击了美国财政部发行国债的承接力。 > [04:22](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=262s) 目前美国债务总额已突破 40 万亿美元大关，沉重的利息偿还规模正演变为严峻的财政压力。 > [05:38](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=338s) 在联合国投票倾向一致的国家之间，彼此互相投资的意愿明显更高，资本早已不再如过去全球化时代那样无国界自由配置。 > [05:38](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=338s) 全球跨境投资占全球 GDP 的比例在 1980 年仅为 27%，到 2008 年一路攀升至 141%，但金融危机之后这一扩张势头戛然而止并长期处于平台停滞期。 > [07:25](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=445s) 我们究竟是正处于一个新时代的黎明，还是在为一个新时代的虚假承诺买单？ > [08:10](https://www.youtube.com/watch?v=xZPPWP0ZKCI&t=490s) 全球贸易失衡加剧与全球风险资产泡沫化的叠加，正构成一个极度动荡且危险的宏观组合。 ## Mentioned Bloomberg Originals, 人工智能 (AI), 美国财政部, 斯科特·贝森特 (Scott Bessent), 唐纳德·特朗普 (Donald Trump), 房利美 (Fannie Mae), 房地美 (Freddie Mac), 联合国 (United Nations), 美国国债, 高带宽内存 (HBM), GPU ## Video description Bloomberg Originals unpacks how the artificial intelligence boom has resulted in a concentration of immense wealth. By drawing investment away from emerging markets and traditional industries, this massive capital drain is pushing global trade imbalances to levels unseen since the 2008 financial crisis. CHAPTERS 00:00 Introduction 00:58 The Circular Flow of AI Capital 01:09 Title Card 02:03 Rising Imbalances 03:02 Shift from Safe Assets to Riskier Investments 04:22 US Debt and Treasury Market Pressures 05:38 Geopolitics and Fragmenting Investment Patterns 06:52 Capital Flow Stagnation and Emerging Market Needs 07:25 AI's Future Valuation 07:45 Double Risks for Asian Economies 08:10 Conclusion #AI #Artificial #Intelligence -------- Like this video? Subscribe: http://www.youtube.com/Bloomberg?sub_confirmation=1 Get unlimited access to Bloomberg.com for just $1.99 your first month: https://www.bloomberg.com/subscriptions?in_source=YoutubeOriginals Bloomberg Originals offers bold takes for curious minds on today’s biggest topics. Hosted by experts covering stories you haven’t seen and viewpoints you haven’t heard, you’ll discover cinematic, data-led shows that investigate the intersection of business and culture. Exploring every angle of climate change, technology, finance, sports and beyond, Bloomberg Originals is business as you’ve never seen it. Subscribe for business news, but not as you've known it: exclusive interviews, fascinating profiles, data-driven analysis, and the latest in tech innovation from around the world. Visit our partner channel Bloomberg News for global news and insight in an instant. [Source](https://www.youtube.com/watch?v=xZPPWP0ZKCI)

Source quotes

[0:00] 外国投资者在美国股票总持有量中的占比已达到 20%。 [0:00] 如果这些科技巨头无法兑现预期的超高回报，原本支撑行业扩张的庞大资金管道可能会瞬间转变为吸干流动性的深渊。 [2:03] 上一次全球经常账户顺差国与逆差国之间出现如此巨大的鸿沟，恰恰是在 2008 年全球金融危机前夕。 [2:03] 每当海外积累了海量的超额储蓄，资本总会自发寻找高回报率但伴随高风险的资产。 [4:22] AI 产业像一台马力全开的抽水机一样吸食着原本可能流入安全资产的资金，这种“巨大的吸吮声”直接冲击了美国财政部发行国债的承接力。 [4:22] 目前美国债务总额已突破 40 万亿美元大关，沉重的利息偿还规模正演变为严峻的财政压力。 [5:38] 在联合国投票倾向一致的国家之间，彼此互相投资的意愿明显更高，资本早已不再如过去全球化时代那样无国界自由配置。 [5:38] 全球跨境投资占全球 GDP 的比例在 1980 年仅为 27%，到 2008 年一路攀升至 141%，但金融危机之后这一扩张势头戛然而止并长期处于平台停滞期。 [7:25] 我们究竟是正处于一个新时代的黎明，还是在为一个新时代的虚假承诺买单？ [8:10] 全球贸易失衡加剧与全球风险资产泡沫化的叠加，正构成一个极度动荡且危险的宏观组合。

Source quotes

[0:00] a nerdy way of saying the money that flows around the world [3:02] huge sucking sound [5:38] all-in

Original transcript

[0:00] So Global capital flows, it's just a nerdy way of saying the money that flows around the world. And directionally, it does tend to be wherever investors expect the greatest return, which right now pretty much means one thing. AI. AI. AI. AI. AI. AI. There's so much frenzy around AI. An extraordinary amount of capital has flowed into a tiny number of powerful, predominantly American hands. in a lot of investments from all of the world. In fact, foreign investors now count for 20% of all US equity holdings. The AI revolution isn't going to come cheap. The more money AI draws, the less is available to other industries and countries. What people get worried about is the imbalances in these flows. And that's where you have kind of potential vulnerabilities in the Giants need to deliver strong returns. And if they don't, that pipeline could suddenly become a drain. In the world of AI, a circular deal often refers to transactions between customers and their suppliers. For us, as economists, we're taking a larger view. South Korea, Japan, Taiwan are the main producers of these semiconductors, GPU, memory chips that are critical for the AI data center build out in the United States. up a huge amount of wealth and savings. They are looking for a way to get a return on it. Some of those savings then find their way back into U.S. equities and corporate bonds. helping finance the very AI boom that created them. That's the circular flow. It's one reason why the AI revolution is taking place so fast. The same loop that speeds up the AI boom also narrows where the money flows. And it's creating imbalances which look like this. In 2025, economies running current account surpluses recorded about $2.5 trillion between them. Deficit economies ran shortfalls, adding up to about the same amount. Typically what we've seen from history is that whenever there's a huge amount of savings built up in foreign countries, investors like to look for investments that have high return but tend to have more risk to it. which is what we saw in the period leading up to the financial crisis. Expand the chart back to the year 2000 and a pattern begins to emerge. The last time the gap between current account surplus and deficit countries was this wide was in 2008. Not saying that the current... inflow in for money is going to lead to the same outcome, but there are parallels in There is, however, one big difference. This time, the money is going somewhere else. So in the early 2000s, foreign capital was pouring into the United States from China from other East Asian countries, from the petro states of the Gulf, and it was parked in US treasuries. Global savings tended to be managed by central banks, and they were very conservative and risk-averse. Things are different now. The money is being managed by investors with more appetite for risk. Hedge funds, sovereign wealth funds, pension funds, all of them are institutions that are willing to embrace more risk are hoping for higher returns. Foreign investors haven't abandoned US treasuries. They're just much more interested in equities. In the early 2000s, those investors largely played it safe, buying treasuries and bonds issued by government-backed institutions such as Fannie Mae and Freddie Mac. Then equities and corporate bonds took off, thanks in part to low interest rates, right up until the global financial crisis. Now equity buying is back. For the US, that could create a large and awkward problem. Essentially, there is just a huge sucking sound that is coming from the AI world for capital that is starting to affect the ability of the U.S. government to sell treasuries. If fewer people choose to lend America money, America has to offer them a better deal. And that adds to the cost of interest payments for Secretary Besant and for U.S. taxpayers. interest on more than $40 trillion of US debt. Since 2016, Donald Trump has said bringing [5:00] on trade deficits is a priority, partly because he thinks U.S. manufacturing has been hollowed out by overseas competition. But on the other hand... Trump has gone all in on AI. He really believes that the US is in an existential competition with China over leadership of this important technology. And the reality is that that bet on AI is also feeding back into global imbalances. And it's coming at a time when the US has become a source of volatility. More on that in just a moment. Even if we drag the AI-shaped elephant from the room, political uncertainty is still affecting where the money goes. What we've seen in the last few years is geopolitics reasserting itself. In the most recent investment data, countries that vote the same way in the United Nations, they're more likely to invest in each other. And that only makes the imbalance worse. putting money to work. in each other's economies in the same way that they did in the past. Which is what this data shows. Cross-border investments have largely plateaued. They grew from 27% of global GDP in 1980 to 141% in 2008. But since the financial crisis, things have slowed down. Thank you. And this is happening despite there being more money, depicted here in bull form, if that wasn't clear, flying around than ever before. The result is capital isn't necessarily getting to the places where logically it should be, or where it's most needed. If you don't have to worry about politics, capital can just flow to wherever the returns are highest. There are developing economies and emerging markets who need capital to invest in infrastructure, factories, roads, electricity networks. And geopolitics is making the allocation harder. Under Trump, the US has also become more protectionist. Meanwhile, AI continues to suck up global capital. The trillion-dollar question today is, are we at the dawn of a new age, or are we investing in the false promise of a new age? And with so much money in AI, what happens if valuations fall sharply and people discover they were more exposed than they'd anticipated? It's very hard to tell how widespread this exposure is. real consequences for people's livelihoods. And in a way, Asia economies can be hit twice, where the amount they've invested into US companies gets devalued. And at the same time, their trade balance and economy, along with it, weakens as well. As global trade imbalances rise, global risk assets are rising as well. Some worry that that is a volatile combination. *music* *music*