Edward Conard

Top Ten New York Times Bestselling Author

  • “Unintended Consequences represents the most cogent and persuasive analysis of the Financial Crisis to date.” - Andrei Shleifer, 1999 John Bates Clark Medal Winner
  • “…serious thinking for serious thinkers. …a thought-provoking blueprint for growing middle- and working-class incomes.” - Mitt Romney, former Governor of Massachusetts
  • “A full-throated defense of economic dynamism.” - The Wall Street Journal
  • “Unintended Consequences is full of substance, it is one of the must-read books of the year, and once I finish it I will be giving it a second read through right away.” - Tyler Cowen, Professor, George Mason University
  • “…serious thinking for serious thinkers. …a thought-provoking blueprint for growing middle- and working-class incomes.” - Mitt Romney, former Governor of Massachusetts
  • “Unintended Consequences should be read by anyone who takes for granted the superiority of progressive taxation and has not thought carefully about the trade-offs involved.” - The New Republic
  • “…challenges misconceptions that distort our economic debates.” - Arthur Brooks, President of the American Enterprise Institute
  • “…a very valuable contribution.” - Larry Summers, former Secretary of the Treasury and director of the National Economic Council, president emeritus, Harvard University
  • “Unintended Consequences offers deep and well-argued analyses on almost every issue.” - The New York Times
  • “Unintended Consequences is far smarter and more thought-provoking than most economics written for the general public” - Greg Mankiw, Harvard University, Former Chairman of the Council of Economic Advisors
  • “…a fresh argument for the productive value of inequality.” - David Autor, Professor of Economics, Massachusetts Institute of Technology
  • “…a comprehensive explanation of the modern economy.” - Julian Robertson, Founder, Tiger Management
Upside of Inequality Oxford Unintended Consequences

Macro Roundup


Friday, June 12, 2026

AI Summary. Open-source artificial intelligence models cost far less per unit of computing than leading proprietary ones, with one Anthropic model priced more than 50x DeepSeek's per token. Researchers say top proprietary models remain four to six months ahead and can sometimes finish complex tasks using fewer tokens, lowering total cost.

Bradley Olson and Tina Li Wall Street Journal
Date Posted:

Token pricing for Anthropic’s new Fable 5 model is 50x that of DeepSeek’s V4 Pro. A Citadel notes cautions, “Even the most powerful technologies must pass through the prosaic discipline of cost curves, capacity constraints and marginal returns.”

Are cheaper open-source models pressuring proprietary AI pricing?

Core argument: Anthropic's Fable 5 model costs more than 50 times as much per token as DeepSeek's V4 Pro, with open-source models priced far below proprietary alternatives on the basic unit of AI computing.

Open-source models cost far less per token, the basic unit of AI computing. Anthropic’s recently-released Fable 5 model is more than 50 times more expensive per token than DeepSeek’s V4 Pro. The top proprietary models from companies like OpenAI, Anthropic or Google remain four to six months ahead of open-source competitors, researchers say. In some cases that means they can complete a complex task using fewer tokens, equating to a lower total cost. The shift to cheaper models appears to have played a role in a recent decline in a widely followed index that tracks AI spending, Citadel Securities said in a report this week. “Even the most powerful technologies must pass through the prosaic discipline of cost curves, capacity constraints and marginal returns,” the report said.

Takeaways by Macro Roundup® AI

  1. Anthropic's Fable 5 model costs more than 50 times as much per token as DeepSeek's V4 Pro, with open-source models priced far below proprietary alternatives on the basic unit of AI computing.
  2. Researchers say top proprietary models from OpenAI, Anthropic and Google remain four to six months ahead of open-source rivals, in some cases completing complex tasks with fewer tokens and thus lower total cost.
  3. Citadel Securities said in a report that the shift to cheaper models appears to have played a role in a decline in a widely followed index tracking AI spending, citing cost curves, capacity constraints and marginal returns.

Related Articles:

  • Four Takeaways from Microsoft’s Satya Nadella at WEF Today — Microsoft's chief executive says the price of AI model usage ("tokens") falls by roughly half every three months. Commentary argues that such rapid deflation pulls demand forward, encourages overbuilding, and commoditizes models, shifting competitive advantage to orchestration and context rather than the model layer.
  • OpenAI Considers Drastic Price Cuts, Anticipating War for Users With Anthropic — OpenAI is considering steep price cuts as it competes with Anthropic for enterprise customers, a move that could erode both companies' margins given they already lose billions of dollars on the computing costs of running AI systems. Anthropic's revenue surged after its Claude Code coding tool spread among software engineers, and it surpassed OpenAI's valuation.
  • Measurements Matter: The Return of SLOC Slop — Measuring developer productivity by software lines of code captures output rather than value, and AI inflates the metric by generating boilerplate, wrappers, tests, and speculative code at near-zero cost. More code also creates liabilities through added maintenance, complexity, attack surface, and debugging load.

AI Summary. Total spending on large language model usage has roughly doubled even as the price of a single token has fallen more than 90%, an illustration of Jevons' paradox. As tokens get cheaper, companies run more AI agents, automate more workflows, and generate more code, pushing aggregate expenditure higher.

Torsten Sløk Apollo
Date Posted:

An index of token expenditure has ~ doubled since late 2025 even as the price of a single token has fallen by ~90% since 2023. As prices fall, firms “run more AI agents, automate more workflows and generate more code,” Sløk notes.

Does falling token pricing raise total AI spending?

Core argument: The Silicon Data Token Expenditure Index, which tracks total spending on large language model usage, has roughly doubled since late 2025, even as the price of a single token has fallen more than 90% since 2023.

The Silicon Data Token Expenditure Index, which tracks total spending on large language model usage, has roughly doubled since late 2025, even as the price of a single token has fallen more than 90% since 2023. This is Jevons' paradox in action. As tokens get cheaper, companies don't spend less but instead run more AI agents, automate more workflows, and generate more code, pushing aggregate expenditure higher even as the unit cost of intelligence collapses.

Takeaways by Macro Roundup® AI

  1. The Silicon Data Token Expenditure Index, which tracks total spending on large language model usage, has roughly doubled since late 2025, even as the price of a single token has fallen more than 90% since 2023.
  2. Cheaper tokens illustrate Jevons' paradox: companies respond by running more AI agents, automating more workflows, and generating more code, pushing aggregate spending higher as the unit cost of intelligence collapses.

Related Articles:

  • OpenAI Considers Drastic Price Cuts, Anticipating War for Users With Anthropic — OpenAI is considering steep price cuts as it competes with Anthropic for enterprise customers, a move that could erode both companies' margins given they already lose billions of dollars on the computing costs of running AI systems. Anthropic's revenue surged after its Claude Code coding tool spread among software engineers, and it surpassed OpenAI's valuation.
  • Chart of the Day: Small Models are Closing the Gap to Frontier AI — Achieving 60% on MMLU, a test of academic reasoning, once required a 540bn-parameter model but was reached by a 3.8bn-parameter model two years later, a 142x reduction in parameters. Models small enough to run on phones or sensors shift inference from data centers to devices, weakening the case for paying cloud providers per token.
  • We’re Using So Much AI That Computing Firepower Is Running Out — Surging AI demand is straining computing capacity: Anthropic's Claude programming interface recorded 98.95% uptime over a 90-day span, below the 99.99% software companies typically promise, as its annualized revenue run rate climbed from $9bn to $14bn to $30bn. The company has capped how much usage customers can consume during weekday peak morning hours.

AI Summary. Net equity issuance by US domestic non-financial companies has turned positive for the first time since 2021, and a pipeline of initial public offerings and secondary offerings points toward further increases. Net shareholder payouts remain positive on strong earnings and dividends, but the author argues that falling earnings or rising issuance would worsen the outlook for US equities.

Will Denyer Gavekal
Date Posted:

The SpaceX lockups will start to expire btw mid-August and mid-December, unlocking ~ half of SpaceX’s total market capitalization at its IPO price as Anthropic and OpenAI go public in H2. Net shareholder payouts remain positive in H1 2026.

Does rising net equity issuance weaken the outlook for US stocks?

Core argument: Net equity issuance by the US domestic non-financial sector turned positive in 1Q26 for the first time since 2021, covering IPOs, secondary offerings, private placements, and stock-based compensation net of buybacks and merger-related retirements.

Flow of funds data showed net equity issuance by the domestic non-financial sector turned positive in 1Q26 for the first time since 2021. This measure includes IPOs, secondary offerings, private placements, and stock-based compensation, net of share retirements through buybacks and mergers. Given the pipeline of upcoming IPOs and secondary offerings, net issuance seems likely to increase further—which is anything but positive for equity markets. For now, US companies continue to generate strong earnings and pay hefty dividends, such that net shareholder payouts—net dividends less net equity issuance—remain positive. But if earnings and dividends were to roll over, or net equity issuance were to rise significantly from here—and that does seem to be the current direction of travel—the outlook for US equity markets in the second half of the year could become more problematic. Additionally, sales of unlocked shares could provide a second wave of liquidity stresses months after the mega-IPOs.

Takeaways by Macro Roundup® AI

  1. Net equity issuance by the US domestic non-financial sector turned positive in 1Q26 for the first time since 2021, covering IPOs, secondary offerings, private placements, and stock-based compensation net of buybacks and merger-related retirements.
  2. Net shareholder payouts—net dividends less net equity issuance—remain positive as US companies continue to generate strong earnings and pay hefty dividends, but a rollover in earnings and dividends or a significant rise in issuance could make the second-half outlook for US equities more problematic.
  3. The pipeline of upcoming IPOs and secondary offerings suggests net issuance is likely to rise further, and sales of unlocked shares could bring a second wave of liquidity stresses months after the mega-IPOs.

Related Articles:

  • US Stock Market To Stop Shrinking For First Time In 23 Years — Goldman Sachs estimates that net supply of US equity — new shares issued less shares removed by buybacks and take-privates — will be roughly flat in 2026 after being negative since 2003, with a larger influx expected in 2027 as lock-up periods on this year's initial public offerings expire.
  • One Hundred Years in the U.S. Stock Markets — Over a century, U.S. publicly traded stocks generated roughly $91tn in shareholder wealth enhancement, with returns strongly positively skewed and dominated by a relatively small number of high-performing firms. Because most possible outcomes fall below the mean in a skewed distribution, the authors argue financial planning should focus on median outcomes or the full distribution.
  • Capital Allocation — Internal financing has been the main source of capital for U.S. corporations, with steady new debt issuance except in recession years, consistent with a stable capital structure. Companies have retired equity on average, and net equity issuance remains negative even after accounting for stock-based compensation.

AI Summary. The International Monetary Fund calculates that global current-account deficits plus surpluses reached 3.7% of world gross domestic product, after fluctuating between 1% and 3% until the early 2000s. The US deficit of $1.1tn is the largest single imbalance, and the Fund estimates a budget deficit of 2% of GDP raises the current-account deficit by 0.5% of GDP.

Greg Ip The Wall Street Journal
Date Posted:
Is Database:
Database

Greg Ip warns that while global imbalances are “narrower” than they were pre-crisis, “they are more persistent and more entrenched in national behavior,” pointing to China’s trade surplus and chronic US fiscal deficits.

How large have global current-account imbalances grown relative to world output?

Core argument: The IMF calculates that global current-account deficits plus surpluses reached 3.7% of global GDP last year after falling steadily from around the global financial crisis; until the early 2000s they fluctuated between 1% and 3%.

The IMF calculates that deficits plus surpluses reached 3.7% of global gross domestic product last year after falling steadily from around the global financial crisis. Until the early 2000s, they fluctuated between 1% and 3%. This is worrisome. Current-account deficits played a part in the crises that swept Latin America in the early 1980s, East and Southeast Asia in the late 1990s, the U.S. in 2007-09, and the eurozone from 2009 on. Today’s imbalances look different. While annual deficits are narrower than 15 years ago, they are more persistent and more entrenched in national behaviour. Take the U.S., whose current-account deficit of $1.1 trillion is by far the largest single imbalance. Last year, tariffs did reduce some imports, but because the AI boom sucked in foreign-made tech equipment, the current-account deficit narrowed only slightly. The bigger contributor to the U.S. current-account deficit is its budget deficit, which sustains excessive U.S. spending and inadequate saving. The IMF estimates a budget deficit of 2% of GDP increases the current-account deficit by 0.5% of GDP.

Takeaways by Macro Roundup® AI

  1. The IMF calculates that global current-account deficits plus surpluses reached 3.7% of global GDP last year after falling steadily from around the global financial crisis; until the early 2000s they fluctuated between 1% and 3%.
  2. The U.S. current-account deficit of $1.1 trillion is by far the largest single imbalance, narrowing only slightly last year as tariffs reduced some imports while the AI boom drew in foreign-made tech equipment.
  3. The IMF estimates that a budget deficit of 2% of GDP increases the current-account deficit by 0.5% of GDP, and the U.S. budget deficit sustains excessive spending and inadequate saving.

Related Articles:

  • Understanding Global Imbalances — The US, China, Germany, and Japan account for roughly two-thirds of global current account imbalances, with the US deficit at 4% of GDP financed by capital inflows into dollar assets. The average duration of a surplus or deficit spell roughly doubled since the 1980s, building large net foreign asset and liability positions.
  • Why Global Imbalances Matter — Countries running persistent trade surpluses, such as China, Germany and Japan, have built their economies around an excess supply of tradeable goods and services, leaving them dependent on deficit countries to absorb it. That dependence gives deficit countries bargaining power and exposes surplus countries to pressure to expand domestic demand.
  • The U.S. Trade Deficit: Myths and Realities — Taxing capital inflows would weaken the dollar but could raise domestic interest rates and depress investment, harming long-term growth, while an interest rate cut absent recession would be inflationary and erode any real depreciation needed to durably improve the trade balance. Fiscal restraint could weaken the dollar and reduce the trade deficit.

AI Summary. Rising gasoline prices have pushed average hourly earnings 0.7% below their year-earlier level after adjusting for inflation, erasing more than a year of wage gains. Inflation eroding wage increases is one reason consumer sentiment sits near record lows despite relatively low unemployment and a strong stock market.

Justin Lahart The Wall Street Journal
Date Posted:
Is Database:
Database

The .7% y/y decline in real wages has brought average hourly earnings back to their level when Trump took office; the downtick is half the size of their cumulative decline over Biden’s four years.

Have rising gas prices erased more than a year of wage gains?

Core argument: Average hourly earnings adjusted for inflation are 0.7% below their year-earlier level, the biggest such gap since February 2023, as surging gasoline prices erase more than a year of U.S. wage gains.

Surging gasoline prices have erased more than a year of Americans’ wage gains. That inflation-adjusted gap—average-hourly earnings falling 0.7% below their year-earlier level—is the biggest since February 2023. The mismatch has quickly knocked Americans’ inflation-adjusted earnings back to where they were in January 2025, when President Trump returned to the White House. (Over the course of President Joe Biden’s four years in office, when prices surged, inflation-adjusted earnings fell 1.4%.) Seeing inflation eat into wage increases is one reason consumer sentiment is hovering around near-record lows, despite relatively low unemployment and a hot stock market.

Takeaways by Macro Roundup® AI

  1. Average hourly earnings adjusted for inflation are 0.7% below their year-earlier level, the biggest such gap since February 2023, as surging gasoline prices erase more than a year of U.S. wage gains.
  2. Inflation-adjusted earnings have fallen back to their January 2025 level, when President Trump returned to the White House; over President Biden's four years in office they fell 1.4%.
  3. Inflation eating into wage increases is one reason consumer sentiment hovers around near-record lows despite relatively low unemployment and a hot stock market.

Related Articles:

  • US Producer Prices Rise at Fastest Pace Since November 2022 — The Bureau of Labor Statistics reported that US producer prices rose 6.5% from a year earlier, the fastest annual pace in more than two years, and 1.1% from the prior month. A core measure excluding food and energy rose 4.9% year over year.
  • US Consumer Sentiment Slides to Record Low on Price Concerns — The University of Michigan's consumer sentiment index fell 5 points to 44.8, a record low and weaker than all economist projections in a Bloomberg survey. Consumers expect prices to rise 3.9% annually over the next five to 10 years, up from 3.5%, and 4.8% over the coming year.
  • Do Voters Punish Inflation or Pay Cuts? Inflation and Real Wages in U.S. Elections — An analysis of U.S. counties estimates that where wages grew faster than prices, voters supported the incumbent party more, while counties with larger real wage losses than average shifted more toward the opposition. A one-standard-deviation larger real wage loss is associated with shifts of roughly 0.10 and 0.19 percentage points in vote share and margin.

AI Summary. McKinsey forecasts the number of quantum computing systems worldwide will rise from scores today to around 5,000 by 2030, though existing machines remain highly error-prone and face formidable engineering challenges. As these systems begin to outperform conventional computers in niche areas, organizations are preparing for the predicted moment when quantum computers can break current cryptography.

Michael Peel Financial Times
Date Posted:

Both Google and IBM are targeting the delivery of “useful” quantum computers by 2030. Major technical obstacles remain. This fall, Google’s CEO said the technology was now “where maybe AI was five years ago.”

Is quantum computing advancing faster than skeptics expect?

Core argument: McKinsey forecasts the number of quantum computing systems worldwide will rise from scores today to roughly 5,000 by 2030.

Industry observers estimate there are scores of quantum computing systems in the world, a number forecast by McKinsey to rise to around 5,000 by 2030. Sceptics point to the significant technical hurdles that remain to making useful quantum computers. Existing machines are still far too error-prone. The “engineering challenges” that the devices’ developers insist can be solved in the next few years are formidable. But as the new machines begin to indicate they can outperform conventional computers in niche areas — what the industry dubs “quantum advantage” — the risk of dismissing the technology’s importance is growing. Organisations are readying themselves for the prospect of Q-Day — the predicted moment when quantum computers are capable of breaking the cryptographic methods on which modern societies rely.

Takeaways by Macro Roundup® AI

  1. McKinsey forecasts the number of quantum computing systems worldwide will rise from scores today to roughly 5,000 by 2030.
  2. Sceptics note that existing quantum machines remain far too error-prone, and the engineering challenges developers say can be solved within a few years are formidable.
  3. Organisations are preparing for Q-Day, the predicted moment when quantum computers can break the cryptographic methods modern societies rely on, as machines begin to indicate "quantum advantage" over conventional computers in niche areas.

Related Articles:

  • New Advances Bring the Era of Quantum Computers Closer Than Ever — Estimates of the quantum computing power needed to run Shor's algorithm and break encryption have fallen from billions of qubits to a million, while existing machines have hundreds. Researchers have announced a design requiring only tens of thousands of qubits and an implementation of the algorithm 10x as efficient as the prior best.
  • U.S. to Award Quantum-Computing Firms $2 Billion and Take Equity Stakes — The U.S. government will award about $2bn to quantum-computing firms under the 2022 Chips and Science Act, taking a minority equity stake in each company. D-Wave said its entire $100m award would be an equity investment, with Rigetti and Infleqtion describing similar structures.
  • An Old Problem Could Stop The U.S. From Leading On Quantum Computers — Many quantum computers depend on dilution refrigerators, produced by a few small firms in the US, Britain and Finland, and on critical minerals such as niobium, concentrated in Brazil, and tantalum, mined in Africa and mostly processed in China. China has announced mass production of a domestic dilution refrigerator and has been buying niobium mines.

Monday, June 15, 2026

AI Summary. JPMorgan projects that share sales will add roughly $1.5tn of stock to the US equity market over the next two years even after buybacks, which would be the strongest net equity issuance since at least the late 1990s. Equity has grown cheaper relative to debt as a funding source.

Lu Wang and Carmen Reinicke Bloomberg
Date Posted:
Is Database:
Database

JPM estimates that between 2006 and 2025, net US equity issuance was ~ negative $430B. This de-equitization trend may be over, as JPM forecasts that ~$1.5T of stock will be added to US equity markets over the next two years.

Is the long era of shrinking stock supply reversing?

Core argument: JPMorgan projects that IPOs, secondary offerings, and other share sales will add roughly $1.5 trillion of stock to the US equity market over the next two years, even after accounting for buybacks.

For the better part of two decades, a defining feature of the US stock market has been scarcity. Year after year, shares disappeared from public hands, with buybacks by S&P 500 companies alone erasing nearly $12 trillion worth. According to JPMorgan, IPOs, secondary offerings, and other share sales are poised to add roughly $1.5 trillion of stock to the US equity market over the next two years, even after accounting for buybacks. If realized, it would mark the strongest period of net equity issuance since at least the late 1990s. The shift toward selling more equity is being driven not only by a need for capital but by the opportunity to raise it on attractive terms. The cost of equity has been getting cheaper vis-à-vis debt since the Federal Reserve raised rates to a two-decade high in 2023, ending a prolonged era of cheap borrowing. Thanks to a $30 trillion share rally since, the advantage has persisted even as the Fed started cutting rates.

Takeaways by Macro Roundup® AI

  1. JPMorgan projects that IPOs, secondary offerings, and other share sales will add roughly $1.5 trillion of stock to the US equity market over the next two years, even after accounting for buybacks.
  2. Such issuance would mark the strongest period of net US equity issuance since at least the late 1990s, following two decades in which S&P 500 buybacks alone erased nearly $12 trillion worth of shares.
  3. The cost of equity has fallen relative to debt since the Federal Reserve raised rates to a two-decade high in 2023, and a $30 trillion share rally has sustained that advantage even as the Fed began cutting rates.

Related Articles:

  • Post-IPO Liquidity Stress Syndrome — Net equity issuance by US domestic non-financial companies has turned positive for the first time since 2021, and a pipeline of initial public offerings and secondary offerings points toward further increases. Net shareholder payouts remain positive on strong earnings and dividends, but the author argues that falling earnings or rising issuance would worsen the outlook for US equities.
  • US Stock Market To Stop Shrinking For First Time In 23 Years — Goldman Sachs estimates that net supply of US equity — new shares issued less shares removed by buybacks and take-privates — will be roughly flat in 2026 after being negative since 2003, with a larger influx expected in 2027 as lock-up periods on this year's initial public offerings expire.
  • Capital Allocation — Internal financing has been the main source of capital for U.S. corporations, with steady new debt issuance except in recession years, consistent with a stable capital structure. Companies have retired equity on average, and net equity issuance remains negative even after accounting for stock-based compensation.

AI Summary. Across a 28-country sample, accumulated trade balances are only weakly correlated with changes in countries' net foreign asset positions, with trade deficits accounting for about one-third of the deterioration in the US position and financial effects entirely explaining Canada's 73 percentage-point improvement relative to its economy.

Stefan Avdjiev, Kristin Forbes, Tsvetelina Nenova and Marjorie Santos Massachusetts Institute of Technology
Date Posted:
Is Database:
Database

Avdjiev, et al decompose the changes in countries’ net international investment positions into the effect of accumulated trade flows, net investment income, and asset revaluation. For the US, only ⅓ of the NIIP deterioration was due to trade flows.

Do trade balances explain changes in countries' net foreign asset positions?

Core argument: Accumulated trade deficits over 2010–24 contributed only about one-third of the deterioration in the U.S. net international investment position, in a sample of 28 countries plus the euro area.

Figure 8 shows accumulated trade flows over 2010-24 (as a share of 2024 GDP) for each of the 28 countries in our sample (plus the euro area as one entity) in blue bars, with the corresponding changes in NIIPs (from Figure 5) in black squares. For a few countries (such as Germany and France), changes in NIIPs largely correspond to accumulated trade surpluses or deficits. However, in most cases, trade flows are only weakly correlated with these changes in imbalances. In the United States, trade deficits contributed to only about one-third of the deterioration in its NIIP, and in Canada, the striking improvement of 73pp in its NIIP (relative to 2024 GDP) entirely reflected financial effects, as accumulated trade balances alone would have generated an NIIP deficit. In other countries, accumulated trade flows would have generated much larger changes in imbalances than actually occurred if they were not offset by financial channels working in the opposite direction. For example, in China, accumulated trade surpluses would have improved the NIIP by 24% of 2024 GDP, rather than the reported 10% (with the difference primarily explained by financial losses and unexplained errors and omissions).

Takeaways by Macro Roundup® AI

  1. Accumulated trade deficits over 2010–24 contributed only about one-third of the deterioration in the U.S. net international investment position, in a sample of 28 countries plus the euro area.
  2. Canada's 73pp improvement in its NIIP relative to 2024 GDP entirely reflected financial effects, since accumulated trade balances alone would have generated an NIIP deficit.
  3. China's accumulated trade surpluses over 2010–24 would have improved its NIIP by 24% of 2024 GDP rather than the reported 10%, with the difference primarily explained by financial losses and unexplained errors and omissions.

Related Articles:

  • Honey, Who Shrunk the U.S. Income Surplus? — Foreign holdings of U.S. financial assets are valued at $69tn versus $41tn of U.S. holdings abroad, a gap that has been mounting for decades. With the net liability position in interest-bearing assets at -$15tn, a 1 percentage point rise in interest rates would subtract $150bn from the U.S. net investment income balance.
  • Understanding Global Imbalances — The US, China, Germany, and Japan account for roughly two-thirds of global current account imbalances, with the US deficit at 4% of GDP financed by capital inflows into dollar assets. The average duration of a surplus or deficit spell roughly doubled since the 1980s, building large net foreign asset and liability positions.
  • Debt, Deficits & Global Imbalances — Taxes or restrictions on short-term capital inflows were argued to be a more direct and potentially superior tool than tariffs for shrinking the trade deficit, since tariffs risk retaliation and supply-chain distortions. Proponents said such measures could reduce reliance on household and government borrowing to sustain employment, if paired with efforts to strengthen domestic demand and tradables capacity.

AI Summary. Asia's $1.5tn trade surplus is larger as a share of world output than at any point since 1945 and is the only large surplus in the global economy. Exchange rates have a meaningful impact on trade balances: China's surplus reached about 10% of its output amid renminbi depreciation, then fell below 2% after a 40% real appreciation.

Brad Setser and Shahin Vallée Foreign Affairs
Date Posted:

Noting the strong connection between Asian countries’ currency depreciation and their trade surpluses since the early 2000s, Setser and Vallée argue “exchange rates still matter,” and urge G-7 “currency diplomacy” to fight chronic undervaluation.

Do exchange rates drive the size of trade surpluses?

Core argument: Asia's trade surplus of $1.5 trillion is larger as a share of world GDP than at any point since 1945, and it is the only large surplus in the global economy.

At $1.5 trillion, Asia’s trade surplus is far larger as a share of world GDP than at any point since 1945. It is also the only large surplus in the global economy. Currency moves were central to the development of the trade imbalances that preceded the 2008 global financial crisis. From 2002 to 2005, for instance, the renminbi was tightly linked to a depreciating dollar even as Beijing’s entry to the World Trade Organization turbocharged Chinese exports. The resulting depreciation is a key reason why China’s surplus reached ten percent or so of its GDP—the first China shock. Conversely, the 40 percent real appreciation in the renminbi from 2005 to 2014 was a big reason why China’s surplus fell back to under two percent of its GDP. Exchange rates still matter, and they have a meaningful impact on the balance of trade.

Takeaways by Macro Roundup® AI

  1. Asia's trade surplus of $1.5 trillion is larger as a share of world GDP than at any point since 1945, and it is the only large surplus in the global economy.
  2. The renminbi's tight link to a depreciating dollar from 2002 to 2005, alongside China's WTO entry turbocharging exports, is a key reason China's surplus reached roughly 10 percent of its GDP in the first China shock.
  3. The renminbi's 40 percent real appreciation from 2005 to 2014 was a big reason China's surplus fell back to under 2 percent of its GDP, indicating that exchange rates have a meaningful impact on the trade balance.

Related Articles:

  • Understanding Global Imbalances — The US, China, Germany, and Japan account for roughly two-thirds of global current account imbalances, with the US deficit at 4% of GDP financed by capital inflows into dollar assets. The average duration of a surplus or deficit spell roughly doubled since the 1980s, building large net foreign asset and liability positions.
  • Why Global Imbalances Matter — Countries running persistent trade surpluses, such as China, Germany and Japan, have built their economies around an excess supply of tradeable goods and services, leaving them dependent on deficit countries to absorb it. That dependence gives deficit countries bargaining power and exposes surplus countries to pressure to expand domestic demand.
  • The U.S. Trade Deficit: Myths and Realities — Taxing capital inflows would weaken the dollar but could raise domestic interest rates and depress investment, harming long-term growth, while an interest rate cut absent recession would be inflationary and erode any real depreciation needed to durably improve the trade balance. Fiscal restraint could weaken the dollar and reduce the trade deficit.

AI Summary. The US government ordered an AI startup to disable access to its most advanced AI models for all foreign nationals after discovering that the guardrails on a newly released model could be bypassed. The company called the response "disproportionate" and warned such an approach could halt all new model deployments by frontier AI developers.

Maggie Eastland, Joshua Green, and Nancy Cook Bloomberg
Date Posted:

The USG ordered Anthropic to disable all foreign nationals from accessing its Mythos AI models following a report that a team was able to “jailbreak” Fable 5, a commercial version of Mythos. In response, Anthropic pulled Fable 5 from all commercial use.

Does blocking foreign access to advanced AI models risk halting new deployments?

Core argument: The US government ordered Anthropic to disable access to its most advanced AI platforms for all foreign nationals, an unprecedented step.

Washington has taken the unprecedented step of ordering the AI startup to disable access to its most advanced AI platforms for all foreign nationals. The US government issued the order after discovering it was possible to “jailbreak,” or bypass the guardrails, of the Fable 5 AI model Anthropic released just days prior. The swiftness of the move triggered widespread concern over what the startup called a “disproportionate” response — and a warning that such an approach could “halt all new model deployments for all frontier model providers.”

Takeaways by Macro Roundup® AI

  1. The US government ordered Anthropic to disable access to its most advanced AI platforms for all foreign nationals, an unprecedented step.
  2. Washington issued the order after discovering that the guardrails of Anthropic's Fable 5 model, released days earlier, could be bypassed through "jailbreaking.".
  3. Anthropic called the response "disproportionate" and warned that such an approach could "halt all new model deployments for all frontier model providers.".

Related Articles:

  • Misanthropic: On Mythos, Bad Human Behaviors and Systems Vulnerabilities — Anthropic reports that its Mythos model has detected thousands of high-severity cyber vulnerabilities, including ones found by chaining multiple obscure software weaknesses, and that these hacking skills are emergent rather than a designed capability. Mythos completed a corporate network attack simulation estimated to take an expert over 10 hours, which no other frontier model had solved.
  • US National Security Agency Using Anthropic’s Mythos For Cyber Attacks — Artificial intelligence company Anthropic is helping the US National Security Agency deploy its Mythos model for offensive cyber operations, with roughly half a dozen engineers embedded inside the agency to guide use of the technology and customize models, according to people familiar with the arrangement.
  • China’s Own Mythos Is Coming — America Must Prepare — The authors argue the US can extend its estimated seven-month lead in artificial intelligence by closing gaps in export controls: halting all AI chip exports to China including less advanced models, curbing smuggling networks, blocking cloud and remote access to controlled chips and US models, and restricting chipmaking equipment.

AI Summary. China's nuclear generation capacity rose 76% (24 GW) between 2016 and 2024, and it has 36 reactors under construction, accounting for more than 49% of all nuclear construction worldwide. Its nuclear fleet is concentrated near eastern population centers along the Pacific coastline and consists mostly of pressurized water reactors.

Slade Johnson and Jonathan Russo Energy Information Administration
Date Posted:

Btw 2016 and 2024, China’s nuclear generation capacity increased 76%. China currently has 36 reactors under construction, 49% of global nuclear construction.

How rapidly is China expanding its nuclear power capacity?

Core argument: China's nuclear generation capacity rose 76%, or 24 GW, from 2016 to 2024, according to International Energy Statistics data.

From 2016 to 2024, China’s nuclear generation capacity increased 76% (24 GW), based on our International Energy Statistics (IES) data. According to the International Atomic Energy Agency’s Power Reactor Information System (PRIS), China added an additional 1.1 GW of nuclear power capacity in 2025 and 2.2 GW in 2026 (through May). China is continuing to build out its nuclear generating capacity and has 36 reactors under construction, accounting for more than 49% of total world nuclear construction, according to PRIS. China’s nuclear fleet is concentrated near population centers in the eastern part of the country, along the Pacific Ocean coastline from the Liaoning province in the north to the Hainan province in the south. According to IAEA’s PRIS, China’s existing nuclear fleet mostly consists of pressurized water reactors.

Takeaways by Macro Roundup® AI

  1. China's nuclear generation capacity rose 76%, or 24 GW, from 2016 to 2024, according to International Energy Statistics data.
  2. China has 36 reactors under construction, more than 49% of total world nuclear construction, according to the IAEA's Power Reactor Information System.
  3. China added 1.1 GW of nuclear capacity in 2025 and 2.2 GW in 2026 through May, according to the IAEA's Power Reactor Information System, with its fleet concentrated near eastern population centers from Liaoning to Hainan and consisting mostly of pressurized water reactors.

Related Articles:

  • Fighting Words: The Energy Transition in 2026 — Estimates based on primary energy overstate how much electricity is needed to replace oil in passenger cars, since electric motors convert 80%-90% of electricity into motion versus about 18% for gasoline cars. Measuring useful final energy net of waste heat avoids this "primary energy fallacy": the US consumed 45 exajoules of useful final energy versus 95 exajoules of primary energy.
  • Power Politics: Energy Self-Sufficiency in a Modern Mercantilist World — China leads in adding power to the grid, with more than 30,000 miles of ultra-high-voltage transmission lines versus zero in the US and lead times for large transformers of 48 weeks against a US average of 143 weeks. This positions China to accommodate new demand sources such as artificial intelligence.
  • How China Raced Ahead Of The U.S. On Nuclear Power — Data published in Nature show U.S. nuclear construction costs rose steeply after the 1960s while China's fell by half during the 2000s and have since stabilized. China's nuclear companies repeatedly build only a handful of reactor designs, which analysts say streamlines licensing, simplifies supply chains, and is essential for scaling efficiently.