Edward Conard

Top Ten New York Times Bestselling Author

  • “A full-throated defense of economic dynamism.” - The Wall Street Journal
  • “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
  • “…a must-read for serious students of economic policy.” - Glenn Hubbard, Dean, Columbia Business School, and former Chairman of the Council of Economic Advisers
  • “…reminds us that inequality sends a signal of what society lacks most, in America’s case, entrepreneurship and risk taking.” - Lawrence Lindsey, CEO, The Lindsey Group, former Director of the National Economic Council
  • “…a fresh argument for the productive value of inequality.” - David Autor, Professor of Economics, Massachusetts Institute of Technology
  • “…serious thinking for serious thinkers. …a thought-provoking blueprint for growing middle- and working-class incomes.” - Mitt Romney, former Governor of Massachusetts
  • “…serious thinking for serious thinkers. …a thought-provoking blueprint for growing middle- and working-class incomes.” - Mitt Romney, former Governor of Massachusetts
  • “There are an amazing number of good ideas and interesting points made in Unintended Consequences. The thinking underlying it, and the obvious depth of understanding of the author, are very impressive.” - Steven Levitt, coauthor of Freakonomics; 2004 John Bates Clark Medal
  • “…a very valuable contribution.” - Larry Summers, former Secretary of the Treasury and director of the National Economic Council, president emeritus, Harvard University
  • “…a comprehensive explanation of the modern economy.” - Julian Robertson, Founder, Tiger Management
  • “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
  • “Unintended Consequences provides a provocative interpretation of the causes of the global financial crisis and the policies needed to return to rapid growth. Whether you agree or not, this analysis is well worth reading.” - Nouriel Roubini, New York University; Chairman, Roubini Global Economics
Upside of Inequality Oxford Unintended Consequences

Macro Roundup


Tuesday, February 17, 2026

AI Summary. Survey data show around 70% of businesses use some artificial intelligence technology, with larger, more productive, and higher-paying firms more likely to adopt it. Firms report little employment effect and a modest productivity boost so far, but predict adoption will raise productivity by about 1.4% and cut employment by about 0.7% over three years.

Ivan Yotzov, Jose Maria Barrero, Nicholas Bloom, Philip Bunn, et al. National Bureau of Economic Research
Date Posted:

A survey of 6,000 Western senior executives finds 70% of firms are using AI, 80% of which report no impact as of yet. Respondents forecast a productivity boost of ~1.4%, reflecting a 0.8% output increase and a 0.7% employment cut over the next 3 years.

How do firms report AI affecting their productivity and employment?

Core argument: Around 70% of surveyed businesses currently use AI technologies, with larger, more productive, and higher-paying firms more likely to be adopters, and adoption expected to increase further.

We outline four key findings. First, AI technologies are currently used by around 70% of businesses, and adoption is expected to increase. Larger, more productive, and higher-paying firms are more likely to be using some AI technology. Second, we show that AI technologies are being actively used by senior survey respondents (typically CEOs, CFOs, and senior finance managers). On average, respondents use AI for around 1.5 hours in a typical working week, and this frequency has increased sharply since the start of 2025. Third, firms estimate that AI has had little impact on their employment so far, and only a modest boost to productivity over the past three years. Finally, firms expect AI to have larger impacts over the medium-term. Over the next three years, firms predict that the adoption of AI will boost productivity by around 1.4%, on average, while reducing employment by around 0.7%. This is in sharp contrast to expectations of employees, who expected higher job creation as a result of AI, along with smaller productivity gains over the next three years.

Takeaways by Macro Roundup® AI

  1. Around 70% of surveyed businesses currently use AI technologies, with larger, more productive, and higher-paying firms more likely to be adopters, and adoption expected to increase further.
  2. Firms predict AI adoption will boost productivity by around 1.4% and reduce employment by around 0.7% over the next three years, while reporting little employment impact and only modest productivity gains so far.
  3. Senior survey respondents, typically CEOs, CFOs, and senior finance managers, use AI for around 1.5 hours in a typical working week, a frequency that has increased sharply since the start of 2025.

Related Articles:

  • Productivity During and Since the Pandemic — Total factor productivity growth since early 2021 has run only slightly above its 2010–19 average, suggesting the 2020 productivity surge reflected cyclical dynamics layered on a continuation of the slow pre-pandemic trend. Productivity has since grown somewhat faster than the authors' analysis predicted, but remains well below the 1995–2004 pace.
  • Everyone’s Looking For A Bubble. No One Sees The Stampede — A proprietary model estimates monthly artificial intelligence revenue grew roughly 18-fold in two years, from $772m to $13.8bn, with AI accounting for an estimated 23% of Google Cloud, 10% of Azure and 5% of Amazon Web Services. The model's ratio of investment to revenue fell from 6.1x to 4.7x, still historically high.
  • Concentrated Growth: The Role Of The IT Sector — Since 1988, total factor productivity in the information technology sector has risen about 178%, versus about 12% across all non-IT sectors combined, while aggregate productivity rose about 40%. The IT sector accounts for roughly 45% of cumulative aggregate productivity growth, once sectoral linkages to downstream industries are weighted in.

AI Summary. Artificial intelligence has yet to appear in employment, productivity, or inflation data, or in profit margins and earnings expectations outside the largest technology firms. With fierce competition pushing large language model prices toward zero for users, value creation depends on sector-level implementation, and the author argues AI will likely enhance labor in some sectors rather than replace it across all.

Torsten Sløk Apollo
Date Posted:

Noting “no signs of AI in profit margins or earnings expectations” for the S&P 493 as of the start of 2026, Sløk observes that, in terms of AI, value creation will be found in how the technology diffuses across the economy.

Why has AI not yet appeared in macroeconomic data?

Core argument: Three years after ChatGPT's release, AI is absent from employment, productivity and inflation data, and shows no signs in profit margins or earnings expectations for the S&P 493, echoing Robert Solow's 1987 productivity paradox.

The Nobel Prize–winning economist Robert Solow said in 1987, “You can see the computer age everywhere but in the productivity statistics.” This observation is the so‑called Solow productivity paradox. The same can be said today: AI is everywhere except in the incoming macroeconomic data. Today, you don’t see AI in the employment data, productivity data or inflation data. Similarly, for the S&P 493, there are no signs of AI in profit margins or earnings expectations. Maybe there is a J‑curve effect for AI, where it takes time for AI to show up in the macro data. Maybe not. Whether there is a J-curve effect depends on the value creation from AI. There is fierce competition between the builders of large language models (LLMs), which is driving the price of LLMs toward zero for end-users. From a macro perspective, the value creation is not the product, but how generative AI is used and implemented in different sectors in the economy. The academic literature is still inconclusive about the potential macro effects of AI [see table in gallery]. The [disparate and inconclusive results in the table suggest that] after three years with ChatGPT and still no signs of AI in the incoming data, AI will likely be labor-enhancing in some sectors rather than labor-replacing in all sectors.

Takeaways by Macro Roundup® AI

  1. Three years after ChatGPT's release, AI is absent from employment, productivity and inflation data, and shows no signs in profit margins or earnings expectations for the S&P 493, echoing Robert Solow's 1987 productivity paradox.
  2. Competition among builders of large language models is driving the price of LLMs toward zero for end-users, so the macro value creation lies in how generative AI is implemented across sectors rather than in the product itself.
  3. The academic literature on AI's potential macroeconomic effects remains inconclusive, and the disparate results suggest AI will likely be labor-enhancing in some sectors rather than labor-replacing in all sectors.

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.
  • Everyone’s Looking For A Bubble. No One Sees The Stampede — A proprietary model estimates monthly artificial intelligence revenue grew roughly 18-fold in two years, from $772m to $13.8bn, with AI accounting for an estimated 23% of Google Cloud, 10% of Azure and 5% of Amazon Web Services. The model's ratio of investment to revenue fell from 6.1x to 4.7x, still historically high.
  • Looking for the Ladder — The article argues that the hiring downturn in occupations most exposed to generative AI reflects interest-rate sensitivity rather than AI itself, since roughly 38% of workers in the most AI-exposed quintile work in information, finance and insurance, and professional and technical services, versus under 2% in the least-exposed quintile.

AI Summary. US payroll growth was revised down by about 403,000 jobs while real GDP stayed strong, including 3.7% growth in the fourth quarter, a combination of high output with lower labor input that the author identifies as productivity growth. The author estimates US productivity rose roughly 2.7%, nearly double the 1.4% annual average of the past decade.

Erik Brynjolfsson Financial Times
Date Posted:

Brynjolfsson argues that downward revisions to job creation while real GDP growth remains steady suggest that AI may be impacting US productivity growth. He estimates labor productivity grew 2.7% last year, compared to a 1.4% mean over the past decade.

Is strong output with weaker job growth a sign of rising productivity?

Core argument: US total payroll growth was revised down by roughly 403,000 jobs even as real GDP stayed strong, including 3.7% growth in the fourth quarter.

Data released this week offers a striking corrective to the narrative that AI has yet to have an impact on the US economy as a whole. While initial reports suggested a year of steady labour expansion in the US, the new figures reveal that total payroll growth was revised downward by approximately 403,000 jobs. Crucially, this downward revision occurred while real GDP remained robust, including a 3.7% growth rate in the fourth quarter. This decoupling — maintaining high output with significantly lower labour input — is the hallmark of productivity growth. My own updated analysis suggests a US productivity increase of roughly 2.7% for 2025, a near doubling from the sluggish 1.4% annual average that characterised the past decade.

Takeaways by Macro Roundup® AI

  1. US total payroll growth was revised down by roughly 403,000 jobs even as real GDP stayed strong, including 3.7% growth in the fourth quarter.
  2. The author estimates US productivity rose roughly 2.7% in 2025, nearly double the 1.4% annual average of the past decade.
  3. The author argues that maintaining high output with lower labour input is the hallmark of productivity growth and a corrective to claims that AI has not affected the US economy.

Related Articles:

  • SF FedViews: Solid Growth, but Small Cracks Have Emerged — Since the second half of 2023, four-quarter US productivity growth has averaged nearly 2.5%, well above the pre-pandemic trend of about 1% to 1.25%, supporting real GDP growth. It remains unclear whether this reflects a higher productivity regime fostered by rising AI use or a temporary post-pandemic reorganization surge.
  • Claude Code is the Inflection Point — Claude Code authors 4% of public GitHub commits, and the authors project it will reach 20%+ of daily commits by the end of 2026 and drive revenue growth that lets Anthropic outgrow OpenAI. They argue falling costs and rising quality of AI tokens have inverted the coder-tool relationship.
  • Looking for the Ladder — The article argues that the hiring downturn in occupations most exposed to generative AI reflects interest-rate sensitivity rather than AI itself, since roughly 38% of workers in the most AI-exposed quintile work in information, finance and insurance, and professional and technical services, versus under 2% in the least-exposed quintile.

AI Summary. State-level measures of exposure to industrial robots show substantial geographic variation across the US, with the highest concentrations in the Southwest, Midwest, and Northwest and the lowest in the Mountain West, parts of the South, and northern regions. The authors find state robot exposure is positively correlated with state minimum wage increases.

Erik Brynjolfsson, J. Frank Li, Javier Miranda, Robert Seamans, et al. Stanford University
Date Posted:
Is Database:
Database
Is Important:
Important

Across US states, a 10pp higher growth rate of the minimum wage over 1992–2021 was associated with an ~8% higher-than-expected installation of industrial robots.

Do higher state minimum wages increase adoption of industrial robots?

Core argument: A 0.0056 increase in robot exposure amounts to roughly 0.0056 more excess robots per 100 baseline workers, equal to 8% of the 0.07 mean state excess robot accumulation over the period.

We construct state-level long-difference measures of exposure to industrial robots by combining robot installation information from [several agencies]. Figure 1a displays our robot exposure measure, [revealing] significant geographic variation in robot exposure across U.S. states, with the color gradient ranging from light blue (minimal change) to dark navy (maximum change). States in the Southwest, Midwest, and Northwest show the highest concentrations of robot exposure, reflecting strong manufacturing sectors and states in the Mountain West, parts of the South, and Northern regions, show minimal robot exposure consistent with less dependence on manufacturing. Figure 1b displays state-level minimum wage changes covering the period 2003–2015. The map reveals substantial geographic variation in minimum wage policy exposure across U.S. states, with the gradient ranging from light blue (minimal change) to dark navy (maximum change). Finally, Figure 1c displays a strong positive correlation in the relationship between the state-level robot exposure and the minimum wage measures. [The .0056 increase in robot exposure is 8% of the 0.07 mean — meaning roughly 0.0056 more excess robots per 100 baseline workers, which is 8% of the average state's excess robot accumulation over the period].

Takeaways by Macro Roundup® AI

  1. A 0.0056 increase in robot exposure amounts to roughly 0.0056 more excess robots per 100 baseline workers, equal to 8% of the 0.07 mean state excess robot accumulation over the period.
  2. State-level robot exposure and state-level minimum wage changes over 2003–2015 show a strong positive correlation across U.S. states in the authors' long-difference measures.
  3. Robot exposure is concentrated in Southwest, Midwest, and Northwest states with strong manufacturing sectors, while the Mountain West, parts of the South, and Northern regions show minimal exposure.

Related Articles:

  • The Mysterious Slowdown in U.S. Manufacturing Productivity — US manufacturing labor productivity growth slowed broadly after the Great Recession, with the four largest firms in each industry averaging –0.5% growth from 2010 to 2022 versus 3.7% from 1987 to 2007, and smaller follower firms showing an identical 4.2 percentage-point decline. The authors find the slowdown spans all industries and firm sizes.
  • What Would It Take to Bring Back US Manufacturing? Part 2: Making American Manufacturing More Productive — The authors argue the United States is a decade away from the scale of automated "lights-out" manufacturing needed to meet its own end-to-end demand at comparable cost. With 98% of US manufacturers being small or medium enterprises, low deployment of mature productivity technologies reflects limited commercial adoption and investment rather than technical readiness.
  • Did California’s Fast Food Minimum Wage Reduce Employment? — The authors estimate that California's $20 fast food minimum wage, up from $16, raised fast food sector wages roughly 8% relative to the rest of the country while employment fell about 2.7% relative to fast food employment elsewhere in the US. Their median estimate implies 18,000 fewer California fast food jobs than otherwise.

AI Summary. Corporate profit margins have trended upward across roughly three decades of company data. The author attributes the increase largely to a shift in market composition, as technology companies with superior unit economics and economies of scale supplant manufacturing and service businesses.

Aswath Damodaran Musings on Markets
Date Posted:
Is Database:
Database

The S&P 500 had a net profit margin of 13.55% in 2025. American software firms had operating margins of 25–30%, in contrast to industrial firms’ single-digit margins.

Are rising corporate profit margins driven by more technology companies?

Core argument: Corporate profit margins tracked across roughly three decades of Damodaran's datasets show a clear upward trend over the period.

I have tracked profit margins for companies for a long time (about three decades) in my datasets, and there is clear evidence that they have trended upwards during the period. The most salient rationale for margin increases, is that the composition of the market has changed, as technology companies supplant old-economy companies, bringing superior unit economics and economies of scale to play. Put simple, a market that gets the largest portion of its value from tech companies will deliver much higher margins than one that gets much of its value from manufacturing and service businesses.

Takeaways by Macro Roundup® AI

  1. Corporate profit margins tracked across roughly three decades of Damodaran's datasets show a clear upward trend over the period.
  2. Damodaran attributes rising margins primarily to a change in market composition, as technology companies supplant old-economy firms and bring superior unit economics and economies of scale.
  3. Damodaran argues that a market deriving the largest portion of its value from technology companies delivers much higher margins than one drawing much of its value from manufacturing and service businesses.

Related Articles:

  • S&P 500 Margin Expansion All Coming from Tech — All of the increase in the S&P 500's operating margin over the past 20 years has come from tech-related sectors — information technology, communication services and consumer discretionary — while operating margins for non-tech companies have stayed near 9%.
  • Bayes and Base Rates — Investment in artificial intelligence infrastructure has reached a scale comparable to some of the largest investment booms in U.S. history, with firms issuing historically high growth forecasts. Among U.S. public companies with $2–5bn in starting sales from 1950 to 2024, none sustained the roughly 108% five-year annual growth rate implied by OpenAI's revenue projection.
  • Chart of the Day: AI Capex, From Bonanza to Burden — Stock price reactions to large artificial intelligence capital spending announcements have shifted from rewarding to penalizing companies, with the valuation multiple markets assign to such spending collapsing and likely turning negative. Technology companies once valued for low debt and low capital intensity are now among the leaders in both.

AI Summary. The share of Americans aged 25-54 with a job has held at multi-decade highs, reaching its highest level since 2001, even as the unemployment rate has risen. This prime-age employment rate has been the best indicator of cyclical conditions since women entered the workforce en masse in the 1980s.

Matt Klein The Overshoot
Date Posted:

Since 2023, the share of Americans aged 25–54 with a job has been at its highest level since 1998–2000. Matt Klein warns against rate cuts, noting many CPI components are rising 2–3 standard deviations faster than pre-pandemic.

Is prime-age employment strength at odds with rising unemployment?

Core argument: The employment rate for Americans aged 25–54 has held at multi-decade highs since the start of 2023 and rose slightly in January to its highest level since April 2001.

The share of Americans aged 25-54 with a job has been stable at multi-decade highs ever since the beginning of 2023, and rose slightly in January to its highest level since April 2001. The so-called “prime age” employment rate has been the single best indicator of cyclical conditions since women entered the workforce en masse in the 1980s. It is therefore extremely odd to have had the unemployment rate rise as much as it has during a period when employment has been so strong. Either way, it is hard to say that things have changed much since mid-2024.

Takeaways by Macro Roundup® AI

  1. The employment rate for Americans aged 25–54 has held at multi-decade highs since the start of 2023 and rose slightly in January to its highest level since April 2001.
  2. The prime-age employment rate has, in the author's assessment, been the single best indicator of cyclical conditions since women entered the workforce en masse in the 1980s.
  3. The rise in the U.S. unemployment rate alongside such strong prime-age employment is odd, and on either reading conditions have changed little since mid-2024.

Related Articles:

  • Trump Wants to Run the Economy Hot. There’s a Good Chance He’ll Succeed — A US tax law expanded deductions for state and local taxes, overtime, tips and seniors retroactively, but withholding tables adjusted only later, so workers get both higher take-home pay and refunds. One analyst estimates this injects nearly $200bn, lifting first-half annualized growth by up to half a percentage point.
  • Is the Market Underpricing the Risk of Fed Hikes? — Futures prices imply essentially no chance the Federal Reserve raises rates over two years and over 90% odds of a cut within 12 months. The author argues this understates hike risk, citing prime-age employment near its multi-decade high and stable pay growth, and warns lower rates could unlock inflationary spending.
  • Data Update 5 for 2026: Risk and Hurdle Rates — A histogram of costs of capital across all publicly traded firms shows what typical, high, and low hurdle rates look like, with the cost of capital best understood as an opportunity cost—the return available elsewhere on equally risky investments. US dollar risk-free rates have remained near 4% for a fourth year.

Wednesday, February 18, 2026

AI Summary. In a randomized experiment on a realistic business analysis task, participants without an AI assistant showed an education gap, with high-education participants scoring 0.548 standard deviations higher overall. Access to the AI assistant raised scores by 1.242 standard deviations for low-education participants versus 0.834 for high-education participants.

Guillermo Cruces, Diego Fernández Meijide, Sebastian Galiani, Ramiro Gálvez, et al. National Bureau of Economic Research
Date Posted:
Is Database:
Database

An experiment in which participants were evaluated on their performance of hypothetical business tasks found that while high-education participants performed better with or without AI, access to AI assistance closed two-thirds of the performance gap.

Does AI assistance narrow performance gaps between education groups?

Core argument: In a randomized experiment, access to an AI assistant raised overall task scores by 1.242 standard deviations for low-education participants versus 0.834 SD for high-education participants, a gap of 0.408 SD significant at the 1% level.

The experimental task presented a realistic hypothetical business scenario - an email from the participant’s boss describing a problem [requiring] analysis of information sources, including text, a figure, and a table. Participants were asked to answer three questions to diagnose the problem and propose a solution. We assess performance in the task on a 0–10 scale along two dimensions, content and writing. Figure 2 presents the results on the impact of having access to the AI assistant on this outcome for each education group. The task without access to the AI assistant is substantially more challenging for low-education participants: their high-education counterparts perform 0.548 standard deviations (SD) higher in the overall score. AI substantially increases task performance for participants of both education groups - overall score rises by 1.242 SD for those in the low education group, whereas the impact of the assistant for high-education subjects is 0.834SD. The treatment effect of access to the AI assistant is 0.408 SD higher for the low-education group [than for] the high-education group, with both results statistically significant at the 1% level.

Takeaways by Macro Roundup® AI

  1. In a randomized experiment, access to an AI assistant raised overall task scores by 1.242 standard deviations for low-education participants versus 0.834 SD for high-education participants, a gap of 0.408 SD significant at the 1% level.
  2. Without the AI assistant, high-education participants scored 0.548 standard deviations higher than low-education participants on the overall score, a task the study describes as substantially more challenging for the low-education group.
  3. The experimental task asked participants to diagnose a problem and propose a solution using text, a figure, and a table from a hypothetical email from their boss, scored 0–10 on content and writing.

Related Articles:

  • Centaurs and Cyborgs on the Jagged Frontier — In a pre-registered experiment at Boston Consulting Group covering 18 realistic consulting tasks, consultants using ChatGPT-4 completed 12.2% more tasks, worked 25.1% faster, and produced 40% higher quality results than those without it. The tool acted as a skill leveler: the lowest-scoring consultants improved 43%, while top performers gained less.
  • Generative AI at Work — A study of 5,179 customer support agents finds that access to a generative AI conversational assistant raised issues resolved per hour by 14% on average, with the largest gains for novice and low-skilled workers and minimal effect on experienced, highly skilled ones. The authors also find AI assistance improved customer sentiment, reduced requests for managerial intervention, and improved employee retention.
  • Concentrated Growth: The Role Of The IT Sector — Since 1988, total factor productivity in the information technology sector has risen about 178%, versus about 12% across all non-IT sectors combined, while aggregate productivity rose about 40%. The IT sector accounts for roughly 45% of cumulative aggregate productivity growth, once sectoral linkages to downstream industries are weighted in.

AI Summary. Forecasters expect US gross domestic product to have expanded 2.7% over the year while employment barely grew, a divergence drawing comparisons to the "jobless recovery" that followed the early-2000s technology stock collapse. Unlike that episode, which began with a recession, this "jobless boom" is occurring without one — a first in the postwar era.

Augusta Saraiva and Enda Curran Bloomberg
Date Posted:
Is Database:
Database

2025’s “Jobless Boom” is the result of a divergence btw strong real GDP growth and slow growth in nonfarm payrolls, a dynamic that has “never happened” this far into an expansion.

Can the economy keep growing while employment barely increases?

Core argument: Forecasters expect Friday's GDP report to show the US economy expanded 2.7% in 2025, a solid pace for a developed country, while employment barely grew.

Forecasters expect Friday’s report on gross domestic product to show the economy expanded 2.7% in 2025, a solid pace by any standard for a developed country. But employment barely grew, and the combination is drawing comparisons to the infamous “jobless recovery” of the early 2000s that followed the tech bubble and collapse. There’s one major difference between then and now that makes the current divergence all the more unusual: The 2000s episode kicked off with a recession. This time, the “jobless boom” is happening without one. That marks a first in the postwar era.

Takeaways by Macro Roundup® AI

  1. Forecasters expect Friday's GDP report to show the US economy expanded 2.7% in 2025, a solid pace for a developed country, while employment barely grew.
  2. The divergence between US growth and hiring is drawing comparisons to the "jobless recovery" of the early 2000s that followed the tech bubble and its collapse.
  3. Unlike the early-2000s episode, which began with a recession, the current "jobless boom" is occurring without one, a first in the postwar era.

Related Articles:

  • What Sweeping Revisions and a Blowout Month Tell Us About the U.S. Job Market — Labor Department revisions lowered previously estimated job gains for the past two calendar years, cutting the combined net change in construction, manufacturing, and retail by 115,000 jobs. Healthcare added about 437,000 jobs and social assistance about 321,000 over 12 months, while all other private-sector jobs combined shrank.
  • Jobless Growth — Labor demand growth may be too weak to keep pace with even modest labor supply growth: the labor market is somewhat weaker than before the pandemic, job growth has turned negative outside healthcare, and firms are focused on using AI to cut labor costs. Economists estimate most employment losses in routine manual occupations since the 1970s occurred during recessions.
  • Our Favorite Charts of 2025 — Wage growth is slowing more for lower-paid workers than for higher earners, limiting those households' ability to keep spending, while hiring is concentrated in select industries such as healthcare and among larger firms as small private firms shed workers. Layoffs remain low, but the number of workers unemployed for 27 weeks or more has only rarely been higher.

AI Summary. The 1951 Federal Reserve–Treasury Accord reestablished central bank independence by granting the Federal Reserve the right to buy government bonds only when it judged prudent and to let long-term interest rates float upward. The author argues the Treasury should control the debt's maturity structure, with the Federal Reserve confining purchases to short-term bonds or lending against collateral.

John Cochrane The Grumpy Economist
Date Posted:

Noting that the 1951 Accord gave the Fed the right not to buy debt, Cochrane proposes a new Accord – which he believes is consonant with Warsh’s views – under which the Fed determines short rates, while the Treasury alone manages the maturity structure.

Should the Treasury, not the Fed, control debt maturity structure?

Core argument: The 1951 Fed-Treasury Accord reestablished Fed independence by giving the Fed the right to buy bonds only when it judged prudent and to let long-term interest rates float upward.

The first Fed-Treasury Accord in 1951 reestablished Fed independence. In the Accord, the Fed won the right to buy bonds only when it thought prudent, and to allow long-term interest rates to float upward. That background is important. A lot of commentary [predicts that] Warsh [will be] Treasury Secretary Bessent’s puppet at the Fed, setting the Fed up to finance deficits without complaint. I think this is exactly the opposite of what is likely to transpire. Warsh has spent over a decade saying that the Fed needs to buy less government bonds, shrink its footprint on markets, let long-term interest rates go where they want to go, and do less overall. The invocation of the Accord, in which the Fed won the right not to buy bonds tells you what’s up. Otherwise, it would be “repeal the accord.” Someone needs to be in charge of the overall maturity structure and its risk-reward consequences for the taxpayer. In my view, this is a job for the Treasury. If the Fed wants to buy lots of Treasury debt in order to provide lots of bank reserves, then the Fed should buy only short-term bonds, or issue reserves via collateralized lending as the ECB does, without buying Treasury debt. If the Fed needs to buy long term debt temporarily for “market function” reasons, then it needs to swap out the interest rate risk with the Treasury, or exchange that promptly for short-term debt.

Takeaways by Macro Roundup® AI

  1. The 1951 Fed-Treasury Accord reestablished Fed independence by giving the Fed the right to buy bonds only when it judged prudent and to let long-term interest rates float upward.
  2. Contrary to commentary predicting Kevin Warsh would act as Treasury Secretary Bessent's puppet and finance deficits without complaint, Warsh has spent over a decade arguing the Fed should buy fewer government bonds, shrink its market footprint, and do less overall.
  3. The Treasury, rather than the Fed, should manage the overall maturity structure and its risk-reward consequences for taxpayers, with the Fed buying only short-term bonds or issuing reserves via collateralized lending as the ECB does.

Related Articles:

  • Warsh Call for Fed-Treasury Accord Stirs Debate in $30 Trillion Bond Market — A nominee to lead the Federal Reserve has backed a new version of a 1951 Treasury-Federal Reserve agreement that would spell out the size of the central bank's balance sheet alongside the Treasury's debt-issuance plans. Analysts argue such a public pact could tie monetary operations to government deficits and resemble a framework for controlling bond yields.
  • Convenience Lost — The authors find that fiscal expansion and the resulting rise in Treasury supply pushed convenience yields to zero on short-term Treasurys and below zero on long-term ones, reducing the revenue the U.S. earns from issuing safe, liquid assets. They argue shortening the debt's maturity structure cannot fully offset this, and that international investors may turn to alternative long-term safe assets.
  • Fiscal Dominance and the Return of Zero-Interest Bank Reserve Requirements — Fiscal dominance requires governments to fund deficits at the margin with non-interest-bearing debt, and the revenue from inflation taxation—both expected and unexpected—is limited. The authors argue a US fiscal dominance episode could end the payment of interest on bank reserves and restore requirements that banks hold large zero-interest reserves against deposits.

AI Summary. Analysts estimate that each 1bp increase in gold's share of US financial portfolios, driven by incremental purchases, raises gold prices by 1.5%, and with gold exchange-traded fund allocations at only about 0.2% of US private sector financial portfolios, further diversification is the main upside risk to a $5,400 gold price forecast.

Daan Struyven, Alexandra Paulus, Lavinia Forcellese, Lina Thomas, et al. Goldman Sachs
Date Posted:

A GS analysis notes upside risk to their current gold price target of $5,400 per troy ounce due to retail buying, given that gold ETF allocations in US private sector financial portfolios are only ~0.2%.

Does further portfolio diversification into gold pose the main upside price risk?

Core argument: Every 1bp increase in gold's share of US financial portfolios, driven by incremental gold purchases, is estimated to raise gold prices by 1.5%.

Private sector diversification is also the key upside risk to our $5,400 December 26 forecast, which only assumes i) that central banks continue to buy at their elevated pace of the past 12 months and ii) a modest boost to ETF holdings from 50bp of Fed cuts. With gold ETF allocations in US private sector financial portfolios at only ~0.2%, there is room for additional diversification into gold. We estimate that every 1bp increase in the gold share of US financial portfolios--driven by incremental gold purchases--raises prices by 1.5%. (This diversification is increasingly expressed via call options, which has also boosted price volatility).

Takeaways by Macro Roundup® AI

  1. Every 1bp increase in gold's share of US financial portfolios, driven by incremental gold purchases, is estimated to raise gold prices by 1.5%.
  2. Gold ETF allocations in US private sector financial portfolios stand at about 0.2%, leaving room for additional diversification into gold, increasingly expressed via call options that have boosted price volatility.

Related Articles:

  • Precious Comment: Raising Our Gold Price Forecast on Sticky Hedges — Analysts forecast a 17% rise in gold prices by end-2026, with emerging-market central bank buying averaging 60 tonnes contributing 14 percentage points and Western exchange-traded fund inflows on 50 basis points of Federal Reserve rate cuts adding 3. Risks are seen as skewed to the upside.
  • The Debasement Trade Roars Back — Currencies of low-public-debt countries such as Switzerland and Sweden are showing increasing correlation with daily moves in gold and silver prices. Sweden's krona has traditionally been highly volatile without safe-haven attributes, a pattern the article attributes to the "debasement trade.
  • Political Pressure on the Fed — Using data on President–Federal Reserve interactions, the author estimates that political pressure to ease monetary policy raises the price level strongly and persistently without boosting real activity, working through inflation expectations more than a typical easing. Such pressure occurred mainly in the 1970s and contributed to inflationary episodes.

AI Summary. The US constraint in copper is processing capacity rather than raw supply: domestic mine output and scrap are exported in large volumes, often to China, to be refined into the copper cathode used in manufacturing. Analysis by Benchmark Mineral Intelligence finds US raw copper sources could cover 146% of annual demand, versus 40% for China.

Camilla Hodgson Financial Times
Date Posted:

The consultancy Benchmark finds that, while the US is constrained in copper processing, it has more than enough copper from domestic and US-owned overseas mines plus scrap to meet demand. China has access to only 40% of its demand from analogous sources.

Is the US copper constraint processing capacity rather than supply?

Core argument: The US can meet 146% of its annual copper demand using raw copper from domestic and overseas mines plus scrap, compared with 40% for China, the world's largest consumer, according to Benchmark Mineral Intelligence.

The US has a sizeable domestic copper mining industry and produces significant quantities of scrap. But a lack of domestic processing means both are exported in substantial volumes, often to China, to be turned into the copper cathode used in manufacturing. Cathode is also imported into the US for use by semi-finished product manufacturers. The lack of domestic processing also means that copper from US-owned mines overseas often does not go to the US, whereas much of the material from Chinese-owned assets overseas does go to China. Even excluding US overseas investments, however, the nation should be self-reliant in copper due to its domestic mine and scrap production, Benchmark found. The [US] can meet 146% of annual demand using raw copper from overseas and domestic mines and from scrap, analysis by Benchmark Mineral Intelligence found. The figure for China — the world’s largest consumer — is 40%.

Takeaways by Macro Roundup® AI

  1. The US can meet 146% of its annual copper demand using raw copper from domestic and overseas mines plus scrap, compared with 40% for China, the world's largest consumer, according to Benchmark Mineral Intelligence.
  2. A lack of domestic processing capacity means US mined copper and scrap are exported in substantial volumes, often to China, to be turned into the copper cathode used in manufacturing, while the US imports cathode for its semi-finished product manufacturers.
  3. Copper from US-owned mines overseas often does not go to the US, whereas much of the material from Chinese-owned overseas assets does go to China, according to Benchmark Mineral Intelligence.

Related Articles:

  • Just-In-Case, The Possibility For Higher Commodity Prices — A decade of underinvestment alongside demand from artificial intelligence and electrification points to large copper deficits in coming years, though global inventories are not especially tight today. US tariff plans on copper drew metal into US warehouses, tightening London inventories while creating a US surplus, and copper's designation as a critical mineral gives policymakers reason to keep it there.
  • Amazon Is Buying America’s First New Copper Output in More Than a Decade — Amazon is buying copper from an Arizona project expected to yield 14,000 metric tons of copper cathode over four years, less than the tens of thousands of tons a single large data center requires for its electrical components. S&P Global estimates artificial intelligence will help lift copper demand 50% by 2040 while mining output dwindles, producing a 25% shortfall.
  • The World’s Copper Squeeze Is Set to Intensify on AI and Defense Spending, S&P Says — S&P Global projects copper demand will rise 50% to 42m tons by 2040 while production peaks near 33m tons in 2030 as ore quality declines at existing mines and new projects face permitting, financing and construction hurdles. The projected 10m-ton gap assumes recycled copper more than doubles to 10m tons.

AI Summary. Germany's population is projected to fall by more than 4mn to about 79mn by 2050, with the number of people aged 20 to 66 down about 12% and pensioners up more than 20%. Germany's pay-as-you-go pension system, which already absorbs about a quarter of the federal budget, is likely to come under increasing strain.

Olaf Storbeck Financial Times
Date Posted:

Ifo Institute for Economic Research revised its forecast of a 1% German population decline over the next 25 years to a 5% decline, due to lower birth rates and a drop in immigration. The former East Germany will be more impacted than the former West Germany.

How will a shrinking population strain Germany's pension system?

Core argument: Germany's population is projected to fall by more than 4mn to about 79mn by 2050, the lowest level since the early 1990s, with growth projected to hover around 0.4% in the long term.

Germany’s growth rate, projected to hover around just 0.4% in the long term, is likely to face further strain as the labour market comes under pressure from demographic decline. The number of people aged 20 to 66 is projected to fall by about 12%, while the number of pensioners is set to rise by more than 20%, Ifo said. Germany’s pay-as-you-go pension system is likely to come under increasing strain. The government already devotes about a quarter of its budget to supporting it, and the Federal Audit Office warned in 2024 that this share would rise further. By 2050, Germany’s population will fall by more than 4mn to about 79mn, the lowest level since the early 1990s. The inflow of migrants by 2030 is now expected to be significantly lower than previously forecast after the government tightened migration policy. Birth rates, too, were overestimated in earlier projections, with the new data pointing to 150,000 fewer births in 2030 than previously expected. According to Ifo, in 2025 Germany attracted only 225,000 net migrants compared with 454,000 in the previous estimate.

Takeaways by Macro Roundup® AI

  1. Germany's population is projected to fall by more than 4mn to about 79mn by 2050, the lowest level since the early 1990s, with growth projected to hover around 0.4% in the long term.
  2. Ifo projects the number of people in Germany aged 20 to 66 will fall about 12% while pensioners rise more than 20%, straining a pay-as-you-go pension system that already absorbs about a quarter of the government budget.
  3. Germany attracted only 225,000 net migrants in 2025 against a previous estimate of 454,000, according to Ifo, and new data point to 150,000 fewer births in 2030 than earlier projections assumed.

Related Articles:

  • Depopulation Globally and in the Asia-Pacific: The Shape of Things to Come — Population aging and fertility below replacement levels are connected to a revolution in family structure in which families grow smaller, fewer people marry, and voluntary childlessness becomes common across countries. As a result, families become less able to bear the demands that may be placed on them.
  • The Fight to Revive Europe’s Shrinking Rural Areas — The estimated population of predominantly rural European Union regions fell by nearly 8mn, or 8.3%, over a decade while urban population rose by more than 10mn, or 6%. Regions covering about 40% of the EU's land area and holding almost one-third of its population are seeing sustained resident declines.
  • Germany Joins EU’s ‘Ultra-Low’ Fertility Club — Germany's fertility rate fell to 1.35 children per woman, below the United Nations' "ultra-low" threshold of 1.4, a level at which falling birth rates become difficult to reverse. Estonia and Austria also dropped below 1.4, joining nine European Union countries including Spain, Greece, and Italy, while France's rate fell to a record-low 1.67 and Finland's to 1.26.