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


Friday, October 31, 2025

AI Summary. Wages at age 50 rise more steeply with cognitive skills measured at age 16 than with cognitive skills measured at age 50; for women, moving from average to two standard deviations above average raises expected hourly wages from £11 to £22 using age-16 skills, versus £12 to £15 using age-50 skills.

Uta Bolt The Economic Journal
Date Posted:
Is Database:
Database
Is Important:
Important

Data from a British birth cohort followed since 1958 show that cognitive skills early in life predict wages at age 50 better than cognitive skills at age 50. Early cognitive skills likely allow for higher educational attainment or on-the-job learning.

Do teenage cognitive skills predict midlife wages better than adult skills?

Core argument: For women, moving from mean cognitive skills at age 16 to two standard deviations above the mean raises expected age-50 wages from £11 to £22, while the same move in age-50 cognitive skills raises wages only from £12 to £15.

Figure 2 shows the relationship between expected wages at age 50 and cognitive skills at ages 16 and 50, respectively. To construct the figure, I first regress wages at age 50 on a cubic polynomial of cognitive skills at 16. Next, I regress wages at age 50 on a cubic polynomial of cognitive skills at 50. I then use the estimated coefficients from the two regressions to predict the expected wage at age 50 at different values of cognitive skills. For both men and women, we can see a much steeper gradient of age-50 wages in cognitive skills at age 16 than in cognitive skills at age 50. Moving from mean cognitive skills at age 16 to two standard deviations above the mean increases the expected wage for women from £11 to £22. In contrast, moving from mean cognitive skills at age 50 to two standard deviations above the mean only increases wages from £12 to £15. For men, the patterns are similar, albeit somewhat less stark. Having high cognitive skills at the end of childhood is more important than having high cognitive skills in the middle of adulthood, as high early cognitive skills allow for higher educational attainment or learning on the job that is associated with higher wages.

Takeaways by Macro Roundup® AI

  1. For women, moving from mean cognitive skills at age 16 to two standard deviations above the mean raises expected age-50 wages from £11 to £22, while the same move in age-50 cognitive skills raises wages only from £12 to £15.
  2. Regressions of age-50 wages on cubic polynomials of cognitive skills show a steeper wage gradient in skills measured at age 16 than in skills measured at age 50 for both men and women, with men's patterns somewhat less stark.
  3. The author argues that high cognitive skills at the end of childhood matter more than in middle adulthood because early skills allow higher educational attainment or learning on the job associated with higher wages.

Related Articles:

  • Standardized Test Scores and Academic Performance at Ivy-Plus Colleges — An analysis of admissions and transcript records at Ivy-Plus colleges finds standardized test scores predict first-year college grades with a normalized slope 4x greater than high school grade point average, controlling for race, gender, and socioeconomic status. The authors find no evidence test scores underpredict college performance for students from less advantaged backgrounds.
  • Do Adults Have the Skills They Need to Thrive in a Changing World? — An international assessment of adult skills describes numeracy proficiency levels ranging from counting objects and adding small whole numbers at the lowest level, to extracting information from tables and graphs and performing basic arithmetic with money and percentages at Level 1, up to analysing complex mathematical information and critically evaluating whether data support a claim at Level 5.
  • Transformative and Subsistence Entrepreneurs: Origins and Impacts on Economic Growth — Only 17% of firms are "transformative" entrepreneurs, defined as those hiring a research-and-development worker at some point; higher IQ and educational attainment strongly predict transformative entrepreneurship, though the likelihood of becoming an entrepreneur overall declines with IQ. The authors argue education subsidies address financial constraints more effectively than uniform startup or research subsidies.

AI Summary. Off-balance-sheet structures such as special purpose vehicles and asset-tied joint ventures are becoming the standard financing route for AI data center deals. Morgan Stanley estimates tech firms will need as much as $800bn from private credit by 2028, with roughly $1.5tn of external financing required across the AI ecosystem.

Carmen Arroyo Bloomberg
Date Posted:

Morgan Stanley estimates $2.9T will be spent on the datacenter buildout globally by 2028; $1.5T will require debt financing, $800B of which will be raised from private credit in deals tied to specific assets, including SPVs.

Are off-balance-sheet vehicles becoming the standard way to finance AI data centers?

Core argument: Morgan Stanley estimates tech firms will need as much as $800 billion from private credit by 2028, and its global head of debt capital markets, Anish Shah, said around $1.5 trillion of external financing is needed across the AI ecosystem.

Off-balance-sheet debt, through a special purpose vehicle or joint venture tied to assets, is becoming the go-to for AI data center deals, with Morgan Stanley estimating that tech firms will need as much as $800 billion from private credit by 2028. “We are in the very early stages of capital being raised,” said Anish Shah, Morgan Stanley’s global head of debt capital markets. Across the AI ecosystem, “around $1.5 trillion of external financing is needed, and issuers are going to tap a broad range of financing sources.” That may require some creativity in who holds the debt. Elon Musk’s xAI for instance, is pursuing a $20 billion SPV deal for its data center where its commitment will just be to pay rent on the Nvidia chips.

Takeaways by Macro Roundup® AI

  1. Morgan Stanley estimates tech firms will need as much as $800 billion from private credit by 2028, and its global head of debt capital markets, Anish Shah, said around $1.5 trillion of external financing is needed across the AI ecosystem.
  2. Off-balance-sheet debt structured through special purpose vehicles or joint ventures tied to assets is becoming the go-to financing route for AI data center deals at firms including Meta and xAI.
  3. Elon Musk's xAI is pursuing a $20 billion special purpose vehicle deal for its data center under which its commitment will be limited to paying rent on the Nvidia chips.

Related Articles:

  • Minsky Moments and AI CapEx — Capital-spending forecasts for the four largest cloud providers rose nearly 50% over the year, and AI data centers are increasingly funded through leveraged and off-balance-sheet structures, including $3.2bn in senior secured notes and a roughly $30bn asset-backed financing. The author argues this marks a speculative-finance phase where borrowers must roll over principal.
  • The Blob: Capital, China, Chips, Chicago and Chilliwack — Data center construction spending has eclipsed office construction, and the specialized power rates most data centers pay fall short of covering the cost of a new natural gas plant, leaving other customers to bear part of the bill. In the PJM grid region, data center demand accounted for 70% of the increase in electricity costs over the past year.
  • Why We Are Not In A Bubble….Yet — The 5 largest US technology companies are worth about 16% of the global public equity market, and the top 10 US stocks nearly 25%, almost $25tn. The author argues valuations are high but not bubble-level, with the "Magnificent 7" median forward price-to-earnings ratio at 27x, roughly half that of the late-1990s leaders.

AI Summary. Substrate, a Bay Area startup, has emerged from stealth claiming its newly invented X-ray lithography tool can produce leading-edge logic chip wafers at 50% below the cost of existing options. Evidence for the claim remains scarce, though third-party reports support the tool's legitimacy.

Dylan Patel, Jeff Koch, Gerald Wong, and Andrew Wagner Semianalysis
Date Posted:

Substrate, a US startup, has invented a new X-ray lithography tool; if its performance claims are borne out, “leading edge wafers [could be] produced at 50% less than existing options,” disrupting both ASML’s and TSMC’s effective monopolies.

Can a new X-ray lithography tool halve advanced chip wafer costs?

Core argument: Substrate, a Bay Area startup recently out of stealth, claims its X-ray lithography tool will produce leading-edge wafers at 50% less cost than existing options, as part of a mission to power next-generation foundries.

Substrate is a recently out-of-stealth Bay Area startup inventing “technology to power next-generation foundries” with the mission to substantially reduce the cost of advanced logic wafers. The first major step towards this mission is a new X-ray lithography (XRL) tool that the company has invented. Their performance claims are stunning. Leading edge wafers will be produced at 50% less than existing options. [This is an] extraordinary claim and thus demands extraordinary evidence. Evidence so far is scarce, so we repeat these claims with some healthy skepticism. But we should also note, external contacts and 3rd party reports are all telling us the same story: the litho tool is legit.

Takeaways by Macro Roundup® AI

  1. Substrate, a Bay Area startup recently out of stealth, claims its X-ray lithography tool will produce leading-edge wafers at 50% less cost than existing options, as part of a mission to power next-generation foundries.
  2. Evidence for Substrate's X-ray lithography performance claims remains scarce, warranting skepticism of an extraordinary claim that demands extraordinary evidence.
  3. External contacts and third-party reports indicate that Substrate's lithography tool is legitimate.

Related Articles:

  • China Seeks To Triple Output Of AI Chips In Race With The US — China is preparing three new fabrication plants dedicated to producing Huawei's artificial intelligence processors, and people familiar with the effort say their combined capacity, once fully ramped up, could exceed current output from comparable lines at SMIC, China's leading chipmaker. Huawei denies having plans to launch its own fabrication plants.
  • TSMC To Speed Up Construction Of US Chip Plants By ‘Several Quarters’ — TSMC says it is accelerating construction of its second and third Arizona plants by several quarters to meet strong U.S. customer demand for smartphone and AI computing chips. The company has committed an additional $100bn to build five more facilities in the state and expects around 30% of its 2-nanometer and more advanced capacity to be located in Arizona.
  • Crises at Boeing and Intel Are a National Emergency — Intel and Boeing, once benchmarks for high-quality, demanding manufacturing, declined because of their own mistakes rather than cheap foreign competition. Because leading U.S. tech firms' software and devices depend on advanced semiconductors fabricated under contract, especially by Taiwan Semiconductor Manufacturing Co., Chinese subjugation of Taiwan could leave the U.S. tech sector at Beijing's mercy.

AI Summary. The US and China agreed to suspend recently announced export controls for a year, with the US cutting 10 percentage points from fentanyl-related tariffs to an average 45% levy and China resuming soybean purchases. China has repeatedly used retaliatory tariffs and rare earth export controls to force Washington back to negotiations.

Joe Leahy and Demetri Sevastopulo Financial Times
Date Posted:

The US and China have agreed to de-escalate the trade war, with China delaying implementation of new export controls on rare earths for a year in exchange for the US postponing expansion of additional American export controls on chips.

Has China established itself as an equal negotiating power with Washington?

Core argument: The US cut 10 percentage points off fentanyl-related tariffs on Chinese goods, reducing the average levy to 45%, while Beijing agreed to resume soybean purchases and both sides suspended recently announced export controls and new shipping levies for one year.

[China and the US] agreed to suspend recently announced export controls for one year as well as new levies on shipping. The US, meanwhile, cut 10 percentage points off tariffs on Chinese goods related to fentanyl, reducing the average levy to 45%, while Beijing agreed to resume soyabeans purchases. Since Trump announced his “liberation day” tariffs in April, Beijing has on at least three occasions blocked Washington from carrying out punitive measures and forced it back to the negotiating table. The first clash came when Trump increased reciprocal tariffs on China to 145%. Beijing matched these and eventually forced Washington to suspend the levies. Then the two sides clashed over China’s export controls on rare earths, the production and refining of which it dominates. The rules threatened to shut down US industry and led to another round of talks. This month, after Washington extended export controls on semiconductors to thousands of subsidiaries of Chinese companies, Beijing announced sweeping new controls on rare earths that again had the US pressing for a truce.

Takeaways by Macro Roundup® AI

  1. The US cut 10 percentage points off fentanyl-related tariffs on Chinese goods, reducing the average levy to 45%, while Beijing agreed to resume soybean purchases and both sides suspended recently announced export controls and new shipping levies for one year.
  2. Since Trump announced his "liberation day" tariffs in April, Beijing has on at least three occasions blocked Washington from carrying out punitive measures and forced it back to the negotiating table, beginning when China matched US reciprocal tariffs of 145%.
  3. China dominates rare earth production and refining, and its export controls threatened to shut down US industry, prompting talks; after Washington extended semiconductor export controls to thousands of subsidiaries of Chinese companies, Beijing announced sweeping new rare earth controls.

Related Articles:

  • Why China Keeps Winning The Trade War — Governments seeking supply-chain leverage face limited visibility into their own supply chains, and China's rare earth licensing requirements for multinationals may be partly intended to give Beijing that visibility and allow more precise targeting of future restrictions. Effective measures must impose asymmetric costs, and keeping foreign manufacturers dependent on drip-fed supply is central to Beijing's strategy.
  • Deflationary Pressure Builds In China — China's producer price inflation for durable consumer goods — the category most exposed to US tariffs — dropped sharply alongside the imposition of substantial US tariffs, consistent with exporters marking down prices to sell goods at home. Because limited warehouse capacity forces exporters to keep goods moving, discounting to generate new demand cuts into their profitability.
  • Industrial Policy and Economic Security — Industrial policy justified on national security grounds requires weighing how much economic inefficiency is worth tolerating, since costs incurred today make a country poorer and therefore less able to project power later. The calculation is difficult because security gains are probabilistic and long-term economic costs are speculative.

AI Summary. Trade diversion signals deep distress in China's export sector, and the author argues a sustained 10% tariff hike would be more credible than larger ones and could counter China's rare earths leverage. Easing up on Chinese exporters instead signals weakness and discourages Europe from coordinating with the US.

Robin Brooks Robin Brooks Substack
Date Posted:
Is Database:
Database

Given that deflation is becoming entrenched in China, Robin Brooks argues that the US has the upper hand in the trade war. He cautions that easing up on China’s exporters while China’s hold on rare earths is undiminished, undermines the US position going forward.

Are smaller sustained tariff hikes more credible than larger ones against China?

Core argument: A 10% tariff hike on China would be more credible than a larger one because it can be sustained indefinitely, and paired with threats of further increases could go a long way to countering China's rare earths leverage, the author argues.

Trade diversion is a sign of deep distress in the export sector, which the US should exploit. A smaller tariff hike of 10% has the advantage that it can be sustained indefinitely, making it more credible. Given all the damage we’re already seeing to exporters, such a hike could go a long way to countering the rare earths threat, especially if coupled with threats of further tariff hikes. This steady - but measured - escalation is how to stare down China. President Trump yesterday did the opposite. He eased up on China’s exporters, even as China’s hold on rare earths is undiminished. If China needed any validation that it’s found a weak spot, this was it. Yesterday’s deal will also reinforce European fears that joining forces with the US - say by imposing secondary tariffs - is a bad idea, because there’s no guarantee Trump won’t suddenly cut a deal and leave Europe hanging. All in all, it’s hard to see how the US put itself in a better position with [recent] concessions.

Takeaways by Macro Roundup® AI

  1. A 10% tariff hike on China would be more credible than a larger one because it can be sustained indefinitely, and paired with threats of further increases could go a long way to countering China's rare earths leverage, the author argues.
  2. Trade diversion signals deep distress in China's export sector that the US should exploit, yet Trump eased up on Chinese exporters while China's hold on rare earths remained undiminished.
  3. The deal with China will reinforce European fears that joining the US in measures such as secondary tariffs is a bad idea, since Trump could suddenly cut a deal and leave Europe hanging, the author argues.

Related Articles:

  • Deflationary Pressure Builds In China — China's producer price inflation for durable consumer goods — the category most exposed to US tariffs — dropped sharply alongside the imposition of substantial US tariffs, consistent with exporters marking down prices to sell goods at home. Because limited warehouse capacity forces exporters to keep goods moving, discounting to generate new demand cuts into their profitability.
  • China’s Youth Unemployment Soars To 2-Year High As Job Crunch Deepens — China's urban unemployment rate for 16- to 24-year-olds excluding students rose to 18.9% from 17.8%, the highest since the National Bureau of Statistics revised its calculation method for the series. A record 12.2m university students graduated into a job market strained by deflationary pressures and external uncertainties, with many struggling to find roles matching their education and skills.
  • China’s “Anti-Involution” Campaign: Global Implications — China's economy has slowed as domestic investment stagnates and consumer spending stays weak, with the shortfall partly offset by rising net exports. The author argues these internal problems are being passed to the rest of the world, since export values are rising about 6% a year while import values are down 2.2%.

Monday, November 3, 2025

AI Summary. Across much of the G10, 30-year government bond yields have risen even as central banks cut rates, which the author reads as a sign markets are on average becoming more averse to holding government debt. Switzerland is the only country whose 30-year yield fell, while Germany and the UK stand out as disproportionate risers.

Robin Brooks Robin Brooks Substack
Date Posted:

Only one country – Switzerland – has seen 30-year yields fall since January 2024 and the start of the cutting cycle. Robin Brooks suggests this implies “there’s a slow-motion run on government debt” underway.

Are markets growing more averse to holding long-term government debt?

Core argument: Across much of the G10, 30-year government bond yields have risen since January 2024 despite cumulative central bank rate cuts, a pattern the author reads as a potential sign that markets are becoming more averse on average to holding government debt.

I compare the rise in 30-year government bond yields across much of the G10 with central bank rate cuts. Central bank cuts should pull down yields across the curve, including at the very long end, but that hasn’t been happening during this easing cycle. [Editor’s note: In fact, since the 2008 financial crisis, there have been episodes where the Fed cut rates to support growth, yet long-term rates rose due to market fears of future inflation, such as the inflationary surge following 2021]. That’s potentially a sign that - on average - markets are becoming more averse to holding government debt. The chart zooms in on the cross-country perspective. The horizontal axis shows cumulative cuts (-) made since January 2024. The vertical axis shows the rise (+) in the 30-year government bond yield since then. Three points are worth making. First, more rate cuts - on average - tend to hold down longer-term yields. This positive relationship is to be expected and normal. Second, the only place that’s seen a fall in its 30-year yield is Switzerland, a reflection of its low debt and safe haven status. Third, Germany and the UK are outliers, in the sense that their 30-year yields look to have risen disproportionately. In contrast, and perhaps to the frustration of many, the US looks ok. The US continues to benefit from the fact that “there is no alternative.”

Takeaways by Macro Roundup® AI

  1. Across much of the G10, 30-year government bond yields have risen since January 2024 despite cumulative central bank rate cuts, a pattern the author reads as a potential sign that markets are becoming more averse on average to holding government debt.
  2. Switzerland is the only country in the comparison whose 30-year yield has fallen, which the author attributes to its low debt and safe haven status.
  3. Germany and the UK stand out as outliers whose 30-year yields appear to have risen disproportionately, while the US looks fine, continuing to benefit from the absence of an alternative.

Related Articles:

  • Across The Rich World, Fiscal Crises Loom — An analysis estimates most rich countries could keep debt stable as a share of their economies with only small primary surpluses—Britain's required surplus is 0.3% of GDP—even if all debt were refinanced immediately at current rates, but the tightening needed to stabilize debt ratios exceeds 2% of GDP in Britain and America and 3% in France.
  • Data Update 4 for 2025: Interest Rates, Inflation and Central Banks! — Interest rates are ultimately driven by macroeconomic fundamentals, with an "intrinsic risk-free rate" defined as the sum of the inflation rate and real GDP growth tracking the observed 10-year Treasury bond yield closely. On this view, central banks' influence comes from preferential access to data on those fundamentals, their capacity to alter them, and their credibility.
  • Global Debt Report 2025 — About 60% of fixed-rate government debt in advanced economies maturing by 2027 — roughly $9tn — was issued before the recent interest-rate tightening, with weighted average yields below 2% versus projected 10-year rates near 3.6%, implying refinancing at nearly double the original cost. France, Spain, the UK, and the US face heightened vulnerability, with maturing debt exceeding 15% of GDP.

AI Summary. Investors have continued buying US government debt despite concerns over tariffs, tax cuts, budget deficits, and pressure on the Federal Reserve, with benchmark 10-year yields down about half a percentage point to roughly 4%. The $30tn Treasuries market has returned about 6%, and US 10-year yields have fallen the most among G-7 nations.

Michael MacKenzie and Ye Xie Bloomberg
Date Posted:

Investors keep “piling” into Treasuries. The 10-year yield has fallen ~50bps this year, to ~4%, by far the largest decline among the G-7 countries.

Are investors still buying US government debt despite deficit and tariff worries?

Core argument: Benchmark US 10-year Treasury yields have fallen about half a percentage point this year to roughly 4%, the largest decline among G-7 nations, despite Wall Street warnings to "Sell America.".

From sweeping tariffs and tax cuts, to yawning budget deficits and persistent attacks on the Federal Reserve — a steady drumbeat of worries had many on Wall Street sounding the same ominous alarm: "Sell America." In spite of it all, though, investors keep piling into US government debt. Benchmark 10-year yields have fallen about a half-percentage point this year, to roughly 4%. In fact, the $30 trillion Treasuries market is up about 6% this year, on its way to the best year since 2020. US 10-year yields have declined the most by far among G-7 nations this year. 

Takeaways by Macro Roundup® AI

  1. Benchmark US 10-year Treasury yields have fallen about half a percentage point this year to roughly 4%, the largest decline among G-7 nations, despite Wall Street warnings to "Sell America.".
  2. The $30 trillion Treasuries market has returned about 6% this year, putting it on track for its best year since 2020.
  3. Investors have kept buying US government debt through sweeping tariffs, tax cuts, wide budget deficits and persistent attacks on the Federal Reserve.

Related Articles:

  • Bank of America Sees US Bill Supply Boost Curbing 10-Year Yield — Bank of America expects the US Treasury to lean more heavily on short-term bill issuance, lifting bills to about 25% of outstanding debt from about 20%. Its strategists estimate the resulting reduction in longer-dated note and bond supply is equivalent to roughly $1tn of 10-year notes, worth at least 31 basis points of lower yields.
  • JPMorgan Is Worried About Who’s Going to Buy All the Bonds — JPMorgan analysts estimate that the Federal Reserve's Treasury holdings have fallen $180bn as it pursues quantitative tightening, while commercial banks' holdings are down $60bn and foreign official holdings down $50bn. The analysts describe simultaneous negative demand from all three buyer types as rare and say they remain concerned about the lack of structural demand for Treasuries.
  • Our Thoughts on Large US Deficits and Their Impact on Bond Yields — Total borrowing in the US is about average despite high government deficits, as high public borrowing has contributed to the need for higher interest rates that have depressed private sector borrowing. Risk premiums on long-dated debt look about average, with no market signs that economic or debt limits are currently being hit.

AI Summary. The authors construct a long-short "AI factor" portfolio and find that firms investing in artificial intelligence shift their risk profiles, with total volatility roughly unchanged but a reallocation from firm-specific risk to market-wide risk. This increased comovement with all sectors of the economy mostly reflects greater upside exposure rather than downside.

Tania Babina, Anastassia Fedyk, Alex He and James Hodson Columbia University
Date Posted:

Btw 2014 and 2018, firms with higher than median investment in AI experienced a large and statistically significant increase in systematic risk as measured by market beta, driven by upside risks “consistent with AI investments creating growth options.”

How does investing in AI change firms' risk profiles?

Core argument: The authors' long-short "AI factor," long firms ranking above the median among those with non-zero AI investments and short all other firms, shows no statistically significant market alpha, though returns to AI-investing firms were consistently positive later in the sample.

We construct an “AI factor": a long-short portfolio with the long side comprised of firms that rank above the median among firms with non-zero AI investments and the short side comprising 20 all other firms. Figure 1 shows the time series of the estimates of the AI factor’s market alpha (α (AI) ) [a measure of “excess return”] in Panel A and beta (β (AI)) [a measure of systematic market risk] in Panel B. Panel A does not show statistically significant alpha, but the returns to AI-investing firms have been consistently positive during the later part of the sample. Panel B shows that the beta of the AI factor has been relatively low in the first few years of the sample, but starts to increase in the middle and becomes especially high at the end (in 2017 and 2018), consistent with the increasing potential to realize AI-related growth options over time. Our evidence points to a significant shift in firms’ risk profiles as they invest in AI. The total volatility of AI-investing firms does not change significantly, but there is a reallocation from idiosyncratic risks to systematic risks (higher market beta). AI-investing firms experience increases in comovement with all sectors of the economy—including their own sector, the tech sector, and other industry sectors. This increased market covariance mostly reflects increased upside, as AI-investing firms experience much larger increases in upside market beta than downside market beta.

Takeaways by Macro Roundup® AI

  1. The authors' long-short "AI factor," long firms ranking above the median among those with non-zero AI investments and short all other firms, shows no statistically significant market alpha, though returns to AI-investing firms were consistently positive later in the sample.
  2. The AI factor's market beta was relatively low in the sample's first years, rose in the middle, and became especially high in 2017–2018, which the authors read as consistent with the increasing potential to realize AI-related growth options.
  3. Total volatility of AI-investing firms does not change significantly, but risk shifts from idiosyncratic to systematic, with greater comovement with their own sector, the tech sector, and other industries, and larger increases in upside than downside market beta.

Related Articles:

  • The Golden Opportunity for American AI — Artificial intelligence is a general-purpose technology comparable to electricity, with potential to lift innovation and productivity. The authors argue the United States needs a national strategy spanning research and development, workforce skilling, AI infrastructure investment, and international partnerships to expand American AI exports, counter China's influence, and shape global standards.
  • Exploring How AI Will Impact the Economy — Artificial intelligence is projected to pass through to productivity faster than past general-purpose technologies, though peak effects remain years away. The argument rests on lower capital spending per unit of labor saved, capital spending concentrated among major software players that can distribute AI tools through platforms businesses already use, and low unemployment with high wage growth raising incentives to automate.
  • The Outlook for Long-Term Economic Growth — Growth theory holds that long-run growth in living standards is determined by growth in the worldwide number of people searching for ideas, and growth accounting for the US suggests temporary boosts from rising educational attainment and idea investment could fade. Potential offsets include China and India, better talent allocation, and artificial intelligence.

AI Summary. Big Tech raised at least $75bn from investment-grade bonds in a single month to fund AI data centers, with banks syndicating another $38bn in loans, versus an average of $37bn issued per year before the pandemic. Because AI-related capital spending growth exceeded growth in total non-residential fixed investment, spending on everything else declined.

John Authers and Richard Abbey Bloomberg
Date Posted:

Morgan Stanley shows AI-related cap-ex growth exceeds total non-residential fixed investment growth for the year. “In other words, capex in everything not related to AI has actually declined.”

Is AI borrowing and capital spending crowding out other investment?

Core argument: Big Tech raised at least $75 billion from investment-grade bonds in September alone to fund AI data centers, against an average of $37 billion issued each year before the pandemic, according to Bank of America data.

How much money is going into building out AI? In September alone, Big Tech raised at least $75 billion from investment-grade bonds to fund AI data centers. Banks are syndicating a further $38 billion in loans tied to Oracle and Vantage facilities. Before the pandemic, these companies were issuing an average of $37 billion each year, Bank of America data shows. Archimedes’ principle holds that the upward buoyant force on an object submerged in a fluid is equal to the weight of the fluid it displaces. How much is the AI splurge going to displace? The extra spending on AI could galvanize the economy, but will it crowd out everything else? AI-related capex is immense, as we’ve discussed — but as Morgan Stanley shows, its growth actually exceeds total non-residential fixed investment growth for the year. In other words, capex in everything not related to AI has actually declined. [Note that H1:2025 AI spend outpaced total non-residential fixed investment]. GDP accounting can reduce this by excluding the AI spending that goes on imports [Exhibit 4 in the gallery]. With chips mostly manufactured outside the US, that makes a big difference.

Takeaways by Macro Roundup® AI

  1. Big Tech raised at least $75 billion from investment-grade bonds in September alone to fund AI data centers, against an average of $37 billion issued each year before the pandemic, according to Bank of America data.
  2. Banks are syndicating a further $38 billion in loans tied to Oracle and Vantage facilities.
  3. Morgan Stanley shows AI-related capex growth exceeded total non-residential fixed investment growth for the year, implying capex unrelated to AI declined, while GDP accounting excludes AI spending on imports, including chips manufactured outside the US.

Related Articles:

  • How America’s AI Boom Is Squeezing The Rest Of The Economy — About a sixth of America's 2% real GDP growth over the past year came from investment in computer and communications equipment and data centres, and one estimate puts the artificial intelligence boom's total contribution to growth at 40%. Average electricity bills rose 7%, attributed in part to data-centre demand, while housebuilding and non-AI business investment slumped and real consumption flatlined.
  • AI Data Center Boom Threatens Trump’s Manufacturing Revival — Bloomberg Economics projects the largest participants could pour as much as $4tn into artificial intelligence infrastructure by the end of the decade, with capital spending by hyperscalers such as Alphabet, Amazon, Meta, and Microsoft rising from almost $400bn to as much as $600bn. It expects AI's contribution to GDP growth could reach as much as 1.5 percentage points.
  • The Blob: Capital, China, Chips, Chicago and Chilliwack — Data center construction spending has eclipsed office construction, and the specialized power rates most data centers pay fall short of covering the cost of a new natural gas plant, leaving other customers to bear part of the bill. In the PJM grid region, data center demand accounted for 70% of the increase in electricity costs over the past year.

AI Summary. Research by finance professors Eugene Fama and Kenneth French identified an "investment factor," the tendency of stocks in companies that invest the most to lag the market, a pattern that held from the 1960s through the early 2000s except during the Nifty Fifty and dot-com booms. That factor has behaved normally recently, with higher-spending large companies tending to underperform slightly.

James Mackintosh Wall Street Journal
Date Posted:

James Mackintosh notes that the historical tendency for the total return of high investment firms to lag those with lower investment noted by Fama and French continues to hold, “with higher-spending big companies tending to underperform a little.”

Are investors still penalizing the heaviest spenders during the AI boom?

Core argument: The "investment factor," named by finance professors Eugene Fama and Kenneth French, has behaved normally this year, with higher-spending big companies tending to underperform a little.

The history of investor views of capex is neatly measured by what finance professors Eugene Fama and Kenneth French dubbed the “investment factor,” the old tendency of stocks in companies that invest the most to lag behind. From the 1960s through to the early 2000s, companies that increased capex the most were pretty consistent laggards. The two big exceptions were the Nifty Fifty boom of the late 1960s and early 1970s, briefly interrupted by a recession, and the dot-com bubble of the late 1990s. In both periods of excess, investors thought companies could invest successfully, and rushed to buy the heaviest spenders before being disappointed. [Contemporarily] there is little sign of a shift in the investment factor, which has behaved perfectly normally this year, with higher-spending big companies tending to underperform a little.

Takeaways by Macro Roundup® AI

  1. The "investment factor," named by finance professors Eugene Fama and Kenneth French, has behaved normally this year, with higher-spending big companies tending to underperform a little.
  2. From the 1960s through the early 2000s, companies that increased capital expenditure the most were pretty consistent stock market laggards, as measured by the Fama-French investment factor.
  3. The two big exceptions were the Nifty Fifty boom of the late 1960s and early 1970s and the dot-com bubble of the late 1990s, when investors rushed to buy the heaviest spenders before being disappointed.

Related Articles:

  • Surviving the AI Capex Boom — Capital spending by the largest US technology companies has risen from 4% to 15% of revenue since 2012, approaching the capital intensity of utilities, with Meta, Microsoft, and Alphabet each set to spend 21–35% of revenue. The authors argue this shift toward asset-heavy models is associated with inferior returns.
  • Minsky Moments and AI CapEx — Capital-spending forecasts for the four largest cloud providers rose nearly 50% over the year, and AI data centers are increasingly funded through leveraged and off-balance-sheet structures, including $3.2bn in senior secured notes and a roughly $30bn asset-backed financing. The author argues this marks a speculative-finance phase where borrowers must roll over principal.
  • Why We Are Not In A Bubble….Yet — The 5 largest US technology companies are worth about 16% of the global public equity market, and the top 10 US stocks nearly 25%, almost $25tn. The author argues valuations are high but not bubble-level, with the "Magnificent 7" median forward price-to-earnings ratio at 27x, roughly half that of the late-1990s leaders.

AI Summary. Six of the world's fifteen most populous countries in 2100 — including three of the top five — are projected to be places with under $7,000 in per capita GDP today, led by Pakistan, Nigeria, and the Democratic Republic of the Congo, with Ethiopia and Tanzania close behind. Over 2bn people are projected to live in those five countries.

Noah Smith Noahpinion
Date Posted:

In 75 years, 20% of the world’s population is projected to live in Pakistan, Nigeria, the DRC, Ethiopia and Tanzania. Noah Smith notes, “More and more of humanity is going to live in a few big countries that can’t manage themselves.”

Will the world's most populous countries be among its poorest?

Core argument: By 2100, six of the world's fifteen most populous countries — and three of the top five — are projected to be places with under $7,000 in per capita GDP (PPP) as of 2025.

By 2100, six out of the fifteen most populous countries in the world — and three of the top five — will be places that as of 2025 have under $7,000 in per capita GDP (PPP). The biggest ones will be Pakistan, Nigeria, and the Democratic Republic of the Congo, with Ethiopia and Tanzania not far behind. Altogether, over 2 billion people — a fifth or more of humanity’s peak population — is projected to live in those five countries. And that percentage will only grow after that, at least unless some sort of radical unexpected change happens to demographic patterns. Fertility is falling everywhere, but in the Big 5 poor countries, it’s not falling much faster than in the world as a whole.

Takeaways by Macro Roundup® AI

  1. By 2100, six of the world's fifteen most populous countries — and three of the top five — are projected to be places with under $7,000 in per capita GDP (PPP) as of 2025.
  2. Pakistan, Nigeria, the Democratic Republic of the Congo, Ethiopia, and Tanzania are projected to hold over 2 billion people, a fifth or more of humanity's peak population, with that share rising thereafter absent radical demographic change.
  3. Fertility is falling everywhere, but in those five poor countries it is not falling much faster than in the world as a whole.

Related Articles:

  • 21st-Century Africa Governance and Growth — Sub-Saharan Africa's GDP per capita has fallen relative to the world average over three decades, from 30% to 22%, reflecting the region's inability to sustain growth. Had it grown at the global pace since 1990, income per capita would be over 40% higher; at emerging East Asia's pace, nearly 3x higher.
  • Changing Sub-Saharan Africa’s Growth Trajectory — Sub-Saharan Africa's working-age population is growing nearly 3% a year, versus close to zero in much of the world, and is projected to exceed China's within about a decade and reach a quarter of the global total. With modest productivity growth near 2%, the region is likely among the world's fastest-growing.
  • Exploring the Implications of Sub-Saharan Africa’s Demographic Explosion — Sub-Saharan Africa is projected to hold a quarter of the world's working-age population within a few decades and to surpass China's working-age population in roughly a decade. Even with modest productivity growth near 2%, the region's growth is estimated at 4.5–5.5%, raising its share of global output from about 3% while remaining small.