The Plutonomy—AI Edition—and the Coming Crisis

Citigroup strategists recognized that the rise of plutonomy in the United States represented a return to the highly unequal, finance-driven economy of the 1920s—a system that ultimately collapsed into the Great Depression. That crash led to the creation of institutions—including a progressive tax system—that helped turn the United States into a more egalitarian society.

One of the defining characteristics of plutonomies is the erosion of the share of national income going to workers. Citigroup strategists noted that labor’s share of GDP in plutonomies had steadily declined since the early 1980s, a trend that has continued ever since. While this would be a serious problem in an economy driven by broadly shared wage growth, in a plutonomy what matters is spending by the rich. Because the rich derive most of their income from financial assets rather than wages, maintaining aggregate demand in plutonomies requires continuously rising asset prices. Domestic asset bubbles attract investments by foreign plutocrats, boosting asset prices still higher. Serial asset-price bubbles are therefore not an unfortunate byproduct of plutonomy—they have been one of the essential features of the U.S. economy since at least the 1990s.

Citigroup strategists expected that the “technology revolution, and financial revolution, are likely to continue,” a prediction that was prescient given the rise of generative artificial intelligence (AI) and financial “innovations,” including cryptocurrencies and prediction markets. Today, we are living in a new plutonomy—the AI edition. In this economy, the very technology that threatens the jobs and livelihoods of average workers is at the same time fueling a financial bubble that the wealthy rely on for their consumption and accumulation of wealth.
These are not separate phenomena, but two sides of the same coin. While the working class faces stagnating wages and an uncertain job market, the financial sector generates massive amounts of paper wealth, further fueling the consumption and asset accumulation of the top one-tenth of the 1%.

Today’s plutonomy did not arise naturally. Technology is not destiny, nor are financial bubbles inevitable. The return of America’s Gilded Age economy is the product of decades of political and economic choices. In the neoliberal period that began almost a half-century ago, Western governments abandoned their commitment to full employment, embracing counter-cyclical monetary policy as the solution to the business cycle (lowering interest rates in recessions and raising them in recoveries). Meanwhile, fiscal policy was hamstrung by both political parties in the name of so-called fiscal discipline—given all the tax cuts for the plutocracy, government supposedly could no longer afford to spend money to improve living standards.

Since the era of Paul Volcker, who was Fed chair from 1979 to 1987, monetary policy has been focused like a laser against labor—raising interest rates whenever labor markets improved under the assumption that rising wages would cause inflation. On the other hand, the Fed always looked favorably on rising profits and asset prices because those induce investment and productivity increases. Implicitly, the Fed assumes that rising wages do not generate more worker effort, and that rising wealth stimulates innovation by the plutocrats.

On the financial side, the New Deal’s reforms were gradually weakened and dismantled, allowing for the emergence of large banking conglomerates. These banks, which take deposits from regular people, are now allowed to engage in a whole host of risky financial dealings with unregulated “shadow banking.” The rise of what the mid-20th-century economist Hyman P. Minsky called money manager capitalism created an insatiable demand for tradeable financial instruments that Wall Street was happy to provide. Each time a bubble driven by this “creative finance” burst, the Fed reliably stepped in to prevent the plutocratic class from taking big losses on their assets.

This combination of a macroeconomic policy framework that favors asset markets, technological disruption, and globalization led to the weakening of U.S. labor and consequently to a continuous drop in its share of national income. The result is an economy in which workers increasingly bear the costs of technological change while the gains are privatized at the top.

Galbraith and the Great Crash

To understand how the contemporary AI-driven plutonomy functions, we must look back to the Great Depression era. The bootstrapping of “fictitious wealth” by today’s techno-feudalists and financiers mirrors the financial architecture that precipitated the end of America’s original plutonomy. John Kenneth Galbraith’s The Great Crash, 1929 offers the clearest explanation of what led to the financial collapse that precipitated the Great Depression. Galbraith’s analysis could easily be mistaken for commentary on today’s speculative bubble that is minting millionaires by the tens of thousands, billionaires by the thousands, and even the world’s first trillionaire.

As Galbraith argued, the economy grew in the late 1920s with rising labor productivity but stagnant wages. Growth was driven by investment and rising nonwage income that boosted inequality. The top 5% received about a third of all personal income (largely in the form of interest, dividends, and rent). Growth relied excessively on investment and luxury consumption, so as investment slowed, effective demand was insufficient to maintain growth.

Further, corporate and banking structures created financial fragility. The corporate structure favored grifters, swindlers, and fraudsters, leading to a wave of corporate larceny. Holding companies and investment trusts used holdings as cash cows to service debt, cutting actual investment spending to pay dividends to artificially boost share prices, adding to deflationary pressure.

Galbraith explained that the pre-Depression riches were generated by excess demand for stocks. But despite the common misperception that average citizens were caught up in the fever, Galbraith estimates that total participation at the peak involved fewer than a million buyers, and most stocks were hoovered up by investment trusts. A newly formed trust would issue shares to other trusts, using the proceeds to buy shares in still other trusts. During 1928, an estimated 186 investment trusts were formed. By early 1929, Wall Street was launching one new investment company each day, with the securities of these companies valued at 11 times greater than their 1927 level. Through the bootstrapping of circular finance—a self-reinforcing cycle in which trusts bought one another’s shares and pushed up each other’s valuations—the trusts drove each other’s stocks ever higher.

It all came to an end in the fall of 1929 when trusts tried to unwind positions in other trusts. A general liquidation was the inevitable consequence of this financial daisy chain: selling positions to support their own shares meant that, according to Galbraith, “[t]hey bought their own worthless stock... The autumn of 1929 was, perhaps the first occasion when men succeeded on a large scale in swindling themselves.”
Galbraith blamed the crash on what he terms “embezzlement” that was largely revealed only with the crash:

At any given time there exists an inventory of undiscovered embezzlement in—or more precisely not in—the country’s businesses and banks. This inventory—it should perhaps be called the bezzle—amounts at any moment to many millions of dollars. It also varies in size with the business cycle. In good times people are relaxed, trusting, and money is plentiful. But even though money is plentiful, there are always many people who need more. Under these circumstances the rate of embezzlement grows, the rate of discovery falls off, and the bezzle increases rapidly. In depression this is reversed… Just as the boom accelerated the rate of growth, so the crash enormously advanced the rate of discovery.

As the journalist and businessman Walter Bagehot had earlier put it, “[e]very great crisis reveals the excessive speculations of many houses which no one before suspected.” After the fact, the tsunami of bezzles that drove the boom is revealed. Financial assets crashed by 85%; GDP fell by half; and unemployment reached 25%. That financial crash is what made the Great Depression so great.

Galbraith insisted that the collapse was implicit in the speculative frenzy that preceded it. He rejected common conjectures about the causes—such as excessively easy monetary policy (Milton Friedman’s claim)—as “obviously nonsense.” It was instead a “pervasive sense of confidence and optimism and conviction that ordinary people were meant to be rich.”

Minsky and Money Manager Capitalism

The Great Depression created an opening for reforming finance capitalism—an economic system where the financial sector has an oversized share of economic activity, employment, and profits—that had failed. The New Deal put in place comprehensive safeguards against the financialization of the economy. Furthermore, as Minsky argued, in the post-war period we had the “Big Bank” (central bank) and “Big Government” (fiscal policy) that stabilized the economy, allowing for a generation-long period of growth without crises.

However, he predicted in the late 1950s that relative stability with government backstops would encourage financial adventurism, such as the increased use of repurchase agreements, negotiable certificates of deposit, the rise of commercial paper and Eurodollar markets, and eventually securitization. This was rational behavior because backstops reduced the perception of risk. The evolution of financial practice appeared safe because in a stable economy most bets paid off. It was profitable to create new financial products to escape the New Deal’s constraints. When pushed too far, market players treated the government backstops as free insurance against loss. (See Gerald Epstein, “From Boring Banking to Roaring Banking,” D&S, July/August 2015.)

The role of finance in the economy gradually increased as regulations were scrapped or “reinterpreted”—so much so that Minsky argued we had entered a new phase he called “money manager capitalism.” The New Deal temporarily replaced finance capitalism with managerial or welfare-state capitalism. In this new era, the financial sector was significantly downsized through regulations, widespread default-driven “clearing of the slate” during the Great Depression, and reduced household and corporate reliance on debt due to wartime savings, growing wages, and robust profits. However, relative stability, financial engineering, and deregulation allowed finance to assume greater importance. Eventually, finance became the “tail” that wags the “dog” of our economy. By the time of the Global Financial Crisis (2008–2009), the financial sector took 40% of corporate profits while accounting for 20% of value added—both figures indicating that it was far too big.

The financial practices that led to the Global Financial Crisis looked remarkably like those that Galbraith identified in the Great Crash of 1929. Investment banks found ways around regulatory constraints while government actively gutted the last remnants of New Deal regulations, including the elimination of the Glass-Steagall Act in 1999 (which separated dangerous investment banking from commercial banking). Instead of the 1920s investment trusts, modern finance created bank holding companies with off-balance-sheet special-purpose vehicles that supposedly created firewalls between FDIC-insured commercial banks and risky speculative activities. When the crisis hit, these firewalls turned out to be more like paper walls, leading the Fed to backstop every part of the financial system.

The New Plutonomy and the AI Bezzle

While the Great Depression wiped out a large swath of plutocrats, the Fed’s actions rescued the plutocracy—setting the stage for today’s AI-driven edition. Just as the 1920s bubble engineered the investment trusts’ bezzle—what Galbraith called “inventory of undiscovered embezzlement … in the country’s businesses and banks”—we are currently dealing with the AI bezzle. The promise of future productivity gains is being transformed into financial wealth for the few, subsidizing the lavish lifestyles of the rich while the technology threatens to permanently disenfranchise workers. At the same time, AI’s enormous energy demands are contributing to both the affordability crisis and climate catastrophe.
By some measures, inequality in the United States is the highest it has ever been. The top 10th of the 1% hold about 14.5% of the nation’s wealth. With wages stagnant, economic growth relies largely on spending by those who receive profits, interest, and rents, and those who avoid income tax by borrowing against capital gains. That is how the plutonomy benefits from rising asset prices, and our economy has never relied more on the plutonomy to keep it pumping along than it does today.

The problem is that this is an inherently fragile base. Traders have long been using algorithms, which account for 60% of trading today. A growing share of trading is now controlled by AI agents that quickly process data. Because these AI agents have access to the same data, they tend to make the exact same moves, amplifying market volatility.
As in previous bubbles, the AI boom is supported by “innovative” finance. The private credit industry has become an important lender to software and AI companies, funding a significant portion of data center buildouts. Private credit funds have filled the gap left by traditional banks, which became subject to stricter capital regulations after the Global Financial Crisis. Unlike banks, private credit offers risky innovations like payment-in-kind options that allow the borrower to add interest to the principal. This is exactly what Minsky called “Ponzi finance”—an inherently unsustainable arrangement in which borrowers take on ever-growing debt because unpaid interest is added to the principal, causing the balance to grow at a compound rate.

These funds serve a purpose similar to the 1920s investment trusts by distributing risks throughout the financial system. But even though banks seem to have lost business to private credit funds, they remain indirectly involved by granting these funds lines of credit, essentially backstopping their lending. This mirrors how banks were intertwined with the shadow banking system leading up to the Global Financial Crisis, as explained above. While that crisis began in the shadow banking system, it quickly spread throughout the regulated banking system, forcing the Federal Reserve to bail out the entire apparatus. Similarly, while the private credit industry lacks direct access to the Fed, its lines of credit from traditional banks provide de facto access to the Fed’s discount window.

The processes that brought on the Great Crash parallel the circular valuation we are witnessing in the tech sector today, where Big Tech companies invest in AI start-ups using cloud computing credits rather than cash. These start-ups then turn around and use those credits to buy services from these Big Tech firms. The magic of circular finance artificially inflates the value of both companies: start-ups boast big investments while tech firms book artificial revenue.

Trouble Ahead: Déjà Vu All Over Again?

This time around, much of the investment in AI infrastructure looks like the 1920s investment in trusts: investment made for a purpose to be revealed later, or perhaps not at all. There is an assumption that AI will prove useful for something other than parlor tricks and hacking, but there is significant disagreement over what and when.

Meanwhile, the financing of the bubble appears to be thoroughly speculative. Circular finance links the balance sheets of various players in the same way that the investment trusts of the 1920s were linked. These linkages ensure that “liquidation” is unavoidable; debtors will have to sell their assets to service their debts, creating an asset-price deflation process where everyone is a seller and nobody wants to buy. That makes history repeat itself.

An article published by Bloomberg recently worried that a “bruising selloff, or ‘chip-wreck,’ in several technology giants was the latest trigger for concern that the AI frenzy... might be overblown.” As AI providers scale back planned investment, the value of chip producers falls. Just as the value of the investment trusts of the late 1920s consisted of shares of other trusts, a significant amount of the worth of AI-related firms resides in the shares of other firms. But the deflation of the AI bubble might look more like the crash of the 17th century’s tulip bulb mania—which took about a week—than the Great Crash’s many months.

The main physical investment during this AI boom is in data centers. The business model of data centers corresponds directly to Minsky’s “Ponzi” classification: it generates no revenue while being built, but expenses add up. Borrowers must continuously borrow to service the principal and interest. That is why payment-in-kind was invented—to allow borrowers to add interest to the principal rather than paying it out of their income. Unlike the dot-com investments in fiber optic cables that could lay underground for decades and still remain usable, the chips in data centers become obsolete very quickly. They may be obsolete before they even come online, meaning guaranteed default on debts.

In some respects, AI represents the ultimate plutonomy experiment: an attempt to achieve mass production without needing mass human labor or mass consumer demand, relying instead on asset appreciation among the wealthiest households. This is highly unlikely to work. While rich folks do have an insatiable demand for luxury goods, they are relatively small in number and what makes luxury goods appealing to plutocrats is that they are relatively rare. Yet, AI is supposed to create an abundance of everything—which necessarily eliminates the snob appeal. And the masses released from exploitation won’t have the income necessary to purchase the abundance AI is meant to create.

Silicon Valley has proclaimed that a basic income guarantee (providing regular, unconditional cash payments to people regardless of income) is the answer to giving displaced humans the income to buy up the abundance created by AI. We have many objections to this, but Pope Leo XIV has raised the strongest argument. As his recent encyclical put it:
…work is not simply an instrument; it expresses and enhances the dignity of our lives. It is a requirement of the human condition, a normal path toward maturity, development and personal fulfilment. In this regard, financial assistance to the poor may at times be necessary in emergencies, but it cannot become the sole response, since the goal is to enable each person to live with dignity through his or her own work.

A basic income guarantee cannot be an adequate replacement for participating in productive life. Humans are not going to tolerate relegation to mere consumerism funded by welfare. Consuming without contributing to production has always been the role of the idle and miserable rich. That will not be our future.

Sources: Walter Bagehot, The Works and Life of Walter Bagehot, edited by Mrs. Russell Barrington, vol. 6. (Longmans, Green, and Co., 1915); J. K. Galbraith, The Great Crash (Houghton Mifflin Company, 1955); Ajay Kapur, Niall MacLeod, and Narendra Singh, “Plutonomy: Buying Luxury, Explaining Global Imbalances,” Citigroup, Industry Note, October 16, 2005; Pope Leo XIV, “The dignity of work at a time of digital transition,” Magnifica Humanitas, 149, The Vatican, 2026; Yeva Nersisyan, “The repeal of the Glass–Steagall Act and the Federal Reserve’s extraordinary intervention during the global financial crisis,” Journal of Post Keynesian Economics, 2015; Yeva Nersisyan and L. Randall Wray, “The global financial crisis and the shift to shadow banking,” European Journal of Economics and Economic Policies, 2010; David Rovella, “Wall Street ‘Chip-Wreck’ Triggers AI Bubble Fear,” Bloomberg, June 23, 2026 (bloomberg.com); L. Randall Wray, “$29,000,000,000,000: A Detailed Look at the Fed’s Bailout of the Financial System,” Levy Economics Institute, One-Pager No. 23, 2011 (levyinstitute.org); L. Randall Wray, “Artificial Intelligence: Friend, Foe, Fraud,” Levy Economics Institute Working Paper No. 1107, 2026 (levyinstitute.org).