US Empire on the Precipice: Looming AI/‘Everything’ Bubble and Dollar Crash Spells Extreme Inflation Ahead
The latest global capitalist meltdown appears to be underway. Runaway gold and silver prices point to a US dollar currency collapse and extreme inflation ahead. Economic data suggests compellingly that the unfolding crisis may amount to the worst one ever, from which capitalism may never recover — necessitating an unprecedented struggle for world socialism and human liberation.
The half-century of ‘neoliberalism’ is going the same way as the postwar ‘Keynesian social democratic mixed economy’ that it usurped during the 1970s and 80s.
Following a couple of decades of rising working class power — the growing strength of both the socialist bloc in the east and, everywhere else, major gains for workers and social democratic parties, including significant levels of state ownership — capital, the (still dominant) part of the world economy that remained privately owned, sank into a deep profitability crisis. The combination of slumping private investment and high wages sparked inflationary crises that made profitable investment even more unaffordable. US inflation hit 6% in 1968, 12% in 1973, and 15% in 1981; while GDP contracted by 0.6% in 1969–70; 3.2% in 1973–5, and 2.2% in 1980.
To survive, the capitalist class turned the crisis into a weapon, taking advantage of the falling demand for labour and going on a worldwide offensive to restore profitability and stability: beating back union power; decimating wages and curtailing other working class gains; and continually privatising state-owned enterprise and infrastructure and plundering public wealth — brutally inflicting defeats both on social democracy and then the socialist bloc.
The US also sought to address its trade deficit; its expenditure on imports outstripping its income from exports, a problem that did not emerge until the start of the 1970s. In 1985 the US succeeded in coercing the world to co-operate with the terms of the Plaza Accord, arranging the orderly depreciation of the US dollar, particularly against the Yen, in order to make US exports more competitive (since the high use of and demand for the dollar, which most oil is priced in, continually inflates its value, making US goods expensive to buy in terms of currency exchange) — one of many manouvres that blunted and undercut the then rising power of Japanese capital.
Successfully staved off for a while, the profitability crisis is back with a vengeance.
The price of silver has exploded from $12 an ounce in September 2020 and $28 on 31 March 2025 to $100 on 23 January 2026; putting it on course to rise month-on-month nine consecutive times, its longest ‘winning streak’ on record. Gold, now above an eye-watering $4,900, has more than trebled over the past six years and nearly doubled in the past year, racing past its inflation-adjusted 1981 record. Silver looks likely to follow: its inflation-adjusted price in today’s money at the start of 1981 would be $147.
Although demand — outpacing supply — for silver has been partly driven by its use in solar panels and AI chips, the exponential rate of this rally leaves no doubt: investors are buying silver and gold as hedges against anticipated galloping inflation; that the value of the US dollar — the dominant global reserve currency — is expected to fall sharply and lose significant appeal.
The US empire is in serious trouble — but so too is everything it preys on.
In the 1960s and 70s, investors sharply reduced their holdings of 10-Year Treasury bonds (‘government’ [nation-state] debt, paid back after 10 years), driving up the rate of interest — the price of borrowing/profit rate for lenders — in response to rising inflation, which erodes the real value of the fixed returns on loans (the interest the state pays back on what it borrows during those ten years).
In the past three years, however, average inflation has been trending downwards, reflected in and followed by a drop in the 2-Year and 3-Month bond yields over the past 18 months, from about 5% to 3.5%. Yet in that time the 30-Year has risen from just below 4% to nearly 5%. The 10-Year dropped from 4.8% in October 2023 to 4.1% in mid-January 2025, but has been trending upwards from troughs of 3.6% in September 2024 and 2% in February 2022.
With inflation falling, unemployment creeping up, and bank lending tightening — the latter by the most on record in the last two weeks of March 2023 — in September 2024 the Federal Reserve, the US central bank, began to cut the interest it pays on the reserves banks hold with it; encouraging banks to reduce reserves and lend to the private market, pushing down market rates and stimulating economic activity. The 10-Year would typically fall in response — by a historical average of 26 basis points — in the months following a first Fed cut. Instead, it remained more than 100 basis points (one percentage point) higher than its pre-cut low even after multiple cuts, an unprecedented divergence from historical behaviour.
Bond markets did respond similarly, though not to the same extent, in the 1960s and 1970s. The Fed dropped interest rates to stimulate economic growth and the 3-Month fell; but bond market ‘vigilantes’ responded by selling off the 10-Year — pushing up the interest of mortgages and business loans, contributing to economy-wide recessions and stock market downturns.
With overall annual inflation presently trending down to low levels — 2.7% in December; 1.7% according to Truflation’s more up-to-date/real-time tracking — something else is going on. In the 1960s and early 70s, the US [federal] state’s debt-to-GDP consistently fell, from 83.1% in 1950 to 41.2% in 1974.
Spending on war declined significantly compared to WWII; outdated or unprofitable capital had been devalued or destroyed, clearing the way for new profitable investment; wartime inventions were converted into consumer goods; soldiers were returning to the workforce; and, with wages rising and prices falling, families were growing in size, further boosting productivity. Most importantly, the state, across the world, took on a huge bulk of the costs of the postwar rebuild, significantly lowering the costs of things the private sector needed to buy to expand and innovate its operations. Inflation, of course, also played its part, since fixed nominal repayments fell — the state paid bondholders back in money worth less than when the debt had been issued.
Today, the private sector is running out of public wealth to privatise; and, apart from in China and a select few other countries, state-owned enterprises (SOEs) are not available to provide the private sector with low-cost goods, equipment and inputs. A (hot) world war has not yet (officially) started; birth rates are dropping faster than predicted (because capital increasingly invests in capital, not labour, and at a slowing rate); large parts of the workforce are retiring; new generations are denied affordable education and training; women are no longer increasing the size of the exploitable labour base; and the economy is increasingly dependent on selling junk food and addictive pharmaceuticals — further decimating the population’s health and productivity.
Since 1974, the US’s debt has been rising and at a generally accelerating rate. According to the European Central Bank, nations with smaller debt burdens have lower interest rates and vice-versa; with state debt rising above 90% a significant inflection point at which investors start to balk. (Other studies have found similar results.)
Whereas private and state US debt combined as a share of GDP rose from around 140% in 1950 to 160% in 1980, it has since exploded to 400% — an unrepayable amount. Around 50% of this debt had been wiped out or repaid in the years following the ‘Great Recession’ of 2008–09 — but it has since bounced back and recently usurped the 2009 record.
Back then, the majority of debt was owned by households, at 100% of GDP (80% of which was held by the wealthier 50%). That figure has fallen substantially, transferred to the (federal) state, which has seen its debt ratio in that time rise from 50% to 124%. With taxation and borrowing unable to meet the state’s spending commitments (despite large cuts in spending per capita to education, welfare, etc) — upon which the profitability and viability of capital and private property are increasingly dependent — the Fed has had to (largely digitally) ‘print’ the difference, lending it into existence by buying mortgage-backed, corporate and government debt from banks and corporations, thereby bailing out the private sector and shifting much of its debt onto the backs of the public.
As the money supply expands, currency obviously tends to lose value — the US dollar’s purchasing power is down by 98% since 1971; 25% against food in just the past five years — and so investors are flocking to gold, silver, stocks, property, land, etc. as hedges against inflation.
(Economic/commodity growth is bound to increase the money supply as more money is needed to mediate more transactions — the Fed’s ‘printing’ is primarily a lagging response, not a driver of economic activity; and banks also lend money into existence, mainly through credit cards. The US M2 money supply has gone from $0.7 trillion in 1971 to $21 trillion in 2025 — a 3,000% increase.)
Total global gold investment demand surged by 84% year-on-year to ~2,175 tonnes in 2025, a record high. Retail-oriented bar and coin demand rose by 16% to a 12-year high. Gold ETFs added 801 tonnes, the second-strongest year ever.
Central banks outside of the US (combined) now own more gold than US Treasuries for the first time in almost 30 years. The US reduced its (monthly) trade deficit by 40% in October 2025 (to $29.4bn), the lowest reading since 2009, partly because of a 3.1% fall in imports — now down by 21.1% from the most recent peak, the worst decline on record other than in 2009 and 2020; mostly from a $14.3bn decline in pharmaceutical preparation chemicals — but mostly because of a $10bn surge in non-monetary (private sector) gold exports (amid total exports of only $7.1bn, meaning other export categories actually fell by $3bn). The implication is that private gold sellers needed to replenish cash.
During the time in which gold has trebled, oil and gas have declined by 30%, with oil having recorded three consecutive annual losses for the first time; the average price of US homes has been flat; and the price of wheat is also down by 30%. That cannot last. The price of oil, food, energy, transport, metals, and (lastly) wages will follow that of gold and silver; since the fall in the demand for the dollar makes imports more expensive and the expansion of the money supply devalues the currency’s purchasing power.
At 1.4, the dollar’s velocity, the average number of times a unit of the currency is spent per year, is flat since the end of 2023, having rebounded from the low of 1.13 in 2020 (after which followed 40-year high inflation, when rising savings were unleashed by the end of the lockdowns); and well down from the peak of 2.2 in 1997.
During a currency collapse, money velocity goes up as the value of money nosedives and is therefore spent as quickly as possible on stockpiling — as the point of holding onto money diminishes — pushing up demand and prices.
Velocity is presently low, however, because a huge chunk of money in the economy is tied up in unproductive, parasitic financial assets, turning much of the economy into an ever-worsening Ponzi scheme. The total stock market as a share of total money is at its highest level in history. Owned by a small fragment of the population, the vast majority of this money is not circulating. With the prices of commodities — especially oil — falling as a result, however, investors are increasingly likely to start selling financial assets in order to take advantage of the improving prospects of making profitable returns from investment in productive capital and innovation (albeit at a likely lower rate than in the past), putting upward pressure on money velocity. The ensuing depreciation of financial assets will therefore weaken and crash a stock market that is more concentrated, expensive and overvalued than ever.
Stock market
Never before has GDP growth been so outstripped by rising stocks. The stock market capitalisation (valuation) of all 3,265 US publicly-listed/traded companies (the Wilshere 5000) is now 224% of GDP, the highest ever; beating the 193% of December 2021 and 135% of March 2000, after which the ‘Dot Com Bubble’ burst, crashing the S&P 500 by 49% and the NASDAQ by 57%. (The S&P 500 is a proxy for the richest 500 US companies; the NASDAQ is the stock market for the tech sector.)
In mid-January 2026, the S&P 500 price-to-book ratio hit 5.6; compared to 5 at the peak of the Dot Com Bubble. That means investors are paying $5.60 for every $1 in physical assets (minus debt) actually owned by the companies they are funding (making the book value a rough proxy for the hard, productive base of the economy).
Similarly, investors are paying an average $3.30 for every $1 of revenue generated, the biggest gap on record. It now takes 1,295 working hours to buy a single unit of the Dow Jones Industrial Average Index, the highest in history and up by more than 500 at the 2017 peak or 2020 bottom.
Doubling since 2022, the market cap ratio to the M2 money supply has hit a record 306%. That is, the total value of all US stocks is now over three times larger than the total amount of dollars held in the US financial system in cash, checking, and savings.
The S&P 500’s market cap to GDP is around 180%, again beating the 160% peak at the turn of the century. It has doubled in just the past three years, something that took 20 years from 1965–85; or 17 in 1997–2014.
The total volume of options traded (betting on price moves) on the S&P 500 has hit $3.5 trillion per day, a six-fold increase since 2020.
The value of all public and private equities to GDP hit 349% in July 2025, up from the previous peaks of 327% in October 2021; 188% in July 2007; 202% in January 2000; and 103% in October 1968.
Tech stocks make up 100% of GDP, an all-time high; that figure having doubled in three years and standing above the 60% it reached before the 2000 Dot Com crash; a figure also applicable to the 2008 Housing Bubble (whereby, in response to falling profitability, the capitalist state gave banks permission to lend poor people mortgages they could not afford to repay, especially once interest rates rose; junk-level debt that banks then sold on, hidden in bundles, in unregulated derivative markets). During the railway mania in the UK in 1845, the market cap of railway stocks was only 10% of GDP; leading to half of all railway companies going bankrupt within four years.
The S&P 500 relative to defensive stocks is now the most expensive since 2000; with healthcare, consumer staples and utilities valuations relative to the broader index falling in value consistently since 2022, as investors reallocate to tech stocks in desperate search for higher returns. Campbell’s Soup is at its lowest price since 2009.
Capitalist decay
Underlying this entire process is the nature of capitalism; or, even more precisely, a simple mathematical reality about the nature of the commodity. In short: making ever-larger profits as required by a capitalist firm or economy of course entails the growth of commodity production — the number of commodities produced — which, via innovation that speeds up production, tends to result in the devaluation of commodities and, in the long run (since devaluation/falling prices initially lowers the cost of investment), less profit per commodity. The more commodity production advances, the more exchange value withers away. (The price of magnetic memory storage, for example — vital to contemporary innovation — declined from $87.5bn per terabyte in 1956 to $93.4m in 1984 and $11 in 2023.)
The result is a tending-to-worsen underproduction of profit relative to the value of capital. As rates of profit fall and investment opportunities dry up, ‘overaccumulated’ capital manifests — gluts of surplus money that cannot be profitably reinvested; factories that cannot be utilised (at least at full capacity); and so on. Debt rises as a partial substitute; and parasitic speculation, whereby value is only siphoned and transferred from one place to another instead of created, rises as competition intensifies over a smaller amount of value per capita. That is, everyone fights for a bigger slice of a shrinking pie.
Karl Marx’s contention that this inherent barrier to innovation and productivity growth tends to grow “more formidable” is illustrated in many ways: by the record high prices of gold, silver and speculative stocks, for example; and the amount of dead cash amassing in US money market funds, soaring above $7.5 trillion in 2025, up by a phenomenal 35% in just three years.
Between 2022 and 2023, US billionaires increased their cash reserves from an average 14% of total assets to 24%. Warren Buffett’s Berkshire Hathaway holdings conglomerate has dumped stocks for 12 consecutive quarters, its longest ever selling streak; its cash reserves exceeding a record $380bn by late 2025.
China
In the face of its worst ever profitability crisis, US capitalism is desperately trying to stave off its own downfall. US capital has long countered falling profitability by frequently exporting overaccumulated capital—investing in production overseas to take advantage of lower levels of development, where capital and labour are therefore cheaper; while also simply expanding the size of the exploitable labour base, the domestic size of which is of course limited. Value is created in the ‘developing’ oppressed nation and then partly shipped to the ‘developed’ imperialist parasite. (According to the most recent studies, ‘Southern’ countries produce 90% of global labour, but receive only ~44 % of global income; and Southern workers receive ~21 % of global wages; although ‘Northern’ workers produce more value per hour since they operate more advanced technology.)
Hence the US’s growing trade deficit. The compulsion to invest overseas makes the cost of importing commodities cheaper than producing them domestically.
With its large industrial basis, and unaffected on domestic territory by global conflict, the US usurped the British Empire in the period before, during and after World War II; and then saw off competition from Japan and the European Union, which has been kneecapped in recent years by the destruction of the Nord Stream 2 gas pipeline, forcing it to reduce its reliance on cheap Russian energy and switch to a more expensive alternative from the US.
Now China is rising as a much more formidable challenger.
With its immense population and a workforce benefitting from the free education of a socialist economy that capitalist countries in the region could not provide, China made for the perfect outlet for the US’s overaccumulated capital. After opening up its economy to private investment since the early 1980s — to tame the aggression of the US, enabling China to import and export — China has industrialised and developed rapidly.
Although the initial transition certainly came as a painful setback for much of China’s working class, decades of hard work have reaped long-term rewards. The high demand for labour has lifted the wealth and skill-level of China’s workers at a rate faster than anywhere else.
Some of China’s domestic businesses have grown into competitors on the world market and its industry has naturally ‘moved up the value chain’, from low-tech to high-tech production. It is the indisputable industrial superpower of the 21st Century; its infrastructural projects are embarrassing the US and western Europe; and it controls more of the world’s most critical resources than anyone else — making it a far more powerful challenger to the US than Japan or the European Union ever were.
China is now the world’s biggest economy when measured in Purchasing Power Parity (PPP) — comparing currencies using the exchange rate that equalises the cost of a common basket of goods — with estimates for 2025 placing China’s GDP (PPP) around $41 trillion and the US’s at $30.5 trillion.
Despite the significant liberalisation of the country’s economy, the stability of the political system and the Communist Party, along with the slower and more controlled pace of privatisation, compared to what happened in the Soviet Union, has allowed China to carefully adapt without losing state-owned industry to the same extent as in Russia and the rest of the ex-socialist bloc.
Energy, raw metals, heavy industry, transportation, banking, telecommunications, infrastructure, construction, defense, and aeropace are all still dominated by enterprises that are at least majority owned by the state. Xi Jinping, the President of the People’s Republic of China since 2013, has also led an economic ‘turn to the left’, implementing a greater degree of state ownership than post-Mao predecessors. Whereas the US political class is led by lawyers who inherited stolen wealth, China’s political leaders are engineers from families who birthed an independent socialist republic.
In 2000, the Fortune Global 500 (FG500) featured 27 state-owned enterprises, climbing to 102 in 2017 and 22% of the total FG500 revenues ($27.7 trillion, up from $12.7 trillion in 2000). China’s number of FG500 SOEs rose from 9 of the 27 in 2000 to 75 of the 102 in 2017. In 2023 SOEs accounted for 97 of China’s 142 FG500 enterprises.
The profit necessity and private plundering and hoarding are therefore much less of a block on production and innovation than in the US. China’s private sector is also much younger, so its overaccumulation of capital is significantly less severe. The proceeds from production are reinvested (in social enterprise) to a much greater degree, instead of sitting on the sidelines in private bank accounts, surplus to requirements and only able to accrue interest.
According to the Penn World Table, in 2019 gross fixed capital formation as a share of gross profit was 42% in China compared to the US rate of 9% and a global average of 14%.
Since 1992, China’s GDP has grown 51-fold compared to the US’s 5-fold; but China’s stock market capitalisation grew 8-fold while the US’s 29-fold.) China’s economy is real and productive; the US’s speculative and debt-based.
China’s economic position has enabled it to offer much more attractive deals and lending terms to nations that were previously dependent on the US. In 2023 China was the largest bilateral trading partner for 145 out of 205 of the world’s economies (about 70%), compared to the US’s 33. (Whether that makes China imperialist is a difficult question to answer but its capital exports as a share of its economy are small; and many of its overseas investments are actually made by SOEs — arguably a kind of exportation of socialism.)
The US dollar still utterly dominates global reserves and transactions, but only 80% of international oil contracts are now priced in US dollars — the ‘petrodollar’ — compared to well over 90% 20–30 years ago. As a share of global reserves, the US dollar has fallen to 56% from 73% in 2000; although its use in foreign exchange trading volume has held at about 88%.
The Chinese renminbi (RMB) is making noticeable inroads. Its share of global reserves roughly doubled from 2016 to over 2% by the end of 2023. It accounted for around 4–4.5%% of global cross-border payments in 2024, up from 0.1% in 2010, making it the fourth most active payment currency after the US dollar, the euro, and pound sterling.
About 28% of China’s goods trade was settled in RMB in 2024–25; and a bit more than half of China’s overall cross-border receipts and payments, a dramatic rise from virtually zero a decade before.
US tariffs aimed at preventing China’s rise have so far failed, with China continuing to increase its share of global exports enormously in 2025 (largely as a result of slashing prices).
Empire strikes back
With the value of the dollar waning, other nations are diversifying their reserves and turning to gold or other currencies. Falling demand in turn compounds the dollar’s devaluation.
New decentralised peer-to-peer blockchain networks are also enabling the circumvention of middle men and cutting down on transaction fees that traditional institutional banks rely on.
The US is therefore increasingly turning to coercion to try to force countries to continue to trade in the dollar and/or privatise state assets and public wealth, as has been the case for example in Iraq, Ukraine, Libya and now Syria and potentially Venezuela. Countries that do not co-operate are sanctioned: US companies that exchange dollars with sanctioned foreign currencies are heavily fined if they are found to have mediated such transactions. Brazil, Russia, India, China and Saudi Arabia, along with others — like Venezuela — have responded by building an alternative trading bloc (BRICS), increasingly shunning the US dollar. With sanctions failing, military action, such as the invasion of Venezuela, is the next course of action. Venezuela is rich with oil and mineral reserves, but successfully attacking it would also be a blow to China, which may lose an export market and a cheap supply of oil.
(Venezuela recovered somewhat from a severe escalation in sanctions in 2017 — which reduced government revenue by a reported 99% — becoming the second fastest growing country in Latin America in 2024 (8.5%) and 2025 (6%).)
Given that capital’s falling profitability makes it increasingly dependent on state subsidies, facilities and contracts, the private sector is increasingly dependent on inventing things to sell to the state — largely by becoming increasingly militarised — and bleeding the public dry through consumer and income tax hikes (as well as fines, and so on).
But the devaluation of the dollar and the increasing demands on the US military — as it overextends itself across the globe in order to defend and extend the US’s private investments — makes the reproduction of the US’s military might increasingly expensive (in real-terms), badly compounding the country’s deficit and debt problem.
This process is obviously unsustainable. For one thing, the US’s reliance on investing overseas makes its economy and military incredibly dependent on imports from none other than China.
China dominates key supply chains for rare earth elements used in advanced defence systems — from fighter jets and missiles to radar and precision guidance — and controls a large share of the world’s production and processing of elements such as dysprosium, samarium, terbium, and yttrium. US import dependence on Chinese rare earths has reached around 70–80%. Beijing has also restricted or banned exports of strategic materials including gallium, germanium, antimony, tungsten and graphite, all vital for semiconductors, batteries and military electronics, disrupted US supply chains and raising costs for defence contractors.
The arms race, furthermore, is accelerating the automation revolution that is making capitalism historically obsolete, as commodities are made increasingly quickly, with less and less labour, reducing the value of wages and in turn the consumer demand for commodities.
Recession
Now, the first extended, full blown recession since 2009 has almost certainly begun.
The US’s unemployment rate crept up from a low of 3.4% in late 2023 to 4.5% in November 2025, despite an absolute 7% increase in GDP. Averaged across 2025, the US added only 44,000 jobs per month, the lowest since 2020 and lower than any year during the 2010s.
After the worst October in two decades, job losses hit 1.2 million in the first 11 months of 2025, up by 54% year-on-year. (The number of jobs created from April 2024 to March 2025 was also revised down by 911,000.) Job cuts have surpassed one million in a single year only five other times since 1993: in 2001, 2002, and 2003, with the Dot Com bust; 2009, during the Housing/Great Financial Crisis (GFC); and in 2020, when the COVID pandemic and lockdowns struck, following a period when Donald Trump had been calling for the Fed to lower interest rates, as US growth in Q4 2018 slipped to 0.7%. (Turkey and Argentina experienced negative growth in 2018 and Germany and the UK flatlined in Q3.)
Average time spent out of work has also hit 24 weeks, up from 16 in the run up to the GFC. The underemployment rate has risen, too, with the number of people employed part-time but wanting full-time work increasing by 980,000 over the year to 10.8%. For recent college graduates, the underemployment rate hit 41.8%.
The unemployment rate for young college educated workers is up by a full 1.6 percentage points in the last year. More broadly, the unemployment rate for new labour market entrants (like young adults looking for their first paycheck) and re-entrants (like stay-at-home parents going back to office work) are both hovering near their highest levels since 2016. Wage growth for the lowest-paid workers is much lower compared to higher-paid counterparts, manifesting in the largest gap in wage growth favouring high earners since the early 2010s.
The US lost an estimated net 145,000 manufacturing jobs in 2025 — making a mockery of Trump’s ‘promise’ of a blue-collar revival — at the fastest rate since early in the pandemic and worse than any period from 2011–19.
Tech sector jobs fell by more than 19,000; adding up to 100,000 since late 2023. It took just over a year before tech employment began to grow again after the GFC; whereas today’s decline continues nearly three years on. If it persists through 2026, it will beat the Dot-Com bust for the longest period of tech job market losses.
The US public sector lost more than 157,000 jobs. Federal employees represent 13% of all public-sector jobs, but the 277,000 jobs eliminated at the national level, plus the 45,000 losses at the state level, easily outstripped small growth at the local level.
The beginning of a recession is unlikely to be officially confirmed until a year after it starts, but a wide range of economic data signals a brewing contraction. According to Moody’s Analytics, 22 US states and the District of Columbia experienced shrinking GDP in 2025; while another 13 have flatlined, leaving only 16 generating or experiencing growth.
Overall, 717 large firms went bankrupt in the first 11 months of 2025, the highest number since 2010. The figure includes 85 consumer discretionary and 32 consumer staples companies; 110 industrial and 46 healthcare companies; and publicly-listed firms with either assets or liabilities over $2m and private firms with liabilities over $10m.
The Institute for Supply Management manufacturing index indicated a contraction in the US manufacturing sector for the 10th consecutive month at the end of December, putting the index at 47.9 (below the 50.1 that indicates growth). The index has only expanded in two of the past 38 months.
Declines in heavy truck sales often reach 20–40% or more before recessions; sales in 2025 were down by around 13–16%. Freight shipping has also approached 2008–09 levels.
Housing starts in the US are expected to come in at 1.3 million for 2025, having dropped from 1.6 million in 2021 to 1.55m in 2022, 1.42 in 2023, and 1.36m in 2024. In August, the delinquency rate of office mortgages packaged into commercial mortgage-backed securities jumped to 11.7%, surpassing the 2008 peak of 10.7%. As recently as December 2022, the rate stood at only 1.6%.
Productivity growth is slowing globally, falling from 1.6% to below 1% in ‘developed nations’ between 2011 and 2022; and from 5.9% to 3.4% in ‘developing nations’.
US GDP growth has been steady at 2.9% and 2.8% in 2023 and -24 (although the figure is of course inflated by ‘value added’ from things like disease-inducing ‘food’, weapons of mass destruction, speculative transactions, bank fees, and gambling; the latter growing from $30bn in 2020 to $72bn in 2024; about 0.5% of GDP).
The average decade-on-decade figure, though, has consistently declined from a high point of around 6% in the 1960s. In the last decade or so it has been propped up by a temporary shale gas boom.
The US’s 10-Year minus 2-year Treasury yield curve has recently uninverted — a historically reliable recession indicator. The curve turns negative when long-term bond yields — usually higher because of the uncertainty that comes with a longer wait and compounding inflation — fall below short-term rates; reflecting investor expectations of weakening growth and central bank rate cuts.
Historically, recessions have followed the exit from the inversion, not the inversion itself, with lead times ranging from 6 to 23 months. The last inversion lasted an unprecedented 26 months, falling to a record ‑1.06 percentage points. Deeper, longer inversions have preceded more severe recessions, suggesting that the next downturn will be worse than 2008–09’s ‘Great Recession’ — the deepest since the Great Depression of 1929–33 — when US GDP contracted by 4.3% from peak to trough and unemployment hit 10%. (10.1% and 14.7% during the much shorter Covid pandemic-lockdown recession.)
Ending postwar US recessions have required an average 6 percentage-point fall in the central bank rate, indicating that the Fed may have to go to zero again, given that it started cutting, very tentatively, from 5.3% in July 2024 (to 3.7% at the end of 2025); and kept rates at zero — for the first time ever — for seven years after 2009 and 20 months after 2020.
Each recession-countering rate cut since 1981 has started from a lower point: 19%; then 9.8% in 1989; 6.1% in 2000; 5.3% in 2008; and 2.4% in (the exceptional circumstances of) 2020. The 5.3% of 2024 bucks the trend, though barely, and followed rate hikes that largely went up to counter 40-year high general inflation (which makes investment and mergers — capital accumulation — more expensive).
Conversely, a recession is the temporary, partial solution to overaccumulation — and renewed investment catalyses recession. As the relative underproduction of profit rises, surplus capital rises; production and supply slows down relative to demand; and prices rise. Interest rates therefore go up, lending tightens, and demand falls as reserves run dry, pulling prices back down, as has been seen in the case of oil and gas; and, even more crucially, (real) wages (relative to overall inflation).
Employee share of income in the US averaged 64% from 1967 to 2002, but fell sharply after the Dot Com crash, the GFC and the short, sharp Covid recession, sinking to around 53–55%, a record low (something that is of course compounded by labour’s shrinking role in the production process as automation advances). This trend has been the biggest factor in overall corporate profit margins (after operating costs) jumping to a record high about 21% after Covid and 20% again now, up from the previous peaks of 18.5% before and after the GFC; and 15.5% in 1997.
(Corporate profits may be at a record high — but for a shrinking number of corporations; and are not only based on bleeding workers dry but the public in general through debt, as illustrated by the fact that the gap between GDP and corporate profits has never been wider.)
As commodity prices fall and capital depreciates, the affordability of investment and innovation improves again, at least for companies with big enough reserves to ride out rising inflation and/or falling demand. At least some surplus money capital pours out of speculation, savings and stocks and back into productive commodity production, as can be seen perhaps most clearly in charts of US money market funds during and after recessions.
The money supply expands, devaluing money and capital, and the effect on the purchasing power of money crushes the ability of firms with slender profit margins or reserves to pay wages, bills and/or debts. Unprofitable capital (including factories) is written off and abandoned or sold off on the cheap to the survivors, resulting in an overall centralisation of capital, a concentration of corporate and monopoly power. Some innovative start-ups do emerge, having taken advantage of falling prices while starting out with low overheads; but to a lesser and lesser extent.
The economy can therefore go through a contraction for some and a boom for others at the same time. Following the fall in the price of oil and borrowing in 2025, at the start of January corporations received $95bn from 55 investment-grade bond deals, the highest weekly volume since May 2020 and the busiest start to a year on record.
“Companies took advantage of strong investor demand for high-quality dollar debt that has pushed borrowing costs close to their lowest level relative to US Treasuries since the global financial crisis… with the cost of borrowing for investment-grade companies at just 0.79 percentage points above government debt, according to Ice BofA data,” The Financial Times reports. “While January has typically been a busy month for new bond issuance, many companies are starting their funding programme even earlier than usual this year to get ahead of an expected issuance glut to finance Merger and Acquisition (M&A) activity and big tech companies’ AI infrastructure. Morgan Stanley forecasts investment-grade bond sales this year of $2.25tn, eclipsing the 2020 record of $1.9tn.”[1]
M&A activity improved by 40% year-on-year in 2025, “as larger firms acquired smaller, leaner competitors to bolster their own operational efficiency”.
The relative boom may not last long — in mid-January the price of oil jumped back up by 10% amid the US’s invasion of Venezuela and concerted uprisings in Iran, both countries of course being critical to global oil supplies and having long had their development obstructed by sanctions, tariffs and blockades that the US continues to reinforce. It did then drop again as Trump responded to a bond market wobble by rowing back on his threat to bomb Iran. Whether oil spikes higher or continues to resume its trajectory downwards before at some point rocketing upwards remains to be seen.
Glorified betting
Desperate not to spook investors, multinational investment bank Goldman Sachs claimed in October that “anticipated investment returns are sustainable”.
A stock market frenzy, however, is itself a bellwether of a weakening real economy, since a shortage of profitable opportunities in commodity production — where new value is created — is chucked instead into the glorified betting of speculation, inflating the value of stocks and generating a false sense of prosperity.
From the start of October to mid-November, furthermore, only 38% of S&P 500 stocks trended upwards. Michael Burry (of The Big Short) has, as he did in 2008 when he made millions of dollars by shorting (betting against) the mortgage market, deregistered his hedge fund and shorted Nvidia, the absurdly overvalued computer chip designer (with a price-to-earnings ratio of 48).
Many commentators and analysts sounding the alarm are drawing comparisons to 2008 and 1929. The situation might well be worse than either. At the end of 2024, US banks had about $483bn of unrealised losses on securities; compared to no more than $150bn around the time of the GFC (or an inflation-adjusted $220bn). That equates to 21% of tier 1 capital instead of the GFC’s 14% (the highest quality capital a bank holds to absorb losses and truly protect depositors and maintain solvency).
To meet cash demands, banks may sell long-duration bonds at a loss, converting paper losses into realised capital losses, weakening buffers. One bank’s forced sales can push down bond prices, creating losses for other banks. Market panic can make liquidity dry up within 90 days even for healthy banks. This threat to the US banking system lies ahead.
The concentration of stocks and general wealth in the US is now on a par with the late 1920s, for the first time since that dark period. From 1979 to 2021, real average wages rose by 1.0% annually; but only 0.6% for the bottom 90%; against average annual inflation rose of 4.1%. In contrast, the top 1% and 0.1% saw their income grow by 2.7% and 4.2% per year. That means income for the top 1% and top 0.1% surged by 206.3% and 465.1%, respectively, while wages for the bottom 90% grew by just 28.7%. The share of total earnings for the top 0.1% rose from about 1.6% to roughly 5.9%.
Just under 50% of all consumer spending is made by the top 10% of earners (who made $251,000 or more in 2024), an historic high and up from 43% in 2020.
The top 10% of households own 93% of the stock market’s value. The top 25 stocks account for 45% of the S&P 500 and the top seven for 31%. Nvidia’s market cap is 7% of all publicly-traded US companies and 4.9% of the MSCI All Country World Index — up from 1% in 2023 — higher than the 4.83% for all of Japan, the world’s fourth largest economy.
Only 10% of households were invested in the stock market when it crashed in 1929 — now 62% of US adults report owning stocks, and not counting retirement funds that are often directly or indirectly invested in the stock market. At least 43% of retail investors (individuals) are buying more stock than their cash can afford.
US household equity holdings have surged to a record 47% of total financial assets, exceeding the 2000 peak by 8 percentage points.
The S&P 500 CAPE Ratio (Cyclically Adjusted Price-Earnings), the S&P 500 price divided by the average of the past 10 years of inflation-adjusted earnings, has risen to 45, beating the 2000 peak of 39.
The CAPE Ratio moves more or less in lockstep with US Households Holding of Equities as a Share of Total Financial Assets. Both peaked together in 2000, followed by the -50% S&P 500 crash. Both peaked again in 2007, followed by the -57% crash.
In the run up to the 1930s Great Depression, all active home mortgages represented 10% of US GDP, rising to 32% in 1930. US mortgage debt today is 70% of GDP, more than $18 trillion in total household debt. (In Australia and Canada, mortgage debt is higher than the country’s entire GDP.)
Households are badly exposed to the coming stock market crash. An estimated 67% of Americans are already living paycheck to paycheck, up from 63% in 2024; spending their entire income on bills and expenses as it comes in, leaving little to no money for savings or emergencies. In 2023, 12.6% of the total US population received food stamp benefits — one in eight people — up from 6.1% 2001.
The rapid expansion of margin debt, the total amount that investors have borrowed to buy securities — which fuelled the 1929 crisis — reached a record high of $1.1 trillion in September, up by 34% on 2024. The private (non-bank) credit market is exploding at a rate of around 15% a year, rising from $0.5 trillion back in 2012.
Drawing comparisons to the 2008 subprime housing crisis, US subprime used-car lender Tricolor filed for bankruptcy in September after it allegedly pledged the same loan portfolios to multiple lenders, exposing private credit providers and major banks such as JPMorgan, Barclays, and Fifth Third to unexpected losses, with its filing listing up to $10bn in assets and liabilities.
‘AI’ bubble
The private credit bubble is only part of a much larger one — an ‘everything bubble’ that since the GFC has for the first time ever engulfed every asset class (debt, equity, reserves). Now the ‘AI Bubble’ has joined the party and outbloated all that came before. One has to ask: what comes after an everything bubble?
The top ‘Magnificent 7’ stocks are all tech sector, and the tech and ‘AI’ sector accounted for a scarcely believable 92% of US growth in the first half of 2025. Since the debut of the advanced research assistant tool Chat-GPT in November 2022, AI-related stocks have added an estimated $17.5 trillion in market value — 75% of the S&P 500’s gains.
In breakneck competition to construct expansive data centres equipped with specialised, cutting-edge chips, Microsoft, Alphabet, Amazon, and Meta reported a combined capital expenditure of $245bn in 2024, a figure surpassing $360bn in 2025 and expected to hit up to $600bn in 2026. According to an MIT report, though, only 5% of integrated AI tools are generating profitable returns.
In the third quarter, venture capital deals with private AI firms dropped by 22% quarter-on-quarter. Amid ongoing problems with reliability and the rocketing costs of training new models AI-tool usage at firms with more than 250 employees dropped from nearly 14% in June to under 12% in August. According to Edge AI & Vision Alliance, training costs have surged by more than 4,300% since 2020, driven mostly by the rising price of chips and staff (engineers and researchers).
Investment has been pouring into AI stocks out of hype, lack of other options, and the hope of innovation-based growth (which is certainly not all hype; MIT found a 14% increase in productivity for customer support agents; the London School of Economics found AI-assisted workers save about 7.5 hours per week).
Oracle’s debt-to-equity ratio is 500%. Chat-GPT owner OpenAI is valued at $500bn but reported a net loss of $13.5bn on $4.3bn in revenue in the first half of 2025. The company does not expect to be profitable until the end of the decade and is now begging tail-between-legs for a state bailout. (Something that Trump might be addressing by effectively taxing old “underperforming” defence contractors by capping executive pay and share buybacks.) Such talk at the start of November sent the NASDAQ down by more than 1% in a single day and wiped $1 trillion off the value of Wall Street’s most valuable tech companies within a week.
Bain & Co. estimates that cloud service providers like Google, Microsoft, and Amazon will have to generate an additional $2 trillion in annual revenue by 2030 to afford all the necessary infrastructure, five times more than the current global market for software subscriptions. In 2024 Amazon, Alphabet, Apple, Meta, Microsoft, and Nvidia combined made less than $2 trillion. US electricity consumption growth stayed at 0.1% in 2005–2020. That has since picked up to 1.7%, but AI is expected to add another 2% of demand per year up to 2030, likely consuming power faster than new power plants can be built. Total generation likely needs to be 27–35% higher by 2030 compared to 2022 levels.
An annual power market auction in July 2024 by the largest US electrical grid operator, PJM Interconnection, which covers parts of 13 states from Illinois to New Jersey, resulted in prices more than 800% higher than the year before; up from $28.92 per megawatt-day to $269.92. Electricity bills for customers rose by about 20% per month.
China, in contrast, thanks significantly to the largely joined up state ownership of its core, essential industry — as opposed to being dominated by many competing, profit-seeking, corner-cutting entities — has solidified its position as the global leader in electricity generation and installed capacity, providing a massive, rapidly expanding power base for its AI and industrial sectors. China’s total power generation in 2025 was 10 trillion kWh, roughly 140% higher than the US’s, with renewable enegy now accounting for 59.1% of total capacity, compared to the US’s 28%.
China’s DeepSeek AI performs at least as well as US equivalents while requiring 50–75% less energy. Instead of one giant model, DeepSeek splits its neural network into many smaller, specialised ‘expert’ networks (e.g., for coding, math, poetry). A ‘generalist’ network directs requests to the most relevant experts, so the entire model doesn’t need to run for every query, drastically cutting down on computation demands.
An economy with less dependence on capital accumulation, it turns out, can develop less capital-intensively, enabling more agile and innovative progress.
The US’s capitalist state is attempting to ride to the rescue — although relatively meekly. The 2024 AI National Security Memorandum recast the success of US AI as the national security priority of our times. The Trump administration’s AI Action Plan aims to accelerate AI adoption within the government and military by pushing changes to regulatory and procurement processes. The One Big Beautiful Bill Act authorised $1bn in AI funding, including $450m for AI applications in naval shipbuilding and autonomy and $250m for advancing the AI ecosystem within the Department of Defense. Another $500m has been allocated for AI-related broadband and infrastructure.
Such figures are not particularly impressive, owing to the US’s overaccumulation and indebtedness. OpenAI is now in talks with sovereign wealth funds in the Middle East to secure a reported $50bn.
These subsidies and bailouts nevertheless heap yet more debt onto the backs of a public increasingly weighed down by the burden of saving the private sector continuously since 2008.
Nvidia, OpenAI, and major data centre operators are propping up one another’s growth through large, incestuous circular investments. As demand far outstrips supply, Nvidia sells its chips (which it designs but are actually manufactured by Taiwan Semiconductor Manufacturing Company) at an extremely high margin. It is left with little choice but to subsidise that demand by investing in AI firms that purchase its hardware. OpenAI has committed to buying 10 gigawatts of compute from Nvidia, at a cost of at least $15bn each; but in return Nvidia is investing $100bn in OpenAI (for non-voting shares). Similar arrangements run through other firms such as CoreWeave and Nebius.
Real economic growth will of course continue to tend to slow relative to inflated valuations. As profitability falls, liquidity reserves and lending dry up, and job losses rise, the Federal Reserve will tend to continue to lower its baseline interest rate — fueling the bubble, since investors will find it easier to borrow to speculate on assets and future growth.
If an ‘external’ factor like tariffs or war do not spook investors first, the intensifying concentration of stocks will eventually shut out too many investors from sufficient returns and a massive sell off will begin, bursting the bubble; which will also happen if a panic-sparking absolute economic contraction takes off. The fictitious money investors have been throwing into this hole will be wiped off the ledger board. A 30% correction is strongly correlated with a protracted, economy-wide recession, but a much higher figure, perhaps accumulated over the course of two or more ‘corrections’ (10% stock market declines) and ‘bear markets’ (20%) over several years, should come as no surprise. (20% contractions usually result in short, shallow contractions; such as 2022’s -1.6% in Q1 and -0.6 Q2; and 2025’s -0.5 in Q1.)
Gita Gopinath, former chief economist of the IMF and now with Harvard, believes that a market correction of the same magnitude of the Dot-Com crash — when the S&P 500 and the NASDAQ Composite crashed by 50% and 75%, respectively — could wipe out about $20 trillion in wealth for US households, nearly 70% of annual US GDP. On that basis, consumption would fall by more than 3% and GDP by two percentage points, easily enough to push the US economy into a deep recession. She also estimates that foreign investors could face wealth losses exceeding $15 trillion, about 20% of the rest of the world’s annual GDP, since US equities make up 60% of the global market. We should not be surprised if Gopinath’s estimate turns out to be wildly conservative.
1929
In contrast with the asset inflation that characterised the aftermath of 2009 and the asset and consumer price inflation after 2020; 1929 resulted in absolute overall deflation — average prices were 27.4% lower in April 1933 than in October 1929.
The Fed and every government since have been desperate to avoid a repeat, as income loses its value relative to (fixed) debt; and demand (including for labour) falls faster in anticipation of falling prices. (The Fed’s hands were tied by the commitment to ‘sound money’ and the gold standard — fixing the currency’s value to a specific amount of gold, limiting money printing and therefore economic growth — which the government finally abandoned in 1971 since retaining it would have allowed another full-on depression.)
The Wall Street Crash impacted the whole world so badly because the US, then a net lender, was the world’s biggest lender. The rest of the world needed to borrow US capital to grow their economies.
As the home of the world’s dominant reserve currency and richest consumer base, a major US recession today will again of course devastate the whole world. Now, though, the US is a net borrower and the world’s largest borrower. About 30% of its debt and government spending is financed by foreign capital inflows; but at the end of 2025, China’s US Treasury holdings were $683bn, down rom $1.07 trillion five years earlier. Japan’s did drop from $1.3 trillion to $1.1 trillion before creeping back up to $1.2 trillion. Now European funders are threatening to disinvest in protest against Trump’s intention to buy or seize Greenland.
Since 2020, official US government debt-to-GDP has been higher than the 119% at the end of WWII. On course to soon hit $40 trillion, US government debt is growing at around 6% per year and the annual government deficit — the gap between income and expenditure — 8.5% (since 2008). As discussed above, as the debt rises to particularly high levels, the risk of lending to the government climbs as the chances of the overstretched tax base losing its capacity to repay therefore naturally increases. Lenders therefore tend to demand higher rates, further undermining their own investment as the government is forced to increase its borrowing to cover the difference, expanding the money supply and fueling inflation.
Towards the end of January, the US’s 10-year Treasury bond jumped from 4.14% to to 4.3%; the 20 from 4.73% to 4.88%; and the 30 from 4.79% to 4.92%. At the same time, Japan’s 40-Year bond broke through 4% for the first time since its introduction in 2007. The 30, having gone above 3% for the first time in 2025 since its introduction in 1999 — also moving above China’s falling 30-Year for the first time (down from 4.5% in 2018 to below 2%) — jumped from 3.6% to 3.9% (up from 1.4% in June 2023 and 0.1% in 2016). Government debt is getting harder to sell — less is being bought than the government wants to sell.
Interest payments grew from 6% of government spending in 2020 to 14% in 2025, making it the fastest growing component of state expenditure. A net 19 cents of every $1 (from 24c gross) in taxes collected now goes toward paying interest on public debt — roughly $7,300 per household — double the level in 2021. Net interest payments ($1.2bn gross minus payments owed to state departments) hit $970bn; compared to $345bn in 2020 and $180bn in 2005.
Because the state is rolling over maturing debt — borrowing new debt to pay off debt due to be repaid now — by issuing short-term debt in order to avoid the higher long-term rates, debt is maturing sooner. In 2026 the US government is due to repay about 30% of its $38 trillion debt, more than $9 trillion — nearly double the $5.2 trillion collected in taxes in 2025. The debt has to be repaid by borrowing at higher rates than the debt originally issued.
The £20bn extra brought in by tariffs is not touching the sides.
This state of affairs is clearly unsustainable. The tax base capital is so dependent on for bailouts, subsidies, contracts and sales is collapsing.
The Trump administration is naturally desperate to bring borrowing rates down and is making moves to deregulate banks — to lower the reserves they have to keep, to invest the difference in Treasuries — and pack the Fed with its own people in order to get its wish sooner rather than later; since the Fed has lowered rates slowly and tentatively given the negative reaction from bond markets that expect debt and inflation to rise as a result. The small downgrades the Fed has made have brought down short-term rates, but long-term rates and inflation have moved higher, since lower short-term rates allow the government to borrow and spend more money into existence. A continuation of that reaction would prevent Trump’s hopes of rolling over short-term debt into low-rate long-term debt.
Trump wants some devaluation of the dollar in order to boost the competitiveness of exports; but not so much that the demand for the dollar falls drastically. The Fed wants to keep interest rates higher than Europe’s to make lending to the US the more appealing option.
Official annual inflation is still near 3%, above the 2% target aimed at preventing deflation and maintaining business stability. Average US inflation hit a 40-year high of 9% in 2022. The average basket of goods is now 25% higher than in 2020. Coffee is up 21% in just the past year and ground beef 15%.
Overall monthly inflation is only ‘low’ because of falling oil prices, and when the latter falls too low to retain profitability, oil production will be cut in order to hike prices, something that would certainly negatively impact the stock market and economic growth — potentially very badly, depending on the extent and pace of cuts.
The average cost of operating existing wells in the US is $41 per barrel, and prices have fallen to around $60, with Trump attempting to bring them even lower by seizing foreign oil tankers and give private US multinationals the ownership of Venezuela’s largely state-owned oil industry (which needs a lot of expensive, long-term modernising that may not be worth the effort, since it provides very heavy, sticky and therfore discounted, low-margin oil; but which is the kind US Gulf Coast states have the ability to refine). Some analyses suggest break-even shale production — shale/tight oil being lighter and runnier than crude — presently averages about $70 per barrel and could rise toward $90–$95 over the next decade as core inventory depletes.
Fossil fuels have been vital for capitalism in the past century: they are extractive-intensive and non-renewable, continuously creating new labour time to exploit — whereas additive renewables, once built, do not. Limitless, abundant energy would be free and therefore unprofitable, meaning capitalism literally cannot wean itself off of fossil fuels (making the end of capitalism necessary for the survival of the human species).
New oil discoveries have tended to decline over time, though, as extraction has had to venture deeper and deeper underground to reach untapped reserves, making it increasingly expensive; while the quality of newly found and extracted oil has also declined, making refinement more and more expensive as well.
For now oil prices are falling and so the Fed is planning to drop its rates and print money to buy debt — possibly, given the absolute (debt-inflated) value of the economy is that much bigger, at an even greater rate than it did in 2008 and 2020. In the latter example the money supply grew by a shocking 40% in less than two years, although exceptional circumstances of course exacerbated the move.
At the start of December 2025 the Fed stopped tightening its balance sheet (as it had done since 2022 in its fight against inflation) and started reversing course. Sacrificing the housing market, instead of reinvesting in maturing mortgage-backed securities, proceeds will go solely into purchasing government debt in a desperate bid to bring down the cost of government borrowing.
The Fed has committed to buying $40bn in Treasuries per month to start with. It aims to suppress the long-term interest rates by making up for the lack of demand from Central Banks and foreign and private investors, thereby lowering the cost of refinancing the US government’s debt. Without making this move, the extra returns bond lenders are likely to demand as debt continues to rise could easily take rise from 4.5% to 6% (see below).
During WWII, the British government did the same, suppressing rates to 3% and then 2.5% for a year or so after — debasing pound sterling to the point that it lost its position as the world’s dominant reserve currency.
The Fed’s coming intervention will again devalue wages and savings, amounting to another, likely much greater round of relative inflation in terms of lost purchasing power and plummeting investment and supply — which would then compel the Fed to about-turn and agressively raise rates, as it did in 1981 (to an all-time high of 19%, which today would ‘feel’ much higher given that wages are relatively much lower and household debt is relatively much higher). The Fed is stuck between a rock and a hard place: we could be heading for very high or even hyperinflation followed by (hyper)deflation.
To play its part in sufficiently reincentivising investment — to whatever degree that remains possible — the capitalist state will have no choice but to limit bailouts to a smaller proportion of the private sector compared to 2008 and 2020 (when it did allow a portion to fail in order to limit debt growth and centralise wealth, re-enabling accumulation for the winners); and making even greater cuts to spending on welfare and public services.
Eventually, in its flailing attempts to counter falling profitability, the capitalist class and its state must bite off more than it can chew, as the costs of its warmongering — on competitors, state ownership, public wealth, labour, and consumers — will become completely unaffordable or completely destructive; stimulating an unstoppable process of socialist radicalisation that it will only continue to provoke; or destroying the habitability of Earth.
Death Knell
Liberals and social democrats will blame Donald Trump’s anti-migrant policies for exacerbating labour shortages; along with his tariffs — taxes on US importers and consumers which are impacting demand and draining reserves — but which aim to improve US capital’s profitability, by i) making domestic capital and exports more competitive; ii) strong-arming foreign corporations to move operations to the US; iii) compelling partners to accept trade deals that improve US margins on imports and exports; and iv) lowering a deficit that has been widening over several decades, in order to pull down interest rates.
China, though, seems largely unaffected, with its trade surplus reaching record highs since the escalation of tariffs in 2022. It controls too many critical resources and the close proximity and smart organisation of its domestic industry is largely not worth uprooting for a less efficient setup in the US or elsewhere. China’s exports to the US have fallen by only 6% so far and countries with entrenched supply chain linkages to China have seen the fastest export growth to the US, which therefore remains dependent on China, if not as directly.
Trump and his policies are symptomatic of deeper structural rot; and largely simply intensify the policies of his predecessors. Every US president since Bill Clinton in the early 1990s has rasied tariffs on Chinese imports. Eventually someone had to up the ante.
The reality is that capitalism unavoidably generates the conditions of its own crises and downfall.
Via bankruptcies and mergers and acquisitions — the recent deals between the likes of Nvidia and Open AI represent a gigantic proto-merger[2] — the ownership of capital and wealth necessarily concentrates in order to offset falling profitability; excluding an increasingly large proportion of the population from both the capitalist class and the proceeds of production. The dwindling capitalist class is therefore pushed back into a corner of its own making, from which it lashes out increasingly aggressively, stimulating an intensifying class struggle.
As capitalists continue to automate their operations in the effort to lower outlay and re-widen profit rates and margins, the contradiction at the heart of the system is increasingly aggravated. No wonder AI is unprofitable — data is produced and rearranged at increasingly breathtaking speed. Not to mention that capital accumulation requires capital-intensive solutions; while the relatively falling supply of skilled labor, due to education costs capital cannot afford to pay for (hence education’s privatisation), is making the skilled labour needed to produce and train AI increasingly pricey (producing, along with other scientists, who have usurped manufacturers as the helmsmen of production, a new and emerging proto ruling class of wealthy workers). Unlike humans, robots and computers cannot have their time exploited through wage labour for profit or buy commodities capitalists need to sell. The fully automated system of production capital itself reaches for is making capitalism historically obsolete and (global) socialism necessary.
Why socialism? Empirical economic data again makes the case clear: that concentration of capital through bankruptcies and mergers is demonstrated by the fact that, despite 50 years of aggressive privatisation, the number of private banks and corporations in the US has roughly halved since the turn of the century; while the overall proportion of startups has also declined. That trend goes back much further — since 1921 the number of US banks has fallen from 21,000 to 4,000 — but the merger trend has tended to accelerate since 1980 (following a brief period in the 1970s when private monopolies werebroken up to bring price down — not just because of government policy in part designed to fight inflation but because general wealth had risen, enabling a lot more people to start businesses competitive businesses. The same trend is evident everywhere, most notably in China’s private sector since the GFC.
The average age of the S&P 500 private enterprise, furthermore, has fallen from around 60 in the 1950s and 35 in the late 1970s to around 15 years today.
A ‘final merger’ in the not-too-distant future evidently beckons — to be enacted by a socialist state, since a capitalist state cannot by definition socialise/deprivatise the whole private economy — necessitating the transition to an economy owned entirely by the public, since no exchange of ownership is necessary in a total monopoly.
This conclusion is reinforced by the fact that the number of currencies in the world is trending towards zero — of the roughly 750 that have existed since 1700, less than 20% still remain — with the US dollar and UK pound both having lost nearly 100% of their purchasing power over the past century (mainly since the computing revolution kicked off; accelerating productivity exponentially), indicating the approaching necessity of a moneyless economy.
Private enterprise itself, furthermore, is increasingly ‘centrally planned’, having countered falling profitability with efficiency gains such as the elimination of ‘internal markets’ (competition between departments) and the introduction of automated, centralised databases collecting data in real-time from barcoded, traceable stock, erasing duplication and other inefficiencies (enabling decentralisation as all data is accessible to all parts of a network). Centrally auto-planning the economy as a whole is the next phase of this evolution.
A long, painful, unavoidable struggle awaits — but so too, at last, does working class and human liberation; not only from several hundred years of capitalist tyranny but thousands of years of private property.
Workers of the world — unite!
