A forces‑for‑2027 forecast looking to the end of the decade
In politics and markets, the decisive contest is often the one being played above the visible struggle. Elections, earnings and summit communiqués are the surface game—the real action lies in the metagame, the contest over rules, incentives and narratives that shape outcomes long after today’s skirmishes are forgotten.
Henry Kissinger, American diplomat and political scientist, warned that most statesmen get lost in the “manifestations of events”—daily crises, tactical choices and headlines—and fail to read the deeper “trend of events” that gives those crises meaning. Investors make the same mistake when they trade every move in yields or volatility but never ask which regime those data points are compounding into over a five‑ or ten‑year horizon.
German‑American entrepreneur Peter Thiel’s line that “competition is for losers” captures the business version of the same instinct. The point is not to fight hardest inside the old game, but to see how the game itself is changing and reposition before everyone else.
That lens is essential for understanding President Donald Trump’s America First Policy, the artificial intelligence (AI) driven economic regime now forming, and the likely path of capital markets into 2027 and beyond. The forcing function is debt. At extreme levels, debt ceases to be a background statistic and becomes the constraint that reshapes policy, politics and empire—much as British historian Edward Gibbon suggested fiscal exhaustion, debasement and obligations outrunning productive capacity were central to Rome’s decline.
The metagame is that Trump appears to understand this constraint. He is responding not with managerial patchwork, but with a modernized version of the sovereign‑development playbook used after the American Revolution—a blend of Hamiltonian statecraft and Henry Clay’s (American statesman) American System aimed at rebuilding domestic industry, energy and strategic capacity. The solution is a run‑hot economy in which earnings growth rises above historical norms, powered by AI‑led productivity gains and pro‑investment policy, producing a re‑rating of the U.S. economy rather than a speculative bubble. To treat tariffs, alliance disputes, energy policy, deregulation, AI capex or digital‑asset legislation as separate stories is to miss the metagame. The real question is what historical phase the U.S. is entering—and which assets benefit if that reading is right.
Debt as forcing function
The global metagame now turns on a blunt fact—the world has piled up more debt than its old growth model can comfortably service. Total global debt sits just above 235 per cent of world gross domestic product (GDP), with public debt near 93 per cent and still trending higher, according to International Monetary Fund estimates. Global public debt alone approached US$100 trillion in 2024.
No major electorate is volunteering for a lost decade of austerity, and no policymaker wants to trigger a synchronized refinancing crisis across sovereign, corporate and household balance sheets. The implication is simple—nominal GDP must outrun the effective cost of servicing the debt. That is why policymakers tolerate a hotter nominal economy, why austerity remains mostly rhetorical, and why the search for a credible productivity engine has become the central macro question. The system needs growth more than purification.
If the world is too indebted for clean liquidation, the policy premium shifts to whatever can raise output per worker, reduce frictions, improve capital efficiency and broaden the tax base without reigniting inflation. AI therefore, stops being a niche tech theme and becomes a macro variable. The question is whether AI‑enabled productivity can reconcile high debt with political stability.
From Pax Americana to America First
To see how this intersects with geopolitics, it is worth recalling what made Pax Americana unusual. After 1945, the U.S. possessed unmatched industrial, military and financial power, yet chose order‑building over annexation. Bretton Woods [1] created a rules‑based monetary system, the Marshall Plan [2] rebuilt former enemies, and American power underwrote global trade and security on terms far more generous than any previous hegemon would have offered. This was restraint at the moment of maximum leverage.
The post‑Second World War system depended not just on American power, but on America’s willingness to absorb costs others preferred not to acknowledge. Over time, surplus economies, allied free‑riders and global capital learned to rely on the American balance sheet and Navy. British economist John Maynard Keynes warned that any system becomes unstable when the burden of adjustment falls too heavily on one side—that is what happened as the U.S. hollowed out parts of its industrial base, accumulated debt and watched allies treat U.S. commitments as an entitlement rather than a bargain.
“America First” is best read as the political expression of a system that had stopped pricing hegemony honestly. It is also, in part, a modern derivative of Clay’s early 19th-century American System, a program of protective tariffs, internal improvements and national development designed to build domestic industry and reduce dependence on foreign powers. Trump’s economic and national security strategy makes sense at the metagame level. The point is not that every tariff is sacred or every alliance expendable. It is that the U.S. can no longer remain the world’s consumer of last resort, security guarantor of first resort and fiscal shock absorber at a discount. Hegemony is being repriced.
That repricing operates through several channels. Trade policy becomes less about abstract efficiency and more about repatriating strategic supply chains. Alliances are subjected to harsher burden‑sharing demands. Energy policy is reoriented towards domestic production rather than imported virtue. Underneath the noise, the message is simple—late‑imperial America has over‑promised against a weakened productive base, and the only way to sustain great‑power obligations is to rebuild that foundation.
The decline and fall of the Roman Empire in the age of Trump
Edward Gibbon’s Decline and Fall of the Roman Empire offered a brutal taxonomy of late‑imperial decay: a state living beyond its means, a weakening stock of productive capital and a people drifting from citizens into dependents. Measured against those categories, contemporary America looks uncomfortably Roman—and Trump’s America First Policy reads as a visceral response to that diagnosis.
Gibbon’s Rome does not collapse in a single catastrophe—it sags. Obligations accumulate faster than wealth, coin is shaved, taxes proliferate and extraordinary levies become routine. The governing class refuses to live within society’s real productive capacity. Debt and debasement are not clever tools but visible marks of a deeper refusal to choose.
Alongside this fiscal drift, the productive core hollows out. Independent citizens give way to dependents of landlords, bureaucracies and patronage networks; local initiative weakens; and more of society lives as recipients rather than producers.
The third element of Gibbon’s framework is psychological. Politics becomes spectacle, elite energy is consumed by factional quarrels, and the public comes to see the state less as something it upholds than as something that exists to uphold it. “Bread and circuses” is not merely a moral jab—it is the structure of a society accustomed to being maintained.
Trump’s instinct starts from the recognition that the U.S. has drifted into its own late‑imperial configuration— federal debt towering over output, shuttered factories, hollowed‑out towns, a stretched military and a capital that can always fund the next foreign deployment but struggles to license the next refinery. America First treats excessive debt not as a technocratic nuisance but as a verdict on a model that imports its goods, outsources its energy, subsidizes dependence and borrows to hide the contradictions.
The conclusion is not that America must shrink to fit its weakened base, but that the base must be rebuilt so that America can remain what it is. Hence the relentlessly material focus of Trump‑era economics—deregulation as a solution to permitting bottlenecks that block wells, mines, ports, grids and factories; onshoring as an effort to drag productive capital back inside the political community; and growth through ships, rigs and fabs rather than abstractions in a spreadsheet.
The same logic informs his rough handling of alliances and trade. For decades, Washington piled up security commitments without asking whether a de‑industrialized, energy‑importing economy could support them indefinitely. Demands for fairer burden‑sharing, rebalanced trade and meaningful borders all reflect one Gibbonian instinct—late empires fail when promises outrun capacity. Better to renegotiate now than to default later, economically or strategically. Trump is not Gibbon—he operates on instinct, not footnotes. But the Gibbonian lens clarifies what that instinct is reacting against—and what it is trying, clumsily, to repair.
From AI to the productivity regime
Into this late‑imperial context comes AI. Much commentary still treats AI as a speculative theme akin to social media or early cloud computing. That is too shallow. AI is emerging as a general‑purpose technology and a macro regime variable: a source of economy wide productivity gains, a driver of physical capex and a justification for a policy mix that favours supply expansion over demand micromanagement.
Supply‑side economics, properly understood, is not a relic. It is the proposition that a nation escapes debt and stagnation not by endlessly manipulating demand, but by expanding its capacity to produce: more output per hour, productive capital, innovation, energy and logistical efficiency. Agentic AI sits near the centre of that thesis.
Unlike earlier software waves that accelerated narrow workflows, agentic systems can initiate, coordinate and complete multi‑step tasks with limited human supervision across coding, research, logistics, operations and customer‑facing functions. That makes AI closer to electrification or the microprocessor than to the latest consumer app. General‑purpose technologies diffuse through the capital stock over years and require complementary investment, retraining, energy systems and organizational redesign.
The plausible bull case is not that AI instantly abolishes labour or produces overnight utopia. It is that agentic AI lifts the long‑run ceiling on productivity and keeps investment demand high through the end of the decade, precisely when debt dynamics require faster nominal and real growth.
Bottlenecks, Taiwan and the cycle’s duration
The strongest argument against the “AI bubble” narrative is not rhetorical but physical. Taiwan remains the narrowest point in the global AI value chain. Taiwan Semiconductor Manufacturing Company Limited (TSMC) dominates leading‑edge semiconductor fabrication, with some estimates putting its share of the most advanced chips used in AI workloads above 90 per cent. That concentration is obviously a geopolitical risk. Any blockade, conflict or coercive disruption around Taiwan would trigger a severe supply shock.
But it is also why this AI cycle is likely to last longer than past tech booms. The hardware constraint means the capex wave cannot be fully front‑loaded. The combination of lithography, materials, packaging, engineering depth and supplier ecosystems clustered around Hsinchu cannot be recreated in a few quarters. This is not fibre in empty office parks. It is an industrial build‑out requiring fabs, substations, turbines, gas infrastructure, transmission, cooling, advanced materials and engineering talent.
The bottleneck stretches the cycle by forcing a sequencing of adoption. Productivity gains arrive in waves as compute, power and enterprise integration expand. The right interpretation is not late‑cycle mania but early‑stage diffusion under hard constraints. Bottlenecked revolutions usually last longer because the constraint rations upside over time.
Energy, security and a new peace dividend
None of this works without abundant energy. Training and deploying large AI systems is power intensive and the supporting ecosystem—data centres, grids, fabs, industrial logistics—is even more so. Energy policy thus becomes metagame strategy. Cheap, secure and scalable power is no longer just a utility issue; it is industrial policy, inflation policy and national‑security policy at once.
Here, America First’s instincts, often mocked, are strategically coherent. A nation that controls hydrocarbons, generation, transmission capacity and critical minerals, enjoys a structural advantage over one that prioritizes scarcity while relying on fragile supply chains. Energy dominance is not a slogan. It is the precondition for AI dominance, manufacturing resilience and credible deterrence.
There is also a second‑order market effect—a harder version of the peace dividend. Not the 1990s fantasy that history has ended, but a rugged peace grounded in deterrence, alliance burden‑sharing and control of chokepoints. If American strategy reduces disorder at the margin—keeping the Strait of Hormuz open, securing semiconductor and energy routes, pushing allies to shoulder more of their own defence and deterring escalation—then capital markets gain something invaluable; fewer catastrophic tails. Markets do not need utopia. A rough peace underwritten by strength and energy abundance is enough to compress risk premia.
The policy triumvirate
The metagame is no longer just an intellectual frame—it is acquiring institutional form. The most important development in American macro policy is that the people shaping money, fiscal strategy and market psychology are increasingly aligned on supply‑side logic.
Kevin Warsh, U.S. Federal Reserve Board Governor, has long argued that the Federal Reserve grew too large, too discretionary and too entangled in fiscal and political tasks that should never have been delegated to a central bank. His critique of post‑2008 quantitative easing was philosophical: an overgrown Fed distorts capital allocation, socializes risk, weakens price signals and tempts politicians to dodge structural reform.
Scott Bessent, U.S. Secretary of the Treasury, brings a complementary worldview, stressing fiscal discipline, energy expansion and a shift from transfer‑heavy programs towards productive investment and supply side reform. Former hedge fund manager Stanley Druckenmiller, though outside office, supplies the Wall Street counterpart. He has long warned that cheap money and reflexive bailouts misprice risk and degrade capital allocation, and he has increasingly emphasised AI infrastructure, chips and enabling hardware over narrative‑driven speculation.
What unites this informal triumvirate is an agreement on first principles: supply‑side reform is essential in a debt‑heavy world, the Fed should be smaller and more focused, and policy should raise the after‑tax, after‑regulation return on real investment in energy, semiconductors, logistics and AI infrastructure. That alignment matters because it changes the regime. Previously, new technologies often advanced against policy that rewarded financial engineering over real capex. In the emerging regime, the bull case is no longer fighting Washington—it is increasingly embedded in it.
Markets, doomers and the AI stack
Wall Street’s narrative remains haunted by the last crisis. In quick succession, investors were sold a private credit crisis, told software was structurally doomed by AI, and handed a Middle East scenario built around a catastrophically shut Strait of Hormuz. Reality was less theatrical. Credit bent but did not break. Software adapted instead of evaporating. The strait stayed open. Risk assets absorbed the shock, repriced and moved on. The professional doom industry did not revise its framework—it merely changed costumes.
Performative pessimism is not prudence. It is click bait for those who confuse anxiety with insight. Serious investors should expect mess, shocks and volatility as part of the game. But they should also recognize that in a regime shaped by AI‑led productivity, supply‑side reform, energy abundance and a rough peace dividend, periodic drawdowns are more likely to be resets inside an extended expansion than preludes to systemic collapse.
The key is to stop treating AI as a single trade and start seeing it as a stack. Behind every frontier model is a data centre; behind every data centre is a chain of chips, servers, memory, networking, power, buildings, cooling and transmission. The investor’s job is not to worship one charismatic CEO. It is to own the bottlenecks the new regime cannot function without.
The hot‑run thesis for 2031
This leads to the question that animates markets—what earnings and multiples can this regime plausibly support? The answer looks less like the post‑Global Financial Crisis stagnation era and more like a modernized version of the late 1980s and 1990s, when growth, productivity and valuation rose together. Then, the S&P 500’s trailing price/earnings ratio (P/E) moved from roughly 10.4 x in 1985 to about 34 x by early 2000, while the peak forward P/E reached roughly 24.4 x in March 2000.
In this story’s sequel, Trump runs the economy hot, nominal GDP grows near seven per cent a year, and S&P 500 earnings grow not one‑for‑one with nominal GDP, but at roughly 1.5‑2 x that rate, powered by AI‑driven productivity and investment incentives such as full capex expensing. If earnings are around US$400 in 2027, that regime produces roughly US$600‑650 by 2031, implying 10.5‑14 per cent annual earnings per share growth.
From there, three valuation paths sketch the range of outcomes. In the base case, the market believes the story but stops short of mania. Earnings reach about US$600 and investors pay around 22 x forward earnings—rich by long‑run standards but below the 2000 forward peak—delivering an S&P 500 in the low to mid 13,000s. In the strong bull case, investors become convinced this is a durable supply‑side boom. Earnings move to US$625‑650 and the market pays roughly 25 x trailing earnings, nudging the index towards 15,000‑16,000.
In a euphoric 1999‑style case, earnings still land in the US$625‑650 band but the multiple stretches towards 30 x, taking the S&P 500 into the 18,000‑19,500 range. Dividend reinvestment lowers the bar. With a two per cent yield reinvested each year, part of total return comes from compounding income rather than price alone, making a 15,000‑16,000 destination easier to reach on a total return basis.
By contrast, a consensus that sees the S&P 500 only at 10,000 by 2031 implies a very different world. From roughly 7,500 today, 10,000 in five years is about six per cent annual price growth. With a flat multiple, earnings growth must also be about six per cent—add a two per cent dividend yield reinvested, and total return rises to roughly eight per cent annually. Implicitly, that view assumes only four to five per cent nominal GDP growth, no structural re‑rating and no belief that supply‑side reform and AI have changed the game.
A repricing, not a bubble
The baseline expectation into 2027 on the hotter thesis looks very different. The metagame is the American system updated for the AI age: run the economy hot, rebuild productive capacity and let productivity do the heavy lifting. On that view, earnings growth should run above long‑run averages, driven by AI‑led productivity, capex and broader operating leverage rather than by financial engineering alone. Multiples can remain elevated or expand as the market discounts a higher‑growth, lower tail‑risk regime rather than a replay of secular stagnation. Leadership broadens from early AI winners into energy, industrials, infrastructure, logistics, semiconductors, selected software and financial rails tied to the new productive architecture.
On that framework, a path approaching 10,000 on the S&P 500 by 2027 and 14,000 by 2029 is aggressive but not deranged. It requires above‑trend earnings growth, multiple support from lower perceived tail risk and sustained conviction that this is a re‑rating of productive capacity rather than a fantasy bubble. The point is not to predict a straight line. It is to recognise that if productivity accelerates while risk premia compress, historical valuation anchors will look too low.
That is why this episode should be understood as a rerating, not a bubble. If Wall Street’s models and beliefs remain anchored in the secular‑stagnation era, it will be true to form for strategists to label earnings above trend as speculative excess. But the only real bubble may be in that mislabelling. Bubbles are built on imagined cash flows, absent capacity and an eventual collision with hard constraints. This cycle is being driven by hard constraints—power, fabs, engineering talent and geopolitical chokepoints—and policy choices designed to ease them. The market is not pricing a hallucination—it is beginning to price a new era.
Looking to the end of the decade
By the end of the decade, the central question will not be whether bears invent another alarming acronym. It will be whether the U.S. successfully translates its structural advantages—deep capital markets, energy resources, leading software and chip design, military reach and institutional flexibility—into a new productivity era. If it does, debt becomes more manageable through growth, earnings rise faster than a generation trained in the 2010s expects, and multiples remain above historical averages because the economy has shifted onto a better nominal and real growth path.
This does not mean every stock is sensible or every moment is safe. It means the dominant error of the era is likely to be underestimating duration. Investors remain haunted by past bubbles and struggle to recognize when a genuine re‑rating is underway because they are still using the wrong mental model.
That is the metagame point. Most participants remain trapped in the manifestations of events—the next sell‑off, geopolitical scare or doom thread. The deeper trend points elsewhere: towards a supply‑side, productivity‑led, AI extended, geopolitically repriced American expansion that can run further into the end of the decade than historical norms suggest.
Thiel’s line lands here as well. Competition is for losers in the sense that the serious task for capital is not to fight harder inside the daily noise, but to recognize when the game itself has changed. Many investors make the mistake of investing in the present. If you only respond to today’s noise, your returns will suffer. Forcing yourself to look out 12 to 24 months compels you, almost by definition, to think about the metagame: what forces are driving the trends, whether the rules are changing, and whether structural shifts are underway. The simple lesson is to look forward rather than stay anchored to the past. The right stance is neither euphoria nor complacency, but recognition—ignore the spectacle and accept that for long‑term investors who are willing to play the metagame, the opportunity in this new America‑system world is not merely to survive the transition, but to finance the rebuild.
[1] 1944 agreement which shaped the global financial system pegging currencies to the U.S. dollar.
[2] American initiative providing US$13 billion in aid to help rebuild Western Europe following the Second World War.
