I believe the defining investment question of the coming years is not simply whether the economy will grow, whether inflation will rise or fall, or whether stocks are expensive. The deeper question is whether the global financial system can absorb the simultaneous collision of enormous government borrowing, declining marginal demand for sovereign debt, a historic capital-expenditure boom in artificial intelligence, changing global capital flows, and an increasingly consequential monetary response.
These forces do not operate independently. They interact.
Government deficits create more bonds. A growing supply of bonds requires sufficient buyers. When traditional buyers become less willing or less able to absorb that supply, yields must rise unless some other mechanism intervenes. Higher yields increase the government's interest burden, which increases deficits, which requires additional borrowing. At the same time, artificial intelligence is generating an enormous investment cycle that is itself competing for scarce global capital. If AI produces the productivity gains expected of it, corporate profits can accelerate dramatically. Yet that same AI investment boom may also intensify the demand for capital and place additional upward pressure on interest rates.
The paradox is that the technology most capable of generating a new productivity boom may simultaneously contribute to the financial conditions that eventually force governments and central banks into a more interventionist monetary regime.
That is the framework I use to understand the environment.
The Sovereign Bond Market Is Becoming the Central Constraint
I start with the Treasury market because the Treasury market is ultimately the pricing mechanism for the world's most important reserve asset and benchmark interest rate.
The basic imbalance is straightforward: the supply of sovereign debt is expanding faster than the natural demand for that debt. The problem becomes particularly acute when geopolitical and structural forces cause foreign capital to become less willing to finance the United States through Treasury securities.
This is not merely an issue of government accounting. It is a capital-allocation problem.
A government can issue debt indefinitely only if investors remain willing to hold it at acceptable yields. When the supply of debt increases faster than the pool of savings available to purchase it, the price of that debt must adjust. For bonds, lower prices mean higher yields.
This creates an uncomfortable feedback loop.
Higher yields make government borrowing more expensive. Higher interest expense increases the deficit. A larger deficit requires more debt issuance. More issuance increases the amount of capital required from investors. If demand does not increase proportionately, yields must rise further.
At some point, monetary policy ceases to be the only variable that matters. Fiscal arithmetic begins to dominate the financial system.
I therefore think it is increasingly important to distinguish between a conventional monetary-dominance environment and a fiscal-dominance environment. In a conventional cycle, the central bank can raise rates to restrain demand or lower them to stimulate demand. In a fiscal-dominance environment, the central bank must also operate within the constraints created by the government's debt burden.
That changes the reaction function.
The Three Broad Ways Out of a Debt Problem
When a sovereign government accumulates an excessive debt burden, I see three broad economic strategies.
The first is austerity: reduce spending, increase taxes, or otherwise shrink the deficit.
The second is growth: attempt to grow nominal GDP rapidly enough that the debt burden becomes more manageable relative to the size of the economy.
The third is monetary debasement: allow inflation and currency depreciation to reduce the real value of outstanding obligations.
The first approach is politically difficult. The second is attractive because it appears to solve the problem without directly imposing austerity. The third is politically easier initially because the adjustment occurs indirectly through prices and purchasing power.
But monetary debasement is not costless.
If the government attempts to inflate its way out of excessive debt, it can reduce the real burden of nominal liabilities. Yet the resulting inflation simultaneously reduces the purchasing power of households and changes the distribution of wealth between debtors and creditors.
The system therefore tends to move toward increasingly unconventional financial repression when conventional fiscal adjustment becomes politically or economically difficult.
That can include regulatory changes designed to encourage banks and financial institutions to hold more government debt, changes to capital requirements, central-bank balance-sheet expansion, Treasury-management operations, and eventually explicit or implicit yield-curve control.
The crucial point is that these mechanisms do not eliminate the underlying scarcity of capital. They redistribute the risk.
The Interest-Expense Trap
One of the most important measures I would monitor is not simply federal interest expense, but interest and interest-like obligations relative to government receipts.
The source material cites a measure of "true interest expense" that includes interest together with major entitlement-related obligations. It places that measure at approximately 105% of receipts during a period when the economy was still relatively strong.
Whether one accepts every component of that methodology or not, the conceptual insight is important.
A government can tolerate a large debt burden when nominal economic growth, tax receipts, and interest costs remain favorably aligned. But once interest-related obligations consume an increasingly large share of receipts, the fiscal system becomes highly sensitive to interest rates.
This is why inflation can temporarily appear to solve a debt problem.
During the pandemic period, the combination of extremely low policy rates, large-scale central-bank bond purchases, and substantial inflation caused nominal receipts to increase while interest costs were temporarily compressed. The cited framework describes true interest expense falling from approximately 120% of receipts during the crisis to roughly 80–85% afterward.
That did not eliminate the structural problem.
It bought time.
This distinction matters enormously for investors. A policy response that buys time is not necessarily a policy response that solves the problem. The financial system can survive a dangerous configuration for years if policymakers repeatedly intervene. But each intervention can increase the size of the imbalance that must eventually be managed.
The system becomes increasingly dependent on nominal growth, inflation, and monetary accommodation.
The Part of the Debt That Cannot Be Printed
There is an even deeper constraint.
It is common to say that the United States cannot default because it issues debt in its own currency. In a narrow sense, that is true for nominal Treasury obligations. But the government's broader obligations are not simply pieces of paper denominated in dollars.
Social Security benefits are indexed. Medicare and Medicaid require actual medical services. Veterans' benefits require healthcare. The government cannot manufacture physicians, pharmaceuticals, hospital capacity, or human labor by creating dollars.
That distinction is economically important.
A government can create nominal currency. It cannot create unlimited real resources.
The ultimate constraint on monetary policy is therefore not the government's ability to create dollars. It is the availability of real goods and services that those dollars attempt to command.
This is why excessive monetary expansion can ultimately produce inflation rather than solvency.
The government can create more financial claims on the economy. It cannot necessarily create more of the underlying economy.
Why Higher Interest Rates Could Persist
I do not assume that higher interest rates automatically produce recession.
That conventional relationship becomes more complicated when the economy itself is undergoing a major capital-investment boom.
The demand for capital can rise substantially because businesses are investing in data centers, computing infrastructure, energy systems, software, semiconductor capacity, electrical infrastructure, and other assets required for artificial intelligence.
In economic terms, the marginal demand for capital is rising.
If the supply of savings does not increase proportionately, the equilibrium real interest rate—the rate that balances the marginal supply and demand for capital—can rise.
This is the concept of r-star.
The source material describes estimates suggesting that the equilibrium real-rate range has risen materially over the preceding several quarters, while the inflation-adjusted policy rate remained below the estimated lower bound of that range.
The implication is significant.
If the economy has a higher structural demand for capital, central banks cannot necessarily force long-term rates permanently lower simply by lowering short-term policy rates.
The long end of the yield curve has its own supply-and-demand dynamics.
This is one of the most important distinctions I make when analyzing monetary policy: the Federal Reserve controls the overnight policy rate far more directly than it controls the long-term cost of capital.
The market ultimately determines the price of duration.
AI Changes the Capital Equation
Artificial intelligence is therefore not merely a technology story.
It is a macroeconomic capital-allocation story.
The AI buildout is creating an extraordinary investment cycle. Capital is flowing into computing, semiconductors, data centers, electricity generation, transmission, cooling, networking, software, and associated infrastructure.
The source material characterizes this as one of the largest capital-expenditure booms in American history and potentially the largest relative to GDP among the major historical infrastructure waves discussed.
I find this particularly important because capital expenditure has two effects that can appear contradictory.
In the short term, enormous capex increases aggregate demand. Companies purchase equipment, construct facilities, hire workers, order components, and finance infrastructure.
In the longer term, the accumulated capital stock can increase productivity.
This is capital deepening.
When businesses possess more productive capital per worker, the productive capacity of the economy can increase. The source material points to equipment, R&D, and software investment relative to employee compensation as a measure of capital deepening, with that ratio reaching an cited high of approximately 22.3%. Historical comparisons in the source associate previous episodes of substantial capital deepening with sustained productivity accelerations in some major cycles.
That creates one of the most bullish possibilities in the entire framework.
If AI investment produces a sustained productivity acceleration, the economy may experience substantially faster potential growth than it has experienced in recent decades.
Productivity Is the Bull Case
The most powerful argument for remaining constructive on risk assets is not monetary debasement.
It is productivity.
If the economy moves from something like a 1–2% trend productivity environment toward a 3–4% trend productivity environment, the consequences could be profound.
Productivity increases the amount of economic output generated from a given quantity of labor and capital. That can raise corporate margins, expand potential GDP, increase corporate profits, and support higher valuations.
The source material estimates that such an acceleration could produce roughly a 50% increase in the trend rate of nominal corporate-profit growth.
That is a powerful investment thesis.
It means that investors can potentially be bullish on equities even while being deeply concerned about the Treasury market.
These are not necessarily contradictory positions.
The equity market represents ownership of productive businesses. Treasury securities represent claims on future government payments. If productive capital becomes dramatically more valuable while government liabilities are increasingly diluted through inflation or monetary intervention, scarce productive assets can outperform nominal fixed-income claims.
This is why I do not treat "bullish equities" and "bearish bonds" as mutually exclusive positions.
They can be two sides of the same macroeconomic regime.
The K-Shaped Economy
The distributional consequences are equally important.
Inflation does not affect every household equally.
A household that owns equities, real estate, businesses, commodities, or other scarce assets can potentially see the nominal value of those assets rise alongside inflation. A household whose balance sheet consists primarily of wages and cash may experience the opposite effect.
This creates a K-shaped economy.
The upper branch consists of households and businesses owning appreciating financial and real assets. The lower branch consists of households whose income and savings fail to keep pace with the rising cost of living.
The problem becomes more complicated when government borrowing absorbs an increasing share of available savings.
The source material cites Treasury debt supply relative to global and U.S. savings at levels substantially above long-run averages and argues that capital that historically flowed toward housing, automobiles, consumer credit, and small businesses is increasingly being absorbed by government financing and the owners of government securities.
This is a capital-allocation issue.
Savings are finite.
Every dollar absorbed by one part of the financial system is a dollar that cannot simultaneously finance another activity.
If government borrowing absorbs more of the marginal savings pool, private-sector borrowers must either pay more for capital or receive less capital.
That can disadvantage smaller businesses and lower-income households relative to asset owners.
The Global Dollar Problem
The United States also exists within a global dollar system.
Foreign borrowers have accumulated enormous dollar-denominated liabilities. At the same time, foreign investors hold substantial quantities of dollar assets.
This produces a complicated relationship between the dollar and U.S. interest rates.
A stronger dollar can increase the burden of dollar-denominated debt outside the United States. Foreign borrowers and institutions may therefore be forced to sell dollar assets to obtain dollars.
A weaker dollar can relieve some of that pressure by reducing the real burden of dollar liabilities and supporting global balance sheets.
But an excessively rapid dollar decline creates a different problem.
It can raise inflation expectations and term premiums, increasing long-term Treasury yields.
This means policymakers can face a difficult optimization problem: a weaker dollar may support global liquidity, but a disorderly decline can undermine confidence in the dollar and increase the inflation premium embedded in long-term interest rates.
The relationship is therefore reflexive rather than linear.
Why Stablecoins Are Not a Simple Solution
One of the most interesting financial questions is whether dollar stablecoins can create a new structural source of Treasury demand.
The logic is superficially compelling.
If stablecoins are backed by short-term Treasury bills, then widespread adoption could create additional demand for Treasury securities. Bill issuance could therefore increase without necessarily requiring conventional investors to absorb the entire supply.
But I see a fundamental limitation.
A stablecoin does not create savings out of nothing.
If a stablecoin becomes popular, the capital backing it must come from somewhere. It may represent existing savings moving from bank deposits into digital dollars. It may represent foreign savings moving into dollar instruments. It may represent a change in the form of existing dollar assets.
But the stablecoin itself does not manufacture the underlying purchasing power.
This is particularly important when considering foreign adoption.
Billions of people may prefer a stable digital dollar to an unstable local currency. But currency demand and savings capacity are not the same thing.
A person can desire dollars without possessing substantial financial assets to invest in Treasury securities.
That distinction separates a potentially revolutionary payments technology from a solution to a sovereign debt problem.
Stablecoins may dramatically improve the transmission and accessibility of dollars around the world. They may expand the dollar's digital reach. They may improve settlement and create new financial infrastructure.
But they cannot automatically create the trillions of dollars of net savings required to resolve a structural Treasury supply-demand imbalance.
The Stablecoin Paradox
There is also a second-order problem.
Suppose stablecoins become heavily backed by Treasury bills and foreign users demand enormous quantities of dollar stablecoins.
The result could be stronger demand for dollars.
A stronger dollar can tighten global liquidity because dollar-denominated borrowers outside the United States must acquire more dollars to service their obligations.
That can produce forced selling of other dollar assets.
Foreign investors hold substantial quantities of U.S. Treasuries and equities. If dollar shortages force them to sell those assets, the resulting market decline can reduce capital gains and therefore reduce tax receipts.
That creates another fiscal feedback loop.
The apparent solution to the Treasury problem could therefore create a liquidity shock that ultimately increases the deficit it was intended to alleviate.
This is why I regard stablecoins as potentially transformative financial technology but not necessarily a silver bullet for sovereign debt.
Financial Repression and the Declining "Moneyness" of Bonds
Another mechanism deserves particular attention: financial repression.
Financial repression occurs when policymakers use regulation, monetary policy, institutional incentives, or other mechanisms to encourage investors to hold government debt at yields below what an unconstrained market might demand.
This can work for a while.
But investors are not passive.
If the expected real return on long-duration government bonds becomes sufficiently unattractive, private investors will seek alternatives.
The source material identifies an important change in Treasury ownership: the Federal Reserve, commercial banks, and foreign official institutions have reduced their relative shares of the market, while the global private non-bank sector has absorbed a much larger proportion of the risk.
This matters because private investors are economic actors.
They care about expected returns.
A central bank can buy bonds for monetary-policy purposes. A commercial bank can hold Treasuries because regulations make them useful. A foreign central bank can hold Treasuries as reserve assets.
A private investor asks a different question:
"What is my expected risk-adjusted return?"
If that investor concludes that the real return is inadequate, the investor can sell.
This is why financial repression becomes progressively more difficult as private investors assume a larger share of the marginal Treasury risk.
The Life-Insurance and Private-Credit Problem
The bond market may also contain a less obvious source of instability through the relationship between private credit and institutional investors.
Life insurers and pension funds have historically represented natural buyers of long-duration assets. Yet if these institutions have accumulated substantial private-credit exposure, their ability to rotate from private credit into Treasuries can be constrained.
The problem is valuation.
If private credit is carried at marks that are not representative of what the assets could actually fetch in a stressed market, selling those assets to purchase Treasuries can force losses to be recognized.
That creates a form of institutional paralysis.
An institution may theoretically prefer Treasuries at higher yields but still be unable to sell existing private-credit assets without damaging its capital position.
If private-credit valuations subsequently deteriorate, the institution may instead need to sell highly liquid assets such as Treasuries and mortgage-backed securities to raise capital.
That can produce a nonlinear—or convex—move in Treasury yields.
In other words, the next Treasury-market adjustment does not necessarily have to be gradual.
When liquidity disappears, price can move much faster than fundamentals alone would suggest.
The Treasury Market and Yield-Curve Control
Eventually, policymakers face a difficult choice.
They can allow market-determined long-term yields to rise.
They can increase regulatory incentives for banks and institutions to hold Treasuries.
They can manipulate Treasury issuance and maturity structure.
They can provide liquidity facilities.
Or they can directly intervene in the bond market through the central bank's balance sheet.
The final mechanism is essentially yield-curve control: the government or central bank establishes a target or ceiling for particular interest rates and commits sufficient balance-sheet capacity to defend that target.
At that point, the bond market is no longer fully determining the price of government debt.
The monetary authority is.
This is an important regime change.
Yield-curve control does not necessarily mean immediate hyperinflation. But it does mean that policymakers have chosen to prioritize the stability of government financing conditions over completely free-market price discovery in the bond market.
That is why I view yield-curve control as a critical potential dividing line between the current environment and a more explicit monetary-debasement regime.
Why Gold Becomes More Important
Gold occupies a unique position in this framework because it is not another government's liability.
It is a scarce asset without a promise from a sovereign issuer to make future payments.
That distinction becomes increasingly valuable when investors begin questioning the real value of nominal claims.
In a conventional inflationary cycle, rising interest rates can make gold less attractive because higher real yields increase the opportunity cost of holding a non-yielding asset.
But in a fiscal-dominance regime, that relationship can break down.
If rising rates are interpreted as evidence of fiscal stress rather than simply tighter monetary policy, gold can rise alongside bond yields.
That is a fundamentally different regime.
The relevant question becomes not "Are rates rising?" but "Why are rates rising?"
If rates rise because the economy is experiencing strong productivity growth and real investment demand, the implications differ from a situation in which rates rise because investors demand compensation for increasing sovereign fiscal risk.
The distinction between nominal yields and real purchasing power becomes essential.
Equities Can Rise While Losing Real Value
One of the most important investment lessons is that nominal performance can be profoundly misleading during monetary debasement.
An equity index can rise dramatically in dollar terms while losing purchasing power relative to gold or another scarce asset.
This is not necessarily a contradiction.
If the supply of dollars expands faster than the supply of productive or scarce assets, the nominal price of those assets can rise even if their relative real value is unchanged.
The source material uses the concept of "stocks up in dollars but down in gold" to describe this phenomenon. It cites the example of the S&P 500 producing a very large dollar gain over an extended period while underperforming gold over the same period.
This is why I believe investors should occasionally evaluate returns in multiple units of account.
The dollar is a measuring instrument, but the measuring instrument itself can change in purchasing power.
If I own an asset that rises 20% while the currency in which it is measured loses 15% of its purchasing power, the nominal return substantially overstates the improvement in my real wealth.
The AI Boom May Become a Bubble—and Still Be Real
I see no contradiction between believing in AI and believing that AI equities can become excessively valued.
Technological revolutions frequently generate both genuine productivity improvements and speculative excess.
The two can coexist.
Railroads transformed the economy and produced speculative bubbles.
Telecommunications transformed the economy and produced speculative bubbles.
The internet transformed the economy and produced speculative bubbles.
A genuine technological revolution does not protect investors from overpaying for its beneficiaries.
That distinction is especially important now because AI is producing a massive capital cycle.
If earnings continue accelerating, equity prices can continue rising even when valuations appear stretched. The fundamental story can remain intact while the investment returns become increasingly dependent on expectations.
This is where I become more selective.
I would rather own the infrastructure required for the AI economy than assume every high-profile AI company will remain an attractive investment at every valuation.
Electricity Is a Bottleneck
One of the clearest infrastructure implications is electricity.
The United States experienced a remarkably long period of limited growth in electricity-generation capacity. At the same time, AI requires enormous quantities of computing power, and computing power ultimately requires electricity.
That creates a bottleneck.
The AI economy is therefore not purely digital.
It requires physical infrastructure.
Data centers require electricity, transmission, cooling, land, construction, networking, semiconductor equipment, and increasingly sophisticated power management.
This makes electrical infrastructure a potentially attractive way to participate in the AI capital cycle without assuming the full valuation risk of the most aggressively priced AI companies.
The broader investment principle is simple:
When a technological revolution accelerates demand for a scarce physical input, the infrastructure supplying that input can become a critical investment opportunity.
The Investment Regime I See
I therefore do not believe the appropriate response to these risks is simply to retreat into cash.
That would ignore the possibility that the same forces creating systemic risk can generate extraordinary gains in scarce assets before the eventual adjustment arrives.
Instead, I think in terms of regime exposure.
The bullish case contains several reinforcing mechanisms.
AI capex increases aggregate demand.
Capital deepening increases productive capacity.
Productivity growth increases corporate profitability.
Higher nominal growth supports revenues and earnings.
Monetary accommodation can increase liquidity.
Currency debasement can raise nominal asset prices.
Scarce assets can outperform nominal fixed-income claims.
This combination can produce an environment in which equities, gold, Bitcoin, and selected infrastructure assets all appreciate substantially before the financial system reaches the point at which policymakers are forced into more aggressive intervention.
That is the paradox of a late-cycle inflationary regime:
The environment can become more dangerous while markets continue rising.
The Portfolio Implication
The source framework describes a portfolio constructed around several broad buckets: cash, gold and gold miners, electrical infrastructure equities, Bitcoin, and diversified large-cap equities. One cited allocation example was approximately 15% cash, 40% gold and gold miners, 15% electrical infrastructure, roughly 5–7% Bitcoin, with the remainder in large-cap equities.
I would interpret that allocation less as a universal prescription than as an illustration of a broader principle.
The portfolio is designed to participate in appreciation while maintaining optionality.
Cash is not necessarily a bearish position.
Cash can be an option.
If volatility rises sharply, cash allows me to buy assets when other investors are forced to sell. If the bond market becomes disorderly, optionality becomes more valuable.
Gold serves a different function. It provides exposure to monetary debasement and declining confidence in sovereign paper.
Equities provide exposure to productivity, nominal growth, corporate earnings, and technological transformation.
Electrical infrastructure provides exposure to the physical bottlenecks created by AI.
Bitcoin provides exposure to a scarce digital asset whose supply is not controlled by a sovereign monetary authority.
The objective is not to predict every policy decision.
It is to construct a portfolio that does not require me to be correct about every policy decision.
Volatility Is Part of the Regime
One of the greatest mistakes investors can make in this environment is interpreting volatility as evidence that the underlying thesis has failed.
Inflationary regimes can produce violent corrections.
When inflation expectations, bond yields, currency values, and monetary-policy expectations are unstable, asset prices can become extremely volatile.
Gold itself can experience significant drawdowns during a long-term secular advance.
Equities can fall sharply even when their long-term fundamentals remain favorable.
Bitcoin can experience much larger fluctuations.
This means position sizing becomes as important as asset selection.
A theoretically correct macro thesis can still produce a disastrous investment result if I use excessive leverage and cannot survive the path between the present and the eventual outcome.
That is why I consider leverage one of the most important risks in the entire framework.
The Most Dangerous Mistake: Excessive Leverage
The future distribution of economic outcomes is unusually wide.
There are simply too many interacting variables for me to assign a single deterministic forecast with confidence.
Fiscal policy can change.
Monetary policy can change.
Global capital flows can change.
AI productivity can exceed expectations.
AI productivity can disappoint.
Inflation can accelerate.
Inflation can collapse.
The dollar can strengthen.
The dollar can weaken.
Treasury demand can deteriorate gradually or suddenly.
Regulators can intervene.
Investors can change their expectations before policymakers change their policies.
When the distribution of possible outcomes widens, leverage becomes increasingly dangerous because leverage converts uncertainty into forced action.
An unleveraged investor can wait.
A leveraged investor may receive a margin call.
That distinction can determine whether an investor survives a regime transition.
The Historical Lesson: Regime Changes Are Not Normal Business Cycles
I also think it is dangerous to assume that the next decade will behave like the previous four decades.
For much of the period from the early 1980s through the pandemic era, investors operated within a relatively familiar framework: declining inflation, declining interest rates, expanding globalization, increasing financialization, and a long secular bull market in bonds.
That environment created enormous tailwinds for duration-sensitive financial assets.
If the structural regime shifts toward fiscal dominance, persistent inflation, rising capital demand, and monetary intervention, those historical relationships may weaken or reverse.
The past cannot simply be extrapolated forward.
This is especially important for investors trained primarily during the post-Volcker era.
A market participant can be extremely sophisticated and still be poorly prepared for a regime that has not existed within his or her investing lifetime.
The Political Economy of Asset Inflation
Monetary debasement also creates a political problem.
When asset prices rise faster than wages, the ownership of assets becomes increasingly important to economic security.
Those who already own capital participate in the inflation.
Those who primarily earn wages experience the inflation as a higher cost of living.
The result can be increasing pressure for redistribution.
This produces a feedback loop:
Inflation raises asset prices.
Asset ownership becomes more valuable.
Wealth becomes more concentrated.
Political pressure for redistribution increases.
Tax policy becomes increasingly focused on capital, corporations, and high-income households.
Those policies alter expected corporate profits and investment incentives.
Markets begin pricing the possibility of redistribution.
This is where macroeconomics, finance, technology, and political economy converge.
AI Creates a New Tax Problem
Artificial intelligence makes the distributional issue even more complicated because AI has the potential to increase productivity while simultaneously reducing the demand for certain categories of labor.
That is the fundamental technological transition.
If machines can perform tasks previously performed by workers, businesses can potentially produce the same or greater output with fewer workers.
From a productivity perspective, that is positive.
From a tax perspective, however, it creates a problem if government revenue depends heavily on labor income.
The source material estimates that roughly 85–90% of federal receipts are connected directly or indirectly to workers through individual income and payroll taxes.
If AI significantly reduces the labor share of national income, governments may eventually have to redesign how they collect revenue.
That could mean greater taxation of corporate profits, capital income, high-net-worth households, or other forms of wealth generation.
The paradox becomes striking.
The more successful AI becomes at replacing labor with capital, the more government may need to shift taxation toward the capital that AI makes more productive.
That creates a potential conflict between technological productivity and fiscal policy.
The AI Profit Paradox
There is an even deeper feedback loop.
AI companies and infrastructure providers justify enormous capital expenditure because they expect future productivity and profits.
Investors fund that capex because they expect those profits to materialize.
The capex itself creates economic growth.
That growth supports corporate earnings.
Higher earnings support higher equity valuations.
Higher valuations create additional wealth.
But if AI eventually reduces labor income substantially, the tax base can become increasingly concentrated in the very corporations and investors benefiting from the technology.
Government may then seek a greater share of those profits.
If taxation becomes sufficiently aggressive, expected future profits fall.
Valuations can contract.
The same productivity boom that created the original investment opportunity can eventually become the target of fiscal redistribution.
I do not view that as an immediate conclusion. I view it as a structural risk that becomes increasingly relevant as AI's share of economic production rises.
Why the Bull Market Could Continue Much Longer Than Expected
None of this means I should assume that the market is about to collapse.
In fact, the opposite may be true.
If AI productivity accelerates corporate profits while fiscal and monetary policy continue supporting nominal demand, risk assets can remain extremely strong.
The source framework anticipates the possibility of substantial bubbles in equities, gold, and Bitcoin before the eventual bond-market reckoning.
This is one of the most important conclusions I take from the framework.
A bearish long-term thesis does not automatically imply a bearish short-term market.
Markets can become dramatically more expensive while the underlying monetary system becomes dramatically more fragile.
That is why timing matters.
The existence of a structural problem does not tell me when the market will recognize it.
The Difference Between a Market Crisis and a Civilizational Crisis
I also distinguish between several different "endgames."
The first is a bond-market crisis.
The second is a fiscal and monetary-policy response.
The third is an inflationary adjustment.
The fourth is a broader political and social response.
These are not the same event.
A Treasury-market crisis could occur without immediately producing economic collapse.
A period of monetary debasement could produce substantial gains in stocks, gold, and Bitcoin.
A later redistribution of wealth could damage valuations without destroying the productive economy.
The danger comes from assuming that all of these stages must occur simultaneously.
They probably will not.
The financial system can move through multiple intermediate regimes.
That is why I prefer to think in terms of probabilities, indicators, and portfolio adaptation rather than a single apocalyptic forecast.
What I Watch
My macro framework therefore centers on a relatively small number of variables.
I watch the Treasury supply-demand balance.
I watch long-term Treasury yields and, more importantly, the decomposition between real yields, inflation expectations, and term premium.
I watch the dollar.
I watch global liquidity.
I watch foreign Treasury ownership.
I watch the behavior of private non-bank Treasury buyers.
I watch credit-market stress.
I watch private-credit valuations.
I watch corporate refinancing costs.
I watch household balance sheets.
I watch the labor share of national income.
I watch corporate profits.
And above all, I watch productivity and capital expenditure.
These indicators allow me to distinguish between two very different environments.
If AI capex is producing genuine productivity gains and corporate profits are accelerating, higher interest rates may be sustainable because the underlying economy is becoming more productive.
If rates are rising while productivity stagnates and fiscal deficits continue expanding, the same level of rates becomes far more dangerous.
The number itself is never enough.
Context determines its meaning.
The Central Investment Principle
I ultimately come back to one principle:
I want to own productive scarcity in an environment where nominal claims are becoming less scarce but less trustworthy.
Productive businesses are scarce.
Energy infrastructure is scarce.
Electricity capacity is scarce.
High-quality technology is scarce.
Gold is scarce.
Bitcoin is scarce.
Human capital remains scarce in many areas.
Government debt, by contrast, can be created in enormous quantities.
Currency can be created.
Financial claims can be created.
The challenge is that the supply of nominal claims can expand much faster than the supply of real resources.
That is the fundamental reason inflationary monetary regimes tend to favor scarce assets.
The Paradox at the Center of the Next Cycle
The deepest paradox I see is that the same forces can simultaneously make me bullish and cautious.
I can be bullish because AI may produce one of the largest productivity accelerations of the modern era.
I can be bullish because capital deepening can increase corporate profits.
I can be bullish because nominal growth can remain strong.
I can be bullish because monetary accommodation can lift asset prices.
I can be bullish because scarce assets can benefit from currency debasement.
And I can be cautious because the Treasury market may be approaching a structural supply-demand problem.
I can be cautious because long-term yields may need to rise to clear the capital market.
I can be cautious because higher yields increase fiscal stress.
I can be cautious because private-credit and institutional balance sheets may amplify volatility.
I can be cautious because AI could eventually disrupt labor income and the tax base.
I can be cautious because governments may respond with increasingly interventionist fiscal and monetary policies.
These are not contradictory beliefs.
They are different stages of the same cycle.
The Framework I Take Forward
I do not believe the appropriate strategy is to predict precisely which day the bond market breaks.
I believe the better strategy is to understand the direction of the structural forces and allow the market to tell me when the regime is changing.
The most important distinction is between capital appreciation and capital preservation.
There will be periods when I want maximum exposure to productive assets.
There will be periods when I want optionality.
There will be periods when gold and Bitcoin offer the strongest expression of monetary debasement.
There will be periods when equities offer the strongest expression of productivity.
There will be periods when cash becomes unusually valuable because volatility creates extraordinary entry points.
The ability to move between these regimes is more important than making one permanent forecast.
That is especially true because the range of possible outcomes is unusually wide.
Conclusion: I Do Not Need to Predict the Future to Invest in It
I believe we are entering an unusual period in which technology, fiscal policy, monetary policy, demographics, capital markets, and geopolitics are colliding simultaneously.
Artificial intelligence may produce an extraordinary productivity boom.
That productivity boom may produce extraordinary corporate profits.
Those profits may produce an extraordinary equity-market expansion.
The resulting capital expenditure may simultaneously increase the demand for global savings.
That increased demand may place upward pressure on interest rates.
Higher rates may increase government interest expense.
Higher fiscal burdens may eventually force policymakers toward financial repression, monetary accommodation, or yield-curve control.
Those policies may weaken the currency and increase demand for scarce assets.
Gold and Bitcoin may therefore rise not merely because investors expect inflation, but because investors increasingly distinguish between nominal financial claims and scarce stores of value.
Meanwhile, AI may reduce labor's share of national income and force governments to rethink how they collect revenue.
That could eventually turn today's technological winners into tomorrow's primary tax targets.
The entire cycle therefore contains a remarkable contradiction: the technology capable of making the economy dramatically more productive may also accelerate the financial, fiscal, and political adjustments that eventually reshape the distribution of wealth.
I do not know exactly where the cycle ends.
I do know that the traditional assumptions of the previous financial regime cannot simply be carried forward without examination.
The most important question for me is no longer simply whether an asset will rise.
It is what that asset is likely to be worth relative to money, productive capacity, and scarce real resources when the monetary regime itself is changing.
That is the lens through which I intend to view the next phase of the global economy.
And that is why I want to remain exposed to the productivity revolution while simultaneously protecting myself against the monetary consequences of financing it.
Reference: https://youtu.be/0-Bw776zKNg?si=CX-KE_jSvejfa4zr