Dollar’s New Dependency: What Happens When the AI Boom Stops?


For much of the past two years, investors have been preparing for a weaker US dollar. The reasoning appeared straightforward: deteriorating fiscal conditions, rising government debt, reserve diversification, and the gradual erosion of American economic exceptionalism.
Instead, the dollar has defied expectations.
The uncomfortable reality for dollar bears is that the United States has discovered a new source of monetary strength. It is not oil. It is not military dominance. Nor is it simply the Federal Reserve's interest rate policy. It is artificial intelligence.
The largest technological investment cycle in a generation is reshaping global capital allocation, influencing monetary policy, and creating a new economic relationship between the United States and Asia. The dollar is becoming one of the biggest beneficiaries of a technological revolution that extends far beyond Silicon Valley.
But there is a fundamental contradiction at the heart of this development. The same forces strengthening the dollar today may be making it increasingly vulnerable tomorrow.
To understand where the dollar is heading in 2027, we must first understand what is actually holding it up.
The Dollar Is No Longer Just an Interest Rate Trade
The conventional framework for understanding currency strength revolves around interest rate differentials, inflation, economic growth, and monetary policy.
When the Federal Reserve maintains higher interest rates than other major central banks, dollar-denominated assets become relatively more attractive. International capital flows toward the United States, increasing demand for the currency. But this framework alone is no longer sufficient.
On November 19, Deutsche Bank revised its 2027 dollar outlook upward. Having previously anticipated depreciation, the bank now expects dollar strength to persist into the first half of 2027, followed by a relatively modest retreat. AI-related investment has become a central component of its revised outlook.
What we are witnessing is more than a conventional high-interest-rate environment. It is a technology-driven capital concentration.
Consider the position of an international institutional investor today.
If a Japanese pension fund wants exposure to the leading AI infrastructure companies, where does it allocate its capital? If a European asset manager wants exposure to frontier AI models, cloud computing, or advanced chip design, which capital market offers the deepest selection?
In many cases, the answer remains the United States.
The world's most important AI companies are disproportionately concentrated within the American financial system. Investors seeking exposure to this technological revolution may therefore need to acquire dollar-denominated assets, regardless of their underlying view of the US currency.
This creates a fundamentally different source of demand.
Investors are not necessarily buying the dollar because they believe in the dollar. They are buying the dollar because they believe in what America is building.
That distinction matters.
The Three Engines of AI-Driven Dollar Strength
The relationship between artificial intelligence and the dollar operates through three interconnected economic channels.

The AI-Dollar Transmission Mechanism
The first engine is economic growth.
Artificial intelligence requires an extraordinary amount of physical investment. Behind every AI model lies a supply chain of semiconductors, data centers, cooling systems, electricity generation, networking equipment, and construction.
Unlike previous software revolutions, the current AI expansion is exceptionally capital-intensive.
This spending supports economic activity long before the promised productivity gains are fully realized. Construction companies receive contracts. Semiconductor manufacturers expand capacity. Energy providers build infrastructure. Technology companies increase investment budgets.
Even if AI monetization takes longer than expected, the physical construction of the AI economy is already generating demand.
The second engine is monetary policy.
There is a widespread assumption that artificial intelligence will ultimately be deflationary. By improving productivity and reducing labor costs, AI could theoretically lower the cost of producing goods and services.
But the transition toward that outcome can be inflationary.
Thousands of data centers competing for electricity, scarce engineering talent, advanced chips, and construction materials put upward pressure on the cost of these resources.
If demand expands faster than productive capacity, the Federal Reserve has less room to ease monetary policy.
The distinction between long-term productivity and short-term investment demand is critical. AI may eventually reduce inflation while simultaneously contributing to inflationary pressures during its construction phase.
That creates an environment in which US interest rates can remain elevated despite technological progress.
The third engine is international equity demand.
Foreign investment in American equities provides another channel of currency support, particularly when overseas investors leave their dollar exposure unhedged.
This changes the nature of monetary transmission. The attractiveness of US assets is no longer determined solely by Treasury yields or monetary policy. It increasingly depends on the expected profitability of American technology companies.
The dollar has effectively acquired exposure to the global AI growth narrative.
From Petrodollars to Semiconductor Dollars
To understand the international dimension of this cycle, it is worth revisiting the economic logic behind the petrodollar system.
For decades, oil-exporting economies accumulated substantial dollar revenues through energy exports. A portion of those revenues was subsequently reinvested in US financial assets, including Treasury securities.
America imported energy. Energy producers accumulated dollars. Those dollars returned to American capital markets.
The AI economy is developing a comparable recycling mechanism, although one with very different participants and risk characteristics. Consider the geography of the semiconductor industry.
The United States dominates much of advanced chip design, cloud infrastructure, and frontier AI development. Taiwan plays an indispensable role in advanced semiconductor manufacturing. South Korea is a major producer of memory chips, including high-bandwidth memory essential to AI computing. Japan provides specialized materials, components, and semiconductor manufacturing equipment.
America owns much of the intellectual property. Asia manufactures much of the physical infrastructure.
A July 2026 research note published by Federal Reserve economists estimated that approximately 90% of the relevant technology equipment imports examined in its AI investment analysis were sourced from East Asia.
This concentration has enormous implications for international capital flows.
The Semiconductor Capital Cycle

The striking feature of this arrangement is that America's external dependence can reinforce its financial dominance.
The United States spends money importing advanced technology equipment. Asian exporters generate foreign currency earnings. Some of those earnings are recycled into American financial markets. The resulting capital flows support asset valuations and financing conditions, helping sustain further American investment.
The dollars leave America through trade and can return through investment.
But unlike oil, which benefits from persistent global consumption, semiconductor demand is exposed to technological obsolescence, corporate investment budgets, and rapidly changing expectations about future profitability.
The AI capital cycle may consequently prove more volatile than the commodity-based recycling systems that preceded it.
The question is not merely how many chips America purchases. It is what the economies selling those chips ultimately decide to do with their earnings.
America's Twin Deficits: A Problem Temporarily Disguised as Strength
There is an apparent contradiction in the current dollar environment.
The United States faces substantial fiscal deficits. Its external financing requirements remain significant. Yet international capital continues to support the currency.
How can a country simultaneously accumulate financial imbalances and maintain a strong exchange rate? The answer lies in the distinction between a country's need for capital and the rest of the world's willingness to provide it.
America operates with two important economic imbalances: a fiscal deficit and a current account deficit.
The fiscal deficit reflects the government spending more than it receives in revenue. The current account deficit reflects an economy whose aggregate investment exceeds national saving, requiring net financing from abroad. These are separate but interconnected problems.
When the United States runs a current account deficit, it can sustain that imbalance because foreign investors are willing to accumulate American assets or otherwise provide net financing. They purchase Treasury securities, corporate bonds, equities, real estate, and other financial claims.
Historically, the depth and credibility of American financial markets have made this financing exceptionally accessible. The AI boom adds another dimension. Foreign capital is being attracted not merely by the safety of government debt but by the possibility of participating in technological progress.
This creates what Enzac Research would describe as a productivity premium on American financial imbalances. As long as investors believe the United States can convert imported capital and technology into superior future economic output, they may tolerate deficits that would be deeply problematic elsewhere.
The Federal Reserve's July 2026 research also highlights the possibility that AI investment could contribute to a persistent deterioration in the US current account through higher technology-related imports.
In other words, the AI boom can strengthen the dollar while simultaneously worsening one of America's underlying economic imbalances.
That is the paradox.
The danger emerges when investors begin questioning whether the future productivity gains justify the scale of today's investment. A deficit financed by confidence is manageable. A deficit facing a loss of confidence is a different proposition entirely.
The Hidden Risk: When Equity Capital Replaces Defensive Capital
Not all international capital flows behave in the same way.
A central bank purchasing US Treasuries for reserve management has a fundamentally different objective from a European investment fund purchasing shares in an AI company.
The former prioritizes liquidity, security, and institutional obligations. The latter is primarily seeking returns.
One is predominantly defensive. The other is growth-sensitive.
Defensive capital | Growth-sensitive capital | |
Primary objective | Capital preservation | Capital appreciation |
Typical assets | Sovereign bonds, reserves | Equities, technology investments |
Main driver | Safety and liquidity | Expected earnings and growth |
Reaction to an AI slowdown | Potentially limited or positive | Potentially significant |
Currency implications | Relatively persistent demand | Greater sensitivity to sentiment |
As marginal demand for American assets becomes increasingly associated with technology investment, the composition of foreign financing deserves greater attention.
A foreign investor holding US equities faces two sources of risk: the underlying asset price and the exchange rate.
Consider a simplified example. A European investor purchases $100 million of American technology stocks without hedging the dollar exposure.
If those stocks decline by 25% while the dollar simultaneously depreciates by 10% against the euro, the investor's combined loss in euro terms becomes 32.5%, before costs.
The AI Financing Problem: Who Pays for the Next Trillion?
There is another development that deserves considerably more attention: the changing financing structure of artificial intelligence.
The first stage of AI investment was largely driven by some of the most profitable corporations in history. Companies with enormous operating cash flows could finance infrastructure expansion without depending excessively on external borrowing.
But as the scale of investment grows, financing requirements become increasingly demanding.
In January 2026, researchers at the Bank for International Settlements identified a potential transition from internally generated corporate cash flows toward greater reliance on debt financing, including private credit. They also emphasized the importance of future earnings in determining whether the investment expansion remains sustainable.
This is where the AI narrative intersects with financial stability. A profitable technology company financing a data center from existing cash reserves presents one type of risk. A leveraged infrastructure developer borrowing against projected future demand presents another.
As debt financing becomes more prominent, the economic viability of AI infrastructure becomes increasingly dependent on utilization rates, financing costs, contractual commitments, and actual revenue generation.
High interest rates complicate the equation. The very monetary conditions that help support the dollar can increase the cost of financing the infrastructure underpinning its strength.
A sufficiently large increase in borrowing costs could make marginal data center projects uneconomic. If companies respond by postponing construction or reducing investment, the economic growth impulse begins to fade. This is a potentially self-limiting mechanism.
AI investment supports growth and interest rates. Higher interest rates strengthen the dollar. But higher financing costs can eventually restrict further investment.
The mechanism that produces dollar strength also contains a potential mechanism for its reversal.
Asia: The Other Side of the American AI Miracle
Much of the discussion surrounding AI focuses on American companies. But the next major development in the foreign exchange market may originate in Japan, South Korea, or Taiwan.
The United States captures a significant portion of the financial valuations associated with AI. North Asia captures a substantial portion of the manufacturing revenue.
These economies matter more to the dollar than many investors realize.
The Federal Reserve's broad trade-weighted dollar index incorporates currencies according to America's trading relationships, providing a different perspective from the conventional DXY index, which is heavily influenced by the euro.
An investor watching only DXY may therefore overlook important changes in the relationship between the dollar and Asian industrial economies.
Japan: The End of Cheap Capital?
For decades, Japan's exceptionally low interest rates encouraged domestic and international investors to seek higher returns overseas.
The yen became a major funding currency for global carry trades. Japanese institutions accumulated foreign bonds and other international assets. Low Japanese yields helped facilitate capital allocation outside the country.
But monetary normalization changes the calculation. If Japanese government bonds offer increasingly attractive domestic yields, Japanese institutions have less incentive to absorb foreign currency risk in pursuit of returns.
This matters because an apparently attractive US Treasury yield can become considerably less attractive once the cost of hedging dollar exposure back into yen is incorporated.
A gradual reallocation toward domestic Japanese assets could reduce one source of international financing for American markets.
The more disruptive possibility is a sudden carry trade unwind. Investors borrowing yen to finance international positions may be forced to reduce those positions if the yen appreciates significantly.
Japan does not need to abandon American assets entirely to influence the dollar. It only needs to become a more attractive destination for its own capital.
South Korea and Taiwan: Manufacturing Power Becomes Financial Power
South Korea and Taiwan occupy a different position.
Their strategic importance is tied to the profitability of semiconductor exports and their role within the global AI supply chain. As export revenues increase, so does the potential amount of capital available for domestic conversion, overseas investment, or reserve accumulation.
The destination of those earnings becomes an important currency variable.
If companies and institutions maintain their earnings in dollar-denominated assets, they help sustain the existing capital recycling system.
If they convert a greater share into domestic currencies, increase domestic investment, or diversify their overseas holdings, the distribution of currency demand changes.
This is why the future of the dollar cannot be understood exclusively through decisions made in Washington.
It must also be understood through corporate treasuries in Seoul, semiconductor factories in Taiwan, and institutional investment committees in Tokyo.
Dollar Strength Does Not Mean De-Dollarization Is Over
One of the most common mistakes in macroeconomic analysis is treating exchange rate appreciation and reserve currency dominance as the same phenomenon.
They are not.
A currency's exchange rate reflects its relative market price. Its reserve status reflects its role within the international financial architecture.
The dollar can appreciate even as central banks gradually diversify their reserve holdings.
According to the IMF's second-quarter 2026 COFER release, the dollar accounted for approximately 56.7% of reported global foreign exchange reserves. The share declined from the previous quarter even though the absolute amount of dollar holdings was broadly unchanged.
The dollar remains the dominant reserve currency
The distinction becomes particularly important in the AI era.
Private investors may increase their exposure to US technology while official reserve managers gradually diversify their holdings. These behaviors are entirely compatible. A sovereign institution may question its long-term dependence on American financial assets. A private fund in the same country may simultaneously increase its investment in American AI companies.
One is managing geopolitical concentration. The other is pursuing investment performance.
The two forces can move in opposite directions for years.
The dollar's current strength should therefore not be interpreted as proof that reserve diversification has failed. Equally, reserve diversification should not be interpreted as evidence that an imminent dollar collapse is inevitable.
America's technological dominance and its long-term monetary vulnerabilities can coexist.
What Happens If the AI Boom Disappoints?
This is the most important question heading into 2027.
The market does not require AI to fail for the dollar's current support structure to weaken.
It simply requires AI to deliver less than what investors have already priced in.
There is a crucial difference between technological success and investment success.
The internet transformed the global economy. That did not prevent the collapse of technology valuations in 2000.
Railways revolutionized transportation. That did not prevent repeated speculative bubbles and financial crises throughout the nineteenth century.
A technology can be transformative while the capital invested in its early development generates disappointing returns.
Artificial intelligence is not exempt from this economic reality. Imagine that major technology companies begin reporting slower-than-expected AI monetization. Revenue continues growing, but not quickly enough to justify the enormous increase in infrastructure expenditure.
Management teams respond by revising future capital expenditure budgets downward.
The implications could extend considerably beyond the technology sector.
The AI Repricing Scenario
AI revenue disappoints expectations
Technology companies reduce future capital expenditure
Economic channel
Weaker investment, growth and potential easing of inflationary pressure
Financial channel
Lower equity valuations, reduced foreign demand and increased currency hedging
Potentially lower real yields
Weaker capital inflows
Structural dollar downside risk
Source: Enzac Research scenario analysis. Safe-haven demand and dollar funding stress could temporarily reverse the currency reaction.
The issue is correlation.
Investors often evaluate US interest rates and foreign equity inflows as separate pillars supporting the dollar. But if both are increasingly driven by the same underlying investment cycle, they may not be independent sources of strength at all.
They could be two expressions of the same risk. There is, however, an important complication. In the initial stages of a severe financial crisis, the dollar could strengthen rather than weaken.
Global investors may liquidate risky assets, seek dollar liquidity, repay dollar-denominated obligations, and accumulate cash. This is consistent with the dollar's historical role as a global funding and defensive currency.
The potential sequence is therefore not necessarily an immediate dollar collapse.
It could be an initial dollar liquidity squeeze, followed by a more persistent depreciation as US growth slows, monetary policy adjusts, and international portfolios diversify.
Timing is everything.
The Implications for Bitcoin, Gold and Global Risk Assets
For cryptocurrency investors, the emergence of the AI-driven dollar presents a complicated environment. A technological investment boom does not necessarily translate into abundant liquidity for every risk asset.
Capital is finite at any given allocation decision. An institutional investor choosing to increase exposure to AI infrastructure may reduce exposure elsewhere.
If American technology companies offer compelling growth expectations while Treasury securities offer attractive yields, alternative assets face competition from both sides of the investment spectrum.
Bitcoin competes with growth assets for speculative capital and with monetary assets for alternative-store-of-value allocations. A strong dollar combined with relatively high real yields can therefore create a difficult environment for crypto, particularly for speculative assets without sustainable earnings or identifiable economic demand.
But what happens if the AI cycle reverses?
The immediate reaction could be negative for Bitcoin. A broad reduction in risk exposure rarely distinguishes perfectly between technology equities, cryptocurrencies, and other volatile assets.
Over a longer horizon, however, the macroeconomic environment could change materially.
Lower real interest rates, potential monetary accommodation, increased concern about US fiscal sustainability, and greater demand for monetary diversification could create more favorable conditions for Bitcoin and gold.
Gold may behave differently from Bitcoin during the initial stress period because of its established defensive role, although it too remains sensitive to real yields, currency movements, and forced liquidation.
The essential distinction is between a liquidity crisis and the monetary response to a liquidity crisis.
The former can destroy risk appetite. The latter can eventually restore it. Investors who confuse these two stages risk positioning themselves correctly for the long-term outcome but incorrectly for the immediate market reaction.
The 2027 Outlook: Three Possible Dollar Regimes
Rather than constructing a single exchange rate forecast, Enzac Research believes it is more useful to identify the economic conditions that could produce different dollar regimes.
Scenario I — American Exceptionalism Continues
Dollar-positive scenario
AI investment remains strong. Corporate earnings validate current expectations. US growth outperforms other developed economies, and the Federal Reserve maintains relatively restrictive monetary conditions.
International capital continues favoring American assets.
Market implications
Asset | Potential impact |
US dollar | Sustained strength |
US technology equities | Supported by earnings |
Treasury yields | Remain relatively elevated |
Emerging market currencies | Financing and exchange rate pressure |
Crypto | Mixed; liquidity conditions remain restrictive |
Scenario II — The Orderly Rebalancing
Gradual dollar depreciation scenario
AI investment continues but growth in expenditure moderates. Productivity improvements become more visible, inflation pressures ease, and monetary divergence between the United States and other economies narrows.
Foreign investors gradually diversify their portfolios without abandoning American markets.
Market implications
Asset | Potential impact |
US dollar | Gradual depreciation |
Asian currencies | Potential appreciation |
US Treasury yields | Moderate decline |
Emerging markets | Selective improvement |
Crypto | Potentially more supportive liquidity conditions |
Scenario III — The AI Capital Expenditure Shock
Financial instability scenario
AI monetization disappoints, infrastructure spending is reduced, highly leveraged projects face refinancing difficulties, and technology valuations experience a material correction.
Foreign equity inflows weaken as monetary policy expectations shift.
Market implications
Asset | Potential impact |
US dollar | Initial defensive strength, followed by potential weakness |
Technology equities | Significant repricing risk |
Credit markets | Wider risk premiums |
Gold | Potential medium-term defensive demand |
Crypto | Initial liquidation; subsequent recovery depends on liquidity |
Our working base case is an eventual orderly rebalancing, although continued AI investment momentum could delay that transition throughout much of 2027.
The greatest risk is not necessarily that the optimistic scenario proves wrong. It is that investors are positioned for a gradual transition when the adjustment ultimately becomes abrupt.
What Investors Should Actually Watch
The most useful indicators for understanding the dollar in 2027 may not be the dollar itself.
By the time the exchange rate begins reflecting structural changes, much of the adjustment may already be underway elsewhere. We would prioritize the following indicators:
Indicator | What it tells us | Warning signal |
Hyperscaler capital expenditure guidance | Strength of the AI investment cycle | Broad downward revisions |
AI-related revenue and free cash flow | Whether investment is being monetized | Spending growth materially exceeding revenue generation |
US 10-year real Treasury yield | Inflation-adjusted return on US government debt | Sustained decline relative to overseas yields |
Treasury International Capital data | Foreign demand for American financial assets | Persistent weakening of equity inflows |
AI infrastructure credit spreads | Financing stress | Rising borrowing costs and refinancing difficulties |
Japanese government bond yields | Incentives for Japanese capital repatriation | Improving domestic returns relative to hedged US returns |
Asian semiconductor export data | Strength of the physical AI supply chain | Falling orders and inventory accumulation |
IMF reserve allocation data | Long-term monetary diversification | Sustained reductions in dollar allocation |
The most consequential signal would not be weakness in any single indicator.
It would be a simultaneous deterioration in AI investment expectations, American real yields, and foreign equity inflows.
That would indicate that the fundamental sources of dollar strength are beginning to weaken together.
The Enzac Assessment: A Strong Dollar Is Not Necessarily a Safe Dollar
The United States has achieved something remarkable. At a time when its fiscal position faces growing scrutiny and its monetary dominance is being challenged, it has positioned itself at the center of the most significant technological investment cycle in decades.
Artificial intelligence has given international investors a compelling reason to continue allocating capital toward America.
It has reinforced economic growth, supported financial valuations, influenced monetary conditions, and created a powerful relationship between Asian manufacturing and American capital markets.
But this strength comes with an increasingly important concentration risk.
The dollar's traditional foundations remain intact: deep financial markets, institutional infrastructure, global settlement networks, and the absence of a fully comparable reserve currency alternative.
What has changed is the marginal source of its strength.
A growing portion of the dollar's cyclical support is tied to expectations about a single technological revolution, the scale of its financing, and its ability to generate future economic returns.
The central question for 2027 is therefore not whether artificial intelligence will change the world. It almost certainly will.
The question is whether the economic returns arrive quickly enough to justify the extraordinary amount of capital already committed to that future.
If they do, American financial exceptionalism could persist considerably longer than dollar bears anticipate.
If they do not, the consequences will extend far beyond Silicon Valley.
They will reach sovereign bond markets, international capital flows, Asian currencies, global liquidity, and the valuation of virtually every major asset class.
The world is financing America's technological future. In doing so, it is also financing the strength of the American dollar.
The dollar has become stronger because the world believes in AI. Its next great vulnerability may be precisely the same belief.




Comments