U.S. artificial intelligence financing is reshaping the dollar’s traditional support system, with private capital flowing into technology companies at a scale that increasingly rivals the role once played by foreign purchases of long-term U.S. debt, Deutsche Bank said in a Sept. 3 foreign-exchange report.
The bank estimated that U.S. companies will spend roughly $800 billion on AI-related capital expenditures this year, while AI venture fundraising has exceeded $400 billion. That combination places technology spending at the center of a wider funding cycle involving private markets, investment-grade bonds and public equity issuance.
Deutsche Bank argued that the changing composition of capital flows could alter the dollar’s behavior. Foreign official institutions have shown less appetite for long-dated Treasury securities amid geopolitical tensions, the report said, while private overseas money has moved toward U.S. technology assets with returns tied more directly to AI growth expectations.
Two of the largest AI laboratories have raised nearly $217 billion combined and have valuations approaching $1 trillion each, according to the report. Meanwhile, Google, Meta, Amazon and Oracle have issued investment-grade debt this year at a pace roughly 10 times their average annual borrowing between 2020 and 2024.
Private capital, bonds and equity converge around AI
The AI funding surge is taking place through three separate channels that are increasingly reinforcing one another.
Private-market AI venture fundraising has passed $400 billion, Deutsche Bank said. More than 90% of the largest transactions were concentrated in the United States, while the annualized pace of fundraising was three times higher than last year’s rate. The concentration gives U.S. technology companies a substantial advantage in attracting global capital, but also increases their role in financing the country’s investment boom.
Debt markets are absorbing a growing part of that demand. Major technology companies historically known for large cash balances have turned more aggressively to bond issuance to fund infrastructure, data centers and computing capacity. The report’s estimate that borrowing by Google, Meta, Amazon and Oracle is running at 10 times its recent annual average suggests AI investment is pushing capital needs beyond internally generated cash flow.
Public equity has added another route. Deutsche Bank said Google completed its first share sale since its 2004 initial public offering in June 2026, raising $85 billion. The report also cited SpaceX’s public listing as the largest IPO on record, valuing the company at close to $2 trillion and allocating about 30% of the offering to retail participants, more than triple a typical IPO allocation.
Together, those transactions illustrate how AI and adjacent technology sectors are drawing funding from venture firms, bond buyers, public shareholders and retail trading accounts at the same time. That differs from a capital cycle driven mainly by bank lending or Treasury demand, since each source carries its own expectations for growth, liquidity and returns.
Foreign equity inflows overtake the familiar Treasury pattern
The funding needs of technology companies are arriving while the United States runs large fiscal and external deficits. Deutsche Bank put the federal budget deficit above 6% of gross domestic product and the current-account deficit near 4% of GDP.
Domestic fiscal saving is limited under those conditions, leaving overseas capital to finance a meaningful share of U.S. spending and investment. In the second quarter of 2026, the United States attracted more than $400 billion in foreign equity inflows in a single quarter, Deutsche Bank said, a record level that exceeded foreign debt inflows.
For years, foreign purchases of U.S. debt were a dominant source of capital-account funding. A larger role for equity and private technology financing could make cross-border flows more sensitive to corporate earnings, AI monetization and market valuations. Deutsche Bank said the dollar may become more correlated with equities as shorter-duration, return-driven technology capital replaces some long-duration official demand for Treasuries.
The report pointed to South Korea as an example of how access to foreign markets can influence these flows. South Korea represents about 2% of global nominal GDP, yet accounted for around 10% of the $740 billion in foreign inflows to U.S. equities last year, according to Deutsche Bank. The bank linked part of that participation to South Korea’s earlier adoption of fractional trading in foreign shares, which lowers the capital required to buy high-priced U.S. stocks.
Tokenization moves toward market infrastructure
Deutsche Bank’s report also connected the AI-driven funding cycle to the development of tokenized financial markets. Tokenization records ownership of assets such as funds, bonds or equities on blockchain-based systems, potentially allowing faster settlement, fractional ownership and round-the-clock trading.
The Depository Trust & Clearing Corporation, which holds about $115 trillion of U.S. assets in custody, completed an initial tokenization pilot in July 2026 involving real transactions with 40 financial institutions, according to the report. The tests included a tokenization exercise involving the SPDR S&P 500 ETF, known as SPY, and a separate trial in which JPMorgan used tokenized assets to meet margin requirements at CME.
DTCC plans to introduce a tokenization service in October 2026. The project follows an SEC no-action letter issued to DTCC in December 2025, which Deutsche Bank said allowed the same security to trade through conventional and on-chain systems while retaining equivalent ownership rights and protections. A January 2026 follow-on decision stated that federal securities laws apply regardless of the issuance format or custody method.
Trading venues are preparing for similar changes. The New York Stock Exchange is working with Securitize on a digital platform designed for 24/7 trading, instant settlement, fractional shares and stablecoin settlement. Nasdaq has outlined an equity-token design with full tokenization capability targeted for 2027, while also planning to offer 23-hour weekday trading.
Liquidity gains also make exits easier
The amount of tokenized real-world assets remains small beside conventional markets. Deutsche Bank estimated tokenized assets at about $40 billion globally, against more than $100 trillion in U.S. assets. Market forecasts cited by the bank range from $2 trillion to $30 trillion in tokenized assets during the 2030s.
Faster trading and settlement could make U.S. securities easier to access across borders, particularly for participants unable to operate during New York market hours or unable to purchase whole shares. The same features could also speed withdrawals from U.S. markets during periods of stress, Deutsche Bank cautioned, making accessibility a two-way channel rather than a one-directional source of inflows.
The report estimated that AI-related revenues could improve the U.S. current account by about one percentage point over the next decade through higher exports of AI-enabled services. That outcome depends on whether companies can convert enormous infrastructure spending into durable revenue and retain pricing power in international markets.
Explore how tokenized markets may amplify these shifts in dollar dynamics in today all markets onto blockchain.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
