The Zhitong Finance App learned that on July 14, large Wall Street banks proved to the market with a series of record results that the beneficiaries of the artificial intelligence boom are far more than Silicon Valley tech giants and chip makers. Goldman Sachs (GS.US) and J.P. Morgan Chase (JPM.US) both handed over historic quarterly responses, and revenue soared 39% to US$20.3 billion and 27% to US$58 billion, respectively. What drives these numbers is the “ubiquitous” presence of AI in financial markets.
Jeremy Barnum, Chief Financial Officer of J.P. Morgan Chase, said bluntly during the earnings call: “The market in these fields is booming, activity is frequent, large-scale IPOs are constantly emerging, the index is highly adjusted, and the Asian market is also extremely active. Many of these are closely related to the subject of artificial intelligence... It's a very, very, very active market environment.”
Stock trading revenue explodes across the board: Goldman Sachs sets industry record with US$7.42 billion
The most impressive data for the quarter came from the stock trading business. Goldman Sachs stock trading revenue increased 72% year over year to US$7.42 billion, setting the highest historical record for stock trading revenue for a single Wall Street bank. The revenue for the past three months alone exceeds the sum of the bank's four quarters for the full year of 2019.
J.P. Morgan Chase was also impressive. Stock trading revenue soared 86% to $6 billion, a record high for the bank. Together, the two companies' share trading revenue was $4.4 billion higher than analysts' expectations. Bank of America's stock trading revenue also increased 70% to $3.6 billion. Citigroup's stock business revenue increased 45% year over year. The five major banks achieved a total profit of US$49 billion in the second quarter, an increase of 39% over the previous year.
Analysts expect that the five major banks, J.P. Morgan Chase, Bank of America, Citibank, Goldman Sachs, and Morgan Stanley, will have total transaction revenue of nearly $39 billion in the second quarter. Goldman Sachs's stock business revenue in a single quarter has exceeded the total for the full year of 2019. Since Trump took office again, the six largest banks have broken previous records for total stock trading revenue every quarter.
Five major banks, J.P. Morgan Chase, Bank of America, Citibank, Wells Fargo, and Goldman Sachs, concentrated on releasing second-quarter earnings reports. According to the data, the five major banks achieved a total profit of 49 billion US dollars, an increase of 39% over the previous year. Wall Street bankers are proving with record financial reports: in this trillion-dollar capital feast, those who collate deals are making huge profits.
Market fluctuations provide continued support for the trading business — geopolitical tension compounded by industrial uncertainty brought about by AI kept the volatility high. Goldman Sachs CEO David Solomon said in June that there is “more greed than fear” in the market, and investors are flocking to issue IPOs.
AI drives large-scale financing to land one after another: from SpaceX to SK Hynix
The explosion in stock trading stemmed directly from the reallocation of global capital around AI themes. Investors expanded their search for AI beneficiaries from US tech giants to Asian markets such as South Korea, Taiwan, and Japan. Soofian Zuberi, President of Global Markets at Bank of America, stated, “People look at transactions in the AI sector and then ask, 'What best reflects this trend outside of the US? ' You'll find that some US clients are diversifying their investments and allocating more money to Asia”. Bank of America's stock trading revenue increased 70% to $3.6 billion.
AI's “capital expenditure supercycle” is driving the investment banking business to unprecedented heights. SpaceX's record IPO was the most notable case of the quarter. Goldman Sachs, as the lead underwriter, collaborated with Morgan Stanley, Bank of America, etc. to facilitate the largest IPO in history. In total, the underwriting team earned approximately $500 million in expenses.
Goldman Sachs and Morgan Stanley each earned approximately $100 million in underwriting revenue from the SpaceX IPO. More companies are preparing to join this wave — Anthropic is expected to go public within the year, and both OpenAI and DeepSeek have plans to go public.
SK Hynix's ADR listing was also remarkable. The memory chip giant raised 26.5 billion US dollars through the issuance of American Depositary Receipts, making it the largest number for foreign companies to go public in the US. As a co-global coordinator, Citi earned more than $70 million from the deal.
Goldman Sachs provided advisory services for $1.2 trillion worth of transactions in the first half of the year, leading its closest competitor by about US$425 billion. Investment bank fee revenue increased 55% year over year to US$3.4 billion, the highest since 2021. J.P. Morgan Investment Bank's revenue increased 30% to $3.3 billion. Bank of America investment banking service fees increased 50% to $2.1 billion.
The “ripple effect” of AI capital expenditure: from data centers to power infrastructure
The diffusion effect of the AI investment boom far exceeds the capital market itself. Solomon pointed out that this is creating a “ripple effect” in the US economy, providing banks with plenty of new opportunities.
Goldman Sachs executives detailed this logic during the earnings call: the AI capital cycle is extending to physical construction — data centers require real estate, power generation, transmission capacity, cooling systems, commodities, and structured finance; suppliers need working capital; infrastructure developers need private equity; technology companies may issue stocks or bonds; and utility companies may acquire assets or restructure portfolios to meet new energy needs. Solomon described this process as a “multiplier effect” — the advisory business is only the starting point, which then spawns financing, risk management, capital market execution, and investment opportunities.
For example, banks such as Goldman Sachs acted as advisors on Alphabet's $90 billion share increase and Dominion Energy's sale to NextEra Energy. According to Dealogic data, global investment banking revenue reached US$61.4 billion in the first half of 2026, an increase of 24% over the previous year.
Data center operator Switch has hired an investment bank to prepare for an IPO, which may raise up to $10 billion, with a valuation close to $80 billion. J.P. Morgan CFO Barnum notes that even non-AI-related companies are generating capital expenses and loan requirements: “Just as data centers will eventually create significant demand for plumbers and electricians, you'll see this impact in less visible places.”
Bank of America has raised nearly $500 billion for AI-related companies since 2025, accounting for 60% of total investment-grade bonds, leveraged financing, and stock market financing. The bank also recently provided a $520 million credit line to OpenAI, which is the first time it has lent to this AI company.
Solomon defined the current situation as an “AI capital expenditure supercycle,” and stated that “every region and every industry in the world faces huge financing needs.” Solomon said Goldman Sachs is preparing for a “three to five-year investment cycle,” which “is still in its early stages.”
Reinventing the banking industry from within: a “two-way flywheel” for AI applications
While profiting from the external AI boom, Wall Street giants are also actively deploying AI technology internally, forming a positive cycle.
J.P. Morgan CEO Jamie Dimon revealed that the bank currently has about 1,000 AI application cases under development, of which about 50 are the bank's biggest efforts in the fields of fraud detection, risk management, marketing, and document review. Dimon said AI has reduced jobs by 30% to 40% in some specific fields, but most of these people have obtained jobs elsewhere.
Goldman Sachs, on the other hand, has set a goal of achieving a 15% increase in efficiency within the next three years through an AI “digital factory”. The bank is collaborating with Anthropic to develop AI agents to gradually automate internal functions such as accounting, compliance processes, and customer entry.
Bank of America's Global Markets President Soofian Zuberi's summary highlights the nature of this two-way relationship: “Artificial intelligence is driving the banking industry forward by helping to streamline processes. And the banking industry is also driving the development of artificial intelligence, because without banking, it would be impossible to fund all these data centers.”
How long will the feast last?
Stephen Biggar, head of financial services research at Argus Research, stated: “The AI-driven capital expenditure supercycle is already benefiting stock issuances, mergers and acquisitions, and debt financing.”
However, under the feast, there are also hidden worries. Recently, technology stocks have fluctuated greatly, and the market has begun to question whether a number of large-scale IPOs under preparation can proceed smoothly. OpenAI is considering delaying the IPO from the original schedule this fall to next year. As a result, the stock prices of Morgan Stanley and Goldman Sachs once fell by more than 4%.
Solomon warned that the AI capital expenditure cycle will be accompanied by “twists and turns” — companies are still uncertain about the content of infrastructure construction, pricing methods, and how to procure this infrastructure. However, he stressed that Goldman Sachs is ready for a three to five year investment cycle, and this cycle “is still in its early stages.”
The AI boom has spread from chip stocks to Wall Street's “all-round revenue engine” — from stock trading to IPO underwriting, from debt financing to M&A consulting, from wealth management to internal efficiency improvements. The question is no longer whether banks can profit from AI, but rather how long this feast can last.