Goldman hits a new high, and bank stocks are also benefiting from the AI boom

TL;DR

· Five major U.S. banks—JPMorgan (JPM), Bank of America (BAC), Citigroup (Citi), Wells Fargo (WFC), and Goldman Sachs (GS)—released their Q2 earnings reports on July 14, with Goldman Sachs’ stock trading revenue hitting a record high.

· AI-related fundraising, underwriting, trading, and data-center lending are starting to flow into bank income statements, but they still can’t fully explain the growth.

· Related tickers: JPMorgan (JPM), Bank of America (BAC), Citigroup (Citi), Wells Fargo (WFC), Goldman Sachs (GS)

On July 14, five major U.S. banks—JPMorgan (JPM), Bank of America (BAC), Citigroup (Citi), Wells Fargo (WFC), and Goldman Sachs (GS)—all disclosed their Q2 earnings reports. Results from Goldman Sachs, JPMorgan, and others beat market expectations. Goldman Sachs’ stock price briefly hit a new intraday high, and bank stocks delivered a split performance.

These earnings are being amplified because a new pricing question has come to the surface: Is the AI investment boom already starting to “feed” Wall Street? In the past, when investors looked at bank stocks, they mainly focused on net interest margins, the credit cycle, and capital returns. Now the market is asking whether banks can become a “toll station” for AI capital expenditures.

This toll station isn’t mysterious. Tech giants and companies build AI data centers, buy chips, arrange financing, go public, and conduct IPOs and M&A. Hedge funds circle around AI stock trading and use leverage to amplify returns. Banks earn underwriting fees, loan-arrangement fees, trading price spreads, and financing interest. AI may not be written directly into banks’ balance sheets, but it does show up in their income statements.

The divide is here too. Goldman Sachs CEO David Solomon emphasized that the trading pipeline, the client network, and the business flywheel still have momentum, and that AI buildout is still in its early stages. JPMorgan CEO Jamie Dimon is more restrained, and the market is also concerned that companies will re-price costs, prices, and returns on investment.

Wall Street takes a cut from AI capital flows

Retail investors may easily interpret AI tailwinds as coming from chip makers, cloud providers, and data-center rental rates, but banks earn in a different layer. As long as AI-related assets are more expensive, more active, and more in need of financing, banks can take a cut from trading and capital flows.

Stock trading revenue is the most direct example. It isn’t the bank itself betting on whether stocks will rise or fall. Instead, it provides execution, hedging, financing, and liquidity for institutional clients. The more volatile AI concept stocks are, the more frequently funds rebalance, and the busier banks’ trading desks become.

In Q2, Goldman Sachs’ stock trading revenue reached $7.42 billion, setting a record and becoming a trigger for the market to re-price investment bank stocks. For companies like Goldman Sachs, where capital-markets business has a high share, changes in trading activity can quickly translate into earnings leverage.

Another link is AI capital expenditures. Tech giants such as Microsoft, Google, and Meta building AI infrastructure will drive data centers, electricity, chips, and private-asset financing. Reuters mentioned that Wall Street banks are viewing the AI supercycle as one of the sources of future trading and financing activity.

Banks may not be the most visible part of the AI industry chain, but they are an easily overlooked second-order beneficiary. The more capital-intensive the AI buildout is, the more it needs financial intermediaries. The more frequently AI assets are traded, the more it benefits the trading business.

Two samples provided by Goldman Sachs and JPMorgan

The second quarter’s clearest example of earnings elasticity is Goldman Sachs. The company’s net revenue was $20.34 billion, net profit was $6.63 billion, and EPS was $20.98. In its official earnings report, Solomon mentioned “One Goldman Sachs” and the business flywheel, and also emphasized that client activity and the trading pipeline still have momentum.

What the market rewards is not just the quarterly profit itself, but Goldman Sachs’ position in the AI narrative. If, over the next few years, continued AI buildout keeps driving IPOs, M&A, equity financing, data-center loans, and trading volumes, then Goldman Sachs’ earnings leverage could stand out more than that of traditional retail banks.

JPMorgan’s sample is more comprehensive. In Q2, its reported revenue was $57.3 billion, and managed revenue was $58.0 billion. Reported net income was $21.2 billion, and net profit excluding major items was $16.9 billion. It has both capital-markets business and consumer finance and corporate credit.

Here, attribution needs restraint. The strong performance from large banks in Q2 didn’t come from AI alone. The recovery in investment banking, active equity-market trading, changes in the interest-rate environment, and hedging demand driven by macro uncertainty are all pushing up revenue. A more accurate role for AI is that, when the trading and investment-banking cycle repairs, it adds another source of high-elasticity demand.

Morgan Stanley should also be watched. It disclosed its earnings on July 15, later than the five large banks, but its business structure is closer to Goldman Sachs. If the market continues to trade the capital-markets activity brought by AI, it will be placed in the same valuation-comparison peer set as Goldman Sachs.

Solomon’s flywheel and Dimon’s brake

Solomon’s narrative is clear: the global client network, strategic trading pipeline, and “integrated Goldman Sachs” are forming a flywheel; AI buildout is still early; and there is room for further capital-markets activity. As long as the AI investment cycle continues, client financing and trading demand will keep flowing back to Wall Street.

This logic matters a lot for valuing bank stocks. The valuations of traditional banks are easily constrained by net interest margins, provisions, and regulatory capital. If investment-banking and trading income is viewed as a structural incremental driver, the market may assign higher weight to Goldman Sachs, Morgan Stanley, and JPMorgan’s capital-markets business.

Dimon’s restraint forms the other boundary. He did not deny the value of AI, and JPMorgan itself is investing in AI and discussing efficiency improvements. But concerns about corporate AI spending are rising: budgets won’t expand indefinitely, and clients will start asking how much revenue growth or cost savings each dollar of investment can deliver.

This is a realistic risk for banks. AI capital expenditures can create trading and financing heat, but they may not continue in a strictly linear way. If companies slow down data-center construction, delay IPO financing, or if the cost of private-asset financing rises, related bank fees and trading revenue could cool off.

Loan quality and trading revenue determine the quality of re-rating

Whether bank stocks can truly treat AI as a new valuation anchor depends on whether AI-related financing can keep flowing into investment-banking revenue, whether data-center loan quality can remain stable, and whether stock trading revenue will mean-revert after volatility fades.

Data-center financing is especially worth monitoring. It is currently fuel for the expansion of AI infrastructure, and also a source of fees for banks and private-credit lenders. But if assumptions about leases, utilization, power costs, or financing costs prove wrong, these assets could shift from being an incremental revenue driver to becoming an exposure to risk.

The traditional credit cycle hasn’t disappeared either. High borrowing costs, energy prices, and geopolitical risks will show up with a lag in credit cards, auto loans, and corporate credit quality. Goldman Sachs is more likely to benefit from trading heat, while banks such as Citigroup and Wells Fargo—more reliant on traditional credit—will likely be priced differently by the market.

The signal from Q2 earnings is that the AI tailwind has spread from tech stocks to financial intermediaries, but it is still in the stage of an “incremental amplifier.” What can support re-rating is not an AI slogan, but the continuous delivery of investment-banking fees, trading revenue, data-center loan quality, and the returns on corporate AI investments.

Click to learn more about positions at Luyin BlockBeats

Welcome to join Luyin BlockBeats’ official community:

Telegram subscription group: https://t.me/theblockbeats

Telegram discussion group: https://t.me/BlockBeats_App

Twitter official account: https://twitter.com/BlockBeatsAsia