Morgan Stanley Reports $20.3B with 41% Rise in Revenue

Morgan Stanley closed the third quarter with a net revenue of $20.3 billion, up from $18.2 billion a year earlier, and earnings of $3.03 per share, surpassing market estimates. The return on tangible equity stood at 22.6%, marking the fifth consecutive quarter above 20%. The figure that set the bank apart from peers, however, came from the equity trading desk: revenue of about $5.8 billion, a 41% rise compared to the same period in 2025.

Shares fell about 3% on Wednesday (14), despite the better-than-expected results. The market closely noted Ted Pick's comments regarding the pace of business after August.

Two Quarters in One

Pick divided the period into two phases. July and August exhibited a strong risk appetite, driven by hyperscaler investment in AI infrastructure. After Labor Day, activity slowed with the resurgence of tensions in the Middle East and rising global interest rates. Nevertheless, the CEO stated that the bank, previously "overweight" in AI before the quarter, is now "doubly overweight" on the topic, both in serving clients across the chain and in the internal use of technology.

The wealth management division sustained the other half of the results. Revenue reached a record $9.1 billion, with $232 billion in new net assets collected during the quarter and total client assets of $10.5 trillion across the wealth and investment management arm. Fixed income was the weak point: revenue fell 7%, to about $2 billion.

Capex as a Revenue Stream

For those following the thesis that AI consumes money without returning, Morgan Stanley offers the other side of the equation. Every round of data center financing, every debt issuance by hyperscalers, and every stock offering related to the chip supply chain passes through desks like Pick's. The bank doesn't need to predict which model will prevail to profit from building the infrastructure.

The optimistic reading has named critics. A Bloomberg Opinion column published on Tuesday (13), on the day of JPMorgan and Goldman Sachs' results, argued that bank CEOs are too optimistic about the AI cycle, recalling the recent decline in chip manufacturers' stocks. Pick himself listed geopolitical conflicts, benchmark interest rates, rising energy prices, and high asset prices as risks for the fourth quarter. The drop after Labor Day is the most concrete evidence that AI-related business flows respond to macro sentiment and do not follow a distinct curve.

From New York to Veldhoven

The same cycle appeared on the other side of the Atlantic the same day. Dutch company ASML reported record net sales of 11.7 billion euros in the third quarter, according to the company’s statement, and stated that the planned supply for 2027 and 2028 will not fully meet customer demand. Some customers have already expressed interest in placing orders for 2029 and 2030. The company raised the floor of its annual projection and now expects sales between 44 billion and 45 billion euros in 2026, up from 43 billion to 45 billion in July.

The combination of the two results shows where capex money is landing. In the United States, it turns into capital markets commissions and trading revenue; in the Netherlands, it becomes a lithography order book with visibility of three to four years. The difference lies in the risk: ASML has contracts and customer advances, while Morgan Stanley's revenue depends on risk appetite returning to the levels of July and August. September numbers show that it does not always return at the same pace.

For the consulting executive selling AI projects to banks, the message from the quarter is less about Morgan Stanley and more about the calendar. The banks profiting from financing infrastructure are the same ones deciding how much to spend on internal adoption. Pick stated that the bank is doubly exposed to the theme; the results show that revenue exposure is most sensitive to a poor market quarter, and it is through this that technology budgets for 2027 will be calibrated.