Morgan Stanley capped a record bank earnings week with records of its own. The muted reaction showed what investors care about now: not the beat, but whether it lasts.
Key points
- Morgan Stanley posted record revenue, profit and stock-trading, and beat on both lines.
- The stock barely moved. It was already at a record, up more than 24% this year.
- Over half of the record $148 billion wealth haul came from one-off IPO stock plans.
- Management would not put a number on its share of the AI financing cycle.
Morgan Stanley just posted the best quarter in its history. Revenue, profit and stock-trading all hit records. The beat was large and it was clean. Yet the shares barely moved. That gap is the story. Investors already knew active markets would help. What they want to know now is whether these numbers can hold when markets calm down.
The beat came from the business, not the accounting
The firm earned $3.46 per share on $21.35 billion in revenue. Both were records. Analysts polled by LSEG had looked for $2.94 and $19.64 billion. So the beat was wide on both lines.
The quality was high. Revenue rose 27% from a year ago. Total expenses rose only 16%. Pay climbed 14%, slower than revenue. Credit costs fell by half. The tax rate even ticked up, from 22.7% to 23.1%. So no tax break created the surprise. This was operating leverage, not a one-time gain.
This was not a rebound from a weak base. The first quarter was also a record. Revenue rose only 4% from there. So the tone from management shifted. Last quarter the message was strength. This quarter it leaned hard on durability, on what Pick calls "higher lows."
Two items help read the year-ago lines. A prior-year deferred-pay gain of $294 million is no longer shown in wealth revenue. A first-quarter severance charge of $178 million sat in pay. Neither changes the story. Both are worth a glance when comparing quarters.
The clearest sign of strength was in stock trading. Buy-side desks had penciled in about $4.4 billion of equities revenue, per StreetAccount. The business brought in $6.3 billion. That is a gap of roughly $1.9 billion in one line.
The trading boom belongs to the whole street
Equities revenue jumped 69% to that record $6.3 billion. But this was not a Morgan Stanley quirk. JPMorgan and Goldman Sachs also posted record profit on booming stock trading. Across six large banks, profit rose 37% from a year ago. The whole group rode the same wave.
Ted Pick named two drivers. One is the AI boom. The other is what he called the "return of geopolitics." Both push clients to trade more. Both also depend on the mood of the market.
Some of the gain looks built to last. Morgan Stanley has spent years on trading tech, risk tools and Asia. Its Asia business now spans Japan, India, Korea, Taiwan and China. A 25% stake by MUFG helps. That breadth is a real edge.
Fixed income was a useful check on the story. Revenue rose 13% from a year ago. But it fell 27% from the first quarter. Commodities cooled after a strong start to the year. So even inside the good quarter, some lines slowed.
Investment banking added to the beat. Fees rose 58% to $2.4 billion. Advisory, equity underwriting and bond underwriting all climbed. Management said the deal pipeline is healthy but still below past peaks. Buyers have led the way. Private-equity sellers are only starting to return.
Still, management did not call $6.3 billion a floor. Pick said the pace depends on asset prices, growth and enough volatility to keep clients active. The platform is bigger. The revenue still leans on the weather outside.
The $148 billion wealth figure needs a second read
Wealth Management pulled in a record $148 billion of net new assets. That is the headline. The mix matters more.
Just over half came from stock-plan assets tied to a few big client IPOs. Those flows rise and fall with deal timing and vesting. They are not steady. Fee-based flows, the kind that earn recurring advice fees, were only $39.1 billion. That was below the $42.8 billion of a year ago.
So the $148 billion should not be read as $148 billion of advice relationships. Much of it is employee stock that may or may not stay. The open question is conversion. Can Morgan Stanley turn these accounts into long-term, fee-paying clients?
The setup is strong. The firm runs stock plans for about 70% of the top 100 private unicorns. It now counts 20 million client relationships. But management would not give a conversion rate or a timeline. It pointed to referral tools and adviser matching instead.
The wealth margin hit 30.5%. Pick called 30% a benchmark the firm has now passed several times. He would not raise the target. He would rather keep investing to win clients than squeeze the margin now. That is a fair trade. It also caps how fast the margin can climb.
Investment Management played a smaller part. Assets under management reached $2 trillion, a record. Long-term inflows were $7.5 billion. Together with wealth, total client assets crossed $10 trillion. That milestone is real. It did not create the beat.
AI pays off today. The $10 trillion is still a forecast.
AI showed up in two ways. First, it drove the trading and deal boom right now. Morgan Stanley helped take SpaceX public and earned a $100 million fee on that deal, matching Goldman. Second, it points to a long financing cycle ahead.
Pick laid out the map. Data-center spending could reach about $850 billion this year. He sees $1.3 trillion in 2027 and $1.5 trillion in 2028. Over a full cycle, he sees a possible $10 trillion of AI compute investment. He thinks the industry is only 10% to 15% of the way in.
But he would not say how much of that Morgan Stanley will capture. Asked for a share, he called it "meaningful" and stopped there. Some money will move straight between big tech firms. Some will need banks. The size is real. The bank's cut is not yet countable.
Capital backs the ambition. The firm added $18 billion of core capital over 10 quarters. It holds 300 to 350 basis points more than it needs. It raised the dividend 15 cents to $1.15 and set a new $20 billion buyback plan. It bought back $1.5 billion of stock this quarter, a modest sum against that base. Management would rather lend the capital to clients than rush it back to shareholders.
What the quarter settled, and what it did not
The reaction told the tale. The stock closed at a record the day before and sat near its 52-week high. It is up more than 24% this year. On the print, it barely moved. A strong beat was already priced in.
So the debate has shifted. No one now asks whether capital markets would reopen. They did. No one asks whether the model can work. It did. The new questions are harder. Can the firm keep the assets these IPOs delivered? Can it hold its trading share when activity cools? Can it turn the AI story into revenue you can measure?
Pick keeps promising higher highs and higher lows. The highs are here. The next few quarters are about the lows.