Newsletter · · Ashutosh Agarwal

Goldman's Blowout Quarter Proves the Reopening Is Real - The Capital-Markets Reopening - Week of July 21, 2026

Capital markets podcast intelligence for the week of July 21, 2026. Five of the six largest US banks reported second-quarter results and the reopening thesis moved from forecast to fact, with Goldman Sachs posting its best quarter ever, JPMorgan its biggest profit in history, and bank CEOs from Morgan Stanley to Wells Fargo going on the record even as SpaceX slipped below its IPO price and a judge froze the Paramount deal.

The Capital-Markets Reopening

Week of July 21, 2026: Goldman's Blowout Quarter Proves the Reopening Is Real

TL;DR

  • The proof arrived, and it was bigger than anyone dared forecast. Five of the six largest US banks reported second-quarter results this week, and the capital-markets machine didn't just work, it printed records. Goldman's equity-trading revenue came in at $7.42 billion against a roughly $5 billion estimate, a near-50% beat; investment banking rose 55%; the stock jumped about 9% to an all-time high. JPMorgan booked its biggest quarterly profit ever, $21.2 billion. Across the five reporters, revenue was up 39% from a year ago. Last week we said "the proof arrives this week." It did.
  • The bosses themselves put their names on it, but every one of them hedged. For the first time in this cycle we heard operators, not pundits. Wells Fargo's CFO called it a "record fee quarter for investment banking fees." Morgan Stanley's CEO Ted Pick said we're only "10 to 15 percent of the way through" an AI investment cycle he thinks tops $3 trillion over three years. And JPMorgan's Jamie Dimon delivered the line of the week: "it's getting close to as good as it gets. We just don't know how long it's going to last."
  • The one crack: the exit door got tighter. SpaceX, last week's "textbook" mega-IPO, slipped below its $135 offer price. A federal judge slapped a 14-day pause on the $110 billion Paramount and Warner Bros. deal even after the Justice Department had cleared it. And Fed Chair Kevin Warsh, testifying to Congress, refused to say inflation is beaten. The revenue engine is roaring; the plumbing is where the risk now lives.

What's new

Ranked by what matters most for a trading book. Where the speaker is a company insider (a CEO or CFO talking about their own firm), it is flagged as such; everything else is analyst, journalist, or investor commentary.

1. Goldman didn't just beat, it lapped the field. The single most actionable fact of the week is how good Goldman's quarter was. On Schwab Network, "Big Bank Earnings Kick Off as IBM Slides, Chip Stocks Rebound" (July 14), senior equities contributor Alex Coffey walked through the line items: equity sales-and-trading revenue of "$7.42 billion" when "the street was expecting about $5 billion... basically almost a 50% beat," fixed-income and commodities trading at "$4.59 billion" versus "$3.76 the street was looking for," investment banking up "55%," advisory up "17%," and underwriting revenue that "over doubled where it was last year." On DHUnplugged, "#810: Big Blew Up" (July 15), the hosts tallied the bottom line: profit of "$6.6 billion for the quarter... $20.98 per share versus $14.48," a stock up "9% today at $1,140, a 52-week and all-time high," and roughly a 25% return on equity. And on CNBC's Squawk on the Street, "9am Hour: Big Bank Earnings Reaction" (July 14), Jim Cramer, who once worked there, was almost rueful: "Fastest EPS growth I've ever seen, 92%. Fastest revenue growth, 39%... Equity's up 72." Why it moves the thesis: this is exactly the operator-grade proof the bull case needed, and it beat even elevated expectations. Note the plain-English version of what happened: Goldman makes money three ways, advising on deals, underwriting new stock and bond sales, and trading, and all three fired at once.

2. The insiders finally spoke, and the tone was "great, but don't get greedy." This is the first week of the cycle with actual company bosses on the record, and the split between their confidence and their caution is the story. On Bloomberg Talks, "Wells Fargo CFO Mike Santomassimo Talks Earnings" (July 14), the CFO (an insider) said Wells "had a record fee quarter for investment banking fees this quarter across the firm," has "grown our balance sheet in the markets business by almost $200 billion since the end of 2024," saw that markets revenue "grow 24% in the quarter," and described deal pipelines as "quite healthy... across really all the products set there, including M&A... everything from healthcare to tech to media to... all industrial sectors." On Squawk on the Street, "11AM Hour: Gary Cohn & Morgan Stanley Earnings" (July 15), CNBC relayed that Morgan Stanley's CFO Sharon Yeshaya (insider) said simply, "we're playing our game and we play our game really well," and that CEO Ted Pick (insider) framed the AI build-out as a runway, not a peak: more than "$3 trillion will be spent over the next three years," it's "plausible" for "$10 trillion" over many years, and "you're basically looking at us being around 10 to 15 percent of the way through the investment cycle." But the caution was just as loud. On the same Squawk segment, JPMorgan's Jamie Dimon (insider) said on the call: "we're in a very healthy, active, exuberant market with very high prices and very high volumes. We benefit from that. We just don't know how long it will continue... it's getting close to as good as it gets." Wells Fargo's Charlie Scharf (insider) echoed it: "We know that such favorable conditions do not go on forever. And so we're being selective about how much and where to grow." Why it matters: when the people running the machine tell you it's firing on all cylinders and in the same breath tell you it can't last, position for the revenue now but respect the "as good as it gets" warning.

3. The other three reporters confirm it's broad, not a one-name fluke. The reopening isn't just a Goldman story. Per DHUnplugged (July 15) and Morning Brew Daily, "JPMorgan Has Historic Q2" (July 15): JPMorgan's markets revenue rose 35% to about $12 billion, including an 86% jump in equities trading to roughly $6 billion, with investment-banking fees of $3.28 billion and a record $21.2 billion profit (helped by a $4.6 billion gain on its Visa stake). Bank of America's net income rose 27% to $9.1 billion, with investment-banking fees up "50% to $2 billion" and equity trading up about 70%. Citigroup's net income jumped 45% to $5.8 billion, its "highest quarterly revenue in a decade," up 14%, with investment banking up 44% and equity trading up 45%, though the stock still fell about 5.5%. Wells Fargo posted a four-year high in fee income ($10.3 billion) but its stock dipped because net interest margin, the spread it earns between lending and deposit rates, is still under mild pressure. Why it matters: the strength shows up at the diversified giant (JPMorgan), the trading house (Goldman), the balance-sheet lender (Bank of America) and the turnaround story (Citi) all at once. That breadth is what separates a durable reopening from a lucky quarter, but note that only Goldman actually rose on its results; the market is rewarding the trading-and-advisory mix and punishing anything that smells like margin pressure.

4. The mega-deal count hit a record, and that's the fee pool refilling in real time. On DHUnplugged (July 15), the hosts flagged that global deals worth more than $10 billion reached a record level in the first half of 2026, making up roughly 43% of all newly announced M&A. The named deals kept coming this week: Uber's $14.8 billion move for Germany's Delivery Hero (on Motley Fool Hidden Gems, "Uber's New Acquisition and GE Aerospace's Search for Parts" (July 16)); GSK's $10.6 billion purchase of New Valent, which closed just 36 days after being announced (on Citeline's Scrip Deals Podcast (July 16)); Eli Lilly's roughly $3.8 billion deal for psychedelics-focused Atai Life Sciences ($2.8 billion upfront plus a $1 billion milestone-based payment), part of what one guest counted as Lilly's 17th deal of 2026 (on Biotech Hangout, "Episode 189" (July 17)); and Solstice's $4.5 billion cash-and-stock takeover of Element Solutions (on Chip Stock Investor, "Solstice's $4.5B Element Solutions Deal: Why It Sold Off" (July 17)). Why it matters: every one of these is advisory and financing fees for someone, and the record share of $10 billion-plus deals is precisely the pull-through the elite advisory franchises live on. The Citeline point is the subtle tell: deals are closing faster because the antitrust regulator is friendlier, which shortens the time between "announced" and "fee booked."

5. Warsh refused to declare victory, and rate odds whipsawed all week. The macro governor on the debt-funded half of the fee pool is Fed Chair Kevin Warsh, who testified before Congress on July 14 and 15. June inflation had just cooled, headline CPI fell 0.4% on the month to 3.5% year-over-year, with core at 2.6%. On Squawk on the Street, "10AM Hour: Warsh Testifies" (July 15), Warsh said that cooler data is "one data point" and he is "not satisfied" the inflation fight is over, noting producer prices still running near 5.5%. On Power Lunch (July 14), the cooler print swung the market's odds of a July rate hike from about 42% down to 12%, though December still showed roughly 42% odds of a quarter-point move. On TD Cowen Insights, "Kevin Warsh And The Future Direction For Rates" (July 14), TD's Gennady Goldberg said he sees no hikes and no cuts through 2026 to 2027, even as the market prices about 45 basis points of hikes by early 2027 and "nearly half the committee" sees a hike as appropriate this year. Warsh has also killed the Fed's forward guidance and its "dot plot," the chart that used to telegraph where rates were headed. Why it matters: the equity-underwriting engine can run on animal spirits, but the debt-issuance and leveraged-finance half runs on the cost of money. With Warsh deliberately refusing to signal his next move, financing volumes now carry more headline risk into every data point.

The debate

The core question is unchanged, but this week it got sharper: is this a durable, multi-year reopening of capital markets, or a fragile head-fake that cracks the moment the AI trade or the rate backdrop turns?

The durable case. This is no longer a promise; it's a P&L. Five banks just posted a 39% jump in revenue, with a record share of mega-deals and trading desks at or near all-time highs. The insiders are telling you there's a long runway: Ted Pick's "10 to 15 percent of the way through the investment cycle," Santomassimo's "record fee quarter" and pipelines that are "healthy across really all the products." The best articulation of the bull case came from an investor who used to run a bank: on Squawk on the Street, "11AM Hour: Former Barclays CEO on Bank Earnings" (July 16), Bob Diamond (now CEO of Atlas Merchant Capital) argued that "where else would you do an IPO for these AI companies other than the US?", that the strength runs from the mega-caps down to middle-market advisory, and that the AI build is a self-reinforcing income loop: investment spending is running about 3% of GDP and set to grow "5x" over three years, versus "2x in 5 years" during the dot-com boom, and "every dollar of investment spending for one institution is income for another institution." On credit, Diamond said he hasn't "seen a blip in credit since the financial crisis" and doesn't "see anything yet systemic brewing," a view echoed by tight spreads: on Bloomberg Surveillance, "TV: July 14th" (July 14), double-B credit spreads were described as sitting at "cycle tights."

The fragile case. The revenue is real; the durability is the question. The exit door, the ability to actually sell new stock and have it hold, got visibly tighter this week. SpaceX, last week's "textbook" IPO, fell below its $135 offer price (noted on Squawk on the Street, "9am Hour: SpaceX Falls Below IPO Price" (July 16)), following SK Hynix's record-then-crash the week before. The deal machine also caught a legal jolt: even Diamond conceded a "correction in credit with higher rates" is probably coming, and Dimon and Scharf both volunteered that "favorable conditions do not go on forever." The deeper worry is on the debt side. On Van Hesser's 3 Things in Credit, A KBRA Podcast, "AI Debt Issuance, Updating the Fed, and Big Bank Credit Color" (July 17), KBRA's Van Hesser tallied roughly $0.57 trillion of debt already issued by the five hyperscalers plus Nvidia and SpaceX, big-tech now 6.5% of the investment-grade bond index, and flagged JPMorgan's own commentary on "aggressive data center underwriting with power supply and tenant risks," alongside a full-year credit-card loss guide of 3.2%. And on Monetary Matters with Jack Farley, "Nick Nemeth: Private Credit Will Blow-up Insurance System" (July 20), the guest laid out the tail risk that private credit's "immense leverage" and "shaky loans" have quietly built run-risk into the retirement system.

My read. The equity-underwriting, advisory and trading engine is running hot, and this week it was confirmed in audited numbers, not anecdotes: the thesis moved from "should" to "did." But the market's own reaction is the tell: only Goldman rose on results, and the two soft spots (SpaceX below issue, a judge freezing a DOJ-cleared merger) both sit in the exit and approval plumbing, not the demand. The reopening is real, and it's getting choosier about price. I'd still favor the elite equity-underwriting and strategic-advisory names, Goldman, Morgan Stanley, over the volume leveraged-finance and pure-sponsor plays, where a friendlier Fed is priced in and any credit wobble bites first. The single biggest swing factor from here isn't demand; it's whether the new-issue market keeps absorbing supply, and whether Warsh's committee stays on hold.

Stocks in play

A note up front: the money-center banks reported this week and got heavy podcast coverage. The two exchanges (ICE, Nasdaq) and the three boutiques (Evercore, Moelis, Jefferies) report later in the season and drew little or no dedicated coverage, those gaps are flagged honestly below rather than papered over.

  • Goldman Sachs (GS). Bull: the cleanest print of the group, equity trading of $7.42 billion (a near-50% beat), investment banking up 55%, underwriting more than double last year, profit of $6.6 billion ($20.98 a share), and a stock that jumped ~9% to an all-time high around $1,140 (Schwab Network, July 14; DHUnplugged, July 15). Bear: the "episodic" businesses (trading, big-deal underwriting) that drove the beat are the first to fade if the new-issue window tightens, and SpaceX slipping below its offer price is a live example. Even Cramer's takeaway was to watch that Goldman leans into its "non-episodic" recurring lines. Catalyst / number to watch: whether the advisory backlog and trading momentum carry into Q3, and how the pipeline reads once the mega-IPO calendar (Anthropic, OpenAI) actually prices.
  • Morgan Stanley (MS). Bull: a top- and bottom-line beat, investment banking up 58% with beats across advisory and both equity and debt underwriting, and a record $148 billion of net-new money into wealth management, half of it minted by the quarter's mega-IPOs, since Morgan Stanley was a lead underwriter on SpaceX (Squawk on the Street, July 15). CEO Ted Pick (insider) thinks the AI capex cycle is only "10 to 15 percent" run. Bear: the stock barely moved because Goldman's blowout the day before had already priced in the good news, and the wealth inflows are partly a one-off function of IPO wealth creation that slows if the IPO window does. Catalyst / number to watch: wealth net-new-asset flows once the SpaceX sugar-high fades, and how "meaningful" a share of the AI capital-raising Morgan Stanley actually captures.
  • Intercontinental Exchange (ICE). Bull: the read-through is favorable, a heavy debt-issuance wave (KBRA's ~$0.57 trillion of hyperscaler/AI bonds) and choppy rate expectations feed rates and fixed-income data and futures volumes. Bear: the mortgage-data segment stays tied to a frozen, high-rate housing market, and higher-for-longer rates don't help. Catalyst / number to watch: Q2 results (not yet reported), specifically the transaction-versus-recurring-data revenue mix. (No ICE-specific podcast coverage this week, an honest gap; the exchange discussion was all IPO listings and the Nasdaq/NYSE venue debate.)
  • Nasdaq (NDAQ). Bull: the listings pipeline behind it is deep, Anthropic targeting an October listing, OpenAI guided to a 2026 IPO, plus a queue that includes Moonshot AI, Jersey Mike's, Cumberland Farms and Reformation (Bloomberg Intelligence, July 20; Market Maker, July 20). Bear: listings revenue is reflexive and the after-market has turned hostile, SpaceX below its offer price and SK Hynix's record-then-crash show how fast fresh supply can meet a tired market. Catalyst / number to watch: whether Anthropic and OpenAI choose Nasdaq, and the Q2 listings-and-data revenue mix. (No Nasdaq-specific earnings coverage this week, the franchise surfaced only through the IPO-pipeline discussion.)
  • Evercore (EVR). Bull: a genuinely busy strategic-deal tape, Uber/Delivery Hero, GSK/New Valent, Lilly/Atai, Solstice/Element, plus a record share of $10 billion-plus deals, is pure boutique-advisory fuel, and Bob Diamond specifically called middle-market advisory "outstanding." Bear: the sponsor-funded half of the deal market stays capped while rates bite, and boutiques have no trading ballast to smooth a quiet quarter. Catalyst / number to watch: Q2 advisory backlog and managing-director productivity. (No Evercore franchise coverage this week; the only mention was an Evercore analyst's Apple research note on Mac OS Ken, July 17, sell-side research, not the advisory business.)
  • Moelis (MC). Bull: a pure-play advisor that captures the same strong M&A pool, with restructuring optionality if the leverage building in AI-linked debt and private credit ever turns to stress. Bear: no trading or lending ballast at all, so it is the most exposed of the seven to any deal-volume air pocket. Catalyst / number to watch: Q2 advisory revenue and any shift in the restructuring mix. (No Moelis-specific podcast coverage this week, an honest gap.)
  • Jefferies (JEF). Bull: leveraged-finance plus advisory leverage to a real reopening, and, with its off-calendar quarter, often the group's earliest tell on where fees are heading. Bear: most directly in the path of any leveraged-finance slowdown, and the AI-debt concentration KBRA flagged is an early yellow flag for the debt machine. Catalyst / number to watch: its next print as the group's leading indicator. (No Jefferies-specific podcast coverage this week, an honest gap.)

Read-throughs

  • Boutique advisors (EVR / MC / JEF). The strategic-deal tape is unambiguously busy, Uber/Delivery Hero ($14.8 billion), GSK/New Valent ($10.6 billion), Lilly/Atai (~$3.8 billion), Solstice/Element Solutions ($4.5 billion), Sky/ITV (£1.6 billion), on top of a record share of $10 billion-plus mega-deals (DHUnplugged, July 15; Motley Fool Hidden Gems, July 16; Citeline Scrip, July 16; Chip Stock Investor, July 17). Two structural tailwinds help the boutiques specifically: deals are closing faster under a lighter-touch FTC (GSK closed New Valent in 36 days per Citeline), and Bob Diamond flagged middle-market advisory as "outstanding" (Squawk, July 16). The catch: boutiques earn on completed deals with no trading cushion, so any air pocket shows up in their results faster than at the diversified banks.
  • Leveraged finance & private credit. Two signals pull in opposite directions. Reassuringly, spreads are calm, double-B credit at "cycle tights" (Bloomberg Surveillance, July 14) and Diamond seeing nothing "systemic brewing." Worryingly, the AI-debt build is getting concentrated fast: KBRA's Van Hesser counted roughly $0.57 trillion issued by the hyperscalers plus Nvidia and SpaceX, now 6.5% of the investment-grade index, with JPMorgan itself flagging "power supply and tenant risks" in data-center underwriting (Van Hesser's 3 Things in Credit, July 17). And the tail risk is louder this week, the argument that private credit's leverage has seeded "run risk" in the insurance and retirement system (Monetary Matters, July 20). One insider data point worth holding onto: on Bloomberg Talks, "BofA CEO Brian Moynihan Talks Deals" (July 16), Moynihan said the banks are winning share back from private-credit funds, where "investor money going into the private capital funds... has been slowed" with "withdrawals and people getting out."
  • PE sponsors monetizing exits. The exit math is the week's asterisk. The IPO pipeline is deep, Anthropic (October target), OpenAI (2026), Moonshot AI ($30 billion Hong Kong listing), Jersey Mike's ($1.1 billion at an $8 billion value), plus Cumberland Farms and Reformation, but SpaceX slipping below its offer price and SK Hynix's record-then-crash show the after-market is no longer a sure thing. On Capital Ideas, "2026 Midyear Outlook" (July 16), the guest noted more than $8 trillion of money-market cash on the sidelines that could fund these deals, but warned that historically hot IPO markets have signaled tops, drawing a late-1990s parallel. Sponsors can monetize into this window; the question is at what discount.
  • Exchange listings & data. A full IPO calendar keeps the listings franchise compounding for Nasdaq, and the AI-driven investment-grade issuance wave feeds ICE and Nasdaq fixed-income data over time, as long as investors keep buying the paper. On Thoughts on the Market, "A Test for Capital Markets: Funding AI" (July 16), Morgan Stanley's team sized the shift: data-center bond issuance has run about $30 billion across 15 deals since fall 2025, and hyperscaler issuance has grown from under 1% of the investment-grade market to over 10%. That's a durable new data-and-transaction stream for the exchanges, and a concentration risk if AI issuance stalls.

What changed vs. last week

Last week was a preview: we said the big banks report "this week" and that Goldman's prior-quarter IB was up 48%, the story was still analyst-and-economist commentary, SK Hynix had just cracked after listing, and the Fed backdrop was "no move until at least December." This week the forecasts became facts. Five updates:

  • From "the proof arrives this week" to a blowout on the board. Last week: an anticipated bank-earnings season with Goldman's 48% Q1 IB growth as the tease. This week: five banks up 39% in revenue, Goldman's IB up 55% and equities up 72%, JPMorgan's biggest profit ever, and, for the first time, CEOs and CFOs (Santomassimo, Pick, Yeshaya, Dimon, Scharf) putting their names on it. The thesis moved from anecdote to audited numbers, and it beat elevated expectations.
  • The exit door tightened, SpaceX joined SK Hynix below the waterline. Last week SK Hynix suffered its worst-ever drop days after listing. This week SpaceX, the "textbook" mega-IPO, fell below its $135 offer price. Two of the marquee listings the reopening thesis leaned on are now trading underwater, a real warning that fresh supply is meeting a choosier market.
  • Paramount escalated from "sued" to "frozen." Last week a coalition of 12 states led by California's Rob Bonta filed suit to block the $110 billion Paramount and Warner Bros. deal despite June's DOJ clearance. This week a federal judge paused the deal outright, ordering a 14-day hold and teeing up a preliminary-injunction ruling around July 22 (Bloomberg Businessweek, "Judge Pauses Deal With Serious Issues" (July 20); Streaming Into the Void (July 20)). A live reminder that even a DOJ-cleared mega-deal can stall, a headwind for advisory close rates on the very biggest transactions.
  • The IPO pipeline firmed to a date. Last week the read was "Anthropic and OpenAI very likely within 6 to 9 months." This week Anthropic is reported scheduling IPO talks with an October listing target and OpenAI is guided to a 2026 IPO (Prof G Markets, July 20; Bloomberg Intelligence, July 20), though OpenAI heads into it under an Apple trade-secrets lawsuit.
  • The Fed's near-term hawkishness eased, but the governor stayed non-committal. Last week: no move seen until at least December. This week: a cooler June CPI knocked July-hike odds from ~42% to ~12%, but Warsh testified he's "not satisfied" inflation is beaten, and forecasters still pencil in the possibility of one hike before year-end. Net: less urgency to hike now, but a Fed chair who refuses to signal keeps financing volumes hostage to every data point.