NextFin News - Bank of America Corp. Chief Executive Officer Brian Moynihan warned Monday that third-quarter trading revenue will be "relatively flat" compared with a year earlier, sending bank shares tumbling as investors absorbed a core-inflation print that ran hotter than expected, oil above $108 a barrel, and President Donald Trump's public clash with AI executives over how fast the technology should advance.
The comments, delivered at the Barclays Global Financial Services Conference in New York, mark the first hard read on third-quarter markets activity from a money-center bank and break the narrative that the trading surge of the first half would carry through the year. Investment-banking fees are projected at roughly $1.6 billion to $1.8 billion, below the roughly $2 billion analysts had penciled in. Bank of America shares slumped as much as 6 percent intraday, the steepest drop since April of last year, before closing down 5.9 percent around $59 a share.
One week earlier, Moynihan had called the near-collapse of the highly leveraged AI hedge fund Situational Awareness a "warning shot" for financial markets. Bank of America was among the fund's prime brokers. Now the same executive is flagging softer trading activity just as the Federal Reserve confronts sticky inflation and the White House tells the AI industry to keep its foot on the gas. The combination raises a question markets would rather not answer this week: what happens when the banking sector cools at the same moment the biggest capital-spending boom in the economy is being told not to slow down?
One Word From a Bank CEO, and the Whole Sector Wobbles
Moynihan's phrase was measured - "relatively flat" - but the market heard it as a warning because trading revenue is the swing factor in bank earnings. In the second quarter, Bank of America's Global Markets division delivered total sales and trading revenue of $7.1 billion, up 33 percent from a year earlier, with equities trading alone surging 70 percent to $3.6 billion. A return to flat growth in the third quarter implies that the surge was a cycle, not a new baseline.
The investment-banking line matters just as much. Fees projected at $1.6 billion to $1.8 billion sit well below the roughly $2 billion consensus, suggesting that the dealmaking recovery Wall Street has been betting on remains uneven. When the largest retail deposit gatherer on Wall Street signals a stall in both trading and advisory activity, investors do not wait for the peer group to confirm it - they reprice the whole sector first and ask questions later.
"These are all warning shots," Moynihan said in August, after the Situational Awareness episode forced a fire sale of equities to Citadel. "Valuations get out, leverage in the system gets there. You have to be careful."
That caution reads differently now. A CEO who warned about leverage in August is warning about revenue in September, and the market is treating the second warning as the more immediate one. Bank of America fell 5.9 percent on Monday; peers followed lower, with major diversified banks down between 1 and 2 percent by the close.
The mechanism is straightforward but unforgiving. Bank valuations embed an assumption that elevated trading activity persists long enough to offset pressure on net interest income from any shift in the rate path. Moynihan's comment attacks the first leg of that assumption. If trading revenue mean-reverts while rates stay higher for longer, the earnings bridge that supported bank multiples this year has one fewer pillar.
The AI Slowdown Debate Just Became a Political Fight
While banks flagged caution, the political arena moved in the opposite direction. On Saturday, September 12, Anthropic Chief Executive Officer Dario Amodei published a 3,815-word essay calling for the industry to "slow the pace at which we improve the capabilities of AI models." "I won't lie to you - there are real dangers," Amodei wrote. "And I think for too long the industry lied to people about the fact that this technology had risks."
The appeal drew immediate support. OpenAI Chief Executive Officer Sam Altman posted that he agreed "we need to pace the frontier." Elon Musk, whose xAI competes directly with both companies, added only: "Dario is right." The unusual alignment of three rival chief executives - at a moment when Anthropic is widely expected to pursue a public listing - is itself a signal that safety concerns have moved from internal risk committees to the open.
President Trump's response came on Monday via Truth Social, and it left no ambiguity about which side the administration is on. "The only control or 'guardrails' that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!" he wrote, adding that the administration "has stopped AI 'people' from doing bad, or potentially bad, 'things,' like Dario (Anthropic!), who is now pretending to be a 'perfect little angel'." In a second post, he warned of a "SICK conspiracy going on against AI and Data Centers," declared that "WHOEVER WINS AI, WINS," and closed with: "Conspiracy Theorists, Treasonists, Traitors, and Leakers, BEWARE!"
The framing matters for markets because it narrows a technical debate into a national-security contest. David Sacks, the White House AI and crypto czar, made the economic logic explicit on X, arguing that Anthropic and OpenAI hold a "duopoly on frontier intelligence" and should not use government to force competitors to slow as well. "Stop pretending you need anyone else's permission," Sacks wrote.
House Speaker Mike Johnson struck a more cautious note, telling a Sunday program that guardrails were necessary but that Congress could not "rush in and have Congress do something that would be detrimental to national security." The tension between those two positions - speed as security versus restraint as responsibility - is now the operating environment for every AI company with a roadmap and a regulator watching it.
Behind the rhetoric is a concrete trigger. Anthropic's September report on AI misuse, covering December 2025 through August 2026, catalogued seven categories of "disrupted misuse" of its Claude chatbot, including cyber operations, influence operations, surveillance, scams, biological misuse, weapons development, and model distillation by competitors. A former Anthropic researcher, Jacob Coxon, resigned last week saying the companies were "gambling with our lives." "The people building AI earnestly believe that it could kill us all by the end of the decade," Coxon wrote. "No other human activity poses this level of danger."
Public sentiment is moving with the executives, not the president. A recent Pew Research Center study found that roughly two-thirds of U.S. adults believe AI is advancing too quickly, and 67 percent said they had little or no confidence that the government could regulate the technology. The gap between that skepticism and the administration's speed-first posture is where policy risk now lives.
Inflation, Oil, and the Fed: The Real Constraint on the Rally
Beneath the earnings and politics, the binding constraint on risk assets is unchanged: the Federal Reserve's reaction function to inflation, now compounded by an oil shock. The Labor Department reported on September 11 that the consumer price index rose 0.4 percent in August and 3.4 percent over the year, in line with forecasts. But core inflation - excluding food and energy - rose 0.3 percent for the month, above the 0.2 percent consensus, leaving the annual core rate at 2.4 percent.
The market's response was to push the odds of a rate increase at the September 17 meeting above 80 percent. That is a stark reversal from the easing expectations that carried equities to record territory earlier in the year, and it changes the discount rate applied to every growth story in the market - including the AI buildout that Oracle and Adobe just reported on.
Oil added fuel. Brent crude rose $3.62, or 3.46 percent, to $108.23 a barrel, and West Texas Intermediate climbed $3.15 to $103.20, after strikes on Saudi Arabian oil infrastructure and renewed attacks on commercial vessels in the Strait of Hormuz over the weekend. An Omani official said a scheduled Monday meeting between Gulf countries and Iran to discuss the strait had been postponed. Energy at these levels feeds directly into the inflation numbers the Fed is watching, tightening the noose around the rate path.
Earnings quality, for the moment, matters less than the rate backdrop. Oracle reported fiscal first-quarter revenue of $19.3 billion, up 30 percent year over year, with adjusted earnings of $1.92 a share against a consensus of $1.74. Cloud infrastructure revenue more than doubled to $7.4 billion, and the company raised full-year guidance to at least $90 billion in revenue and $8.10 in adjusted earnings per share. Shares jumped roughly 8 percent. Adobe, by contrast, posted third-quarter revenue of $6.76 billion and non-GAAP earnings of $6.13 a share, both ahead of estimates, and raised full-year guidance to $26.58 billion to $26.63 billion in revenue and $24.45 to $24.50 in earnings - yet its shares slipped about 4 percent on Friday after current-quarter revenue guidance came in roughly in line.
Two strong prints, two different tape reactions. The market is no longer rewarding the beat; it is discounting the guidance against a higher rate path. That is the second-order effect of the inflation print: good news is no longer good enough unless it comes with pricing power that can survive a hiking cycle.
Cyclical Cooling Meets a Structural Boom - and They Compound
The right read separates the two forces rather than blending them. Bank of America's trading warning is cyclical. Trading revenue mean-reverts; a single quarter of flattish activity after a strong first half does not break the bank model, and history offers multiple cycles in which a hot first half gave way to a softer third quarter before activity returned. The evidence for a cyclical call is the seasonality of client activity, the dependence on volatility regimes, and the fact that one executive's read - however influential - is not a sector-wide dataset.
The AI-policy collision, by contrast, is structural. A regime in which the executive branch defines AI development speed as a national-security asset, in which frontier labs are publicly divided over pacing, and in which misuse categories now include biological and weapons applications, is not a sentiment swing. It will not revert on its own. The rules of the game for the largest capital-spending cycle in the economy are being rewritten in public, in real time.
The risk is not that either force dominates; it is that they compound. A cyclical cooling in bank trading revenue reduces the financial system's appetite for risk exactly when a structural boom in AI infrastructure needs continued access to cheap capital. If the Fed hikes into that junction - which the August CPI and $108 oil now make more likely - the two trades that carried this year's rally face the same headwind from opposite directions.
What to Watch
The falsifying signal for the sector-wide read is simple and dated: when JPMorgan Chase and Goldman Sachs report third-quarter trading revenue, a double-digit year-over-year increase at both banks would indicate that Moynihan's comment was specific to Bank of America rather than an industry-wide stall. Conversely, flat or declining trading revenue across the money-center peers would confirm that the first-half surge has run its course.
On the policy side, watch whether the administration converts rhetoric into action - an executive order, procurement restrictions, or export-control changes tied to AI development pace would move the debate from statements to binding constraints. On the inflation side, the September 17 Federal Open Market Committee decision and its updated rate projections are the pivot: a hike with hawkish forward guidance extends pressure on both bank multiples and growth valuations; a hold with a pause signal would relieve it.
Scenarios for the weeks ahead split cleanly, and they point in different directions across time horizons. In the short term, sentiment is hostage to the Fed meeting and oil headlines - a hawkish hike or a spike past $115 a barrel would extend pressure across both value and growth. Over the medium term, third-quarter earnings season will confirm or refute Moynihan's read; a broad trading stall would force a sector re-rating even if the macro stabilizes. In the long term, the structural question is whether AI capital spending can keep compounding if the cost of capital stays elevated - and whether a political mandate for speed can coexist with the safety constraints the industry's own leaders are demanding.
The base case remains navigable: trading revenue proves merely cyclical, the Fed hikes once and pauses, and AI capital spending continues at pace. The upside case requires core inflation to cool in the next two prints, oil to stabilize below $100, and bank trading to rebound on renewed volatility, allowing both halves of the market to participate. The downside case stacks the headwinds: trading weakness spreads across the money-center peers, oil pushes toward $115 on further Hormuz disruption, and the Fed signals more than one hike - a combination that would test the year's gains across both value and growth.
The market is being asked to price a banking sector that is cooling at the same moment the political class is demanding the AI boom stay hot. Those two trades do not sit comfortably in the same portfolio, and the Fed's meeting next week will decide which one gets the benefit of the doubt.
Explore more exclusive insights at nextfin.ai.
