The AI Earnings Mirage: What IBM’s Historic Crash and Record Bank Profits Reveal About the Market’s Dual Bubble

On a mid-July afternoon, the stark divergence of the American financial system was laid bare in two seemingly unrelated corporate disclosures. On one side, Wall Street’s banking titans were reporting profits that defied historical precedents. On the other, International Business Machines Corp. (IBM), a cornerstone of global technology for over a century, suffered a catastrophic 25% single-day stock plunge—the worst trading session in its 115-year history.

This striking juxtaposition caught the attention of Steve Hanke, a professor of applied economics at Johns Hopkins University. Known widely in international policy circles as the “money doctor” for his decades of advising heads of state and treasury departments, Hanke observed that this divergence points to a structural mispricing in global markets. He argues that we are witnessing a “dual bubble” forming in AI markets—one defined not just by how much investors are willing to pay for earnings, but by the fundamental sustainability of those earnings themselves.

The day before our conversation, IBM’s preliminary second-quarter numbers sent shockwaves through the technology sector. The company’s revenue of $17.2 billion missed the Wall Street consensus of $17.9 billion by a seemingly modest 3.7%. Adjusted earnings per share (EPS) came in at $2.93, falling short of the $3.02 analysts expected. More critically, the preliminary disclosure revealed that IBM’s revenue grew by just 1%, far below the 5% expansion the market had priced in.

The reaction was swift and merciless. Investors wiped out roughly $40 billion in IBM’s market value in a single session, a collapse steeper than Enron’s drop on the day the Securities and Exchange Commission opened its accounting inquiry.

IBM Chief Executive Officer Arvind Krishna made no attempt to sugarcoat the disappointment. In an unusually candid letter to stakeholders, Krishna acknowledged the execution failures. Current market conditions required “our teams to execute perfectly,” he wrote, “and this quarter we faltered.” He emphasized that his message offered “not excuses, but … realities.”

The sudden drop prompted immediate soul-searching across the tech industry. The New York Times’ DealBook questioned if the IBM miss was a “canary in the tech coal mine,” while Richard Waters, the Financial Times’ West Coast editor, described it as a stark “warning to the IT sector.” To many, it felt like the tangible arrival of the “SaaSpocalypse”—a theoretical market panic where artificial intelligence rapidly displaces traditional software services, disrupting established enterprise business models.

Traditionally, market bubbles are valuation-driven. In these scenarios, such as the dot-com crash of 2000, asset prices outstrip corporate profits, pushing price-to-earnings (P/E) ratios to unsustainable heights. However, Hanke and other market strategists point to a different, more insidious phenomenon: an “earnings bubble”. Here, the corporate profits themselves are temporarily inflated by cyclical factors, loose credit, or unsustainable capital expenditure, making valuations look deceptively reasonable even while the underlying fundamentals are highly fragile.

Peter Berezin, chief global strategist at BCA Research, has spent months arguing that the current artificial intelligence trade is “primarily an earnings bubble rather than a valuation bubble.” Berezin notes that these dynamics historically cluster in highly cyclical, boom-and-bust sectors—such as pre-2008 banking, pandemic-era work-from-home stocks, and resources, airlines, and semiconductors. The latter is currently the bedrock of the massive AI infrastructure buildout.

The danger of an earnings bubble lies in its invisibility. Analysts are notoriously poor at anticipating when these profit cycles will turn. This creates a distinct “detection lag”, where Wall Street only downgrades its profit estimates after a stock has already collapsed. Berezin pointed out in late May that analysts struggle to predict these inflection points because stock prices typically plunge well before formal profit projections are revised downward.

This lag was on full display following IBM’s crash. Bank of America (BofA) and UBS lowered their estimates only after the 25% drop had occurred. BofA cut its price target from $330 to $280, while maintaining a Buy rating, arguing IBM remains “well positioned” once execution issues are resolved. UBS held its target at $236 but trimmed its 2026 EPS forecasts. Meanwhile, HSBC downgraded the stock to Reduce, and Goldman Sachs warned that the disappointing results would “fully validate the software bear case scenario.”

While tech faltered, the banking sector painted a picture of extraordinary abundance. JPMorgan Chase posted a net income of $21.2 billion—the highest quarterly profit for any bank in U.S. history. Goldman Sachs reported an 84% surge in net earnings attributable to common shareholders to $6.4 billion, on total revenues of $20.34 billion, up 39%.

To Hanke, these record banking profits are not an isolated triumph but a window into the monetary mechanics driving asset prices. He argues that the liquidity inflating these markets is generated primarily by private commercial banks rather than the Federal Reserve alone. This dynamic recalls the famous observation by midcentury economist John Kenneth Galbraith: “The process by which banks create money is so simple that the mind is repelled.” Reflecting on his only meeting with Galbraith, Hanke agreed with the sentiment, noting, “Although my orientation is not the same as Galbraith’s, I thought he was a great man and had many admirable qualities.”

When asked if record bank profits indicate that credit is flowing freely enough to inflate both asset prices and the corporate earnings that justify them, Hanke responded: “What you’re saying is that markets are getting “mugged by reality.””

Even JPMorgan CEO Jamie Dimon signaled caution. While celebrating earnings that were “close to as good as it gets” during a call with analysts, Dimon warned of excessive “exuberance” in the markets, echoing Hanke’s concerns about overextended optimism.

The prevailing bull case for tech giants like Nvidia and Alphabet relies on their massive cash flows and forward P/E ratios of around 22x for the S&P 500—well below the 25x-plus peak of the dot-com era. But this defense ignores whether these earnings are being artificially propped up by “circular AI investment”—where tech companies invest in one another’s platforms, creating a feedback loop of capital expenditure that inflates short-term revenues across the sector.

IBM’s sudden crash suggests that the market’s tolerance for earnings disappointment is wearing thin. Whether this is an isolated operational slip or the beginning of a broader repricing of the AI earnings narrative remains to be seen. At press time, IBM shares were down 2% in intraday trading, leaving investors to ponder whether the corporate earnings backing the tech boom are as robust as they appear.