Earnings season is back, and the big Wall Street banks get the first turn at the microphone.
Today, JPMorgan Chase opened with a record $21.2 billion profit for second-quarter 2026, giving financial television something shiny to wave around before anyone asks what is still hiding on the balance sheet.
The headline numbers matter. So do net interest income, trading revenue, credit-card delinquencies and loan loss reserves.
The big number to keep an eye on? Banks are still sitting on more than $300 billion in unrealized losses from bond holdings and troubled loans.
While improving, banks continues to sit on over $300 billion in unrealized losses from bond holdings and bad loans. (Source: Barchart)
Unrealized gains have improved by half since the 2023 bank crisis that never happened.
From March 10, 2023, to May 1, 2023, a run on banks took out Silicon Valley Bank (SVB), Signature Bank and First Republic Bank – three of the largest five bank failures in U.S. financial history.
Following the pandemic surge in spending, inflation roared, and the Federal Reserve responded, slowly at first, by hiking overnight rates. When rising rates hit bank portfolios like a dropped piano.
The failure of Silicon Valley Bank was a textbook case of an asset-liability mismatch and unhedged interest-rate risk. Flush with tech startup venture capital, bank leadership failed to manage basic risk metrics – even operating under the San Francisco Fed’s nose without a Chief Risk Officer (CRO) for roughly eight months leading up to the collapse.
Across the industry, if banks can hold long bonds to maturity, the losses stay on paper. If depositors get wind of the impropriety, the losses force disaster.
Bank stress matters. A market wobble can quickly morph into something larger – a grey swan – when trust in the banking system erodes.
Unrealized losses, bad-loan reserves, commercial real estate exposure, private-credit commitments and AI-related capital raises backed by bridge loans or bond deals – all come under scrutiny or, worse, go bad all at once.
At the time, we outlined the scenario in our special report Anatomy of A Bust. You can find it in the site’s special report archives.
If you still want exposure to financials, especially during this epic financing boom during the AI buildout, insurance may be the safer, if duller, niche.
As Warren Buffett put it, insurance float is “money that doesn’t belong to us but that we can invest for Berkshire’s benefit.” The customer pays first. The claim comes later. Sometimes it never comes at all.
Good insurers price risk, collect premiums, invest conservatively and try to earn an underwriting profit before anything fancy happens.
Bad underwriting can turn float into gasoline, especially after hurricanes, wildfires, litigation spikes or badly priced long-tail liabilities.
A disciplined insurer with strong reserves, conservative investments and sober risk management can become a quiet compounder. It will not promise to reinvent civilization by Thursday. It collects premiums, invests carefully and lets time do the heavy lifting.
Today’s Grey Swan Pro reveals an insurance company with sizeable underwriting profits in several niche markets that has outperformed the financial sector as a whole by nearly 6X over the past 20 years — details here.
~ Addison
P.S. This week on Grey Swan Live!, we return to one of our favorite themes: Argentina’s economic turnaround. While DOGE met the Washington establishment and the establishment won, Mieli’s Argentina has truly taken a chainsaw to the administrative state.
We’ll get the latest from our man on the ground, Joel Bowman, author of Notes From the End of the World.
