Tom Michaud joined Keefe, Bruyette & Woods 40 years ago as a credit analyst, when the firm employed 75 people. He has been its chief executive for 15. In the spring of 2023, when the regional banking system was running on rumor and panic, he went to Capitol Hill and testified about deposit insurance, and in his conversation with American Banker's Kate Berry this week he returned to the subject with the flatness of a man who has watched the moment pass. "The financial system bent but it did not break," he said. And because it did not break, the fix is not coming. He is, by his own account, nervous about whether reform will pass at all.
The crisis that stopped
The sequence is worth reconstructing, because it is the reason everything since has been calm. In March 2023, Silicon Valley Bank failed, then Signature, then First Republic, three institutions falling in a line that looked, for two weeks, like the beginning of something much larger. Michaud credits a single decision with stopping it: the emergency exemption that insured all deposits at the failed institutions, removing the question that had been pulling deposits out of every midsize bank in the country, which was whether payroll money was safe. The contagion stopped. The system held.
The lesson Michaud drew is specific. The magic of banking, as he puts it, runs on deposit funding and payments, and when depositors decide it is time to run, insurance and capital are the only things standing between a rumor and a collapse. The 2023 crisis proved the mechanism, down to the hour. It also proved the limits of the standing insurance: the $250,000 cap left the operating accounts of thousands of businesses exposed, and only an emergency measure, not the law, caught them.
The deterrent that disappears when it works
That is the trap at the center of the deposit reform debate. The case for reform was never stronger than in the weeks after the crisis. It was never more demonstrable, never more politically alive. But the emergency worked, the panic subsided, and the urgency evaporated, because urgency runs on visible failure and the failure had been prevented. The deterrent succeeded, and success is why the deterrent was never made permanent.
The legislation exists. The Main Street Depositor Protection Act, reintroduced in March 2026 by Senator Bill Hagerty, Senator Angela Alsobrooks, and Representative Frank Lucas, would direct the FDIC to set insurance limits for noninterest-bearing business transaction accounts somewhere between $250,000 and $5 million, covering the payroll and operating money that runs through accounts far above the current cap. And there it sits. The bill's own history is a record of retreat: the first version set a flat $10 million cap, and the reintroduction scaled the ceiling down to $5 million, leaving the number to the FDIC, in an explicit attempt to buy consensus that has not yet been bought. Taxpayer advocacy groups have called it reckless and a solution in search of a problem, citing possible costs in the tens of billions through assessments and premiums. The FDIC itself has cautioned that expanded coverage would thin the reserve ratio and require banks to start reporting account data they do not currently separate. Treasury Secretary Scott Bessent has urged raising the cap to correct a competitive imbalance favoring the biggest banks, and the community banks and credit unions that would benefit are nonetheless fighting each other over parity of coverage. Big banks resisted the earlier, larger version. Community banks and credit unions, nominal allies, argue with each other over parity of coverage. The FDIC itself cautions that expanded coverage would thin the reserve ratio. A bipartisan bill with a real coalition behind it is nonetheless not moving, and Michaud, whose firm's business is the banking system, expects nothing soon.
The pattern is almost a law of its own. Insurance that prevents a panic looks like an unnecessary expense in the years the panic does not come. The years it does come, it is too late to pass anything.
The 12-to-1 business
Michaud's other warnings are the professional habits of a man who has watched banks lever capital 12-to-1 through every cycle since 1983. Interest rate risk, he says, is one of the supreme risks inside a bank, and 2023 taught that held-to-maturity accounting was not the safe harbor everyone believed. His structural observation about the industry is the one that connects to the reform fight: the big banks hold most of the nation's deposits but do not make most of the nation's small-business loans. The institutions that actually finance the payroll of the Main Street economy, the regionals that delivered more than half of the government's COVID assistance, are the ones whose depositors sit closest to the cap.
That asymmetry is the whole argument in one sentence. The system's deposit base has concentrated in banks that do not need it, while the banks that need it operate their customers' payroll money one rumor away from the statutory limit. The reform would fix the asymmetry. The calm is what lets the asymmetry stand.
The system he watches now
Michaud's view of the current system has a specific edge. The economy in August 2026 is strong, loan growth is robust, and the deposits that matter to a bank analyst are growing again, in an era he considers the end of zero and low interest rates. But he watches the risk migrate, and his example is precise: the margin call that forced a prominent hedge fund to sell its leveraged stock portfolio to Citadel, a stress event that happened outside the banking system, in the private credit and leverage complex that now runs parallel to it. The banks that survived 2023 are safer than they were; the risks have not disappeared, they have moved to places with thinner supervision and no deposit insurance at all. His point about the institution of last resort cuts the same way: when support is truly needed, it will end up being the banks, because they are the system's shock absorbers by construction.
That is why he keeps returning to deposit insurance modernization as the unfinished business. A bank system that levers capital 12-to-1, with interest rate risk as one of its supreme internal risks, runs on depositor confidence, and confidence is a public good the system pays for with insurance. The cap is the seam where the public good ends and the panic begins. Michaud's pessimism about reform is the pessimism of someone who has watched the seam hold once and watched Washington conclude the seam no longer matters.
The firm that rebuilt
There is a second story in the interview, older and quieter, and Michaud tells it the way people tell things that still weigh. On September 11, 2001, KBW's offices sat on the 88th and 89th floors of the South Tower, above the point of impact. Sixty-seven employees were killed, roughly a third of the firm's New York staff, including the son of the chief executive and a co-CEO. Most of the research and trading departments died that morning. The firm has never stopped commemorating it: a flag at headquarters whose stripes are made of the names, a memorial plaque at the Central Park Zoo, a moment of silence on the trading floor every year at 9:03, the minute the second plane hit.
The survivors rebuilt the firm, and Michaud's explanation of why is the most important sentence in the interview. They did not want the terrorists to win, he says, and they did not want the life work of the people who died to end on that day. The rebuilding was not sentiment. The firm paid every family the employee's salary through January 2002 plus the year-end bonus, extended health coverage for years, donated about $15 million in commissions to a family fund, and, as an employee-owned firm, paid out roughly $40 million to the families of those who died. Individual portraits of all 67 colleagues hang in the September 11 Memorial and Museum. KBW went public in 2006 and merged with Stifel in 2013. The rebuild is complete, and the obligation behind it never ended.
Protection has to be built before the proof
The two stories are the same idea at different scales. A firm survives its worst day because the people left refuse to let the worst day be the ending. A banking system needs the same refusal, applied in advance: deposit protection built before the panic, kept through the calm, renewed precisely when nothing appears to need it. The system bent in 2023 and did not break, and that is why the reform is stalled. The proof of the problem is also the proof the problem is survivable, and survivable problems do not get fixed until the next time they are not.
Michaud's pessimism is the honest read on all of this. The calm after the bank runs is exactly what reform was supposed to preserve, and the calm is what is delaying it. The people who rebuilt KBW did not wait to see whether the firm would be needed again. Deposit insurance reform is waiting for exactly that.
Primary sources
- American Banker's interview with Tom Michaud by Kate Berry for the account of his career, his testimony, his views on deposit insurance reform and its politics, the 2023 crisis sequence, and his recollections of September 11 and the firm's rebuilding.
- The Main Street Depositor Protection Act's terms and sponsors for the announcement on Representative Lucas's office site, and the opposition from taxpayer advocacy groups for the letter covered by VitalLaw.
- The New York Daily News' reporting for the details of KBW's losses and memorials.