Jackie Reses and Kris Dickson on the SVB Failure and Depositor Confidence

As Silicon Valley Bank’s failure unfolds in March 2023, Harry Stebbings speaks with Lead Bank chair and CEO Jackie Reses and CFO Kris Dickson about what happened, depositor confidence and possible outcomes. The discussion took place before the guests knew the resolution; their predictions and financial assessments should be read in that historical context.

Reses describes Lead Bank as a Kansas City community bank that also supplies infrastructure to fintech companies and works with venture firms. She has worked in financial services and served on a San Francisco Federal Reserve advisory council. Her position at another bank is relevant to her comments on banking and deposit flows.

Dickson says she joined Lead Bank after a decade working on the Lehman Brothers post-bankruptcy parent estate, first as chief accounting officer and then CFO. She says the estate had liquidated and distributed more than $129 billion through the end of 2022. She brings that resolution experience to the discussion while noting that the cases differ.

Dickson starts with banking’s basic risk-management challenge: deposits can be withdrawn while funds have been invested in assets with different yields and liquidity. A bank must balance earnings with the ability to meet withdrawals, and both rates and depositor behavior can change quickly.

She says SVB was unusually concentrated in the venture-backed technology ecosystem. During the deposit surge of 2021, with rates near zero, it invested heavily in long-dated securities, largely mortgage-backed securities. Holding them to maturity signaled that the bank did not intend to sell them to meet liquidity needs.

As rates rose in 2022 and venture funding slowed, Dickson says deposits fell and the market value of the securities dropped. She cites approximately $91 billion in held-to-maturity securities with a market value of $76 billion at year-end. The unrealized gap was not itself the same as a realized loss, but became important as withdrawals accelerated.

Dickson describes a sequence in which depositors withdrew, asset sales generated losses and an attempted capital raise failed to restore confidence. She says depositors sought more than $40 billion in a day, after which regulators took control. This is her contemporaneous reconstruction of a rapidly developing event.

Reses criticizes the sequencing of SVB’s capital-raising announcement. In her view, announcing a planned raise while the bank was already facing questions about deposits intensified fear. She says a depositor’s incentive not to be last amplified messages moving quickly through the technology community.

Asked whether the failure was systemic or a series of mistakes, Dickson points to the FDIC’s report of roughly $620 billion in aggregate unrealized securities losses across banks in the fourth quarter of 2022. She cautions that those losses were not necessarily distributed evenly and that individual balance sheets needed closer examination.

Dickson says SVB’s investment decisions and concentrated customers left it exposed when rates and funding conditions changed. Whether it could have ridden out losses without a sudden withdrawal wave cannot be known from the interview. Reses adds that better communication might not have stopped people from withdrawing when they perceived a nonzero risk.

Harry asks whether venture capitalists deserve blame. Reses says advising individual companies to protect deposits can be rational for fiduciaries, even if collective withdrawals worsen a run. She also thinks companies should examine earlier decisions to concentrate cash and payroll relationships in one bank.

Reses expects founders and venture investors to revisit controls over where funds are held and whether backup banking relationships exist. She warns that not every fintech interface is itself a bank, and says depositors should understand the underlying institution and its risks.

Harry asks whether deposits will concentrate at the largest banks. Reses expects some movement toward institutions viewed as safer but argues local and regional banks can be important too. She credits post-2008 capital controls with improving banks’ resilience in general, while recognizing that the present episode could still undermine confidence.

Dickson cites far more US bank failures after Lehman than in the preceding five years and says the immediate danger is a wider crisis of confidence. She argues that excessive concentration in a few large institutions could weaken local credit. Her figures and characterization reflect the information available in March 2023.

Reses says local banks help fund businesses and communities, while fintech firms can make credit easier to access through software. She notes her experience building banking and lending products at Square, and her current interest as Lead Bank’s CEO. In her account, banks remain behind many consumer-facing fintech products.

Harry asks whether depositor psychology matters more than the underlying health of a bank. Dickson says the two interact. She contrasts Lehman’s leveraged mortgage exposure with SVB’s long-duration government-backed securities and concentrated deposit base, and says earlier rebalancing might have reduced losses but cannot establish the counterfactual.

Reses explains that the FDIC often takes control over a weekend, giving it time to evaluate a bank’s assets, liabilities and deposits before business resumes. She says SVB employees were working with the agency to assemble the information required for a resolution.

She describes insured deposits, administrative costs and claims against assets as parts of a complex resolution process. She speculates that an advance payment on uninsured deposits could provide liquidity before the final recovery. Her description and the percentages being discussed at the time were not an announcement of a final outcome.

Dickson explains that a liquidation without a buyer can take a long time: Lehman’s estate remained unresolved years after its 2008 bankruptcy. She calls that an extreme case and does not expect the same duration for SVB. Based on a preliminary look at year-end assets, she thinks most deposits should ultimately be covered, while acknowledging timing and sale prices are uncertain.

Asked when depositors could recover funds, Dickson says assets must be sold without unnecessarily sacrificing value. Reses adds that a buyer taking over deposits and continuing operations would be her preferred outcome. At recording, neither knows whether a buyer will emerge or what depositors will receive.

Reses regards SVB’s client relationships as potentially attractive to a buyer, while recognizing that the economics, regulatory constraints and a possible gap between assets and liabilities would matter. She mentions several large banks as possibilities, not as confirmed bidders.

Dickson notes that in a prolonged liquidation, claims might trade, as Lehman claims did. Reses says regulators have strong incentives to find a buyer but cautions that different banks face different balance-sheet and regulatory constraints. Their statements are expectations, not information about completed negotiations.

On contagion, Reses says depositor psychology and concentration in a customer base can matter alongside balance-sheet fundamentals. She cites a public statement by First Republic about its diversification and calls it strong. That was her view based on information available then, not a verified guarantee of its safety or a current assessment.

Dickson warns that many depositors lack the financial expertise to analyze a bank, and even a confident depositor may fear being left behind if others withdraw. Both guests see a timely, credible resolution as important for public confidence, while acknowledging they do not know what regulators will announce.

Reses identifies her preferred immediate outcome: a buyer takes over deposits, depositors are paid in full and funds are available on Monday. She describes this as an optimal scenario, not a confirmed result. She notes that the harm from delayed access could reach employees, schools, farms and businesses beyond venture investors.

The worst case, she says, would be no buyer, an inadequate advance payment and companies unable to meet payroll. Some firms and investors were trying to arrange short-term payroll financing, but she cautions that not everyone could obtain it immediately. She hopes regulators will avoid that outcome.

Asked about moving funds to a neobank, Reses says the application may be easy to use but founders should identify the bank holding the deposits and understand the arrangement. She favors having diversified banking relationships and backup accounts; she also describes Lead Bank’s own depositor base as diversified, a claim from its CEO.

Harry asks whether withdrawals can become stuck. Reses says some SVB transfers appeared approved but had not arrived when the bank was taken over, while some ordinary payments continued to move. The status of a particular transfer was not established in the interview.

When founders ask about moving company money into personal accounts, both guests urge caution and recommend keeping corporate funds separate. They explicitly say they are not lawyers; they suggest opening an appropriate business account and consulting relevant professional advisers rather than commingling funds.

Reses warns against rumors circulating on social media and messaging services. She says only a small group at the FDIC, SVB and potential acquirers would know the details, and that even a senior contact had not disclosed confidential information to her. She urges listeners to consider the source and recognize uncertainty.

Asked to predict the following Wednesday, Reses guesses that a buyer will emerge and mentions possible firms, while explicitly saying some names came from rumor and should not be trusted. She also expects depositors and company boards to reconsider banking relationships and controls. These were predictions at the time, not confirmed outcomes.

She suggests contingency planning around access to accounts and records, including retaining information needed if systems become inaccessible. Her wider point is that boards should consider bank access and information security together when preparing for interruptions.

Dickson agrees that a buyer could provide continuity and help avoid a confidence crisis. Both guests hope depositors will be made whole, but frame that as an expectation rather than a known outcome. Harry thanks them for joining at short notice.