Summary
The official causal record is institutional, not a shortcut to individual guilt. Federal and California reviews found that SVB's board and management failed to manage risk and that supervisors did not respond effectively enough. Those findings do not by themselves adjudicate a named person's civil or criminal liability.
Credit quality and liquidity are different questions. Long-duration government and agency securities may carry limited credit risk while still losing market value when rates rise. A bank that must sell them to meet withdrawals can turn an unrealized valuation gap into an immediate capital and liquidity problem.
Uninsured deposits were both funding and concentration risk. SVB served technology, venture-capital and life-science clients whose balances and cash needs were correlated. The control failure was not serving those sectors; it was treating a concentrated, digitally mobile funding base as though ordinary historical runoff assumptions were enough.
Supervision identified weaknesses but did not compel timely repair. Portfolio transition, staffing, judgment, tailoring and escalation processes all mattered. Counting findings is not proof of supervision; closure evidence and enforceable deadlines matter.
Resolution protected service continuity but allocated costs. Authorities protected all depositors after a systemic-risk determination, while shareholders and certain unsecured creditors were not protected. The Deposit Insurance Fund cost tied to uninsured-depositor protection was assigned to a special assessment on banking organizations, not erased.
Durable reform needs replayable evidence. Boards and supervisors should be able to reconstruct duration decisions, hedge changes, deposit concentrations, liquidity tests, collateral readiness, risk-leadership vacancies, finding escalation and contingency actions from contemporaneous records—not from explanations assembled after failure.
The official record supports a joined cause, not a one-line morality tale
The Federal Reserve Board's announcement of its Silicon Valley Bank review set out four conclusions. SVB's board and management failed to manage their risks; supervisors did not fully appreciate the vulnerabilities as the firm grew; identified problems were not fixed quickly enough; and regulatory tailoring and a less assertive supervisory stance impeded effective oversight. The release also said the bank had 31 unaddressed safe-and-soundness warnings at failure, three times the peer average.
Those conclusions allocate institutional responsibility across the bank and its primary federal supervisor. They do not establish that every director saw the same information, that every officer made the same decision, or that a particular act met the elements of negligence, breach of fiduciary duty, securities fraud or a criminal offence. Governance analysis must preserve that boundary. A supervisory finding can prove that a control was deficient within the agency's process without resolving damages, intent or person-specific liability.
The failure also resists a simple “safe assets became unsafe” explanation. Treasury and agency mortgage-backed securities can have low expected credit loss and still carry substantial duration risk. When market rates rise, the present value of fixed cash flows falls. Accounting classification may avoid recognizing some unrealized changes in current equity, but it does not create cash. If depositors demand funds and available liquidity is limited public evidence, sale or pledge capacity becomes decisive.
The same joined-cause discipline applies to the run. Digital transfers and concentrated networks increased speed, but technology did not create the underlying asset-liability mismatch. The March disclosure and attempted capital raise were catalysts, but they acted on pre-existing vulnerability. Accountability begins by separating trigger from root conditions and then identifying who controlled each condition before the trigger arrived.
Rapid growth changed the control problem before the portfolio label changed
The full Federal Reserve review of supervision and regulation describes extraordinary growth, a concentrated business model, reliance on uninsured deposits, foundational management weaknesses and an inadequate supervisory response. SVB Financial Group moved above $100 billion in assets and transitioned from the regional banking organization portfolio to the Large and Foreign Banking Organization portfolio, but its risk architecture and supervisory treatment did not mature at the same speed.
Growth is not wrongdoing. A bank can add clients, deposits and assets safely if capital, liquidity, data, governance and operational capacity scale with them. The control question is whether each growth threshold changes requirements before exposure accumulates. A firm entering a larger supervisory portfolio should not wait for the transition to finish before producing large-bank-quality interest-rate, liquidity and concentration evidence. Nor should supervisors treat the transfer as an administrative hand-off in which old findings lose urgency or new examiners restart discovery.
SVB's deposit surge during the technology financing boom created a balance-sheet allocation decision. Cash arriving faster than suitable loan demand can be held in cash, invested at shorter maturities, placed in longer securities, hedged, or offset with more stable funding. Each choice changes earnings and resilience. Extending duration can improve reported yield when rates are low, but it creates sensitivity to rate increases and limits flexibility if deposit behavior changes. A board should see that trade as a funding decision, not merely an investment decision.
The supervisory report also challenges threshold thinking. Legal categories determine which rules apply, but risk does not wait at a statutory line. A bank approaching a threshold can already have the complexity, concentration and run exposure associated with a larger institution. Evidence-based supervision should stage resources and expectations using forward growth, business-model volatility and unresolved weaknesses, while still applying the law actually in force.
Public filings disclosed pieces of the risk but did not assemble the failure path
SVB Financial Group's 2022 Form 10-K is a primary company record of the balance sheet before failure. It reported approximately $91.3 billion of held-to-maturity securities at amortized cost with a fair value of about $76.2 billion at year-end. It also disclosed estimated uninsured deposits, deposit outflows, interest-rate sensitivity and liquidity sources. Those disclosures matter, but the presence of figures in a filing does not prove that management, the board, depositors or supervisors joined them into an executable stress scenario.
Held-to-maturity accounting is not concealment by definition. The classification reflects an intent and ability to hold securities to maturity, and the accounting treatment differs from available-for-sale securities. The governance test is whether the ability assertion remains credible under the bank's funding profile. If concentrated depositors can leave together, a portfolio intended to be held may nonetheless sit beside a liquidity need that forces sale, borrowing or capital action.
Decision-useful disclosure therefore needs bridges. One bridge connects the fair-value gap to capital under plausible forced-sale amounts. Another connects deposits to beneficially distinct decision-makers rather than only account count. A third maps deposit runoff to same-day liquidity, collateral that is already operationally available, borrowing capacity and settlement timing. A fourth explains how hedge positions and deposit pricing change earnings and economic value under the same rate path.
Investors also need changes, not only endpoints. A declining hedge position, a risk-officer vacancy, worsening stress results or repeated limit exceptions may be material in combination even if each can be described separately. The board should receive the integrated picture earlier than external investors, challenge it and record the reasons for any accepted exception. Disclosure cannot substitute for control, but a disclosure process fed by the same governed data can expose contradictions before a crisis.
Duration risk was a deliberate balance-sheet exposure that required a named owner
The Federal Reserve's released supervisory history and examination materials show how risk signals moved through ratings, examinations, Matters Requiring Attention and Matters Requiring Immediate Attention. They include interest-rate and liquidity work, governance findings, portfolio-transition records and late-stage readiness reviews. The materials make clear that supervision is a sequence of judgments, not a single annual inspection.
Inside a bank, asset-liability management committees often recommend duration, funding and hedge positions, while treasury executes and risk provides independent challenge. That structure works only if decision rights are explicit. The record should identify who approved the duration limit, who could reject a breach, what model and deposit assumptions were used, how exceptions were priced, and when the board risk committee was informed. Collective committees must not dissolve accountability into minutes that record discussion but no owner.
Economic value of equity and net interest income tests answer different questions. A position may appear favorable to near-term earnings while producing severe economic-value sensitivity, or a hedge may reduce one measure while increasing volatility in another. Management should not select the metric that makes the existing position appear safest. The risk appetite needs paired limits, common scenarios and an escalation rule when the measures diverge.
Hedge removal deserves the same evidence. Hedges have costs, accounting effects and basis risk; keeping every hedge is not automatically prudent. Yet reducing protection as rates rise must be supported by a documented view of assets, liabilities and depositor behavior, tested against adverse paths. A credible file would show alternatives considered, expected earnings benefit, tail loss, liquidity effect, approvals and triggers for reinstatement. Without it, hindsight debate cannot distinguish an informed risk decision from unmanaged exposure.
Deposit concentration must be measured by correlated behavior, not account totals
The California Department of Financial Protection and Innovation's review of its oversight and regulation of SVB identified rapid growth, high uninsured deposits, slow remediation and the accelerating effects of digital banking and social media. It also published confidential supervisory information with federal approval, allowing the public to see ratings, examinations and warnings that ordinarily remain closed.
SVB's clients were not interchangeable retail accounts. Technology and life-science companies may raise capital from overlapping venture investors, bank through shared professional networks and increase cash burn when financing conditions tighten. A venture fund can influence many portfolio companies at once. Counting each legal account as independent therefore understates common decision channels. Concentration analysis should connect accounts by ownership, sponsor, sector, funding stage, geography, service provider and communication network, subject to lawful data use.
Uninsured status is also not a complete stability score. A corporate operating account may be uninsured because payroll and vendor payments exceed the statutory limit, not because its owner is chasing yield. Its balance can still move immediately when safety is questioned. Conversely, some large balances may be operationally sticky. The correct control uses observed behavior and contractual context, while stress assumptions remain conservative when evidence is weak.
The board needs distributional views: the largest balances, clusters controlled or influenced by the same actors, the portion that can leave through digital channels in hours, and the liquidity needed if several clusters act together. Limits should apply to correlated sources, not merely to a single depositor. Breaches should change funding plans and asset duration before management tries to solve concentration through last-minute relationship calls.
Liquidity stress must model the first day, not only a thirty-day average
The Federal Reserve OIG's material-loss review found that SVB was vulnerable to the business cycles of its concentrated science and technology customers, held long-maturity securities with substantial unrealized losses, and suffered from management, board and supervisory weaknesses. It described roughly $40 billion leaving and additional withdrawal requests of about $100 billion that the bank could not meet.
A stress test that assumes outflows unfold gradually can be internally consistent and operationally useless. Digital commercial clients can initiate very large transfers without branch queues. Liquidity governance must therefore include intraday and same-day scenarios, cutoff times, payment-rail capacity, collateral movement, discount-window readiness and the people authorized to act outside normal hours. A source of liquidity that cannot be operationalized before settlement is not available for the critical horizon.
Collateral readiness is evidence-intensive. Securities must be identified, valued, legally unencumbered, pledged where needed and technically transferable. Staff need tested access and counterparties need current documentation. The bank should run small operational transactions before crisis, record exceptions and test weekend escalation. Stating that assets are “available” is limited public evidence if teams discover during a run that transfer procedures, data or approvals are incomplete.
Stress models also need reverse testing. Instead of asking whether the bank survives the chosen scenario, reverse stress asks what outflow empties usable liquidity and which events could produce it. That threshold can then be compared with actual depositor clusters. If a handful of connected decision-makers can cross it, the risk is not remote merely because historical daily runoff was lower. The board should either reduce the exposure or hold credible resources against it.
The capital raise became a trigger because the repair sequence was not confidence-safe
The DFPI's order taking possession records the immediate legal and liquidity basis for closure. It states that SVB announced an approximately $1.8 billion loss from an investment sale and a capital raise on March 8, received approximately $42 billion of withdrawal requests on March 9, ended that day with a negative cash balance of about $958 million, could not meet its cash letter and was insolvent.
The order is an official state action, not a complete causal trial. Its role is to establish possession and receivership under the circumstances then present. It does not adjudicate every earlier management decision or later civil claim. Still, it shows why contingency sequencing matters. Selling securities, disclosing a realized loss and seeking capital may be individually rational actions, yet together they can signal that the institution has less flexibility than depositors assumed.
A confidence-safe plan should be rehearsed before pressure peaks. It should compare private capital, asset sales, secured borrowing, Federal Reserve facilities, deposit pricing and strategic transactions; identify disclosure requirements; and test how each sequence affects capital, collateral and depositor interpretation. Legal, treasury, communications and resolution teams need the same numbers. A public announcement cannot outrun the operational ability to answer predictable questions.
Communication is not a substitute for solvency or liquidity. Management cannot talk away a fair-value gap and a concentrated run. But inaccurate, incomplete or poorly sequenced communication can accelerate loss of confidence. The control objective is not to suppress material information. It is to ensure that truthful disclosure is paired with a credible, executable funding plan and a clear statement of what has already been completed rather than what management hopes to arrange.
Federal and state supervision shared the bank but neither could outsource escalation
GAO's preliminary review of the March 2023 bank failures found that risky business strategies, weak liquidity and risk management contributed to SVB and Signature Bank's failures, and that regulators raised concerns but did not take adequate action. GAO calculated that SVB's assets grew 198 percent from 2019 through 2021, far above the median growth of its peer group.
SVB was a California-chartered Federal Reserve member bank. DFPI and the Federal Reserve therefore brought different authorities and institutional processes to a shared risk. Joint examinations can reduce duplication, but shared supervision creates a hand-off hazard: one authority may assume the other owns escalation, or each may wait for a coordinated position. The bank can then respond to the narrower request while the combined vulnerability continues.
One joint issue register should show the finding, legal basis, originating authority, severity, bank owner, due date, validation method and escalation stage. Differences between authorities should remain visible rather than averaged into a weak compromise. If either supervisor considers the risk urgent, the record should explain why formal action did or did not follow. Confidentiality rules can govern access without preventing accountable coordination.
Supervisors also need continuity when staff or portfolios change. A transfer package should distinguish verified remediation from management promises, preserve overdue dates and state which risks are worsening. New teams may reconsider judgments, but they should not reset the clock merely because organisational ownership changed. For a fast-growing bank, delay is itself a risk decision and should require approval at a level commensurate with the potential loss.
Finding counts matter only when closure standards and escalation clocks are real
GAO's later report on supervisory escalation weaknesses examined how the Federal Reserve and FDIC move from concerns to stronger action. It found weaknesses in escalation processes and made recommendations concerning guidance, tracking, quality assurance and the use of more forceful tools. The report is important because it moves beyond the existence of warnings to the machinery that should convert warnings into correction.
An MRA is not remediation. Closure should require evidence that the control is designed, implemented, operating for a meaningful period and independently validated. Management's revised policy or target date is an input, not proof. If a finding concerns interest-rate risk, validation should reproduce exposure calculations, inspect assumptions, sample committee decisions and show that breaches trigger action. If it concerns liquidity, validation should include operational tests of collateral and funding.
Escalation clocks should start when risk is identified, not when wording is finalized. The clock can allow documented extensions for genuinely complex repair, but extensions should not be automatic. Severity should rise when the bank misses a deadline, when exposure grows, when related findings accumulate, or when independent testing contradicts management. A repeated concern with a fresh label is still an aged problem.
Quality assurance should examine supervisory decisions, not only file completeness. Reviewers should ask whether the team challenged optimistic assumptions, used specialists, compared peers, joined deposit and asset risks, and considered formal tools soon enough. Supervisory culture matters, but measurable controls make culture inspectable. The proof of a more assertive regime is not a new statement of intent; it is a record of earlier intervention in comparable cases.
Risk leadership vacancies are control events, not ordinary hiring delays
The official reviews discuss weaknesses in risk management and the period in which the firm lacked a permanent chief risk officer. A vacancy in a senior control role does not prove that no risk work occurred, and appointing a CRO does not transfer the board's duty. The accountability question is whether independent challenge retained authority, expertise and access while balance-sheet risk was changing rapidly.
A bank should classify a CRO departure during heightened exposure as a control event. The board should approve interim responsibilities, identify conflicts, set a short appointment timetable and commission targeted assurance over the areas most dependent on challenge. Committee quorums and limit approvals should be reviewed. If the interim owner is also responsible for business performance or treasury execution, additional independent review is necessary.
Succession should be judged by capability, not title continuity. The role needs authority over models, risk appetite, issue escalation and information reaching the board. It should have direct access to the risk committee and protected ability to disagree with the CEO and CFO. Compensation and performance evaluation should not reward acceptance of exposure merely because it supports current earnings.
The board must also receive risk information it can use. Dense packs that report compliance with management-selected metrics can obscure the economic question. Directors need trends, alternative assumptions, limit exceptions, near misses, reverse-stress thresholds and unresolved supervisory concerns. Minutes should record challenge, requested action and follow-up. Attendance and presentation are not evidence that the board held management accountable.
Congressional testimony is evidence of positions, not a substituted judgment
The Senate Banking Committee's hearing with former bank executives preserves prepared testimony, member statements and the public questioning of SVB's former chief executive. It is a primary congressional record of the explanations and disputes after failure. Witness testimony should be attributed as testimony; lawmakers' characterisations should be attributed as political oversight positions.
That boundary prevents two opposite errors. Management's explanation cannot displace supervisory findings merely because it was delivered under oath, and a forceful question does not become an adjudicated fact merely because it was asked in Congress. The value of the hearing is that it exposes competing accounts: management emphasized an unprecedented run and rate environment, while members pressed decisions on securities, hedging, governance, compensation and supervision.
An internal accountability review should use the same adversarial discipline with stronger access to records. For each consequential choice, reviewers should reconstruct the information available then, plausible alternatives, decision authority and dissent. Later market outcomes must not be smuggled into what decision-makers could know, but known rate paths, limit breaches, deposit changes and supervisory warnings cannot be excused as unforeseeable after they appeared.
The result should be a decision ledger, not a blame list. A ledger supports fair person-specific review if needed, while also exposing system defects that would survive personnel changes. It can show whether committees lacked data, incentives favored earnings, risk lacked authority, the board accepted exceptions, or supervisors delayed escalation. Those are repairable mechanisms; rhetoric alone is not.
Civil accountability must preserve allegation, duty and judgment boundaries
The Congressional Research Service's legal analysis of potential director and officer liability explained the statutory and state-law routes that could become relevant after failure, including fiduciary duties, the business-judgment rule, federal standards, agency enforcement powers and compensation recoupment proposals. It also noted reported investigations and emerging claims. The analysis did not decide that any person was liable.
That distinction remains essential. Official causal reviews, congressional testimony, investor complaints, receiver claims and any enforcement investigation have different parties, burdens and remedies. A complaint alleges; a motion ruling may decide only legal sufficiency; a settlement can impose obligations without findings; a judgment establishes what it actually adjudicates. Corporate and individual responsibility must not be inferred from job title or institutional failure alone.
The control lesson is nevertheless concrete. Directors and officers need contemporaneous evidence that decisions were informed, in good faith, within risk appetite and responsive to expert and supervisory warnings. The business-judgment rule is not a documentation trick, and a complete file cannot make an unsafe decision safe. But missing alternatives, ignored limits and unresolved dissent make it harder to show that risk was genuinely considered.
Compensation governance should also connect risk outcomes across time. Revenue and net-interest-income targets can reward duration or funding concentration before losses emerge. Deferred awards, malus and clawback mechanisms should use clearly defined triggers and lawful process. They should not depend solely on criminal misconduct, because severe control failure can occur without a crime; nor should they presume misconduct whenever a bank fails.
Closure and receivership separated urgent continuity from final loss allocation
The FDIC's initial receivership announcement created a Deposit Insurance National Bank and transferred insured deposits, while saying uninsured depositors would receive an advance dividend and receivership certificates. That was the ordinary legal baseline available at closure, before the later systemic-risk action changed depositor treatment.
The sequence matters. On Friday, firms faced uncertainty about funds exceeding insurance limits; payroll, supplier and operating accounts could be affected even when the underlying businesses were viable. Public-sector continuity therefore involved more than protecting financial investors. Authorities had to maintain payment access, preserve loan servicing, value assets and decide whether a broader intervention was necessary without promising a recovery the statute did not yet provide.
A resolution plan for a concentrated commercial bank should identify critical services and data before failure. It should map deposit ownership, payment files, sweep arrangements, credit lines, borrower obligations, third-party technology, correspondent access and customer communication. Records must be exportable and reconcile to the general ledger. A receiver should not have to discover over a weekend which systems are essential to thousands of operating businesses.
Continuity does not mean preserving shareholders, executives or every contract. It means maintaining legally authorized functions while claims are sorted and assets protected. Clear communication must distinguish insured access, estimated dividends, later policy decisions and final recoveries. Blurring those stages can create expectations that constrain future resolutions.
Systemic-risk action protected depositors but did not make the loss disappear
The Treasury, Federal Reserve and FDIC joint statement on March 12 said the Treasury Secretary, acting on recommendations from the FDIC and Federal Reserve boards and after consultation with the President, approved action enabling all SVB depositors to be protected. It stated that shareholders and certain unsecured debtholders would not be protected, senior management had been removed, and Deposit Insurance Fund losses from supporting uninsured depositors would be recovered through a special assessment on banks.
That was a statutory systemic-risk determination, not an ordinary expansion of the $250,000 insurance limit for every bank. It addressed a judgment that least-cost treatment could have serious adverse effects on economic conditions or financial stability. The decision should therefore be evaluated through its legal process, contemporaneous evidence, alternatives and allocation—not through slogans about a universal guarantee or a costless rescue.
Protecting depositors reduced immediate continuity risk and potential contagion. It may also affect future expectations: large depositors and banks could infer that similar protection will recur. That moral-hazard question does not negate the emergency judgment, but it requires counter-controls. Supervisors can impose stronger concentration and liquidity expectations; insurance pricing can reflect exposure; reporting can make uninsured funding visible; and Congress can determine whether selected business accounts warrant different coverage.
Accountability requires publishing enough decision logic to distinguish exceptional action from an implicit standing promise. It also requires tracing costs until receivership ends. Estimates change as assets are sold, litigation proceeds and recoveries arrive. A point-in-time estimate should be dated and sourced rather than treated as the final economic cost.
The bridge bank bought time for continuity and a more orderly sale
The FDIC's bridge-bank announcement transferred all deposits, substantially all assets and qualified financial contracts to Silicon Valley Bridge Bank, N.A. Depositors gained full access, borrowers were told to continue payments, a new chief executive was named and the FDIC explained that the structure was intended to stabilize the institution and preserve value while a resolution was arranged.
A bridge bank is a control mechanism, not merely a new name on the door. It creates temporary governance, keeps systems operating and allows more time to market a franchise than a rushed weekend liquidation. Its success should be measured against service availability, deposit retention, asset-value preservation, operational incidents, customer complaints, sale competition and ultimate recoveries.
The bridge also carries public-sector execution risk. The FDIC becomes responsible for a live commercial bank with complex technology, employees, contracts and client needs. Access controls must change without breaking service. Authority over credit decisions, vendor payments, communications and asset disposition must be clear. Temporary management needs records that separate pre-failure conduct from post-receivership decisions.
For SME continuity, the critical evidence is granular: whether payroll files ran, credit commitments were addressed, borrower payments posted, customer support worked and data remained accurate. Aggregate statements that the bank opened are important but incomplete. Resolution preparedness should predefine these service-level tests for institutions whose clients depend on specialized banking relationships.
The sale illustrates why resolution cost cannot be read from one headline number
The FDIC's sale announcement to First-Citizens Bank & Trust Company said the buyer assumed all deposits and loans, purchased about $72 billion of assets at a $16.5 billion discount, and entered a loss-share arrangement on commercial loans. About $90 billion of securities and other assets remained in receivership, and the FDIC received equity-appreciation rights with potential value.
Those components make simplistic cost claims unreliable. A purchase discount is not automatically the receiver's final loss; retained assets may produce recoveries, loss sharing changes future cash flows, and appreciation rights have value. Conversely, continuing administration and litigation can add costs. The exact Deposit Insurance Fund impact remains contingent until the receivership is terminated.
Resolution governance should retain a deal model that can be replayed. It should compare bids with liquidation, identify assumptions, discount rates, asset marks, operational-continuity effects, loss-share scenarios and counterparty capacity. Independent review should test whether urgency distorted valuation. Later outcomes should be compared with the original ranges so future resolutions learn from forecast error.
The acquiring bank's success is not proof that the failed bank's controls were adequate, and customer continuity does not erase losses borne elsewhere. Resolution closes an operating crisis while redistributing assets, liabilities and risk. Accountability follows each transfer: what moved to the buyer, what remained with the receiver, what protection depositors received, and who funded the residual cost.
The special assessment made cost transfer explicit
The FDIC's final rule on the systemic-risk special assessment set a base tied to estimated uninsured deposits, excluded the first $5 billion at a banking-organization level and established an anticipated collection period subject to adjustment. The agency explained that the assessment was designed to recover Deposit Insurance Fund losses associated with protecting uninsured depositors at SVB and Signature Bank.
This is not the same as saying taxpayers or depositors experienced no economic consequence. The statutory statement that losses would not be borne by taxpayers identifies the direct fiscal allocation. Assessed banking organizations bear a charge, and economic incidence may ultimately be shared through earnings, pricing or other adjustments. Claims about who “paid” should distinguish legal assessment from downstream economic effects that require separate evidence.
The assessment design also reflects an accountability judgment: larger concentrations of uninsured deposits were used as the base, with an exclusion intended to reduce effects on smaller organizations. Any proxy is imperfect. A bank receiving inflows during the crisis may still pay because of its uninsured-deposit base; a bank's contribution to risk may differ from the formula. The rulemaking process, not retrospective intuition, defines the legal obligation.
Future proof should compare collected assessments with actual receivership losses and make final adjustments transparently. A special assessment that is announced but not reconciled would leave cost allocation incomplete. Banks should also report the charge separately enough that boards and stakeholders can understand the consequence of system-wide protection.
Receiver claims are allegations until adjudicated or resolved
The FDIC chair's memorandum supporting authority to sue former SVB officers and directors set out the receiver's institutional position. It asserted mismanagement of held-to-maturity and available-for-sale securities, removal of hedges, breaches of internal risk metrics and an imprudent dividend, and sought authority for professional-liability claims.
That memorandum is not a judgment. Its assertions must be labelled as the FDIC's proposed claims, and any filed complaint would remain allegations unless admitted, settled with findings, or adjudicated. The public causal reports can provide context, but they cannot silently satisfy the elements of a receiver action against each defendant. Defenses, governing state law, causation and damages remain legal questions for the relevant process.
The memorandum nonetheless identifies the evidence architecture a bank should preserve: portfolio purchases, risk limits, hedge changes, dividend approvals, board materials and the relationship between bank and holding company decisions. Records should include dissent, alternatives and downstream capital and liquidity effects. Retention must begin before distress, with legal holds imposed promptly when failure or litigation becomes plausible.
Fair accountability also protects institutional learning. If every risk debate is written as an accusation, decision-makers may sanitize minutes or avoid candid challenge. Records should be factual, specific and clear about uncertainty. They can support both robust defense of informed decisions and action against genuine breaches. The objective is not automatic liability; it is a reliable account that a court or regulator can test.
Deposit insurance reform cannot substitute for bank and supervisory discipline
The FDIC's Options for Deposit Insurance Reform examined limited, unlimited and targeted coverage, including potentially higher protection for business payment accounts. It discussed financial-stability benefits, depositor protection, moral hazard, pricing, fund adequacy and implementation challenges. The report described options for Congress; it did not itself change statutory coverage.
SVB demonstrates the appeal of targeted protection because business operating balances can exceed ordinary limits and their interruption can affect payroll and suppliers. It also demonstrates the design problem. Defining an eligible payment account, preventing evasion, pricing additional protection and verifying actual use are difficult. Unlimited coverage could reduce run incentives while weakening depositor discipline and expanding the fund's exposure.
Whatever coverage Congress chooses, risk must remain with accountable decision-makers. A higher guarantee should not license banks to fund longer duration with more concentrated deposits. Deposit-insurance pricing, capital, liquidity, concentration limits and supervision should respond to the new incentives. Depositors also need clear tools to understand coverage and structure accounts without implying that sophisticated monitoring can replace prudential control.
Reform evidence should include distributional effects. Policymakers should test which firms gain continuity, which banks pay, whether deposits migrate toward or away from smaller institutions, and how behavior changes in calm periods and stress. A policy that looks stable in aggregate may create new concentration. The durable answer is a joined system in which insurance, supervision, resolution and bank governance reinforce rather than offset one another.
The durable lesson is that speed must belong to control as well as withdrawal
SVB's depositors could coordinate and move money faster than the bank and its supervisors could repair accumulated weaknesses. The answer is not to slow lawful access to funds or blame technology for revealing a confidence problem. It is to make risk information, escalation, collateral movement and resolution preparation operate at comparable speed.
Management controls the original balance-sheet choices. The board controls risk appetite, leadership and challenge. Supervisors control ratings, findings, deadlines and formal escalation. Resolution authorities control closure, continuity and loss allocation within law. Congress controls the statutory perimeter and deposit-insurance design. None can transfer its duty to another by pointing to a different part of the causal chain.
Accountability also requires restraint. Official reviews justify strong conclusions about institutional failure, but they do not establish every private allegation or individual claim. Emergency depositor protection maintained continuity, but it did not make costs vanish. A successful sale stabilized services, but it did not validate the failed strategy. Reform announcements show intent, but only later comparable interventions can prove effectiveness.
The prospective test is observable. When a rapidly growing bank combines long-duration assets, correlated uninsured funding, thin same-day liquidity, model exceptions, control-leadership gaps and aged findings, can its board and supervisors see the joined exposure and force action while choices remain? If the evidence answers yes before confidence breaks, then SVB will have changed who controls the clock.