Beginner Level
What Is It?
The Silicon Valley Bank (SVB) collapse in March 2023 was the largest bank failure since 2008, triggered by a rapid depositor run that exposed mismatches between long-duration bond assets and short-term deposits in a rising rate environment.
Origin
SVB had invested heavily in long-duration Treasury and agency securities when rates were low. As the Federal Reserve raised rates, bond values fell, creating unrealized losses. When depositors, concentrated in tech startups, began withdrawing, SVB was forced to sell bonds at losses, confirming insolvency.
Why It Matters
SVB demonstrated that duration risk in bank balance sheets remained dangerous and that social media and mobile banking could accelerate bank runs from weeks to hours. It triggered a broader regional banking crisis and forced emergency regulatory intervention.
Intermediate Level
Market Mechanics
The bank run propagated through venture capital networks—VCs advised portfolio companies to withdraw, creating self-fulfilling panic. SVB's held-to-maturity bond portfolio had over $15 billion in unrealized losses. Forced sales crystallized these losses, rendering the bank insolvent.
How It Behaves
Modern bank runs exhibit viral dynamics—social coordination among depositors accelerates withdrawal velocity. Unrealized losses become realized under stress. Contagion spreads to similar institutions (Signature Bank, First Republic). Systemic risk emerges from correlated depositor behavior.
Key Data to Watch
- Unrealized losses on held-to-maturity securities
- Deposit concentration metrics
- Uninsured deposit ratios
- Social media sentiment and coordination signals
- Bank CDS spreads and equity price action
- VC and startup ecosystem funding stress
Advanced Level
Institutional Behavior
Regulators had classified SVB as not systemically important, exempting it from stricter oversight. The FDIC, Federal Reserve, and Treasury invoked "systemic risk" exceptions to guarantee all deposits, not just insured ones, establishing new intervention precedents.
Professional Use Cases
- Bank duration risk monitoring
- Deposit concentration and runoff analysis
- Regional bank stress identification
- Regulatory capital requirement assessment
- Post-crisis distressed bank opportunity evaluation
AI Interpretation in Systems Like Arkhe
- Risk Agent: Monitors bank bond portfolio duration and unrealized loss exposure
- Macro Agent: Tracks Fed rate path impact on bank balance sheets
- Sentiment Agent: Detects depositor coordination and social media panic signals
Key Takeaways
SVB proved that bank runs could go from concern to collapse within 48 hours in the social media age. It demonstrated that held-to-maturity accounting could mask systemic duration risk and that depositor concentration in correlated industries amplifies fragility.