SCFL
SCFL Reconstruction · Cluster Case

2023 Regional Bank Failures

Silicon Valley Bank · Signature Bank · First Republic Bank — three distinct institutions, three distinct seam profiles, one shared precursor geometry. Applying Standard Coherence Fidelity Layer (SCFL) analysis retroactively to the largest U.S. banking crisis since 2008.

📅 SVB: March 10, 2023 📅 Signature: March 12, 2023 📅 First Republic: May 1, 2023 💰 Combined assets at failure: ~$530 billion

Three Banks — Three Seam Profiles

SVB

Silicon Valley Bank · Assets: $209B

Primary seam: Asset-Liability Mismatch

Bond portfolio yield: 1.79% avg

10-yr Treasury at failure: ~3.9%

Failure window: 48 hours

SIGNATURE

Signature Bank · Assets: $110B

Primary seam: Deposit Concentration

Uninsured deposits: ~90% of total

Crypto-related deposits: ~30% at peak

Cash as % of assets: 5% vs 13% avg

FIRST REPUBLIC

First Republic Bank · Assets: $229B

Primary seam: Duration + Deposit Risk

Uninsured deposits: >66% of total

Loan book growth: $58B → $135B (5 yrs)

Q1 2023 deposit outflow: $72B

SVB Structural Integrity Scan — Illustrative Reconstruction
⚠️
SVB SCFL Verdict: Asset-Liability Seam — Precursor Window Opened 2020–2021 SVB's primary seam was measurable in its balance sheet structure from 2020 onward: a $21B bond portfolio locked at 1.79% average yield purchased when rates were near-zero, held against a deposit base serving venture-backed startups. When the Fed raised rates aggressively beginning March 2022, both sides of the seam activated simultaneously — bond values fell and deposit outflows accelerated. The 48-hour run in March 2023 was not the crisis. It was the terminal event of a multi-year structural geometry.

SVB — Phase Map

2018–
2020
Nominal Coherence — Tech Boom Banking
SVB grows rapidly as VC-backed startup banking partner. Deposit inflows from tech funding surge. Model is operationally coherent: deposit growth funds short-duration lending and securities. Rates near zero; asset-liability mismatch is not yet structurally dangerous.
Phase I · Nominal
2020–
2021
Seam Initiation — Duration Lock at Zero-Rate
SVB deploys a flood of pandemic-era deposit inflows into long-duration mortgage-backed securities and Treasuries at historically low yields — the $21B portfolio averaging 1.79%. This locks the asset side of the balance sheet into a rate environment that will not persist. The liability side (startup deposits) remains short-duration and rate-sensitive. The structural seam is set.
Phase II · Seam Initiation
Mar 2022–
Jan 2023
Seam Activation — Rate Hikes vs. Bond Book
The Fed begins the most aggressive rate hiking cycle in decades. Every hike increases the unrealized loss in SVB's bond portfolio. Simultaneously, VC funding dries up — startups draw down cash reserves held at SVB rather than raising new rounds. Both jaws of the seam are moving: asset values falling, deposit outflows accelerating. The 10-year Treasury reaches ~3.9% against SVB's 1.79% portfolio yield.
Phase II-B · Dual Seam Activation
Mar 8–9
2023
Disclosure Cascade — Loss Realization + Capital Raise Announcement
SVB announces it sold $21B in securities at a $1.8B after-tax loss and intends to raise $2.25B in new capital. The announcement is interpreted by VC firms as a liquidity signal. Peter Thiel's Founders Fund and others advise portfolio companies to withdraw. Within hours, $42B in withdrawal requests are queued — more than the bank can process.
Phase III · Cascade Trigger
Mar 10
2023
Terminal Event — FDIC Receivership in 48 Hours
Trading in SVB shares halted Friday morning. Capital raise and acquisition search both fail. California regulators close the bank and place it in FDIC receivership. Largest U.S. bank failure since Washington Mutual (2008). The 48-hour window from announcement to closure is among the fastest institutional collapses in modern banking history.
Terminal Event

Shared Structural Seams — Across All Three Failures

Deposit Architecture
Uninsured Deposit Concentration
All three banks held disproportionate shares of uninsured deposits: Signature ~90%, First Republic >66%, SVB heavily concentrated in venture-backed startups. Uninsured depositors have no FDIC backstop and will exit on the first credible signal of distress.
Failed — All Three
Rate Sensitivity
Asset-Liability Duration Mismatch
SVB and First Republic both locked long-duration assets at near-zero rates while their deposit bases remained short-duration. When the Fed raised rates, the seam activated on both sides simultaneously.
Failed — SVB, FRB
Sector Concentration
Client-Base Coherence
SVB was concentrated in VC-backed tech; Signature in crypto platforms; First Republic in high-net-worth California/NY. Each bank's deposit stability depended on a single sector's health — making each vulnerable to sector-specific contagion, not just systemic events.
Failed — All Three
Regulatory Coverage
Oversight Threshold Gap
The 2018 Crapo Act raised enhanced oversight thresholds from $50B to $250B in assets. All three banks fell in this gap — too large to fail quietly, too loosely supervised to be caught early. The regulatory coherence layer was structurally absent.
Absent Layer
Contagion Transmission
Social Media Run Acceleration
SVB's 48-hour collapse and Signature's weekend closure were both accelerated by digital communication: VC advisories spread via text and Slack; social media amplified withdrawal urgency. The velocity of information transmission compressed the available intervention window to near zero.
New Variable
Liquidity Buffer
Cash Reserve Inadequacy
Signature held only 5% of assets as cash vs. a 13% industry average. SVB's liquid assets were consumed within hours of the run beginning. First Republic's $30B rescue deposit did not rebuild the structural buffer — it papered over a seam for seven weeks.
Failed — All Three

Cluster Collapse Mechanics

  • Sequential Trigger — SVB as Seam Activator for Signature SVB and Signature did not fail for the same reasons — but Signature failed because SVB failed. SVB's closure created a weekend window during which Signature's 90%-uninsured deposit base, already structurally fragile, received an unambiguous contagion signal. The structural seam preceded the trigger by over a year; the trigger simply set it off.
  • False Recovery Signal — First Republic's $30B Rescue The March 16 deposit rescue by eleven major banks was read by markets as a stabilization. By SCFL geometry, it was a signal that the structural problem had been recognized without being resolved. The $72B Q1 deposit outflow, disclosed just seven weeks later, confirms that the rescue bought time — not recoverability.
  • Velocity Compression — Digital Communication as Collapse Accelerant All three collapses were faster than any comparable historical event. The mechanism was not new: bank runs have always been driven by information asymmetry. What changed was transmission speed. VC advisories via Slack and text reached thousands of depositors simultaneously; social media amplified the signal beyond the tech community. The intervention window was measured in hours, not days.
  • Regulatory Seam — The Crapo Gap All three banks benefited from the 2018 removal of enhanced oversight for banks between $50B and $250B in assets. This regulatory seam allowed the structural vulnerabilities — duration mismatch, deposit concentration, sector exposure — to accumulate without external coherence checks. The supervisory layer that should have constrained the geometry was structurally absent.
SCFL Key Finding — Cluster Case

The 2023 regional bank failures are not a single collapse — they are three distinct seam profiles with one shared structural property: each bank's depositor base was held in place by confidence rather than by structural coherence. When confidence transferred, the structural seam activated.

The SCFL precursor geometry places the measurable seam initiation for all three banks in 2020–2021 — before rate hikes began, before FTX collapsed, and before SVB's capital raise triggered the run. The rate environment did not create the seams; it activated them.

This is the canonical signature of a pre-state geometry failure: the structural conditions for collapse were fully established years before any public signal appeared. The market saw deposit growth and revenue. The SCFL would have seen seam shear.

Reconstruction Author: Ronald Brogdon, Team Leader — Upstream Coherence Measurement Stratum (UCMS)
ORCID: https://orcid.org/0009-0009-0507-2971 · measurement@coherencemanagement.org

This reconstruction applies the Standard Coherence Fidelity Layer (SCFL) retroactively to publicly available historical record. No proprietary operator computations are disclosed herein. See also: DOI 10.5281/zenodo.21664081 (SVB, Signature, First Republic precursor geometry) and the UCMS Zenodo canon.

← Back to Cases