AI-Generated Supervisory Letter

As a test, I used ChatGPT to generate a supervisory letter given a set of facts. I didn’t use the bank’s name, but the findings corresponded to those of an August 21, 2021, examination of Silicon Valley Bank. The query also used Uniform Bank Performance Report (UBPR) data as of June 2021.

Here is the query:

Assume you are a federal bank examiner who has just conducted a liquidity target exam of a $160 billion bank. As part of your review and of your ongoing monitoring, you found the following:

ILST based on historical simulation. Uses capital stress testing and internal stress data, supplemented by high level historical data. The external data consists primarily of retail-oriented banks. Internal liquidity stress testing is based primarily off of business lines.

Liquidity limits based on static metrics. The bank’s contingency funding plan assumes that Fed Funds borrowing lines are listed as still available under “Severe” stress events. Second, the firm identifies the types of contingent funding by source but does not identify available amount based on active contracts or internal firm limits. For example, when compared with the firm’s ILST, the CFP assumes far more repo funding capacity than is currently available through master repurchase agreements.

The bank is not yet subject to the Liquidity Coverage Ratio and has not segmented deposits by operational/non-operational.

NII was 2.09% vs. 2.00% for peer group. PD & nonaccrual 0.16% vs 0.90 for PG Leverage Ratio 6.96% vs. 8.56% for PG Risk based capital ratio is 15%. Average total assets increased from $73.6Bn to $135.6 Bn YOY NIM down 72 bps YOY LT assets to total assets 49.23% vs. 32.08 for peer group. Unrealized (positive) appreciation on AFS & HTM securities is 2.03% (positive). Investment securities (mostly Treasuries and GSE) 61.8% of total assets. Core deposits are 83.35% of total assets vs PG median of 73.64% More than 90% of deposits are above the $250,000 FDIC limit. A large share of deposits come from venture capital-backed companies.

Based on this information, draft a Supervisory Letter and include matters requiring attention and/or matters requiring immediate attention as necessary.

Here is the response:

Supervisory Letter

To: [Board of Directors / CEO – Bank Name]
From: [Examiner Name], [Agency Name]
Date: [Insert Date]
Subject: Liquidity Target Examination – Supervisory Findings and Required Actions

---

Overview

As part of the targeted liquidity examination conducted in [Month Year], the supervisory team evaluated the Bank’s liquidity risk management framework, stress testing methodologies, contingency funding planning, and funding profile. This review included assessment of Internal Liquidity Stress Testing (ILST), the Contingency Funding Plan (CFP), liquidity limits, deposit composition, investment portfolio structure, and key financial metrics.

The examination identified several weaknesses in liquidity risk management practices that, if unaddressed, could expose the Bank to heightened funding risks under stress conditions. While the Bank maintains strong credit quality and above-peer capital ratios, its liquidity framework exhibits vulnerabilities due to concentrated funding sources, optimistic stress assumptions, and lack of segmentation of deposit stability.

Summary of Key Findings

1. Internal Liquidity Stress Testing (ILST)

• ILST methodology is based primarily on historical simulation and leverages capital stress testing and internal stress data supplemented with limited external data.
• External data used for benchmarking is primarily from retail-oriented institutions, which may not reflect the Bank’s unique wholesale and venture-capital-funded deposit base.
• ILST is based largely on business-line input without adequate central challenge or model validation for extreme outlier scenarios.

2. Liquidity Limits

• Liquidity limits are based on static metrics and do not dynamically adjust for changes in the Bank’s funding composition or balance sheet size.

3. Contingency Funding Plan (CFP)

• CFP assumes that Federal Funds borrowing lines remain available even under “Severe” stress scenarios.
• CFP lists contingent funding sources but does not identify available amounts under active contracts or internal limits.
• CFP assumptions for repo funding capacity significantly exceed amounts available under current master repurchase agreements.

4. Regulatory Requirements

• The Bank is not currently subject to the Liquidity Coverage Ratio (LCR) but has not proactively implemented LCR-aligned monitoring or operational/non-operational deposit segmentation.

5. Funding Profile and Concentrations

• Core deposits represent 83.35% of total assets; however, more than 90% of deposits exceed the $250,000 FDIC insurance limit.
• A significant portion of deposits is sourced from venture capital-backed companies, which historically have demonstrated high runoff risk during stress events.
• The Bank’s long-term assets represent 49.23% of total assets versus a peer median of 32.08%, increasing liquidity transformation risk.

6. Balance Sheet Changes

• Total assets nearly doubled year-over-year (from $73.6B to $135.6B), driven largely by securities growth (investment securities comprise 61.8% of assets, mostly Treasuries and GSE obligations).
• Net interest margin declined 72 basis points year-over-year despite NII being slightly above peer median.

---

Matters Requiring Immediate Attention (MRIAs)

1. Contingency Funding Plan – Unrealistic Assumptions

• The Bank must immediately revise CFP assumptions to reflect actual, contractually available capacity for contingent funding sources, including master repurchase agreements and unsecured borrowing lines, under severe stress conditions.
• Assumptions that Federal Funds lines remain available in severe stress scenarios must be removed unless supported by binding contractual agreements that survive stress triggers.
• Updated CFP must be submitted to this office within 60 days.

2. Deposit Concentration Risk

• Given the high concentration of uninsured deposits from venture-capital-backed firms, the Bank must develop and implement an enhanced liquidity monitoring and stress testing framework that specifically models rapid outflows of these deposits under idiosyncratic and market-wide stress events.
• A revised ILST incorporating these assumptions must be provided within 90 days.

---

Matters Requiring Attention (MRAs)

1. ILST Methodology Enhancement

• Broaden external benchmarking data to include wholesale- and technology-sector-focused banks.
• Strengthen governance over ILST assumptions by implementing a centralized independent review and challenge process.
• Review scenario severity to ensure coverage of both market-wide and idiosyncratic extreme stress events.

2. Dynamic Liquidity Limits

• Replace static liquidity limits with limits that adjust based on changes in funding composition, deposit concentrations, and balance sheet growth.

3. Operational/Non-Operational Deposit Segmentation

• Even though the Bank is not currently subject to LCR, develop and maintain an internal process to segment operational versus non-operational deposits for enhanced liquidity risk monitoring.

---

Supervisory Expectations and Next Steps

Management and the Board are expected to address the above MRIAs and MRAs promptly. Written responses detailing corrective actions, responsible parties, and target completion dates should be submitted to this office by [date – generally 30 days from letter issuance].

Failure to address these matters in a timely and effective manner may result in further supervisory action.

---

Sincerely,
[Examiner Name]
[Title]
[Agency Name]

ChatGPT also offered “to prepare an appendix for this letter that includes a table mapping each weakness to a specific regulatory citation or supervisory guidance (e.g., SR 10-6, SR 14-1, SR 12-17) so it’s aligned with examiner documentation standards. That way, the letter would be defensible and tied directly to policy.”