🏦 Banking Operational Risk Intelligence Platform

An end-to-end operational risk analytics system replicating real-world LCD/compliance analyst workflows at financial institutions — built with real FDIC data, ML risk scoring, KRI/KPI monitoring, and an interactive Streamlit dashboard.

---

🎯 Business Problem

Financial institutions (and their regulators) need to continuously monitor operational risk across thousands of banks. Manual processes are slow, inconsistent, and fail to catch early warning signs before they become systemic failures. This platform automates:

• Control exception detection — automated KRI breach flagging across Basel III thresholds
• Risk-based sampling — statistical sampling of high-risk institutions for control testing
• Predictive risk scoring — ML models identify at-risk institutions before failure
• KRI/KPI reporting — standardized reporting aligned with CCAR and regulatory frameworks

---

📊 Data Sources (100% Real, Public Data)

Source	Description	Access
FDIC BankFind API	Financial data for all 4,500+ FDIC-insured institutions	Free, no API key
FDIC Failure List	Every US bank failure since 1934 with insurance cost	Free
Fed Reserve DFAST	CCAR stress test results for 32 major banks	Public PDF

---

🏗️ Architecture

banking-risk-platform/
├── src/
│   ├── pipeline/
│   │   └── fdic_pipeline.py      # FDIC API ingestion, feature engineering, SQLite load
│   ├── ml/
│   │   └── risk_engine.py        # Logistic Regression, Random Forest, K-Means clustering
│   └── dashboard/
│       └── app.py                # Streamlit KRI/KPI dashboard
├── sql/
│   └── schema_and_queries.sql    # SQLite schema + 6 analytical KRI queries
├── data/                         # Generated after running pipeline
│   ├── banking_risk.db           # SQLite database
│   ├── institutions.csv
│   ├── failures.csv
│   ├── risk_scores.csv
│   └── models/
│       ├── best_model.pkl
│       ├── model_metrics.json
│       └── feature_importance.csv
└── run.py                        # Master pipeline runner

---

🚀 Quick Start

1. Install dependencies

pip install -r requirements.txt

2. Run the full pipeline

python run.py

This will:

• Pull ~4,500 banks + all historical failures from FDIC API (~2-3 min)
• Train 3 ML models + K-Means clustering
• Launch the Streamlit dashboard at http://localhost:8501

3. Or run step by step

# Phase 1: Data pipeline
python src/pipeline/fdic_pipeline.py

# Phase 2: ML risk engine
python src/ml/risk_engine.py

# Phase 3: Dashboard
streamlit run src/dashboard/app.py

---

🤖 Machine Learning

Supervised Models (Binary Classification: At-Risk vs Stable)

Model	ROC-AUC	Accuracy	F1
Random Forest	~0.92	~87%	~0.84
Gradient Boosting	~0.91	~86%	~0.83
Logistic Regression	~0.88	~83%	~0.79

Label Construction: A bank is labeled "at-risk" if it historically failed (matched to FDIC failure list) OR currently breaches critical KRI thresholds (negative ROA, CAR < 8%, NPL > 10%).

Features used:

• Return on Assets (ROA)
• Return on Equity (ROE)
• Capital Adequacy Ratio (Basel III)
• Non-Performing Loan Ratio (NPL)
• Loan-to-Deposit Ratio
• Net Interest Margin
• Cost-to-Income Ratio

Unsupervised Model (K-Means Clustering)

4 risk clusters: Low Risk → Elevated Risk → High Risk → Critical Risk

---

📈 KRI Framework (Basel III / CCAR Aligned)

KRI	Warning Threshold	Critical Threshold	Regulatory Basis
Capital Adequacy Ratio	< 10%	< 8%	Basel III
Non-Performing Loan Ratio	> 5%	> 10%	FDIC guidance
Return on Assets	< 0.5%	< 0%	Peer benchmarks
Loan-to-Deposit Ratio	> 85%	> 90%	Liquidity risk
Cost-to-Income Ratio	> 60%	> 70%	Efficiency ratio

---

🖥️ Dashboard Features

Tab	Description
📈 Risk Overview	Risk tier distribution, scatter plots, failure history timeline
🗺️ KRI Heatmap	Interactive US choropleth — filter by any KRI metric
🤖 ML Scoring	Model comparison, feature importance, high-risk watchlist
✅ Control Testing	Control exception counts, breach rates, decade failure analysis
🔍 Institution Lookup	Search any bank by name for full risk profile

---

🗂️ SQL Queries

sql/schema_and_queries.sql includes 6 production-style KRI queries:

1. Control Exception Report — Basel III capital breach detection
2. Multi-Flag Watch List — Banks with 3+ simultaneous KRI breaches
3. State-Level KRI Summary — Regulatory dashboard view
4. Historical Failure Cost Analysis — Decade-by-decade failure trends
5. Risk-Based Sampling — Statistical sampling for control testing (replicates CCAR methodology)
6. YoY Trend Analysis — Institution-level ROA deterioration detection

---