Liquidity & Resource Management Dashboard

A production-style analytics dashboard for a global markets / trading floor resource management use case. Built with Python, Pandas, Plotly Dash, and Dash Bootstrap Components.

---

Business Problem

Trading desks generate thousands of positions daily across repos, bonds, and equity financing. Risk managers and senior stakeholders need to:

• Monitor liquidity usage against approved limits in real time.
• Track secured funding costs by desk, product, and counterparty.
• Identify overborrow positions where notional exceeds collateral value.
• Evaluate what-if scenarios before executing new trades.
• Run stress tests to see how portfolios behave under shock conditions.

Without a unified view, teams rely on fragmented spreadsheets that create blind spots and slow down risk decisions.

---

Target Users

Role	Primary Need
Traders	Real-time liquidity headroom, funding cost
Risk Managers	Utilization rates, overborrow, limit breaches
Resource Management	Collateral efficiency, secured funding allocation
Senior Stakeholders	Executive KPI summary, stress scenario outcomes

---

Key Features

• 4-tab Dash app — Executive Overview, Desk Drilldown, What-If Scenario, Stress Testing
• KPI cards — Market value, utilization, funding cost, overborrow, high-risk count
• Interactive filters — Date range, desk, product, counterparty, currency, risk level
• Scenario upload — Upload hypothetical CSV and compare against base case
• Stress sliders — Shock haircuts, rates, market values, and limits interactively
• Risk flags — Auto-assigned High / Medium / Low based on utilization thresholds
• 5,500+ synthetic positions across 90 trading days

---

Financial Metrics

Metric	Formula
Liquidity Usage	market_value × liquidity_weight
Utilization Rate	total_liquidity_usage ÷ liquidity_limit
Collateral Adjusted Value	market_value × (1 − haircut)
Daily Funding Cost	notional × funding_rate ÷ 365
Annual Funding Cost	notional × funding_rate
Overborrow Amount	max(0, notional − collateral_adjusted_value)
Collateral Efficiency	collateral_adjusted_value ÷ annual_funding_cost

Risk Flags

Flag	Threshold
High Risk	Utilization ≥ 90%
Medium Risk	Utilization ≥ 75%
Low Risk	Utilization < 75%

---

Tech Stack

Component	Technology
Language	Python 3.11
Data	Pandas, NumPy
Dashboard	Plotly Dash 2.x
UI Components	Dash Bootstrap Components
Database (optional)	PostgreSQL / SQLite + SQLAlchemy
Containerization	Docker, Docker Compose
Testing	Pytest

---

Data Model

positions             — 5,500+ rows, 90 days, 5 desks
liquidity_limits      — Per-portfolio limits and risk thresholds
funding_rates         — Product × collateral × currency rate matrix
counterparties        — 15 counterparty profiles with credit ratings

---

Project Structure

liquidity-resource-dashboard/
├── app.py                    # Main Dash application
├── requirements.txt
├── Dockerfile
├── docker-compose.yml
├── data/
│   ├── sample_positions.csv
│   ├── sample_funding_rates.csv
│   ├── sample_liquidity_limits.csv
│   └── sample_counterparties.csv
├── src/
│   ├── data_generator.py     # Synthetic data generation
│   ├── data_loader.py        # CSV / DB loading with caching
│   ├── metrics.py            # Core financial calculations
│   ├── scenario_engine.py    # What-if and stress logic
│   ├── risk_flags.py         # Risk flag assignment
│   └── charts.py             # Plotly figure factories
├── sql/
│   ├── schema.sql            # PostgreSQL / SQLite schema
│   └── sample_queries.sql    # Analytical SQL examples
└── tests/
    ├── test_metrics.py
    └── test_scenario_engine.py

---

How to Run Locally

# 1. Clone the repo
git clone https://github.com/yourname/liquidity-resource-dashboard.git
cd liquidity-resource-dashboard

# 2. Create a virtual environment
python -m venv .venv
source .venv/bin/activate        # Windows: .venv\Scripts\activate

# 3. Install dependencies
pip install -r requirements.txt

# 4. Generate synthetic data
python src/data_generator.py

# 5. Launch the dashboard
python app.py

Open http://localhost:8050 in your browser.

---

How to Run with Docker

# Build and start
docker compose up --build

# Open http://localhost:8050

---

Running Tests

pytest tests/ -v

---

What-If Scenario CSV Format

Upload a CSV with these columns to Tab 3:

desk,product_type,market_value,notional,haircut,funding_rate,liquidity_weight,maturity_days
Repo Desk,Repo,50000000,49000000,0.02,0.055,0.95,7

---

Future Improvements

• Live DB integration — swap CSV loader for SQLAlchemy + PostgreSQL
• Authentication — add Dash BasicAuth or OAuth2
• Real-time data — connect to market data feed via WebSocket
• Alerting — email / Slack notification when a portfolio breaches 90%
• Power BI embed — publish aggregated KPIs to a Power BI workspace
• Databricks — run heavy ETL jobs on Databricks, serve results via Delta Lake

---