Market Risk Dashboard

Project Objectives

• Deliver an interactive market risk dashboard focused on Brazilian equities and REITs.
• Automate price collection, cleaning, and incremental storage so the app stays fast after the first run.
• Offer risk analytics, portfolio optimisation, and Monte Carlo retirement planning in a single workflow.

Tech Stack

• Language: Python 3.12
• Core libraries: Streamlit, Plotly, Pandas, NumPy, SciPy, yfinance, python-dateutil
• Tooling: Git for version control

Getting Started

1. Clone the repository:

   git clone https://github.com/your-user/riskcontrol.git
   cd riskcontrol

2. (Optional) Create and activate a virtual environment.
3. Install dependencies:

   pip install -r requirements.txt

4. Launch the dashboard:

   streamlit run app.py

Key Features

Data collection and cache

• Downloads price history from yfinance with an optional rolling-window filter.
• Caches each ticker independently via st.cache_data, so previously downloaded assets are re-used instantly even when the watchlist changes.
• Ships with a default ticker list and a sidebar input to add any B3 symbol (suffix .SA).

Return analytics

• Interactive price history and daily return charts.
• Histogram of returns to assess the normality assumption.
• Drawdown curve to monitor peak-to-trough losses.
• Rolling annualised volatility for 21-day and 63-day windows.

Risk indicators

• Annualised volatility.
• Parametric VaR (95% and 99%).
• Historical VaR and CVaR / Expected Shortfall (95%).
• Sharpe ratio that honours the user-defined risk-free rate.
• Correlation heatmap to compare asset co-movements.

Portfolio and optimisation

• Manual weights with automatic normalisation plus an annual risk-free rate input.
• Portfolio summary with expected return, volatility, and Sharpe.
• Markowitz optimisation to maximise Sharpe or minimise volatility.
• Efficient frontier chart highlighting the optimal and minimum risk portfolios.

Financial planning (Monte Carlo)

• Detailed inputs: current and retirement ages, income, expenses, extra contributions, contribution growth, inflation, withdrawal rate, number of simulations, and RNG seed.
• Simulates annual real returns with rebalancing and inflows that can grow over time.
• Fan chart for wealth percentiles (p10, p25, p50, p75, p90) plus the final wealth distribution.
• Key outputs: probability of reaching the target wealth implied by the withdrawal rate, target level, median projection, and estimated annual contribution.

Interactive interface

• Tabs for Return Analysis, Risk, Correlation, Portfolio, and Financial Planning.
• Contextual guidance through expanders in each section.
• Plotly visualisations embedded directly in Streamlit.

Dashboard Previews

• Risk indicators dashboard - Consolidated volatility, VaR, CVaR, and Sharpe metrics overlaid with rolling volatility.
• Historical drawdown - Peak-to-trough history for the selected assets.
• Optimised portfolios - Side-by-side comparison of maximum Sharpe and minimum volatility allocations.
• Efficient frontier visualisations - Scatter and segmented frontier views with highlighted max Sharpe and min vol portfolios.
• Portfolio tab overview - Efficient frontier table, optimisation highlights, and sidebar controls.
• Monte Carlo planning - Probability of achieving goals and wealth distribution outcomes.

Repository Structure

• app.py: main Streamlit flow that wires together data, indicators, visuals, and simulations.
• models/data_extraction.py: handles downloads, rolling window definition, and cache persistence.
• models/indicators.py: volatility, VaR (parametric and historical), CVaR, Sharpe, and correlation calculations.
• models/portfolio.py: return/volatility helpers, weight normalisation, optimisation, and efficient frontier sampling.
• models/visualizations.py: interactive charts and tables (prices, returns, risk metrics, correlation, frontier, simulations).
• models/simulation.py: Monte Carlo engine for retirement planning scenarios.