Applied Quantitative Finance: NVIDIA (NVDA) Stock Analysis

This repository contains an advanced, interactive, and modular notebook focused on the quantitative financial analysis of NVIDIA Corporation (ticker: NVDA). The notebook demonstrates how to combine fundamental financial data, time-series analysis, volatility modeling, and risk simulation using Python-based tools and libraries.

---

Project Structure & Scope

1. Fundamental Dashboard

This section gathers essential company-specific data using yfinance. It includes:

• CAPM Beta and company metadata
• Shareholding Structure (major, institutional, and mutual fund holders)
• Dividend & Stock Split History
• Financial Statements (Income Statement, Balance Sheet, Cash Flow)
• Recent News and headlines

All data is retrieved programmatically and displayed in an interactive tabbed view for easy navigation and comparison.

---

2. Exploratory Data Analysis (EDA)

❯ Configuration

• Stocks Analyzed: NVDA, MSFT, AMD, INTC, ADBE
• Time Period: 2020-01-01 to 2025-06-05
• Frequency: Daily (1d)
• Source: Yahoo Finance

❯ Data Prepared

• Adjusted Close Prices (dividend & split-adjusted)
• Simple Returns:
  $( R_t = \frac{P_t}{P_{t-1}} - 1)$
• Log Returns:
  $( r_t = \ln\left(\frac{P_t}{P_{t-1}}\right))$

Each return series is analyzed structurally and visually for missing values, distributional properties, and volatility.

---

Time-Series Modeling

3. Volatility Models

❯ AR(1)-GARCH(1,1)

• Fitted with Student’s t-distribution
• Captures volatility clustering and heavy tails

❯ EGARCH(1,1)

• Captures asymmetry in volatility shocks
• Incorporates leverage effects

❯ Outputs

• Daily conditional volatility forecast
• 10-day ahead simulations with confidence bounds
• Inter-model comparison based on RMSE, MAE, AIC/BIC

---

Risk Management

4. Monte Carlo Simulation & VaR/CVaR

• Parametric simulation of return paths
• 1% and 5% Value at Risk (VaR) estimation
• Conditional VaR (CVaR) for tail-risk modeling
• Kupiec Test for statistical backtesting

Results are visualized with interactive plots and explained with financial intuition.

---

Diagnostics & Validation

• Residual Diagnostics: Ljung-Box, Q-Q plots
• Normality Tests: Jarque-Bera, Shapiro-Wilk
• Model Validity: Distribution fits and volatility forecasts
• Rolling Windows: Robustness checks over time

---

Visualization Tools

All plots are generated with plotly, allowing:

• Interactive zoom and pan
• Hover tooltips
• Annotated charts for model interpretation
• Dynamic subplots with shared x-axis for comparative view

---

Key Libraries Used

• pandas, numpy
• yfinance
• statsmodels
• arch
• scipy, matplotlib, plotly

---

Use Cases

This notebook is ideal for:

• Financial Econometrics Projects
• Academic Thesis & Research
• Quantitative Risk Analysis
• Algorithmic Trading Pipeline
• Time-Series Forecasting & Backtesting

---

File Structure

📦 Applied_Quantitative_Finance_NVDA.ipynb
├── 📊 Fundamental Dashboard (Tabs)
├── 🧮 Return Calculation & EDA
├── 📈 GARCH/EGARCH Modeling
├── 🔁 Forecasting & Simulation
├── ⚠️ Risk Metrics (VaR, CVaR)
├── 🧪 Diagnostics & Backtesting
├── 📈 Technical Analysis

---

Reference

This notebook is based on methods and examples presented in:

Aleš Kresta (2024). Applied Quantitative Finance in Python: Selected theories and examples. ResearchGate

Please cite this work accordingly if you use or modify the notebook for academic or professional purposes.