📈 Predictive Financial Analytics Dashboard

A professional-grade stock market analytics system built entirely with Python, Pandas & NumPy — no ML libraries, no shortcuts. Features real-time feature engineering, manual predictive modeling, portfolio optimization via Monte Carlo simulation, and 10 auto-generated financial visualizations.

🔍 Overview

This project demonstrates mastery of data wrangling, numerical computation, and statistical analysis applied to real-world financial data. It pulls historical OHLCV data for 5 major stocks (AAPL, MSFT, GOOGL, AMZN, TSLA) and runs a full analytics pipeline:

Raw CSV Data  →  Cleaning  →  Feature Engineering  →  Prediction  →  Portfolio Optimization  →  Visualizations

Everything from RSI calculation to linear regression to the efficient frontier is implemented from scratch using only Pandas and NumPy arrays.

✨ Features

📥 1. Data Acquisition & Cleaning

Loads multiple CSVs of historical OHLCV stock data
Handles missing values with forward fill, backward fill, and linear interpolation
Normalizes and standardizes data using NumPy (Min-Max & Z-score)
Aligns all stocks to a common date index automatically

⚙️ 2. Feature Engineering

Simple & Exponential Moving Averages (MA20, MA50, EMA20)
Daily returns, cumulative returns, annualized volatility
Bollinger Bands — dynamic support/resistance levels
RSI (Relative Strength Index) — momentum oscillator
Sharpe Ratio — risk-adjusted return metric
Momentum indicators

📊 3. Time-Series Analysis

Rolling statistics: mean, std, skewness, kurtosis
Full correlation matrix across all stocks
Trend detection using MA crossover signals
Autocorrelation analysis with pure NumPy

🔮 4. Predictive Modeling (No ML Libraries)

Linear Regression implemented manually using NumPy OLS formula: β = (X'X)⁻¹ X'y
Moving Average Forecast for short-term price prediction
Model evaluation: MSE, RMSE, MAE — all computed with NumPy

💼 5. Portfolio Optimization

Monte Carlo simulation — 10,000 random portfolio scenarios
Mean-variance optimization using NumPy covariance matrices
Identifies the optimal portfolio (maximum Sharpe Ratio)
Identifies the minimum variance portfolio

📉 6. Visualizations (10 Charts)

Price charts with Bollinger Bands & moving averages
RSI with overbought/oversold zones
Daily returns distribution histograms
Cumulative returns comparison
Rolling volatility over time
Correlation heatmap
Efficient frontier scatter plot
Sharpe ratio bar chart
Risk vs Return scatter
Short-term price forecasts

📁 Project Structure

financial_dashboard/
│
├── 📂 data/                      # Raw stock CSV files
│   ├── AAPL.csv
│   ├── MSFT.csv
│   ├── GOOGL.csv
│   ├── AMZN.csv
│   └── TSLA.csv
│
├── 📂 output/                    # All generated results
│   ├── 📂 plots/                 # 10 generated PNG charts
│   ├── correlation_matrix.csv
│   ├── monte_carlo_results.csv
│   └── {TICKER}_features.csv
│
├── 📄 data_loader.py             # Module 1: Data loading & cleaning
├── 📄 features.py                # Module 2: Feature engineering
├── 📄 time_series.py             # Module 3: Time-series analysis
├── 📄 predictor.py               # Module 4: Predictive modeling
├── 📄 portfolio.py               # Module 5: Portfolio optimization
├── 📄 visualizer.py              # Module 6: All visualizations
├── 📄 main.py                    # Pipeline entry point
├── 📄 download_data.py           # One-time data downloader
├── 📄 requirements.txt
└── 📄 .gitignore

⚙️ Setup & Installation

Prerequisites

Python 3.10+
Git

1. Clone the repository

git clone https://github.com/YOUR_USERNAME/financial_dashboard.git
cd financial_dashboard

2. Create & activate virtual environment

# Windows
python -m venv virtualEnviroment
virtualEnviroment\Scripts\activate

# macOS / Linux
python -m venv virtualEnviroment
source virtualEnviroment/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Download stock data (one-time)

python download_data.py

▶️ How to Run

python main.py

Expected terminal output:

=======================================================
  STEP 1: Loading & Cleaning Data
=======================================================
  Loaded AAPL:  1006 rows | columns: ['Close', 'High', 'Low', 'Open', 'Volume']
  Loaded AMZN:  1006 rows | columns: ['Close', 'High', 'Low', 'Open', 'Volume']
  Loaded GOOGL: 1006 rows | columns: ['Close', 'High', 'Low', 'Open', 'Volume']
  Loaded MSFT:  1006 rows | columns: ['Close', 'High', 'Low', 'Open', 'Volume']
  Loaded TSLA:  1006 rows | columns: ['Close', 'High', 'Low', 'Open', 'Volume']

=======================================================
  STEP 2: Engineering Features
=======================================================
  AAPL  Sharpe Ratio: 0.8312
  AMZN  Sharpe Ratio: 0.4483
  GOOGL Sharpe Ratio: 0.6419
  MSFT  Sharpe Ratio: 0.7844
  TSLA  Sharpe Ratio: 1.1065

=======================================================
  STEP 3: Time-Series & Correlation Analysis
=======================================================
Correlation Matrix:
        AAPL   AMZN  GOOGL   MSFT   TSLA
AAPL   1.000  0.624  0.691  0.777  0.511
AMZN   0.624  1.000  0.664  0.679  0.454
GOOGL  0.691  0.664  1.000  0.773  0.428
MSFT   0.777  0.679  0.773  1.000  0.472
TSLA   0.511  0.454  0.428  0.472  1.000

=======================================================
  STEP 5: Portfolio Optimization (Monte Carlo)
=======================================================
  Optimal Portfolio (Max Sharpe Ratio):
    Expected Return : 50.31%
    Risk (Std Dev)  : 42.06%
    Sharpe Ratio    : 1.1962
    Weights:
      AAPL:  21.61%
      AMZN:   0.85%
      GOOGL:  1.64%
      MSFT:  30.96%
      TSLA:  44.94%

📊 Sample Output

Terminal Results Summary

Stock	Sharpe Ratio	Interpretation
TSLA	1.1065	✅ Excellent risk-adjusted return
AAPL	0.8312	✅ Good
MSFT	0.7844	✅ Good
GOOGL	0.6419	🟡 Acceptable
AMZN	0.4483	🟡 Below average

Optimal Portfolio Allocation

Stock	Weight
TSLA	44.94%
MSFT	30.96%
AAPL	21.61%
GOOGL	1.64%
AMZN	0.85%

🖼️ Visualizations

All charts are auto-generated and saved to output/plots/ when you run main.py.

📌 Efficient Frontier — Monte Carlo Portfolio Optimization

10,000 randomly simulated portfolios coloured by Sharpe Ratio. The red star marks the optimal portfolio.

📌 Correlation Heatmap

Pearson correlation between daily returns of all 5 stocks. AAPL–MSFT are most correlated (0.77).

📌 Cumulative Returns — All Stocks (2020–2024)

TSLA dramatically outperformed all others despite extreme volatility.

📌 Sharpe Ratio Comparison

Risk-adjusted performance across all 5 stocks. TSLA leads at 1.10.

📌 Risk vs Return

Annualized return plotted against annualized volatility per stock.

📌 Rolling Volatility

30-day annualized rolling volatility — TSLA shows significantly higher risk.

📌 AAPL — Price & Moving Averages with Bollinger Bands

📌 AAPL — RSI (14-period)

Highlights overbought (>70) and oversold (<30) zones.

📌 AAPL — Daily Returns Distribution

📌 AAPL — 10-Day Price Forecast

Short-term forecast using Moving Average model built with pure NumPy.

🛠️ Tech Stack

Tool	Purpose
Python 3.10+	Core language
Pandas	Data loading, cleaning, time-series operations
NumPy	All numerical computation, OLS regression, Monte Carlo
Matplotlib	All chart generation
yfinance	One-time historical data download only

⚠️ No scikit-learn, no statsmodels, no ML frameworks — everything is implemented from scratch.

📐 Key Algorithms Implemented from Scratch

# OLS Linear Regression — pure NumPy
β = (X'X)⁻¹ X'y
beta = np.linalg.pinv(X.T @ X) @ X.T @ y

# Sharpe Ratio
sharpe = (mean_excess_return / std_excess_return) * sqrt(252)

# Portfolio Variance
variance = weights @ covariance_matrix @ weights.T

# RSI
RS = avg_gain / avg_loss
RSI = 100 - (100 / (1 + RS))

# Bollinger Bands
upper = MA(20) + 2 * rolling_std(20)
lower = MA(20) - 2 * rolling_std(20)

📜 License

This project is licensed under the MIT License — feel free to use, modify, and distribute.

🙋 Author

Built by [Your Name] 🔗 LinkedIn • GitHub

⭐ If you found this useful, give it a star on GitHub!

Name		Name	Last commit message	Last commit date
__pycache__		__pycache__
data		data
output		output
README.md		README.md
data_loader.py		data_loader.py
features.py		features.py
main.py		main.py
portfolio.py		portfolio.py
predictor.py		predictor.py
requirements.txt		requirements.txt
time_series.py		time_series.py
visualizer.py		visualizer.py

Latest commit

History

Folders and files

Repository files navigation