This project aims to forecast the closing prices of some of India’s most influential companies over the next 30 days using Meta’s Prophet time series forecasting model. In today's volatile economic environment shaped by global events, policy changes, and market cycles this project provides a robust tool for short-term forecasting and long-term investment analysis.
The study combines deep historical analysis with machine learning to uncover patterns, volatility trends, and stock-specific behaviors between January 1, 2020 and April 10, 2025. It also evaluates how a diversified portfolio of these companies would have performed over the last 5 years and projects what the next month may look like for investors.
- Python
- Prophet (by Meta)
- Pandas, NumPy for data manipulation
- Matplotlib, Seaborn, Plotly for visualization
- yFinance for stock data retrieval
- Jupyter Notebook for analysis workflow
The following Indian companies were selected for their significance in the stock market and influence in the economy:
- Tata Consultancy Services (TCS)
- Infosys
- HDFC Bank
- Bharti Airtel
- State Bank of India (SBI)
- Tata Motors
- Reliance Industries
- Sharp decline in early 2020 due to the COVID-19 crash.
- 2021–2023 marked a recovery phase, with most companies rebounding steadily.
- In early 2025, many stocks (TCS, Infosys, Tata Motors) faced corrections, possibly triggered by Trump-era tariffs and economic policy shifts.
- TCS maintained the highest stock price, peaking near ₹4500 before correcting in 2025.
- Infosys and Airtel showed strong long-term growth and steady uptrends.
- Tata Motors and SBI saw strong upward momentum post-2021 but remained volatile.
- HDFC exhibited the most stable growth, ideal for low-risk portfolios.
- Reliance grew consistently, supported by diversification into retail and digital sectors.
- High volatility across all stocks in early 2020 during the pandemic crash.
- Post-2021: Most companies showed stabilized return patterns.
- Tata Motors showed the highest daily return swings, indicating risk.
- HDFC, TCS, and Infosys displayed relatively stable return profiles.
- Massive volatility spikes in 2020 (6–7%) for most companies.
- Post-2022: Volatility for most stocks converged around 1–2%, indicating market stabilization.
- Tata Motors and SBI had periodic volatility spikes due to market sensitivity.
- HDFC maintained the lowest rolling volatility, highlighting its conservative nature.
- Annualized Return: 20.03%
- Annualized Volatility: 20.74%
- Sharpe Ratio: 0.97
- Strong long-term growth across the portfolio.
- Portfolio stayed resilient during crashes and corrections.
- Equal weighting helped maintain balance between aggressive and defensive stocks.
- Noticeable portfolio correction in early 2025 aligns with observed stock-specific dips.
- Tata Motors delivered the highest return, especially from 2022 to 2024.
- Airtel and Infosys outperformed the Nifty 50 index consistently.
- Reliance and TCS grew steadily but closely tracked the index.
- HDFC underperformed the index but offered the most consistent returns.
- SBI showed strong mid-period growth but higher fluctuations.
- The Nifty 50 served as a reliable benchmark for market trends and relative performance.
Prophet, developed by Meta, is designed for accurate forecasting of time series data with strong seasonal effects and multiple trend changepoints. Features include:
- Automatic detection of weekly, monthly, and yearly seasonality
- Handles missing values and outliers
- Provides interpretable trend and uncertainty intervals
- Allows for future forecasting even with limited historical data
- Trained a separate Prophet model for each company using the
ds(date) andy(closing price) format. - Generated 30-day future forecasts from April 10, 2025.
- Visualized forecasts and decomposed trends using
plot_components()to understand seasonal behavior.
- HDFC, Infosys, and TCS are ideal for low-volatility, long-term investment strategies.
- Tata Motors and SBI offer high reward but come with higher risk.
- Forecasts can be used for short-term positioning and swing trading based on the next 30 days of expected movement.
- Prophet performs well in handling complex, real-world time series data like stock prices.
- A data-driven strategy significantly enhances portfolio decision-making in unpredictable market conditions.
- Clone the repository
git clone https://github.com/Mindmasterparav/Finance.git