This project provides a comprehensive framework for analyzing stock momentum and risk using advanced data processing, clustering, and visualization techniques. Momentum investing capitalizes on the continuation of existing price trends, where stocks that have performed well in the past are expected to continue performing well in the future. By identifying high-momentum stocks and their associated risk levels, this project provides actionable insights that can be leveraged by traders, investors, and portfolio managers to enhance decision-making processes.
The Fama-French 5-factor model enhances the analysis by accounting for multiple sources of return, including market, size, value, profitability, and investment factors. Incorporating this model allows the project to contextualize momentum within broader market dynamics and adjust for factors that could influence returns beyond momentum alone.
By calculating rolling factor betas, the project captures the sensitivity of stocks to these factors over time. Shifting these betas forward aligns them with future returns, providing a forward-looking risk-adjusted view of stock performance.
Securities that have performed well in the past continue to perform well in the near term. This project uses momentum as a signal for potential future performance, allowing investors to make informed buy or sell decisions. The Relative Strength Index (RSI) is a popular momentum indicator that measures the speed and change of price movements.
By calculating the average Relative Strength Index (RSI) over a three-month period, we identify stocks with high and low momentum, providing a quantitative basis for momentum classification.
ATR is used alongside RSI to capture volatility, providing a broader view of a stock's behavior. High ATR values suggest higher volatility, which is crucial for understanding the potential risk associated with a stock.
Clustering helps in segmenting stocks into distinct groups based on shared characteristics. This project uses KMeans clustering on features like RSI and Average True Range (ATR) to identify patterns and group stocks with similar risk profiles.
Understanding the risk profile of stocks is crucial for constructing portfolios that align with an investor’s risk tolerance and investment objectives. By categorizing stocks into 'Green' (high momentum, lower risk) and 'Red' (low momentum, higher risk) the project offers a clear framework for risk assessment.
Stocks are classified into risk groups based on their momentum scores. The top 20% in momentum are labeled as 'Green,' indicating strong performance and potential for continued growth. Conversely, the bottom 20% are labeled as 'Red,' signaling potential underperformance or higher risk.
Liquidity is a critical consideration in trading and investment. Highly liquid stocks allow for easier entry and exit positions, reducing the impact of market movements on trades. By filtering for the most liquid stocks based on rolling average dollar volumes, the project ensures that the identified momentum stocks are practical for real-world trading strategies.
data.py
Downloads historical stock data for S&P 500 stocks. Preprocesses and structures the dataset to include necessary features like price and volume. Ensures data consistency and handles missing values, saves for further analysis in other modules.
technical_indicators.py
Calculates advanced technical indicators such as RSI, ATR, and moving averages and enhances the stock data by adding features that capture momentum and volatility It incorporates technical analysis into the momentum assessment framework, providing a richer dataset for clustering and risk analysis.
filter_liquid.py
Aggregates stock data to a monthly frequency for a smoother analysis. Calculates rolling average dollar volumes to measure stock liquidity and retains only the most liquid stocks for analysis. Ensures that selected stocks are practical for trading based on their liquidity. Outputs a refined dataset focusing on high-liquidity stocks.
RollingOLS.py
Downloads Fama-French 5-factor data to incorporate macroeconomic factors and calculates rolling regression betas for stocks, measuring sensitivity to each factor. Normalizes the factor betas to standardize their scale. Integrates factor exposures into the overall risk assessment of stocks and prepares the data for use in risk-adjusted momentum analysis.
momentum_tracker.py
Identifies stocks with the highest and lowest momentum using RSI. Sorts stocks into momentum-based risk categories: 'Green', 'Red', and 'Grey'. Applies clustering to segment stocks into distinct risk groups then provides actionable insights into stock momentum and risk profiles.
returns.py
Computes monthly returns for stocks over specified time horizons. Calculates cumulative returns to assess long-term performance. Adjusts returns data for corporate actions like splits and dividends and analyzes returns to identify trends and patterns over time.
clusters.py
Applies KMeans clustering to group stocks based on specified features, analyzes clusters to identify patterns in stock behavior. Visualizes and assesses the resulting clusters using scatter plots. Generates visual representations of clusters over various dates to track changes over time.