09. Demo: Computing Performance

Part 1

Exploring the Investment Universe

Learn about essential aspects of structuring an investment universe using function libraries, with a focus on understanding historical prices and returns.

Key Components:

• Importing Libraries:
  • Essential code components include the function library and Pandas to analyze data effectively.
  • Use relative paths to access saved functions correctly.
• Understanding Historical Prices and Returns:
  • Generate historical prices and returns using the create_historical_prices function with one momentum.
• Ticker Analysis:
  • Identify changes in the number of stock tickers over time, impacting returns.
  • Observe the variation from 353 to 498 tickers, suggesting market evolution.
• Survivorship Bias:
  • Recognize the effect of excluding removed tickers (usually due to underperformance) on data, resulting in higher average returns.
• Performance Calculation:
  • Compute benchmark performance using the compute BM performance function.
  • Steps include computing daily mean returns, cumulative returns, CAGR, and Sharpe Ratio.
  • Use plotting to visualize cumulative and annual returns, providing insights into performance consistency.

This understanding aids in more informed analyses within a structured investment universe.

Part 2

Introduction to RSI Indicator Creation

This demonstration focuses on the creation of the RSI (Relative Strength Index) indicator to evaluate trading signals' quality. The key process involves:

• Historical Data Preparation:

  • Utilize pre-existing historical prices and total returns.
  • Isolate one-day returns for a single stock (e.g., Apple) to keep it illustrative.

• RSI Calculation Methodology:

  • Define gains and losses: Gains are returns > 0, losses < 0.
  • Calculate rolling means for gains and losses over a specified window.
  • Use forward fill to handle missing values, ensuring seamless data continuity.
  • Compute the RSI as gain-to-loss ratio, normalizing between 0-100.

• Implementation and Analysis:

  • Apply the calculate RSI function across multiple data sets to evaluate all tickers.
  • Assess the relationship between RSI and forward returns; note the weak correlation typical in investment data.

• Practical Insights:

  • Examine RSI value bins for stocks with values below 30 or above 70; these bins can indicate potential underpricing or overpricing.
  • Calculate average returns for different bins, focusing on bins with RSI < 30 for possible higher yield.

These steps provide a foundational understanding of implementing RSI in trading algorithms, setting the stage for future explorations and optimizations.