/Volatility-Prediction-using-ML

Prediction of Portfolio Volatility with ML

Primary LanguageJupyter Notebook

Volatility Prediction using ML

  • Volatility Prediction using Voting Ensemble

    • Each Estimator's hyper parameters are optimized by Genetic Search
      • ElasticNet
      • SVR
      • RandomForest Regressor
      • XGBoost Regressor

  • goal : predict next month's volatility and find max-sharpe weights of portfolio at the predicted volatility

  • Train : 6 Months Close Price + Various technical indicators + Average Standard Deviation of portfolio

    • BWI
    • ATR
    • Relative Volatility Index
    • Standard Deviation
    • Sigma
    • True Range

  • Target : next month volatility of portfolio