Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Comparative Analysis of Econometric and Machine Learning Approaches for Forecasting Bitcoin Return Volatility

Persson Podvorec, Filip LU (2024) NEKN02 20241
Department of Economics
Abstract (Swedish)
The fundamental principle of investing, applicable at both systematic and individual levels, is the art of balancing the portfolio between expected returns and acceptable risk exposure. Risk management involves predicting the direction of the underlying asset’s movement and taking appropriate actions accordingly. Traditionally, statistical models such as GARCH have been used to achieve this objective. However, with advances in machine learning, new algorithms may prove more effective by identifying more complex patterns. This paper compares the accuracy of GARCH and LSTM models in predicting future Bitcoin return volatility using the root mean squared error (RMSE) on out-of-sample data. The results indicate that the LSTM model outperforms... (More)
The fundamental principle of investing, applicable at both systematic and individual levels, is the art of balancing the portfolio between expected returns and acceptable risk exposure. Risk management involves predicting the direction of the underlying asset’s movement and taking appropriate actions accordingly. Traditionally, statistical models such as GARCH have been used to achieve this objective. However, with advances in machine learning, new algorithms may prove more effective by identifying more complex patterns. This paper compares the accuracy of GARCH and LSTM models in predicting future Bitcoin return volatility using the root mean squared error (RMSE) on out-of-sample data. The results indicate that the LSTM model outperforms the GARCH model, suggesting significant potential for machine learning applications in enhancing risk management and forecasting practices. (Less)
Please use this url to cite or link to this publication:
author
Persson Podvorec, Filip LU
supervisor
organization
course
NEKN02 20241
year
type
H1 - Master's Degree (One Year)
subject
keywords
Bitcoin, Forecast, LSTM, GARCH, Risk Management, Machine Learning
language
English
id
9156818
date added to LUP
2024-08-12 15:59:09
date last changed
2024-08-12 15:59:09
@misc{9156818,
  abstract     = {{The fundamental principle of investing, applicable at both systematic and individual levels, is the art of balancing the portfolio between expected returns and acceptable risk exposure. Risk management involves predicting the direction of the underlying asset’s movement and taking appropriate actions accordingly. Traditionally, statistical models such as GARCH have been used to achieve this objective. However, with advances in machine learning, new algorithms may prove more effective by identifying more complex patterns. This paper compares the accuracy of GARCH and LSTM models in predicting future Bitcoin return volatility using the root mean squared error (RMSE) on out-of-sample data. The results indicate that the LSTM model outperforms the GARCH model, suggesting significant potential for machine learning applications in enhancing risk management and forecasting practices.}},
  author       = {{Persson Podvorec, Filip}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Comparative Analysis of Econometric and Machine Learning Approaches for Forecasting Bitcoin Return Volatility}},
  year         = {{2024}},
}