Inference and Prediction of Cryptocurrency Market Returns
(2020) NEKN02 20201Department of Economics
- Abstract
- The potential for making profits investing in cryptocurrencies, the hedging benefits and the role in global economy, make it relevant to study the determinants of cryptocurrencies and to analyze different returns prediction models. Previous studies have focused one or some cryptocurrencies, this study analyzes the cryptocurrency market as a whole and finds the determinants of the cryptocurrency market and a returns prediction model using a machine learning approach. Evaluating the immediate impact of features, divided in cryptocurrency market data, information demand, financial markets, exchange rates and macroeconomics, it was found that the most important determinants of the cryptocurrency market returns is the cryptocurrency market... (More)
- The potential for making profits investing in cryptocurrencies, the hedging benefits and the role in global economy, make it relevant to study the determinants of cryptocurrencies and to analyze different returns prediction models. Previous studies have focused one or some cryptocurrencies, this study analyzes the cryptocurrency market as a whole and finds the determinants of the cryptocurrency market and a returns prediction model using a machine learning approach. Evaluating the immediate impact of features, divided in cryptocurrency market data, information demand, financial markets, exchange rates and macroeconomics, it was found that the most important determinants of the cryptocurrency market returns is the cryptocurrency market data. For prediction of the next-day returns, the USD-CNY exchange rate emerged as the most important determinant. Different returns prediction models are evaluated using Lasso, Regression Tress, Random Forest and Boosting. Random Forest presents the best prediction accuracy and can be used to predict the cryptocurrency market returns. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9023391
- author
- Alarcon Valdez, Emilia Felicidad LU
- supervisor
- organization
- course
- NEKN02 20201
- year
- 2020
- type
- H1 - Master's Degree (One Year)
- subject
- keywords
- Cryptocurrency Market, Machine Learning, Prediction, Inference, Determinants
- language
- English
- id
- 9023391
- date added to LUP
- 2020-08-29 11:11:54
- date last changed
- 2020-08-29 11:11:54
@misc{9023391, abstract = {{The potential for making profits investing in cryptocurrencies, the hedging benefits and the role in global economy, make it relevant to study the determinants of cryptocurrencies and to analyze different returns prediction models. Previous studies have focused one or some cryptocurrencies, this study analyzes the cryptocurrency market as a whole and finds the determinants of the cryptocurrency market and a returns prediction model using a machine learning approach. Evaluating the immediate impact of features, divided in cryptocurrency market data, information demand, financial markets, exchange rates and macroeconomics, it was found that the most important determinants of the cryptocurrency market returns is the cryptocurrency market data. For prediction of the next-day returns, the USD-CNY exchange rate emerged as the most important determinant. Different returns prediction models are evaluated using Lasso, Regression Tress, Random Forest and Boosting. Random Forest presents the best prediction accuracy and can be used to predict the cryptocurrency market returns.}}, author = {{Alarcon Valdez, Emilia Felicidad}}, language = {{eng}}, note = {{Student Paper}}, title = {{Inference and Prediction of Cryptocurrency Market Returns}}, year = {{2020}}, }