Forecasting Bitcoin volatility using machine learning techniques

IF 5.4 2区经济学 Q1 BUSINESS, FINANCE

Journal of International Financial Markets Institutions & Money Pub Date : 2024-10-19 DOI:10.1016/j.intfin.2024.102064

Zih-Chun Huang , Ivan Sangiorgi , Andrew Urquhart

引用次数: 0

Abstract

This paper studies the Bitcoin volatility forecasting performance between popular traditional econometric models and machine learning techniques. We compare the 1-day to 2-month ahead forecasting performance of the Long Short-Term Memory (LSTM) and a hybrid Convolutional Neural Network-LSTM (CNN-LSTM) model to the traditional models. We find that neural networks outperform Generalised Autoregressive Conditional Heteroskedasticity (GARCH) models for all forecasting horizons. Furthermore, the LSTM model outperforms the Heterogeneous Autoregressive (HAR) model and by integrating the Markov Transition Field (MTF) into the CNN-LSTM model, we achieve superior forecasting results in the short-term, particularly for the 7-day forecasts.

查看原文本刊更多论文

利用机器学习技术预测比特币波动性

本文研究了流行的传统计量经济学模型与机器学习技术之间的比特币波动率预测性能。我们比较了长短期记忆（LSTM）和混合卷积神经网络-LSTM（CNN-LSTM）模型与传统模型的 1 天至 2 个月的预测性能。我们发现，在所有预测期限内，神经网络都优于广义自回归条件异方差（GARCH）模型。此外，LSTM 模型优于异质自回归（HAR）模型，而且通过将马尔可夫转换场（MTF）整合到 CNN-LSTM 模型中，我们在短期内取得了优异的预测结果，尤其是在 7 天预测方面。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Journal of International Financial Markets Institutions & Money Multiple-

CiteScore

6.60

自引率

10.00%

发文量

142

期刊介绍： International trade, financing and investments, and the related cash and credit transactions, have grown at an extremely rapid pace in recent years. The international monetary system has continued to evolve to accommodate the need for foreign-currency denominated transactions and in the process has provided opportunities for its ongoing observation and study. The purpose of the Journal of International Financial Markets, Institutions & Money is to publish rigorous, original articles dealing with the international aspects of financial markets, institutions and money. Theoretical/conceptual and empirical papers providing meaningful insights into the subject areas will be considered. The following topic areas, although not exhaustive, are representative of the coverage in this Journal. • International financial markets • International securities markets • Foreign exchange markets • Eurocurrency markets • International syndications • Term structures of Eurocurrency rates • Determination of exchange rates • Information, speculation and parity • Forward rates and swaps • International payment mechanisms • International commercial banking; • International investment banking • Central bank intervention • International monetary systems • Balance of payments.