Bitcoin Returns Research Articles

Forecasting Bitcoin returns using machine learning algorithms: impact of investor sentiment

PurposeThis study aims to assess the predictive performance of various factors on Bitcoin returns, used for the development of a robust forecasting support decision model using machine learning techniques, before and during the COVID-19 pandemic. More specifically, the authors investigate the impact of the investor's sentiment on forecasting the Bitcoin returns.Design/methodology/approachThis method uses feature selection techniques to assess the predictive performance of the different factors on the Bitcoin returns. Subsequently, the authors developed a forecasting model for the Bitcoin returns by evaluating the accuracy of three machine learning models, namely the one-dimensional convolutional neural network (1D-CNN), the bidirectional deep learning long short-term memory (BLSTM) neural networks and the support vector machine model.FindingsThe findings shed light on the importance of the investor's sentiment in enhancing the accuracy of the return forecasts. Furthermore, the investor's sentiment, the economic policy uncertainty (EPU), gold and the financial stress index (FSI) are the top best determinants before the COVID-19 outbreak. However, there was a significant decrease in the importance of financial uncertainty (FSI and EPU) during the COVID-19 pandemic, proving that investors attach much more importance to the sentimental side than to the traditional uncertainty factors. Regarding the forecasting model accuracy, the authors found that the 1D-CNN model showed the lowest prediction error before and during the COVID-19 and outperformed the other models. Therefore, it represents the best-performing algorithm among its tested counterparts, while the BLSTM is the least accurate model.Practical implicationsMoreover, this study contributes to a better understanding relevant for investors and policymakers to better forecast the returns based on a forecasting model, which can be used as a decision-making support tool. Therefore, the obtained results can drive the investors to uncover potential determinants, which forecast the Bitcoin returns. It actually gives more weight to the sentiment rather than financial uncertainties factors during the pandemic crisis.Originality/valueTo the authors’ knowledge, this is the first study to have attempted to construct a novel crypto sentiment measure and use it to develop a Bitcoin forecasting model. In fact, the development of a robust forecasting model, using machine learning techniques, offers a practical value as a decision-making support tool for investment strategies and policy formulation.

Predicting volatility of bitcoin returns with ARCH, GARCH and EGARCH models

The ınvestment decisions of institutional and individual investors in financial markets are largely influenced by market uncertainty and volatility of the investment instruments. Thus, the prediction of the uncertainty and volatilities of the prices and returns of the investment instruments becomes imperative for successful investment. In this study we seek to identify the best fit model that can predict the volatility of return of Bitcoin, which is in high demand as an investment tool in recent times. Using the opening data of weekly Bitcoin prices for the period of 11.24.2013–03.22.2020, their logarithmic returns were calculated. The stationarity properties of the Bitcoin return series was tested by applying the ADF unit root test and the series were found to be stationary. After reaching the average equation model as ARMA (2.2), it was tested whether there was an ARCH effect in the ARMA (2,2) model. As a result of the applied ARCH-LM test, it is reached that the residuals of the average equation model selected have ARCH effect. Volatility of Bitcoin return series after detection of ARCH effect has been tried to predict with conditional variance models such as ARCH (1), ARCH (2), ARCH (3), GARCH (1,1), GARCH (1,2), GARCH (1,3), GARCH (2,1), GARCH (2,2), EGARCH (1,1) and EGARCH (1,2). While the obtained findings indicate that the best model is in the direction of GARCH (1,1) according to Akaike info criterion, it was found that GARCH (1,1) model does not have ARCH effect as a result of the applied ARCH-LM test. Thus, our empirical findings highlight an ample guide on appropriate modeling of price information in the Bitcoin market.

The ripple effects of CBDC-related news on Bitcoin returns: Insights from the DCC-GARCH model

Central bank digital currencies (CBDCs) have emerged as a potential substitute for current payment methods, and, as such, major announcements, events and policy discussions regarding CBDCs have the potential to influence cryptocurrency returns. In light of this, the present study undertakes an in-depth analysis of the CoinMarketCap data between August 1, 2017 and April 1, 2022 by implementing the dynamic conditional correlation-generalized autoregressive conditional heteroskedasticity (DCC-GARCH) model. The study reveals a noteworthy influence of news and events related to CBDCs on Bitcoin returns. Precisely, CBDC uncertainty index and CBDC attention index have resulted in significant fluctuations in Bitcoin returns, indicating that positive news can result in significant Bitcoin returns. The findings suggest that future expectations of investors regarding cryptocurrencies are shaped by CBDC-related news and events.

All topics are not created equal: Sentiment and hype of business media topics and the bitcoin market

Using topic modeling analysis, we identify six bitcoin-related topics - Investing, Bitcoin Market, Regulation, Private Equity, Cybersecurity, and Macroeconomy - discussed in the Wall Street Journal and document differences in their sentiment (polarity) and hype (intensity of coverage) over the 2013–2021 period. Importantly, we find the topics differ not only in strength (significant vs. insignificant) but also in direction (negative vs. positive) and shape (linear vs. non-linear) of their relations to bitcoin returns, trading volume, and volatility.

Is Bitcoin Resilient for Indian Investors? The E-GARCH Analysis

This paper explores the financial asset capabilities and hedge alternative properties of Bitcoin by investigating several aspects of its volatility in relation to Nifty50 and Indian exchange rates (USD/INR & EUR/INR). This study delves into the volatility dynamics of the returns of Bitcoin. An asymmetric GARCH model (E-GARCH) is used to examine whether Bitcoin may play crucial role in risk management and ideal for risk-averse investors in anticipation of leverage effect. This paper also examines Bitcoin as an investment and hedge alternative to Nifty50 and exchange rates. The findings suggest that Bitcoin does not attributes the properties of safe hedge and an investment alternative by Indian investors

Transfer‐entropy‐based dynamic feature selection for evaluating Bitcoin price drivers

AbstractDespite the growing literature in cryptocurrency forecasting and their price drivers, the relationship between their price and other financial time series is an ongoing matter of debate. This study proposes a three‐step methodology to cover these arguments. First, we conduct an ad hoc analysis using transfer entropy (TE) to study the causal relationship between Bitcoin (BTC) returns and a vast array of financial time series. Then, we utilize variables with a significant amount of information flow toward BTC returns to forecast multi‐step‐ahead BTC returns. Finally, we use explainable artificial intelligence post hoc analysis methods to discover the contribution of each input feature to the overall forecasting. The results indicate a significant change in the information flow pattern in the first days of the COVID‐19 pandemic outbreak. Additionally, our proposed TE‐based feature‐selection method outperforms both benchmarks, a nonfeature‐selection model, and backward stepwise regression.

Bitcoin volatility and the introduction of bitcoin futures: A portfolio construction approach

We evaluate the introduction of Bitcoin futures on Bitcoin return and volatility using realized volatility and GARCH model pre-and post-futures. We also assess the portfolio construction implications by building two portfolios containing the top 25 S&P stocks, one without futures and one with. GARCH and realized volatility show mixed results. We provide that futures make Bitcoin riskier and more vulnerable to fluctuations over time. However, Bitcoin futures improve the portfolio's volatility and returns profile. Our findings offer implications regarding portfolio strategies implemented by risk-averse and risk-seeking investors and managers as we show how Bitcoin futures can hedge investments.

Cryptocurrency return dependency and economic policy uncertainty

This study investigated whether economic policy uncertainty (EPU) affects the degree of Bitcoin return dependency in the cryptocurrency (crypto) market by calculating the Pearson correlation between Bitcoin returns and the average returns of 19 cryptocurrencies to determine the return dependency. First, we found that an increase in global EPU strengthens the dependency effect. Among the country-level EPU indexes, we found that the increased EPU of crypto-friendly countries (United States, United Kingdom, Japan, and South Korea) is the main enhancer of dependency. By contrast, the EPU index of crypto-unfriendly countries (China and Russia) exhibits a statistically nonsignificant effect. Finally, the empirical results remain similar if we replace Bitcoin with Ethereum to measure dependency.

Do bitcoin news information flow and return volatility fit the sequential information arrival hypothesis and the mixture of distribution hypothesis?

The financial asset return volatility and information field have continued to compare both hypotheses: sequential information arrival hypothesis (SIAH) and the mixture of distribution hypothesis (MDH). However, numerous former studies have not found an appropriate information indicator but just used trading volume as an indirect proxy. The study examines the relationship between Bitcoin return volatility and information flow instead of the trading volume. We apply a text and web mining to get all related 24,316 news items for Bitcoin from 64 news websites. Next, we apply a sentiment analysis of natural language processing (NLP) to generate information flow data to replace the traditional trading volume. Finally, we appropriate vector autoregressive (VAR) models to catch the lead-lag relationship and Spearman Correlation to test contemporaneous nexus. The study results show that Bitcoin return volatility is affected by the negative information flow and parallels SIAH; the positive information flow impacts Bitcoin return volatility and matches MDH. The empirical result benefits investors in making proper investment decisions in Bitcoin, and the gist of the paper fills the gap in academic literature because the aspect of information is still absent in academia.

Do NFTs act as a good hedge and safe haven against Cryptocurrency fluctuations?

In this study, we test the hedge and safe haven properties of NFTs against Bitcoin and Ethereum market fluctuations in the backdrop of COVID-19 and the Russia-Ukraine war. We employ daily returns of Bitcoin, Ethereum, and four NFTS, namely Decentraland, Cryptopunks, Cryptokitties, and SuperRare, from 04 April 2018 to 7 July 2022. For analytical purposes, we estimate dynamic hedge effectiveness and wavelet quantile correlations. The dynamic hedge effectiveness results confirm the NFTs' hedge ability against Bitcoin. The WQC results show that NFTs offer short, medium, and long-run hedge properties and short run safe haven property against Bitcoin. Further, we show that NFTs are at best a short to medium term diversifier against Ethereum fluctuations.

The Relationship Between Bitcoin Returns, Volatility, And Volume Using Asymmetric Garch Modelling

In 2020-2021 where the world experienced a Covid-19 pandemic, the price of Bitcoin increased and there was a fairly high price spike at the turn of the year 2020-2021 and had touched an all-time high (ATH) at the end of 2021. Bitcoin is one of the currencies crypto that is volatile when compared to the exchange rates of widely used currencies. In addition, Bitcoin price movements are difficult to predict. The samples used are Euro, Pound Sterling, Yuan, Yen, Ruble, Franc, and Bitcoin. The data used is in the form of Bitcoin price data and some of these currencies in 2020-2021. This study aims to find the relationship between volatility with Bitcoin trading volume, return with Bitcoin trading volume, and return with Bitcoin volatility for forecasting purposes. The research method used is the Augmented Dickey-Fuller (ADF) stationarity test, then using the ARMA model after that using the EGARCH model to find the relationship. The results of this study indicate that there is a positive relationship between volatility and Bitcoin trading volume. Just like the return and trading volume of Bitcoin there is a positive relationship. However, Bitcoin returns, and volatility have a negative relationship. For further research can use different crypto currency assets with different time periods.
 Keywords: Volatility, Volume, Returns, Bitcoin, GARCH

Extrapolative beliefs about Bitcoin returns

Using survey data from Sentix which records whether investors are bullish, bearish, or neutral about the future price of Bitcoin, we examine how investors form expectations about Bitcoin returns. We use a model for expectations with an exponential decay of weights on past returns and find that investors extrapolate from Bitcoin’s past returns in a way that puts more weight on more recent returns. Compared to institutional investors, individuals display a higher degree of extrapolation, consistent with the perception that nonprofessionals are more prone to behavioral biases. The extrapolative component of expectations is a significant predictor of Bitcoin returns in the short term, however, its explanatory power is small. Understanding how investors in the cryptocurrency market form their beliefs is important because variation in beliefs could affect Bitcoin price dynamics and potentially lead to excess volatility, momentum, and reversal.

A hybrid approach for forecasting bitcoin series

Bitcoin price prediction is a substantial challenge for cryptocurrency investors. This study offers an innovative scheme to predict Bitcoin returns and volatilities using a hybrid model that incorporates the autoregressive fractionally integrated moving average (ARFIMA), empirical wavelet (EW) transform, and local linear wavelet neural network (LLWNN) approaches to produce an ARFIMA-EWLLWNN model. Our methodologies integrate the advantages of the long-memory model, EW decomposition technique, artificial neural network structure, and backpropagation and particle swarm optimization learning algorithms. The experimental results of the optimized hybrid approach outperform some classic models by providing more accurate out-of-sample forecasts over longer horizons. The model proves to be the most appropriate Bitcoin forecasting technique. Moreover, the implemented method produces smaller prediction errors than other computing techniques.

Exploring the relationship between digital trails of social signals and bitcoin returns

PurposeThis study aims to empirically investigate the linkages between digital trails of social signals (content and profile features of bitcoin-related tweets) and bitcoin price return using a VAR-BEKK-GARCH model.Design/methodology/approachBitcoin-related tweets were collected every hour for six months from September 1, 2020, to February 29, 2021. The analysis involved two steps: first, examining tweet content, profiles, sentiment and emotions; and second, investigating the relationship between social signal volatility and hourly bitcoin price return.FindingsResults indicate that bitcoin price changes can impact the sentiment expressed in tweets about bitcoin, and vice versa. While sadness exhibits a bidirectional volatility spillover with bitcoin, fear and anger display a one-period lag. Quartile analyses reveal that only fear in the second quartile shows a bidirectional spillover effect with bitcoin, while all other emotions except sadness demonstrate a unidirectional spillover effect in all remaining quartiles.Originality/valueThe study uses a novel two-step approach to analyze volatility spillovers between social signals and bitcoin price returns. Findings can guide investors and portfolio managers in making better allocation decisions and assist policymakers and regulators in reducing the adverse effects of bitcoin’s volatility on financial system stability.

Are All Text News Just a Noise for Investors? Impact of Online Texts on Bitcoin Returns

Abstract The paper demonstrates the power of alternative data. Relying on the indicators obtained by mining online publicly available news articles, authors analyze their impact on Bitcoin returns. This research shows that in the first quarter of 2022 Bitcoin returns could be explained by the sentiment of information obtained from news published on online portals. However, we find negative relation between Bitcoin news sentiment and its returns. Such result can be explained as anomaly of researched period which is characterized by inception of global political crisis caused by the war in Eastern Europe and turmoil on crypto market. Our research also confirms that the news about Ethereum, Bitcoins’ investment substitute, affected Bitcoin's returns as well. On the other hand, the obtained results show that there is no relation between the lexical readability of the news (i.e., the clarity with which the text is written, measured by the fog index) and the returns on Bitcoin in the analyzed period. Collected evidences speak in favor of Bitcoin’s market inefficiency. In this paper we also demonstrate that returns forecasts based on online news are more accurate in comparison to those generated by ARMA-GARCH model, a conventional financial tool for predicting returns.

Forecasting bitcoin volatility: exploring the potential of deep learning

This study aims to evaluate forecasting properties of classic methodologies (ARCH and GARCH models) in comparison with deep learning methodologies (MLP, RNN, and LSTM architectures) for predicting Bitcoin's volatility. As a new asset class with unique characteristics, Bitcoin's high volatility and structural breaks make forecasting challenging. Based on 2753 observations from 08-09-2014 to 01-05-2022, this study focuses on Bitcoin logarithmic returns. Results show that deep learning methodologies have advantages in terms of forecast quality, although significant computational costs are required. Although both MLP and RNN models produce smoother forecasts with less fluctuation, they fail to capture large spikes. The LSTM architecture, on the other hand, reacts strongly to such movements and tries to adjust its forecast accordingly. To compare forecasting accuracy at different horizons MAPE, MAE metrics are used. Diebold–Mariano tests were conducted to compare the forecast, confirming the superiority of deep learning methodologies. Overall, this study suggests that deep learning methodologies could provide a promising tool for forecasting Bitcoin returns (and therefore volatility), especially for short-term horizons.

How media coverage news and global uncertainties drive forecast of cryptocurrencies returns?

This paper aims to investigate the impact of global financial, economic, and gold price uncertainty indices (VIX, EPU, and GVZ) and investor sentiment based on media coverage news on the returns of Bitcoin and Ethereum during the COVID-19 pandemic. We adopt an asymmetric framework based on the Quantile-on-Quantile approach, which examines the quantiles of the cryptocurrency returns, investor sentiment, and the various uncertainties indicators. The empirical findings suggest that the COVID-19 pandemic has significantly impacted cryptocurrency returns. Specifically, (i) the results demonstrate the predictive power of Economic Policy Uncertainty (EPU) during this period, as evidenced by a strong negative association between EPU and cryptocurrency returns across all quantiles; (ii) the correlation between cryptocurrency returns and the VIX index was negative but weak, across various quantile combinations of Ethereum and Bitcoin returns; (iii) an increase in COVID-19 news negatively affected Bitcoin returns across all quantiles; (iv) Bitcoin and Ethereum cannot be relied upon as effective hedging tools against global financial and economic uncertainty during the COVID-19 pandemic. Studying the behavior of cryptocurrency during uncertainty like pandemics is extremely important because it provides investors with insights on diversifying their portfolios and hedging their risks.

Uncovering Information Linkages between Bitcoin, Sustainable Finance and the Impact of COVID-19: Fractal and Entropy Analysis

This study aimed to uncover the impact of COVID-19 on the leading cryptocurrency (Bitcoin) and on sustainable finance with specific attention to their potential long memory properties. In this article, the application of the selected methodologies is based on a fractal and entropy analysis of the econometric model in the financial market. To detect the regularity/irregularity property of a time series, approximate entropy is introduced to measure deterministic chaos. Using daily data for Bitcoin and sustainable finance, namely DJSW, Green Bond, Carbon, and Clean Energy, we examine long memory behaviour by employing a rescaled range statistic (R/S) methodology. The results of the research present that the returns of Bitcoin, the Dow Jones Sustainability World Index (DJSW), Green Bond, Carbon, and Clean Energy have a significant long memory. Contrastingly, an interdisciplinary approach, namely wavelet analysis, is also used to obtain complementary results. Wavelet analysis can provide warning information about turmoil phenomena and offer insights into co-movements in the time–frequency space. Our findings reveal that approximate entropy shows crisis (turmoil) conditions in the Bitcoin market, despite the nature of the pandemic’s origin. Crucially, compared to Bitcoin assets, sustainable financial assets may play a better safe haven role during a pandemic turmoil period. The policy implications of this study could improve trading strategies for the sake of portfolio managers and investors during crisis and non-crisis periods.

Are non‐fungible token coins a good hedge against the stock market volatility?

AbstractWe test the hedge property of non‐fungible token (NFT) coins against equity market fluctuations and compare it with the hedge property of Bitcoin. We employ daily the returns of Bitcoin; three NFT coins, namely Theta, Enjin Coin and Decentraland, and three equity market indices: S&P 500, NASDAQ and CAC 40, ranging from 18 January 2018 to 12 January 2021. We estimate the hedge effectiveness of the three NFT coins and Bitcoin against stock market fluctuations. Our results suggest that NFT coins are a better hedge against equity market fluctuations than Bitcoin.

Time–frequency analysis of cryptocurrency attention

ABSTRACT We present a wavelet analysis of retail investor attention and the daily returns of Bitcoin, Ethereum, and Litecoin at five selected crypto exchanges that identifies the fractal dynamics of the short- and long-term persistent processes. The investors’ attention is proxied by the Search Volume Index provided by Google at daily frequency. We detect significant temporal cyclical movements and coherence between cryptocurrency returns and retail investor attention at long investment horizons: from the beginning of 2017 to the middle of 2018 and, to a lesser degree, in 2019. Investment horizons that dominated in 2017 and 2018 were mainly driven by retail investor attention rather than by uncertainty, risk, or stock markets. Therefore, we do not confirm that cryptocurrencies can be considered a safe-haven asset in times of crisis because there is no significant negative comovement between the returns of cryptocurrencies and stock returns or economic uncertainty. Furthermore, the phase shift analysis indicates that attention can serve as a leading indicator for the cryptocurrency returns, particularly in 2017 and 2018. Therefore, retail investors are encouraged to use the Search Volume Index as an early warning indicator in case of sudden changes in the cryptocurrency returns to maximise profits or minimise losses.