Figure - available from: The Journal of Supercomputing
This content is subject to copyright. Terms and conditions apply.
Source publication
Bitcoin is widely recognized as the first decentralized digital cryptocurrency based on blockchain technology. Its unique properties make it a leading contender in the realm of digital currencies, and it continues to maintain a dominant position in the short-term. However, the high volatility of bitcoin prices poses significant challenges for predi...
Similar publications
In this study, the method employed for predicting future Bitcoin prices is centred around simulating consistent growth in the cryptocurrency's value, particularly during periods characterised by near-all-time highs (ATH). The objective is to provide a qualitative estimation of Bitcoin's potential price trajectory over a specific time horizon.
Disc...
Citations
... The MaxSharpe strategy's aim to optimize risk-adjusted returns is more effective with ESG criteria, although it may involve higher risks for the returns achieved, consistent with Zhao et al. (2023). In the Post-COVID period, Portfolio 1 showed remarkable performance with the highest return at approximately 1,429.22%, ...
... The MaxSharpe strategy's aim to optimize risk-adjusted returns is more effective with ESG criteria, although it may involve higher risks for the returns achieved. This observation is consistent with Zhao et al. (2023), who noted that optimization strategies require careful risk management. ...
... For example, the use of moving averages (50-day and 200-day) as trading signals has been shown to improve returns and manage risk effectively(Natashekara & Sampath, 2024). Furthermore, a novel cryptocurrency price time series hybrid prediction model using machine learning with MATLAB/Simulink has demonstrated potential in enhancing the accuracy of price predictions, thereby optimizing portfolio performance(Zhao et al., 2023).2.3 ENVIRONMENTAL AND SOCIAL CONSIDERATIONSEnvironmental and social considerations are becoming increasingly important in the context of cryptocurrency investments. The energy consumption of cryptocurrency mining operations has raised concerns about their environmental impact. ...
Objective: This study investigates how Environmental, Social, and Governance (ESG) criteria can be integrated into cryptocurrency portfolio strategies, evaluating their performance across different market conditions and time periods. Theoretical Framework: This research is based on Modern Portfolio Theory (MPT) and principles of ESG investing. The study uses Markowitz's mean-variance optimization and the triple bottom line approach to understand the benefits of ESG integration in investment strategies. Method: The research involves a comparative analysis of various cryptocurrency portfolio strategies, including Buy-and-Hold, Simple Moving Average (SMA), MinVar, and MaxSharpe. Data was collected daily from October 1, 2016, to September 31, 2021. The study uses mean-variance analysis to assess risk-return profiles, incorporating ESG factors into the evaluation framework. Results and Discussion: The results show that the Buy-and-Hold strategy consistently yielded the highest returns across most portfolios. However, during volatile periods, strategies like MinVar and MaxSharpe provided better risk-adjusted returns. The discussion contextualizes these results within the theoretical framework, highlighting how ESG integration enhances risk management and aligns investments with sustainable development goals (SDGs). Research Implications: This research suggests that integrating ESG criteria into cryptocurrency portfolios can improve risk management and align investments with sustainability goals. These findings have practical implications for investment strategy development and sustainable finance practices. Originality/Value: This study offers a unique analysis of cryptocurrency portfolio strategies that incorporate ESG criteria. Its findings are relevant for influencing sustainable investment practices and optimizing cryptocurrency portfolios in line with ESG principles.
... Notably, individual time-series forecasting models often encounter challenges in concurrently deciphering linear and nonlinear data associations, culminating in compromised forecasting accuracy [12]. For instance, while the ARIMA's capacity to decipher nonlinear patterns has been acknowledged as limited, isolated neural network models are reported to require vast datasets to avert overfitting and entail meticulous data preprocessing [13,14]. Additionally, a significant portion of prevailing research seems to fall short in cohesively melding demand forecasting with supply balance strategies, resulting in production designs that manifest reduced adaptability and versatility. ...
... It is a preprocessing step which extracts features from the raw data. Feature engineering in Machine learning consists of 4 main processes: Feature creation, Transformation, Feature extraction and feature selection [24,25] . ...
Cryptocurrency emerged in the market as an asset with significant market capitalization; attracting traders, investors and researchers alike. The nature of cryptocurrency is very much volatile and dynamic which is the key challenge for the researchers for prediction of the cryptocurrency prices. In recent years, machine learning techniques along with deep learning techniques have witnessed promising results in various financial forecasting domains. This research paper presents a comprehensive investigation of Utility cryptocurrency price movement (XRP and Chainlink) using Deep Learning techniques. The study aims to compare the price using different methodologies. The research focuses on long short-term memory (LSTM), gated recurrent units (GRU). Historical price data of XRP and Chainlink are employed to train and evaluate the models using different evaluation metrics such as Root Mean Square Error (RMSE), Mean Absolute Error (MAE), R2 score, Regression Score, (MGD), (MPD). This research contributes to the growing body of knowledge concerning cryptocurrency price prediction by shedding light on the effectiveness of time series models, sentiment analysis, and their hybridization. The objective is to populate findings that have significant implications for different stakeholders like investors, traders,, and financial institutions seeking to make informed decisions in the highly volatile cryptocurrency market.
