Muhammad Nabil Rateb, Sameh Alansary, Marwa Khamis Elzouka, Mohamad Galal
Abstract Sentiment analysis is a powerful tool for extracting valuable insights from social media data. In this paper, more than one million tweets spanning three months (March, June, and December 2022) regarding three cryptocurrencies: Bitcoin (BTC), Ethereum (ETH), and Binance Coin (BNB) during the Russian-Ukrainian War are considered. Two models, a convolutional neural network with long short-term memory (CNN-LSTM) and a support vector machine (SVM) with GloVe and TF-IDF features, are trained on a labeled dataset of more than fifty thousand tweets about Bitcoin labeled as (positive, negative, and neutral). A pretrained model (Pysentimento) for sentiment analysis is also employed to compare the performances of the three models. The models are tested on the labeled dataset and then evaluated on the unlabeled tweets, revealing that Pysentimento's level of accuracy outperforms the other two models. Google Trends, along with the opening and closing prices, and the volume of the three cryptocurrencies, in addition to the results of Pysentimento sentiment classification, are employed to apply the Pearson correlation coefficient and conduct price prediction analysis using the SARIMA model. It is found that Bitcoin may appeal to those seeking stability and a known record of accomplishment, while Binance Coin and Ethereum may attract investors looking for more diverse opportunities. Sentiment analysis using machine learning is found to provide invaluable information for cryptocurrency price forecasting and trading strategies, especially in the context of geopolitical events and market volatility.
Abstract The Bitcoin market has experienced unprecedented growth, attracting financial traders seeking to capitalize on its potential. As the most widely recognized digital currency, Bitcoin holds a crucial position in the global financial landscape, shaping the overall cryptocurrency ecosystem and driving innovation in financial technology. Despite the use of technical analysis and machine learning, devising successful Bitcoin trading strategies remains a challenge. Recently, deep reinforcement learning algorithms have shown promise in tackling complex problems, including profitable trading strategy development. However, existing studies have not adequately addressed the simultaneous consideration of three critical factors: gaining high profits, lowering the level of risk, and maintaining a high number of active trades. In this study, we propose a multi-level deep Q-network (M-DQN) that leverages historical Bitcoin price data and Twitter sentiment analysis. In addition, an innovative preprocessing pipeline is introduced to extract valuable insights from the data, which are then input into the M-DQN model. A novel reward function is further developed to encourage the M-DQN model to focus on these three factors, thereby filling the gap left by previous studies. By integrating the proposed preprocessing technique with the novel reward function and DQN, we aim to optimize trading decisions in the Bitcoin market. In the experiments, this integration led to a noteworthy 29.93% increase in investment value from the initial amount and a Sharpe Ratio in excess of 2.7 in measuring risk-adjusted return. This performance significantly surpasses that of the state-of-the-art studies aiming to develop an efficient Bitcoin trading strategy. Therefore, the proposed method makes a valuable contribution to the field of Bitcoin trading and financial technology.
Dennis Koch, Vahidin Jeleskovic, Zahid Irshad Younas
This paper introduces a unique and valuable research design aimed at analyzing Bitcoin price volatility. To achieve this, a range of models from the Markov Switching-GARCH and Stochastic Autoregressive Volatility (SARV) model classes are considered and their out-of-sample forecasting performance is thoroughly examined. The paper provides insights into the rationale behind the recommendation for a two-stage estimation approach, emphasizing the separate estimation of coefficients in the mean and variance equations. The results presented in this paper indicate that Stochastic Volatility models, particularly SARV models, outperform MS-GARCH models in forecasting Bitcoin price volatility. Moreover, the study suggests that in certain situations, persistent simple GARCH models may even outperform Markov-Switching GARCH models in predicting the variance of Bitcoin log returns. These findings offer valuable guidance for risk management experts, highlighting the potential advantages of SARV models in managing and forecasting Bitcoin price volatility.
Based on the dataset of Bitcoin Price dataset, this paper studied Bitcoin price prediction by using support vector machine model, random forest model, neural network model, XGBoost model and LightGBM model. The models were evaluated by MSE, RMSE, MAE, MAPE and R². First, we divided the Bitcoin Price dataset into a training set and a test set according to a ratio of 7:3, with 70 as the training set and 30 as the test set. We take the stock price change (return) as the target variable, the other variables as the input variables, and use the training set to train the model and the test set to test the model. After model comparison, we found that XGBoost's MSE, RMSE, MAE, MAPE and R² are all optimal, and its prediction effect is also the best. The performance of the other four models ranges from good to different, including LightGBM, random Forest, support vector machine and neural network. Among them, the MSE of the neural network is dozens of times that of the other four models, so it performs the worst. The XGBoost model performs well in dealing with high-dimensional sparse data and nonlinear relationships, while LightGBM and random Forest are suitable for dealing with large-scale data. Support vector machines are suitable for dealing with high-dimensional data and nonlinear relationships, while neural networks require more tuning and optimization to take advantage of their advantages. In summary, the research results of this paper can provide value for the prediction of Bitcoin price in the future, and also provide a certain reference for selecting a suitable machine learning model.
David Alaminos, M. Belén Salas, Manuel Á. Fernández-Gámez
In recent years cryptographic tokens have gained popularity as they can be used as a form of emerging alternative financing and as a means of building platforms. The token markets innovate quickly through technology and decentralization, and they are constantly changing, and they have a high risk. Negotiation strategies must therefore be suited to these new circumstances. The genetic algorithm offers a very appropriate approach to resolving these complex issues. However, very little is known about genetic algorithm methods in cryptographic tokens. Accordingly, this paper presents a case study of the simulation of Fan Tokens trading by implementing selected best trading rule sets by a genetic algorithm that simulates a negotiation system through the Monte Carlo method. We have applied Adaptive Boosting and Genetic Algorithms, Deep Learning Neural Network-Genetic Algorithms, Adaptive Genetic Algorithms with Fuzzy Logic, and Quantum Genetic Algorithm techniques. The period selected is from December 1, 2021 to August 25, 2022, and we have used data from the Fan Tokens of Paris Saint-Germain, Manchester City, and Barcelona, leaders in the market. Our results conclude that the Hybrid and Quantum Genetic algorithm display a good execution during the training and testing period. Our study has a major impact on the current decentralized markets and future business opportunities.
In an attempt to assess the appropriateness of the best-practice lexicon-based approaches as opposed to novel learning-based models to extract the sentiment of textual content in the context of the cryptocurrency market, the current study provides further insights into the association between digital activity and price movement of cryptocurrencies. Using a sample of Bitcoin and Ethereum trade data, this study compares the performance of Harvard IV-4 and BERT models in conjunction with the well-known machine learning classifiers. It examines to what extent learning-based sentiment models can enhance the price movement prediction, compared to lexicon-based approaches, and whether the prediction is improved or impaired by introducing different features as input to the classifiers. Results indicate that the contribution of the selected learning-based model varies across the two cryptocurrencies, and predictions are better in the absence of trade volume as an input feature to the classifiers.
Bitcoin, a decentralized digital currency, has gained widespread acceptance and recognition in recent years. The prediction of Bitcoin prices is a challenging task due to its relatively young age and high volatility. Therefore, this study explores the accuracy of price prediction for Bitcoin using machine learning models and makes comparsion on the outcome of different models, Linear Regression, Long Short-Term Memory, and Recurrent Neural Network. This study utilizes the closing price of Bitcoin in USD from a Kaggle dataset as the independent variable. The study also adopts Mean Absolute Error (MAE) as the measurement indicators, and comparative performance analysis is conducted under various circumstances. The experimental results demonstrate that LR performs poorly in Bitcoin price prediction, while LSTM and RNN outperform LR. Further analysis reveals that LSTM performs better during price apexes, while RNN performs better during price recessions. Graphical representations illustrate the strengths and weaknesses of each model under different market scenarios. Through comparison, the article provides an insight for other researchers to choose corresponding machine learning models under different circumstances to predict bitcoin price.
Andrea Colombo, Teodoro Baldazzi, Luigi Bellomarini, Andrea Gentili · 5 authors
Recent extensions of Datalog that consider the temporal dimension as a first-class citizen have unlocked the possibility of using its temporal variants, such as DatalogMTL, to model and reason about complex financial domains. Very relevant ones are crypto-activity markets, which, according to the recent Markets in Crypto-Assets Regulation (MiCAR) of the EU, are described by white papers published by crypto-assets issuers. In particular, the issuers publish semi-structured information about the assets they are willing to offer. Then, the assets are implemented in decentralized finance contexts (i.e., in a blockchain) as executable scripts known as smart contracts. However, these scripts are often criticized for their complexity, which makes them challenging to understand and communicate. On the other hand, in our experience, the availability of a declarative and executable representation of a crypto-activity market fosters a better understanding of that market as well as improved transparency, reproducibility and, as a consequence, increased fairness. These characteristics are of major interest to the financial authorities for example for supervision purposes. In this paper, we study the problem of automatically translating textual descriptions of crypto-assets, written according to the MiCAR specifications, into DatalogMTL programs that represent and capture the respective crypto-activity market. To this end, we opt for a machine translation approach and leverage a Large Language Model. We discuss promising techniques and preliminary experimental results.
Dayan A. Perera, Jessica Lim, Shuta Gunraku, Wern Han Lim
This research introduces an innovative approach to forecasting cryptocurrency prices by combining user-generated content (UGC) and sentiment analysis with quantitative data. The primary goal is to overcome limitations in existing methods for market forecasting, where accurate forecasting is crucial for informed decision-making and risk mitigation. The paper suggests a robust prediction methodology by integrating sentiment analysis and quantitative data. The study reviews prior research on sentiment analysis and quantitative analysis of cryptocurrency and stock price prediction. It explores the integration of machine learning and deep learning techniques, an area not extensively explored before. The methodology employs Long Short-Term Memory (LSTM), Recurrent Neural Network (RNN), Bidirectional LSTM and Gated Recurrent Unit (GRU) models to capture temporal dependencies. Prediction accuracy is assessed using metrics including Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and a confusion matrix. Results show that GRU models excel in prediction, while RNN models outperform in predicting price movements; with an emphasis on the significance of a suitable data preprocessing pipeline towards improving model performance. In summary, this study demonstrates the effectiveness of integrating sentiment analysis and quantitative data for cryptocurrency price forecasting using UGC data.
It has been an explosive start to 2024 in terms of the share market performance. Driven by a wave of enthusiasm for tech heavyweights like Meta and Nvidia, USA’s S&P 500 index of large American firms is up 5% and has crossed the 5,000 mark for the first time ever. On February 22, Japan's Nikkei 225 broke its own record – which it had established in 1989. In Australia, despite some volatility due to speculation on the direction of Reserve Bank interest rates, its share market also has boomed. Given such share market performance, this article asks if it is time to think about investing exclusively in shares. To answer this, the article will consider two fundamental questions that affect investors in capital markets: (1) what is meant by investment risk vs. return, and (2) can investors optimise their risk-return relationship by holding a single asset type (like stocks), or by holding a diversified portfolio of different asset classes? It will also consider a third question:(3) can using cryptocurrencies such as bitcoin help investors to better diversify their portfolio?
Abstract. This article presents an analysis of the historical dynamics of the cryptocurrency market based on time series data using the OHLCV dataset. The study presents the results of calculations of the main cryptocurrency market momentum technical indicators. Using intelligent computational methods, the paper assesses patterns and trends in the data of major cryptocurrencies. The study emphasizes the importance of technical analysis in understanding the volatile landscape of digital currencies. Key words: time-series, data analysis, cryptocurrency market, momentum indicators, technical analysis indicators, OHLCV.
We investigate whether investors rely more on technical trading language to rationalise price movements in the absence of substantive information. Compared to equity markets, cryptocurrency markets are characterised by high volatility, often occurring without clear explanation from new information. We apply a machine-learning-based vocabulary of technical trading terms to comments from cryptocurrency- and equity-related subreddits on Reddit.com and analyse how investors use technical talk in different market conditions. We find a U-shaped relationship between technical talk and Bitcoin returns, with higher usage during extreme price movements, while technical talk on equity subreddits is concentrated around median market returns. Technical talk increases in cryptocurrency markets when news is scarce but rises in equity markets alongside greater news availability. Our results suggest that technical talk provides an important communication channel for social media users to describe price variation when information is scarce.
F Farnuod Ahmadi, Abbas Toloie Eshlaghy, Reza Radfar
Cryptocurrencies have been widely identified and established as a new form of electronic currency exchange, carrying significant implications for emerging economies and the global economy. This research focused on the "examination and comparison of the efficiency of MLP and SimpleRNN algorithms in predicting cryptocurrency prices" using the Python programming language. Price predictions for Bitcoin, Ethereum, Binance Coin, Cardano, and Ripple were made using two deep learning algorithms (including the MLP algorithm and the SimpleRNN algorithm) over the period from 2017 to 2023. The results of cryptocurrency price prediction using deep learning algorithms were satisfactory; and the comparison of predictions across all cryptocurrencies indicated minimal differences between the algorithms studied, suggesting that they were efficient and had low error rates. Based on the obtained results regarding Bitcoin price prediction, the best algorithm was SimpleRNN; for Ethereum price prediction, the best algorithm was MLP; for Binance Coin price prediction, the best algorithm was SimpleRNN; for Cardano price prediction, the best algorithm was MLP; and for Ripple price prediction, the best algorithm was MLP.
The work carried out a comparative analysis of scientific publications regarding the possibility of predicting the direction of the cryptocurrency exchange rate using the data of open numerical indicators, based on the results of which it can be concluded that due to the volatility of the cryptocurrency market and the need for accurate forecasting, there is a need to create an aggregated indicator that will take into account the current price exchange rate asset, parameters of simple indicators, trading volume, etc. In addition, this indicator will be a parameter for the application of a multi-criteria analysis model in the process of supporting decision-making for cryptocurrency trading. A software decision support system for cryptocurrency traders on the Trading View platform has also been developed, which allows the cryptocurrency trader to get the value of the current situation of the cryptocurrency market in the form of a value using the method of weighting coefficients and selected indicators. Among the selected indicators: RSI, MA, CCI, Stochastic Oscillator, OBV, ADX, CMF to determine the moment of opening a position, and Fibonacci Retracement, Ichimoku Cloud to determine the closing of positions. Taking into account all the indicators and the coefficients determined for them, the obtained range of values is from 0 % to 100 %. If the value of the indicator exceeds the threshold of 20 %, it means that it is necessary to inform the trader about a possible entry point. That is, a value of 20 % to 40 % is weak performance, 40 % to 60 % is medium performance, 60 % to 80 % is strong performance, and a value greater than 80 % will not be overlapped by new pyramiding values for a better overall indicator success rate. The value of the indicator determines the potential effectiveness of opening positions, and thanks to the RSI indicator, the direction of opening positions is determined. The direction of the position is divided into long and short. An indicator has been developed for the TradingView platform, which, unlike existing simple indicators, collects data from open access and calculates a potential point for opening a position. Obtaining the numerical value of a single indicator saves the trader time to review and analyze a collection of indicators and time to decide on opening a position, as the cryptocurrency market is known for its sudden volatility, where a decision must be made quickly.
Researchers and experts have developed various techniques, models, and methods to predict the price movements of cryptocurrencies, especially Bitcoin.However, among the many techniques studied in the literature, there is still a lack of focus on mining, creating, and developing datasets with specific patterns for predicting the next cryptocurrency price movement.This is an exciting reason to conduct further research.A web-based Patterned Dataset Application and a Telegram bot were constructed to address this issue.These tools read the price position of each cryptocurrency and predict the next price direction based on the last position indicated by the Patterned Dataset Application.The experiment's results show that when the Patterned Dataset Application shows a diamond crash position, it is time to make a purchase; conversely, when it shows a diamond moon position, it is time to make a sale.It is hoped that by utilizing the Patterned Dataset Application, potential losses can be minimized, and there is more potential for profit in cryptocurrency trading.Even though the initial data source comes from Indonesia's most prominent digital cryptocurrency trading market, according to coinmarketcap, namely Indodax, the results of this patterned dataset application can often describe the same cryptocurrency conditions globally.The novelty of this research is to produce a new way of predicting the next cryptocurrency price movement using patterned datasets.At the end of this paper, it will be proven that hypothesis 1 and hypothesis 2 on the results of the patterned dataset are true.
This paper delves into the intriguing realm of cryptocurrency price prediction, with a specific focus on Zcash (ZEC), employing a cutting-edge deep learning approach.The study introduces two crucial features, "close_off_high" and "volatility", then systematically analyzes the correlations between these variables and the price of ZEC.By investigating the predictive accuracy of three prominent neural network architectures-Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and the Transformer model-the study discerns that LSTM and GRU models outperform the others in forecasting ZEC's price movements.Furthermore, the paper scrutinizes the influence of different activation functions on model performance, shedding light on the effectiveness of the linear activation function in this context.The research also addresses common challenges in predictive modeling, such as overfitting and multicollinearity.Moreover, it candidly acknowledges the limitations associated with solely focusing on a single cryptocurrency, recognizing that broader research efforts and interdisciplinary collaboration are required for a more comprehensive understanding of the ever-evolving cryptocurrency landscape.As the cryptocurrency market continues to evolve rapidly, this study provides invaluable insights for investors, offering a rational perspective on cryptocurrency investment.It underscores the importance of utilizing appropriate models and embracing interdisciplinary cooperation to navigate the complex and dynamic world of cryptocurrency.By bridging the gap between the cutting-edge world of deep learning and the financial market, this research paves the way for enhanced future investigations and more informed investment decisions.
This paper comprehensively examines Large Language Models (LLMs) in sentiment analysis, specifically focusing on financial markets and exploring the correlation between news sentiment and Bitcoin prices. We systematically categorize various LLMs used in financial sentiment analysis, highlighting their unique applications and features. We also investigate the methodologies for effective data collection and categorization, underscoring the need for diverse and comprehensive datasets. Our research features a case study investigating the correlation between news sentiment and Bitcoin prices, utilizing advanced sentiment analysis and financial analysis methods to demonstrate the practical application of LLMs. The findings reveal a modest but discernible correlation between news sentiment and Bitcoin price fluctuations, with historical news patterns showing a more substantial impact on Bitcoin’s longer-term price than immediate news events. This highlights LLMs’ potential in market trend prediction and informed investment decision-making.