2008 yılında temelleri atılmış olan Kiripto para kavramı, 2017 yılı Aralık ayı itibari ile 19.060 ABD dolarına ulaşmış ve tanınırlığını arttırmıştır. Bitcoin ve sayıları 2700’ü bulan diğer kripto paralar hızlı kazanç elde etmek isteyen yatırımcıların dikkatini çekmeyi başarmıştır. Bu kapsamda kripto paraların fiyatının nasıl ve ne yönde değişeceği birçok kesim tarafından araştırma konusu olmuştur. Bu çalışmanın amacı, Bitcoin, Ethereum, IOTA ve Ripple gibi farklı altyapısal özellikleri olan kripto paraların gelecek fiyatını geçmişte gerçekleşen fiyatlardan hareketle tahmin etmektir. Çalışmada Deng Ju-Long tarafından 1980’li yıllarda ortaya atılan gri sistem teorisi ile fiyat tahminlemesi yapılmıştır. Çalışmada kullanılan geçmiş fiyatlar 11 günlük süreci kapsamaktadır. Literatüre göre kısa sayılabilecek bu süre modelin diğer modellere görece üstünlüğünü göstermektedir. Elde edilen sonuçlara göre GM(1,1) model ve Rolling-GM(1,1) model sonuçlarının birbirine çok yakın hata oranlarıyla tahmin yaptıkları ve yapılan tahminlere ait hata oranlarının çok düşük olduğu görülmüştür.
Bitcoin, as the most popular cryptocurrency, has received increasing attention from both investors and researchers over recent years. One emerging branch of the research on bitcoin focuses on empirical bitcoin pricing. Machine learning methods are well suited for predictive problems, and researchers frequently apply these methods to predict bitcoin prices and returns. In this study, we analyze the existing body of literature on empirical bitcoin pricing via machine learning and structure it according to four different concepts. We show that research on this topic is highly diverse and that the results of several studies can only be compared to a limited extent. We further derive guidelines for future publications in the field to ensure a sufficient level of transparency and reproducibility.
Tri Wijayanti Septiarini, Muhammad Rifki Taufik, Mufti Afif, Atika Rukminastiti Masyrifah
Abstract The objective of this study were (i) to construct the classical statistic and artificial intelligent model for predicting bitcoin cryptocurrency, and (ii) to compare the predicting performance by using root mean square error (RMSE) and mean square error (MSE) as forecasting evaluation tool. The observation data used in this study were collected during January, 5 2017 to October, 1 2019 (in total 1,000 daily observation data). The statistical method used in this study were ARIMA (Autoregressive Moving Average) and Exponential Smoothing. The artificial intelligent model were used in this study were fuzzy time series and ANFIS (Adaptive Neuro Fuzzy Inference System). The partitions data set were of 75%-25% of training and testing, respectively. The cryptocurrency investigated was bitcoin (BTC) which is the top three of most widely traded cryptocurrency. The forecasting results show that the classical method has the smallest value of RMSE and MSE which is exponential smoothing with 9749.81 for MSE and 98.74 for RMSE. However, the performance of forecasting method cannot be guaranteed from either classical or modern forecasting method. Analyzing with different method can be considered for future study, for example machine learning, neural network, modified fuzzy time series, etc.
This article analyzes the relationship between Bitcoin and the stock market by using a vector autoregressive model. To enhance the impulse response signal, the Sliding Window technique is applied. Study results show the relationship between Bitcoin and the stock market. First, the S&P 500 has a relatively significant effect on Bitcoin, while the influence caused by the S&P 500 is weak. In addition, after involving the Sliding Window technique, the effects caused by the standard deviation of the S&P 500 and the mean of the Dow Jones are remarkably strong on the mean of Bitcoin and the standard deviation of the S&P 500 has a comparatively significant effect on the standard deviation of Bitcoin as well. Generally, the S&P 500 and the Dow Jones indexes have an advantageous effect on Bitcoin. Financial investment can be made based on this model and conclusion.
Algorithmic trading is well studied in traditional financial markets. However, it has received less attention in centralized cryptocurrency exchanges. The Commodity Futures Trading Commission (CFTC) attributed the $2010$ flash crash, one of the most turbulent periods in the history of financial markets that saw the Dow Jones Industrial Average lose $9\%$ of its value within minutes, to automated order "spoofing" algorithms. In this paper, we build a set of methodologies to characterize and empirically measure different algorithmic trading strategies in Binance, a large centralized cryptocurrency exchange, using a complete data set of historical trades. We find that a sub-strategy of triangular arbitrage is widespread, where bots convert between two coins through an intermediary coin, and obtain a favorable exchange rate compared to the direct one. We measure the profitability of this strategy, characterize its risks, and outline two strategies that algorithmic trading bots use to mitigate their losses. We find that this strategy yields an exchange ratio that is $0.144\%$, or $14.4$ basis points (bps) better than the direct exchange ratio. $2.71\%$ of all trades on Binance are attributable to this strategy.
Otabek Sattarov, Azamjon Muminov, Cheol Won Lee, Hyun Kyu Kang · 8 authors
The net profit of investors can rapidly increase if they correctly decide to take one of these three actions: buying, selling, or holding the stocks. The right action is related to massive stock market measurements. Therefore, defining the right action requires specific knowledge from investors. The economy scientists, following their research, have suggested several strategies and indicating factors that serve to find the best option for trading in a stock market. However, several investors’ capital decreased when they tried to trade the basis of the recommendation of these strategies. That means the stock market needs more satisfactory research, which can give more guarantee of success for investors. To address this challenge, we tried to apply one of the machine learning algorithms, which is called deep reinforcement learning (DRL) on the stock market. As a result, we developed an application that observes historical price movements and takes action on real-time prices. We tested our proposal algorithm with three—Bitcoin (BTC), Litecoin (LTC), and Ethereum (ETH)—crypto coins’ historical data. The experiment on Bitcoin via DRL application shows that the investor got 14.4% net profits within one month. Similarly, tests on Litecoin and Ethereum also finished with 74% and 41% profit, respectively.
Cryptocurrencies, such as Bitcoin, are becoming increasingly popular, having\nbeen widely used as an exchange medium in areas such as financial transaction\nand asset transfer verification. However, there has been a lack of solutions\nthat can support real-time price prediction to cope with high currency\nvolatility, handle massive heterogeneous data volumes, including social media\nsentiments, while supporting fault tolerance and persistence in real time, and\nprovide real-time adaptation of learning algorithms to cope with new price and\nsentiment data. In this paper we introduce KryptoOracle, a novel real-time and\nadaptive cryptocurrency price prediction platform based on Twitter sentiments.\nThe integrative and modular platform is based on (i) a Spark-based architecture\nwhich handles the large volume of incoming data in a persistent and fault\ntolerant way; (ii) an approach that supports sentiment analysis which can\nrespond to large amounts of natural language processing queries in real time;\nand (iii) a predictive method grounded on online learning in which a model\nadapts its weights to cope with new prices and sentiments. Besides providing an\narchitectural design, the paper also describes the KryptoOracle platform\nimplementation and experimental evaluation. Overall, the proposed platform can\nhelp accelerate decision-making, uncover new opportunities and provide more\ntimely insights based on the available and ever-larger financial data volume\nand variety.\n
Yunan Ye, Hengzhi Pei, Boxin Wang, Pin‐Yu Chen · 7 authors
Portfolio management (PM) is a fundamental financial planning task that aims to achieve investment goals such as maximal profits or minimal risks. Its decision process involves continuous derivation of valuable information from various data sources and sequential decision optimization, which is a prospective research direction for reinforcement learning (RL). In this paper, we propose SARL, a novel State-Augmented RL framework for PM. Our framework aims to address two unique challenges in financial PM: (1) data heterogeneity -- the collected information for each asset is usually diverse, noisy and imbalanced (e.g., news articles); and (2) environment uncertainty -- the financial market is versatile and non-stationary. To incorporate heterogeneous data and enhance robustness against environment uncertainty, our SARL augments the asset information with their price movement prediction as additional states, where the prediction can be solely based on financial data (e.g., asset prices) or derived from alternative sources such as news. Experiments on two real-world datasets, (i) Bitcoin market and (ii) HighTech stock market with 7-year Reuters news articles, validate the effectiveness of SARL over existing PM approaches, both in terms of accumulated profits and risk-adjusted profits. Moreover, extensive simulations are conducted to demonstrate the importance of our proposed state augmentation, providing new insights and boosting performance significantly over standard RL-based PM method and other baselines.
The cryptocurrency market is amongst the fastest-growing of all the financial markets in the world. Unlike traditional markets, such as equities, foreign exchange and commodities, cryptocurrency market is considered to have larger volatility and illiquidity. This paper is inspired by the recent success of using deep learning for stock market prediction. In this work, we analyze and present the characteristics of the cryptocurrency market in a high-frequency setting. In particular, we applied a deep learning approach to predict the direction of the mid-price changes on the upcoming tick. We monitored live tick-level data from $8$ cryptocurrency pairs and applied both statistical and machine learning techniques to provide a live prediction. We reveal that promising results are possible for cryptocurrencies, and in particular, we achieve a consistent $78\%$ accuracy on the prediction of the mid-price movement on live exchange rate of Bitcoins vs US dollars.
In recent years, Bitcoin is rising and become an attractive investment for traders. Unlike stocks or foreign exchange, Bitcoin price is fluctuated, mainly because of its 24-hours a day trading time without close time. To minimize the risk involved and maximize capital gain, traders and investors need a way to predict the Bitcoin price trend accurately. However, many previous works on cryptocurrency price prediction forecast short-term Bitcoin price, have low accuracy and have not been cross-validatedThis paper describes the baseline neural network models to predict the short-term and the long-term Bitcoin price change. Our baseline models are the Multilayer Perceptron (MLP) and the Recurrent Neural Networks (RNN) models. Data used are Bitcoin's blockchain from August 2010 until October 2017 with 2-days period and the total amount of 1300 data. The models generated are predicting both for short-term and long-term price change, from 2-days until 60-days.The result shows that long-term prediction has a better result than short-term prediction, with the best accuracy in Multilayer Perceptron when predicting the next 60-days price change and Recurrent Neural Networks when predicting the next 56-days price change. Multilayer Perceptron outperforms Recurrent Neural Networks with accuracy of 81.3 percent, precision 81 percent, and recall 94.7 percent.
High frequency Bitcoin price series are often non-linear and non-stationary and hence forecasting the price of Bitcoin directly or by transformation using statistical models is subject to large errors. This paper presents an ensemble model using variational mode decomposition (VMD) and Generalized additive model (GAM) to forecast intraday Bitcoin price. To evaluate the performance of the constructed model, it is compared with an ensemble of empirical mode decomposition (EMD) and GAM. The results showed that VMD-GAM model performed better than the EMD-GAM ensemble model in terms of three evaluation metrics (root mean square error, mean absolute percentage error, and bias) used.
Cryptocurrency merchandising is growing as an attractive area of investment. Bitcoin is much preferred over other cryptocurrencies in the world and hence is becoming more popular. But, the bitcoin price is extremely volatile. So, the forecasting of its price is highly desirable. As nature inspired-machine learning is being used extensively for time series analysis and prediction, it can be explored for bitcoin prediction as well. Also, as bitcoin is gradually increasing as a promising virtual asset, its volatility needs to be measured. This paper unveils the consequence of using ChebyShev Ploynomial Neural Networks (CHPNN) for Bitcoin pricing process. The evolutionary algorithms: Particle Swarm Optimization (PSO) and Differential Evolution (DE) are utilized for training the model. This study analyses the performance of the model through three different error measures: Root Mean Square Error (RMSE), RRSE (Relative Root Square Error) and SSE (Sum of Squares Error). It shows that DE-CHPNN predicts better day-ahead price of bitcoin.
We report on the use of sentiment analysis on news and social media to analyze and predict the price of Bitcoin. Bitcoin is the leading cryptocurrency and has the highest market capitalization among digital currencies. Predicting Bitcoin values may help understand and predict potential market movement and future growth of the technology. Unlike (mostly) repeating phenomena like weather, cryptocurrency values do not follow a repeating pattern and mere past value of Bitcoin does not reveal any secret of future Bitcoin value. Humans follow general sentiments and technical analysis to invest in the market. Hence considering people's sentiment can give a good degree of prediction. We focus on using social sentiment as a feature to predict future Bitcoin value, and in particular, consider Google News and Reddit posts. We find that social sentiment gives a good estimate of how future Bitcoin values may move. We achieve the lowest test RMSE of 434.87 using an LSTM that takes as inputs the historical price of various cryptocurrencies, the sentiment of news articles and the sentiment of Reddit posts.
Over the past decade, the blockchain technology and its Bitcoin cryptocurrency have received considerable attention. Bitcoin has experienced significant price swings in daily and long-term valuations. In this paper, we propose a partial differential equation (PDE) model on the bitcoin transaction network for predicting bitcoin price. Through analysis of bitcoin subgraphs or chainlets, the PDE model captures the influence of transaction patterns on bitcoin price over time and combines the effect of all chainlet clusters. In addition, Google Trends Index is incorporated to the PDE model to reflect the effect of bitcoin market sentiment. The experiment shows that the average accuracy of daily bitcoin price prediction is 0.82 for 362 consecutive days in 2017. The results demonstrate the PDE model is capable of predicting bitcoin price. The paper is the first attempt to apply a PDE model to the bitcoin transaction network for predicting bitcoin price.
This paper studies how to forecast daily closing price series of Bitcoin,\nusing data on prices and volumes of prior days. Bitcoin price behaviour is\nstill largely unexplored, presenting new opportunities. We compared our results\nwith two modern works on Bitcoin prices forecasting and with a well-known\nrecent paper that uses Intel, National Bank shares and Microsoft daily NASDAQ\nclosing prices spanning a 3-year interval. We followed different approaches in\nparallel, implementing both statistical techniques and machine learning\nalgorithms. The SLR model for univariate series forecast uses only closing\nprices, whereas the MLR model for multivariate series uses both price and\nvolume data. We applied the ADF -Test to these series, which resulted to be\nindistinguishable from a random walk. We also used two artificial neural\nnetworks: MLP and LSTM. We then partitioned the dataset into shorter sequences,\nrepresenting different price regimes, obtaining best result using more than one\nprevious price, thus confirming our regime hypothesis. All the models were\nevaluated in terms of MAPE and relativeRMSE. They performed well, and were\noverall better than those obtained in the benchmarks. Based on the results, it\nwas possible to demonstrate the efficacy of the proposed methodology and its\ncontribution to the state-of-the-art.\n
This paper investigates the relations between multiple measures of investor sentiment and the returns, volatility, trading volume, and liquidity. Using both data outside and inside market, we find that the Bullishness from socio-finance model are significant related to future realized volatility and trading volume, similar to Tweet, which is thought to capture information of well-informed investors in Bitcoin market
With the volume of activities associated with trading, it has become a very tedious task. The advent of the algorithmic trading has brought with it some positive change such as reduced latency and increase in liquidity in the Financial Market. The Algorithmic Trading also came with some high demands for the technological know-how and the resources to run it. This has put the retail trader in a seemingly disadvantaged position as these algo-programs are carefully guided secrets by those that have access to it. Crypto-currency can no longer be ignored as the concept is forming the bedrock for future transactions. Although highly publicized, the concept of these smart contracts is not really known. Looking into the future where the cryptocurrencies dominates over the traditional currencies, it has become imperative to give the individual trader/ retailer an additional tool to demystify the “black-Box” of the trading crypto-pairs with algorithmic trading strategy. The techniques employed are: Long Short-Term Memory (LSTM), Auto-regressive integrated moving average (ARIMA), Moving Average (MA), Cumulative Moving Average (CMA), and Artificial Neural Networks (ANN). The models performance will be measured via correlation, Mean Percentage Error (MPE), Percentage Error (MAPE), Mean Square Error (RMSE) standard deviation and Sharpe ratio (for the trading models).