We study recurrent patterns in volatility and volume for major cryptocurrencies, Bitcoin and Ether, using data from two centralized exchanges (Coinbase Pro and Binance) and a decentralized exchange (Uniswap V2). We find systematic patterns in both volatility and volume across day-of-the-week, hour-of-the-day, and within the hour. These patterns have grown stronger over the years and can be related to algorithmic trading and funding times in futures markets. We also document that price formation mainly takes place on the centralized exchanges while price adjustments on the decentralized exchanges can be sluggish.
The recent 50% drop in the price of the flagship cryptocurrency Bitcoin reinforces the persistent anxiety among cryptocurrency investors. Can alternative assets hedge Bitcoin risk? This study investigates the ability of equities, commodities, bonds, currencies, and VIX futures to hedge Bitcoin. Our in-sample analysis shows that the USDX, Gilt, Australian dollars, wheat, cocoa, cotton, sugar, copper, and lean hog can hedge Bitcoin, and the out-of-sample analysis reveals that the DAX, Dow-Jones, Nikkei, S&P 500, Brent, and WTI futures can be effective hedging instruments. We use a wavelet-based dynamic hedging model to account for heterogeneous investors in the Bitcoin market. For a short-term horizon, soybean futures reduce the variance in the in-sample hedged portfolio, and cotton futures offer the highest out-of-sample utility. Copper futures are the best for in-sample hedging in a long-term horizon, whereas live cattle futures have the best out-of-sample performance. These results show that conventional assets can hedge wild swings in Bitcoin.
Dimitrios Koutmos, Timothy King, Constantin Zopounidis
Abstract Are cryptocurrencies useful minimum‐variance hedging instruments? This paper develops a two‐step analytical framework to explore this question across time. First, it estimates dynamic optimal weights, calibrated when investing between the aggregate market and a respective sampled cryptocurrency. This is performed separately for 11 major cryptocurrencies using the dynamic conditional correlation approach of Engle. Second, using a fractional regression approach, it uncovers linkages between optimal weights in cryptocurrencies and sources of economic uncertainty. Overall, this paper makes the following important findings. First, optimal weights in cryptocurrencies all rose rapidly during the COVID‐19 pandemic. In all, bitcoin showed to be the leading cryptocurrency in terms of hedging effectiveness during this recent time period. Second, most cryptocurrencies exhibit zero or negative betas consistently across time, thus making them natural hedging instruments for investors seeking to reduce their portfolio's comovement with the market. Finally, cryptocurrencies serve as better hedges for economic uncertainties arising from equity and commodity markets. They are relatively less effective for uncertainties arising from risks in the banking industry and firm default risk. This paper contributes broadly to the asset pricing literature since our two‐step approach herein can tractably be extended to other asset classes or other econometric measures of systematic risk.
Lykke Øverland Bergsli, Andrea Falk Lind, Péter Molnár, Michał Polasik
Since Bitcoin price is highly volatile, forecasting its volatility is crucial for many applications, such as risk management or hedging. We study which model is the most suitable for forecasting Bitcoin volatility. We consider several GARCH and two heterogeneous autoregressive (HAR) models and compare them. Since we utilize realized variance estimated from high frequency data as a proxy for true volatility, we can draw sharper conclusions than studies which use only daily data. We find that EGARCH and APARCH perform best among the GARCH models. HAR models based on realized variance perform better than GARCH models based on daily data. Superiority of HAR models over GARCH models is strongest for short-term volatility forecasts.
Understanding the dependence and risk spillover among hedging assets is crucial for portfolio allocation and regulatory decision making. Using various copula and conditional Value-at-Risk (CoVaR) measures, this paper quantifies the dependence and risk spillover effects between three traditional and emerging hedging assets: Bitcoin, gold, and USD. Furthermore, we investigate these effects at various short- and long-term horizons using a variational model decomposition (VMD) method. The empirical results show that there is strong negative dependence between gold and USD, but Bitcoin and gold are weakly and positively connected. Secondly, risk spillovers exist only between Bitcoin and gold and between gold and USD. The risk spillover effect between Bitcoin and gold are not stable, that is, if Bitcoin or gold faces the downward or upward risk, both the downward and upward risk of another asset have the chance to increase. The negative risk spillover between gold and USD is stable, especially in long-term horizons. Finally, the risk spillover between Bitcoin and gold as well as between gold and USD are asymmetric at downward and upward market environment.
Fahad Mostafa, Pritam Saha, Mohammad Rafiqul Islam, Nguyet Nguyen
Cryptocurrencies are currently traded worldwide, with hundreds of different currencies in existence and even more on the way. This study implements some statistical and machine learning approaches for cryptocurrency investments. First, we implement GJR-GARCH over the GARCH model to estimate the volatility of ten popular cryptocurrencies based on market capitalization: Bitcoin, Bitcoin Cash, Bitcoin SV, Chainlink, EOS, Ethereum, Litecoin, TETHER, Tezos, and XRP. Then, we use Monte Carlo simulations to generate the conditional variance of the cryptocurrencies using the GJR-GARCH model, and calculate the value at risk (VaR) of the simulations. We also estimate the tail-risk using VaR backtesting. Finally, we use an artificial neural network (ANN) for predicting the prices of the ten cryptocurrencies. The graphical analysis and mean square errors (MSEs) from the ANN models confirmed that the predicted prices are close to the market prices. For some cryptocurrencies, the ANN models perform better than traditional ARIMA models.
Daniel Traian Pele, Niels Wesselhöfft, Wolfgang Karl Härdle, Michalis Kolossiatis · 5 authors
This research provides insights for the separation of cryptocurrencies from other assets. Using dimensionality reduction techniques, we show that most of the variation among cryptocurrencies, stocks, exchange rates, commodities, bonds, and real estate indexes can be explained by the tail, memory and moment factors of their log-returns. By applying various classification methods, cryptocurrencies are categorized as a separate asset class, mainly due to the tail factor. The main result is the complete separation of cryptocurrencies from the other asset types, using the Maximum Variance Components Split method. Additionally, we show that cryptocurrencies tend to exhibit similar characteristics over time and become more distinguished from other asset classes (synchronic evolution).
Abstract We study the problem of the intraday short-term volume forecasting in cryptocurrency multi-markets. The predictions are built by using transaction and order book data from different markets where the exchange takes place. Methodologically, we propose a temporal mixture ensemble, capable of adaptively exploiting, for the forecasting, different sources of data and providing a volume point estimate, as well as its uncertainty. We provide evidence of the clear outperformance of our model with respect to econometric models. Moreover our model performs slightly better than Gradient Boosting Machine while having a much clearer interpretability of the results. Finally, we show that the above results are robust also when restricting the prediction analysis to each volume quartile.
One of the notable features of bitcoin is its extreme volatility. The modeling and forecasting of bitcoin volatility are crucial for bitcoin investors’ decision-making analysis and risk management. However, most previous studies of bitcoin volatility were founded on econometric models. Research on bitcoin volatility forecasting using machine learning algorithms is still sparse. In this study, both conventional econometric models and a machine learning model are used to forecast the bitcoin’s return volatility and Value at Risk. The objective of this study is to compare their out-of-sample performance in forecasting accuracy and risk management efficiency. The results demonstrate that the RNN outperforms GARCH and EWMA in average forecasting performance. However, it is less efficient in capturing the bitcoin market’s extreme events. Moreover, the RNN shows poor performance in Value at Risk forecasting, indicating that it could not work well as the econometric models in explaining extreme volatility. This study proposes an alternative method of bitcoin volatility analysis and provides more motivation for economic researchers to apply machine learning methods to the less volatile financial market conditions. Meanwhile, it also shows that the machine learning approaches are not always more advanced than econometric models, contrary to common belief.
Purpose This paper aims to demonstrate a dynamic cointegration-based pairs trading strategy, including an optimal look-back window framework in the cryptocurrency market and evaluate its return and risk by applying three different scenarios. Design/methodology/approach This study uses the Engle-Granger methodology, the Kapetanios-Snell-Shin test and the Johansen test as cointegration tests in different scenarios. This study calibrates the mean-reversion speed of the Ornstein-Uhlenbeck process to obtain the half-life used for the asset selection phase and look-back window estimation. Findings By considering the main limitations in the market microstructure, the strategy of this paper exceeds the naive buy-and-hold approach in the Bitmex exchange. Another significant finding is that this study implements a numerous collection of cryptocurrency coins to formulate the model’s spread, which improves the risk-adjusted profitability of the pairs trading strategy. Besides, the strategy’s maximum drawdown level is reasonably low, which makes it useful to be deployed. The results also indicate that a class of coins has better potential arbitrage opportunities than others. Originality/value This research has some noticeable advantages, making it stand out from similar studies in the cryptocurrency market. First is the accuracy of data in which minute-binned data create the signals in the formation period. Besides, to backtest the strategy during the trading period, this study simulates the trading signals using best bid/ask quotes and market trades. This study exclusively takes the order execution into account when the asset size is already available at its quoted price (with one or more period gaps after signal generation). This action makes the backtesting much more realistic.
This study examines the volatility of nine leading cryptocurrencies by market capitalization—Bitcoin, XRP, Ethereum, Bitcoin Cash, Stellar, Litecoin, TRON, Cardano, and IOTA-by using a Bayesian Stochastic Volatility (SV) model and several GARCH models. We find that when we deal with extremely volatile financial data, such as cryptocurrencies, the SV model performs better than the GARCH family models. Moreover, the forecasting errors of the SV model, compared with the GARCH models, tend to be more accurate as forecast time horizons are longer. This deepens our insight into volatility forecast models in the complex market of cryptocurrencies.
In recent years, the attention of investors, practitioners and academics has grown in cryptocurrency. Initially, the cryptocurrency was designed as a viable digital currency implementation, and subsequently, numerous derivatives were produced in a range of sectors, including nonmonetary activities, financial transactions, and even capital management. The high volatility of exchange rates is one of the main features of cryptocurrencies. The article presents an interesting way to estimate the probability of cryptocurrency volatility clusters. In this regard, the paper explores exponential hybrid methodologies GARCH (or EGARCH) and through its portrayal as a financial asset, ANN models will provide analytical insight into bitcoin. Meanwhile, more scalable modelling is needed to fit financial variable characteristics such as ANN models because of the dynamic, nonlinear association structure between financial variables. For financial forecasting, BP is contained in the most popular methods of neural network training. The backpropagation method is employed to train the two models to determine which one performs the best in terms of predicting. This architecture consists of one hidden layer and one input layer with N neurons. Recent theoretical work on crypto-asset return behavior and risk management is supported by this research. In comparison with other traditional asset classes, these results give appropriate data on the behavior, allowing them to adopt the suitable investment decision. The study conclusions are based on a comparison between the dynamic features of cryptocurrencies and FOREX Currency’s traditional mass financial asset. Thus, the result illustrates how well the probability clusters show the impact on cryptocurrency and currencies. This research covers the sample period between August 2017 and August 2020, as cryptocurrency became popular around that period. The following methodology was implemented and simulated using Eviews and SPSS software. The performance evaluation of the cryptocurrencies is compared with FOREX currencies for better comparative study respectively.
A model is proposed for Bitcoin prices that takes into account market attention. Market attention, modeled by a mean-reverting Cox-Ingersoll-Ross processes, affects the volatility of Bitcoin returns, with some delay. The model is affine and tractable, with closed formulae for the conditional characteristic functions with respect to both the conventional and a delayed filtration. This leads to semi-closed formulae for European call and put prices. A maximum likelihood estimation procedure is provided, as well as a method for changing to a risk-neutral measure. The model compares very well against classical and attention-based models when tested on real data.
Saiful Izzuan Hussain, Nurulkamal Masseran, Nadiah Ruza, Muhammad Aslam Mohd Safari
Abstract Extreme value theory(EVT) has been used to study the frequency and probability related to extreme situations in finance. This approach focuses on the extreme values and able to provide a better estimation for risk models. In this study, Generalized Pareto Distribution (GPD) is employed to model daily extreme returns in the Bitcoin market from 2017 to 2019. These periods have witnessed three phases of extreme volatility for the cryptocurrency market. The returns level for the Bitcoin range between 17.011 and 18.746. The results demonstrate heavy tail and finite tail distribution characteristics for the tails. The findings provide a better understanding of the tails’ behaviour in the cryptocurrency market and help investors to make a financial decision.
Recently, the world of cryptocurrencies has experienced an undoubted increase in interest. Since the first cryptocurrency appeared in 2009 in the aftermath of the Great Recession, the popularity of digital currencies has, year by year, risen continuously. As of February 2021, there are more than 8525 cryptocurrencies with a market value of approximately USD 1676 billion. These particular assets can be used to diversify the portfolio as well as for speculative actions. For this reason, investigating the daily volatility and co-volatility of cryptocurrencies is crucial for investors and portfolio managers. In this work, the interdependencies among a panel of the most traded digital currencies are explored and evaluated from statistical and economic points of view. Taking advantage of the monthly Google queries (which appear to be the factors driving the price dynamics) on cryptocurrencies, we adopted a mixed-frequency approach within the Dynamic Conditional Correlation (DCC) model. In particular, we introduced the Double Asymmetric GARCH–MIDAS model in the DCC framework.
This study is an integrated survey of GARCH methodologies applications on 67 empirical papers that focus on cryptocurrencies. More sophisticated GARCH models are found to better explain the fluctuations in the volatility of cryptocurrencies. The main characteristics and the optimal approaches for modeling returns and volatility of cryptocurrencies are under scrutiny. Moreover, emphasis is placed on interconnectedness and hedging and/or diversifying abilities, measurement of profit-making and risk, efficiency and herding behavior. This leads to fruitful results and sheds light on a broad spectrum of aspects. In-depth analysis is provided of the speculative character of digital currencies and the possibility of improvement of the risk–return trade-off in investors’ portfolios. Overall, it is found that the inclusion of Bitcoin in portfolios with conventional assets could significantly improve the risk–return trade-off of investors’ decisions. Results on whether Bitcoin resembles gold are split. The same is true about whether Bitcoins volatility presents larger reactions to positive or negative shocks. Cryptocurrency markets are found not to be efficient. This study provides a roadmap for researchers and investors as well as authorities.
In this paper, the Heston–Nandi futures option pricing model is applied to Bitcoin futures options. The model prices are compared to market prices to give an indication of the pricing performance. In addition, a multivariate Bitcoin futures option pricing methodology based on a multivatiate GARCH model is developed. The empirical results show that a symmetric model is a better fit when applied to Bitcoin futures returns, and also produces more accurate option prices compared to market prices for two out of three expiry dates considered.
Christoph J. Börner, Ingo Hoffmann, Jonas Krettek, Tim Schmitz
Abstract Cryptocurrencies (CCs) have become increasingly interesting for institutional investors’ strategic asset allocation and will therefore be a fixed component of professional portfolios in the future. However, this asset class differs from established assets primarily in that it has a higher standard deviation and tail risk. The question then arises whether CCs with similar statistical key figures exist. On this basis, a core market incorporating CCs with comparable properties enables the implementation of a tracking error approach. A prerequisite for this is the segmentation of the CC market into a core and a satellite, with the latter comprising the accumulation of the residual CCs remaining in the complement. Using a concrete example, we segment the CC market into these components based on modern methods from image/pattern recognition.
Christoph J. Börner, Ingo Hoffmann, Lars M. Kürzinger, Tim Schmitz
This paper evaluates and assesses the risk associated with capital allocation in cryptocurrencies (CCs). In this regard, we take a basket of 27 CCs and the CC index EWCI$^-$ into account. After considering a series of statistical tests we find the stable distribution (SDI) to be the most appropriate to model the body of CCs returns. However, as we find the SDI to possess less favorable properties in the tail area for high quantiles, the generalized Pareto distribution is adapted for a more precise risk assessment. We use a combination of both distributions to calculate the Value at Risk and the Conditional Value at Risk, indicating two subgroups of CCs with differing risk characteristics.
Purpose This study examines the inter-linkages between Bitcoin prices and CEE stock markets (Hungary, the Czech Republic, Poland, Romania and Croatia). Design/methodology/approach The dynamic contemporaneous nexus has been analyzed using both the multivariate DECO-GARCH model proposed by Engle and Kelly (2012) and quantile on quantile (QQ) methodology proposed by Sim and Zhou (2015). Our study is implemented using the daily data spanning from 6 September 2012 to 12 August 2019. Findings First, the findings show that the average return equicorrelation across Bitcoin prices and CEE stock indices are positive, even though it is found to be time-varying over the research period shown. Second, the Bitcoin-CEE stock market association has positive signs for most pairs of quantiles of both variables and represents a rather similar pattern for the cases of Poland, the Czech Republic and Croatia. However, a weaker and primarily negative connectedness is found for Hungary and Romania, respectively. Furthermore, the interconnectedness between the co-movements in the Bitcoin market and stock returns changes significantly across quantiles of both variables within each nation, indicating that the Bitcoin-stock market relationship is dependent on both the cycle of the stock market and the nature of Bitcoin price shocks. Practical implications The evidence documented in this study has significant implications for divergent economic agents, including global investors, risk managers and policymakers, who would benefit from a comprehensive knowledge of the Bitcoin-stock market relationship to build efficient risk-hedging models and to conduct appropriate policy reactions to information spillover effects in different time horizons. Originality/value This paper is the first study employing both the multivariate DECO-GARCH model and QQ methodology to shed light on the nexus between Bitcoin prices and the stock markets in CEE countries. The DECO model uses more information to compute dynamic correlations between each pair of returns than standard dynamic conditional correlation (DCC) models, declining the estimation noise of the correlations. Besides, QQ approach allows us to capture some nuanced features of the Bitcoin-stock market relationship and explore the interdependence in its entirely. Therefore, the main contribution of this article to the related literature in this field is significant. 研究目的 本研究旨在探討比特幣的價格與中東歐股市(匈牙利、捷克共和國、波蘭、羅馬尼亞和克羅地亞) 之相互聯繫. 研究設計/方法/理念 研究使用恩格爾與凱利(2012)(Engle and Kelly (2012)) 提出的多變量DECO-GARCH模型及Sim 與Zhou(2015)(Sim and Zhou ( 2015)) 研製的分位數-分位數方法來分析動態同期的聯繫。我們的研究使用由2012年9月6日至2019年8月12日期間取得的每日數據來進行. 研究結果 首先、研究結果顯示、跨比特幣價格與中東歐股價指數的平均回報當量關聯是正相關的，即使在研究期間被發現是隨時間而變化的。第二、比特幣與中東歐股市之聯繫在大多數兩變數分位數對而言出現正相關跡象，而且，這聯繫在波蘭、捷克共和國及克羅地亞而言表現一個頗相似的模式。唯就匈牙利而言、這聯繫則較弱、而羅馬尼亞則主要是負聯繫。研究結果亦顯示: 比特幣市場內的聯動與股票回報間之內在關聯會在每個國家內跨兩個變數的分位數而顯著地改變，這顯示比特幣-股市關係是取決於股市的週期和比特幣價格衝擊的本質. 實際的意義 本研究所記載的證據、對不同的經濟行為者而言極具意義 (這包括國際投資者、風險管理經理和政策制定者)，因他們會受惠於對比特幣-股市關係的全面認識，他們可建立有效的風險對沖模型、及在不同時間範圍對資訊溢出效應進行適當的政策反應. 研究的原創性/價值 本文為首個研究使用多變量DECO-GARCH模型和分位數-分位數(QQ)方法、來解釋比特幣價格與中東歐國家之股市的關係。這DECO模型使用比標準動態條件關係模型更多資訊，來計算每對回報間之動態關係，這能減少估測雜訊，而且，QQ方法讓我們可以取得比特幣-股市關係的一些細微特徵及全面地探索其相互依賴性。因此，本文的主要貢獻是在這學術領域內有關的文獻上.
The purpose of this paper is to select the best GARCH-type model for modelling the volatility of Bitcoin, Bitcoin Cash, Litecoin, Dogecoin and Ethereum. GARCH (1,1), IGARCH(1,1), EGARCH(1,1), TGARCH(1,1) and CGARCH(1,1) are used on the cryptocurrencies closing day return. We select the model with the highest Maximum Likelihood and run an OLS regression on the conditional volatility to measure the day-of-the-week effect. The findings show that EGARCH(1,1) model best suits Bitcoin, Litecoin, Dogecoin and Ethereum data and that the GARCH(1,1) model suits best Bitcoin data. The results show a significant presence of day-of-the-week effects on the conditional volatility of some days for Bitcoin, Bitcoin Cash and Ethereum. Wednesday has a significant negative effect on Bitcoin conditional volatility. Friday, Saturday and Sunday are found to be significant and positive on Bitcoin Cash conditional volatility. Finally, Saturday is found to be significant and positive on Ethereum conditional volatility.
Purpose It is crucial to find a better portfolio optimization strategy, considering the cryptocurrencies' asymmetric volatilities. Hence, this research aimed to present dynamic optimization on minimum variance (MVP), equal risk contribution (ERC) and most diversified portfolio (MDP). Design/methodology/approach This study applied dynamic covariances from multivariate GARCH(1,1) with Student’s- t -distribution. This research also constructed static optimization from the conventional MVP, ERC and MDP as comparison. Moreover, the optimization involved transaction cost and out-of-sample analysis from the rolling windows method. The sample consisted of ten significant cryptocurrencies. Findings Dynamic optimization enhanced risk-adjusted return. Moreover, dynamic MDP and ERC could win the naïve strategy (1/N) under various estimation windows, and forecast lengths when the transaction cost ranging from 10 bps to 50 bps. The researcher also used another researcher's sample as a robustness test. Findings showed that dynamic optimization (MDP and ERC) outperformed the benchmark. Practical implications Sophisticated investors may use the dynamic ERC and MDP to optimize cryptocurrencies portfolio. Originality/value To the best of the author’s knowledge, this is the first paper that studies the dynamic optimization on MVP, ERC and MDP using DCC and ADCC-GARCH with multivariate- t- distribution and rolling windows method.