Laura Alessandretti, Abeer ElBahrawy, Luca Maria Aiello, Andrea Baronchelli
Machine learning and AI-assisted trading have attracted growing interest for the past few years. Here, we use this approach to test the hypothesis that the inefficiency of the cryptocurrency market can be exploited to generate abnormal profits. We analyse daily data for $1,681$ cryptocurrencies for the period between Nov. 2015 and Apr. 2018. We show that simple trading strategies assisted by state-of-the-art machine learning algorithms outperform standard benchmarks. Our results show that nontrivial, but ultimately simple, algorithmic mechanisms can help anticipate the short-term evolution of the cryptocurrency market.
Abstract We establish that cryptocurrency returns are driven and can be predicted by factors that are specific to cryptocurrency markets. Cryptocurrency returns are exposed to cryptocurrency network factors but not cryptocurrency production factors. We construct the network factors to capture the user adoption of cryptocurrencies and the production factors to proxy for the costs of cryptocurrency production. Moreover, there is a strong time-series momentum effect, and proxies for investor attention strongly forecast future cryptocurrency returns.
In recent years, Tether issuances (or 'grants') have increased significantly, which correlated broadly with a significant rise in Bitcoin valuation. This paper examines the impact of cryptocurrency issuances on subsequent cryptocurrency returns. It is argued that as Tether is the undisputed 'stable coin', the minting of new Tether acts similarly to monetary expansion in cryptocurrency markets, inflating the prices of Bitcoin. We construct a VAR model and show contrary to investor expectations, Tether issuances do not impact subsequent Bitcoin returns, however, they do impact traded volumes. We also document an increase in Tether trading following a subsequent decrease in Bitcoin returns. This illustrates investor preferences for lower volatility crypto-assets in periods following negative Bitcoin returns.
Abstract In December 2017, both the Chicago Board Options Exchange and the Chicago Mercantile Exchange introduced futures contracts on bitcoin. We investigate to what extent they provide useful information for the price discovery of bitcoin. We rely on the information share methodology of Hasbrouck (1995, J Finance , 50, pp. 1175–1199) and Gonzalo and Granger (1995, J Bus Econ Stat, 13, pp. 27–35) and find that the spot price leads the futures price. We attribute this result to the higher trading volume and the longer trading hours of the globally distributed bitcoin spot market, compared to the relatively restricted access to the US‐based futures markets.
ABSTRACT We offer a general equilibrium analysis of cryptocurrency pricing. The fundamental value of the cryptocurrency is its stream of net transactional benefits, which depend on its future prices. This implies that, in addition to fundamentals, equilibrium prices reflect sunspots. This in turn implies multiple equilibria and extrinsic volatility, that is, cryptocurrency prices fluctuate even when fundamentals are constant. To match our model to the data, we construct indices measuring the net transactional benefits of Bitcoin. In our calibration, part of the variations in Bitcoin returns reflects changes in net transactional benefits, but a larger share reflects extrinsic volatility.
ABSTRACT This paper investigates whether Tether, a digital currency pegged to the U.S. dollar, influenced Bitcoin and other cryptocurrency prices during the 2017 boom. Using algorithms to analyze blockchain data, we find that purchases with Tether are timed following market downturns and result in sizable increases in Bitcoin prices. The flow is attributable to one entity, clusters below round prices, induces asymmetric autocorrelations in Bitcoin, and suggests insufficient Tether reserves before month‐ends. Rather than demand from cash investors, these patterns are most consistent with the supply‐based hypothesis of unbacked digital money inflating cryptocurrency prices.
This paper examines the movement of cryptocurrencies’ return based on price. This volatility can spread to others of the same kind. Currently, the more cryptocurrencies are traded in market, the more chances are available for investors. The author wonders whether contagion risk among these cryptocurrencies happens or not in the event of crashing. We also introduce one empirical evidence of the mutual influence on these cryptocurrencies using Copulas approach. The findings show that all pairs have the structure dependence with Kendall-plots, particularly strong left tail dependence with Chi-plots. It also means the existence of contagion risk among these cryptocurrencies. The three methodologies namely Kendall-plots, Chi-plots and Copulas estimation produce consistent results. Therefore, the investors should carefully perform portfolio diversification to avoid contagious phenomenon.
This paper studies the liquidity of Bitcoin using the time series daily data over the period 1/1/2014 to 12/31/2015. Based on the available data for Bitcoin, five liquidity measures are chosen to compare the liquidities among five Bitcoin exchanges and the liquidities of different sizes of stocks. The results suggest that the liquidity of Bitcoin depends on the choice of the Bitcoin exchanges, and Bitfinex, one of the Bitcoin exchanges, provides the highest liquidity for Bitcoin trading. Moreover, the results indicate that, on average, stocks are more liquid than Bitcoin.
The long-term dependence of Bitcoin (BTC), manifesting itself through a Hurst\nexponent $H>0.5$, is exploited in order to predict future BTC/USD price. A\nMonte Carlo simulation with $10^4$ geometric fractional Brownian motion\nrealisations is performed as extensions of historical data. The accuracy of\nstatistical inferences is 10\\%. The most probable Bitcoin price at the\nbeginning of 2018 is 6358 USD.\n
Financial portfolio management is the process of constant redistribution of a\nfund into different financial products. This paper presents a\nfinancial-model-free Reinforcement Learning framework to provide a deep machine\nlearning solution to the portfolio management problem. The framework consists\nof the Ensemble of Identical Independent Evaluators (EIIE) topology, a\nPortfolio-Vector Memory (PVM), an Online Stochastic Batch Learning (OSBL)\nscheme, and a fully exploiting and explicit reward function. This framework is\nrealized in three instants in this work with a Convolutional Neural Network\n(CNN), a basic Recurrent Neural Network (RNN), and a Long Short-Term Memory\n(LSTM). They are, along with a number of recently reviewed or published\nportfolio-selection strategies, examined in three back-test experiments with a\ntrading period of 30 minutes in a cryptocurrency market. Cryptocurrencies are\nelectronic and decentralized alternatives to government-issued money, with\nBitcoin as the best-known example of a cryptocurrency. All three instances of\nthe framework monopolize the top three positions in all experiments,\noutdistancing other compared trading algorithms. Although with a high\ncommission rate of 0.25% in the backtests, the framework is able to achieve at\nleast 4-fold returns in 50 days.\n
This paper empirically examines interdependencies between BitCoin and altcoin markets in the short- and long-run. We apply time-series analytical mechanisms to daily data of 17 virtual currencies (BitCoin + 16 alternative virtual currencies) and two altcoin price indices for the period 2013–2016. Our empirical findings confirm that indeed BitCoin and altcoin markets are interdependent. The BitCoin-altcoin price relationship is significantly stronger in the short-run than in the long-run. We cannot fully confirm the hypothesis that the BitCoin price relationship is stronger with those altcoins that are more similar in their price formation mechanism to BitCoin. In the long-run, macro-financial indicators determine the altcoin price formation to a slightly greater degree than BitCoin does. The virtual currency supply is exogenous and therefore plays only a limited role in the price formation.
This article examines the pricing efficiency of Bitcoin Investment Trust. We investigate the deviation between prices and net asset values and find that there is a significant and persistent premium with an average of 44%. Such evidence points to pricing inefficiency of the currently available trust and encourages practitioners to introduce better instruments such as Exchange Traded Funds as alternatives to investors interested in having exposure to bitcoins and the digital currencies market.
Cryptocurrencies have become increasingly popular since the introduction of bitcoin in 2009. In this paper, we identify factors associated with variations in cryptocurrencies' market values. In the past, researchers argued that the "buzz" surrounding cryptocurrencies in online media explained their price variations. But this observation obfuscates the notion that cryptocurrencies, unlike fiat currencies, are technologies entailing a true innovation potential. By using, for the first time, a unique measure of innovation potential, we find that the latter is in fact the most important factor associated with increases in cryptocurrency returns. By contrast, we find that the buzz surrounding cryptocurrencies is negatively associated with returns after controlling for a variety of factors, such as supply growth and liquidity. Also interesting is our finding that a cryptocurrency's association with fraudulent activity is not negatively associated with weekly returns-a result that further qualifies the media's influence on cryptocurrencies. Finally, we find that an increase in supply is positively associated with weekly returns. Taken together, our findings show that cryptocurrencies do not behave like traditional currencies or commodities-unlike what most prior research has assumed-and depict an industry that is much more mature, and much less speculative, than has been implied by previous accounts.
Hélder Sebastião, António Portugal Duarte, Gabriel Guerreiro
This paper analyses the price discovery process in the USD/Bitcoin market since the Mt.Gox bankruptcy until the aftermath of the hack attack on Bitfinex (01-Mar-2014 until 30-Nov-2016).The Geweke feedback measures, estimated pairwise using hourly returns, show that there is a positive relationship between the total feedback and market share, measured by trading volume, that most of the information is transmitted between exchanges within an hour, at least for the main four exchanges (Bitfinex, Bitstamp, BTC-e and ItBit), while lagged feedback runs mainly from the major exchange.Other minor exchanges seem to react to price information with some delay and are thus considered as merely satellite exchanges.Bitfinex stands out as the most important exchange in transmitting information to the market: the relative importance of the lagged feedback from Bitfinex to the market is 18.29% while the lagged feedback from the market to Bitfinex accounts only for 0.60% of the total feedback.The volatility in the major exchange in each pair is the main factor explaining the feedback measures, sustaining the claim that the information-based component of volatility increases with the relative dimension of the exchange.
This thesis is a descriptive statistical analysis of cryptocurrency market and its relation within cryptocurrencies and across asset classes, using correlation functions, orthogonalized impulse response functions and OLS regressions. Consistent with Wang (2014), bitcoin does not suffer from a liquidity trap, even though bitcoin is a decentralized system. This thesis concludes that bitcoin has a lead effect on only 2 out of 8 of the top cryptocurrencies, endowing diversification benefits within cryptocurrency market. This paper provides evidence on cryptocurrency market’s and US equity market’s impulse response dynamics which are insignificant, consistent with Gangwal’s (2016) results that adding cryptocurrencies to a diversified portfolio will yield to a higher Sharpe ratio. Lastly, the study reports bitcoin momentum factor having an impact on banking and financial industries’ excess returns.
In this paper, we rigorously investigate the benefit of utilizing an active investment strategy\nbased on momentum when investing in cryptocurrencies. We also examine how including\ncryptocurrencies in a more traditional asset allocation can optimize an investment portfolio.\nFirst, we create strategies with the use of exponential moving averages and simple average\nfilters to generate a trading signal. Second, we provide evidence that the active strategies\nreceive positive return, but significantly less than the passive buy-and-hold\nalternative/benchmark. Third, we find evidence that including a portion of cryptocurrency in\na portfolio with more traditional assets will improve the risk-adjusted return, due to low\nhistorical correlation. And fourth, we look at and evaluate the extreme volatility and risk\nrelated to cryptocurrencies and the suggested cryptocurrency bubble. Our results have\nimportant implications for portfolio managers and first-time investors alike.
A Company's Brand image is an intangible asset. Though Initial Public Offerings (IPOs) try to capture Brand Value, a company's share value is a result of several factors like performance, initial capital, investor identity etc. Moreover, time needed for a company to be listed runs into several months. Therefore, immediate capitalization of Brand Value is not possible. Initial Coin Offerings on the other hand deliver a wide range of possibilities not provided by IPOs. Most important among them is Brand Tokenization and Monetization. This paper explores Brand Tokenization and Monetization through ICOs (Initial Coin Offerings). Firstly, the concept of Brands and cryptocurrencies are explained. Then the concept of ICOs is discussed. I envisage a scenario where a company tokenizes its Brand and attempts to monetize it. We then evaluate the advantages that can accrue from such a venture. The paper concludes as to how Brand Tokenization and monetization can be realized through cryptocurrencies and its impact on future businesses.