Scaling properties of extreme price fluctuations in bitcoin markets
Contents:
Some features of the site may not work correctly. DOI: Stanley and B. Podobnik Published Mathematics, Economics Physica A-statistical Mechanics and Its Applications Detection of power-law behavior and studies of scaling exponents uncover the characteristics of complexity in many real world phenomena. The complexity of financial markets has always presented challenging issues and provided interesting findings, such as the inverse cubic law in the tails of stock price fluctuation distributions.
Motivated by the rise of novel digital assets based on blockchain technology, we study the distributions of cryptocurrency price fluctuations.
We consider Bitcoin… Expand. View PDF on arXiv. Save to Library. Create Alert. Launch Research Feed. Share This Paper.
Background Citations. Methods Citations. Results Citations. Figures and Tables from this paper.
Citation Type. Has PDF. Publication Type. More Filters. Chaos and order in the bitcoin market. View 1 excerpt, cites background. Research Feed. Exponentially decayed double power-law distribution of Bitcoin trade sizes. View 1 excerpt, cites methods. Time-varying properties of asymmetric volatility and multifractality in Bitcoin. The evolution of Bitcoin exchanges in terms of trading volume. The dark grey area right scale represents the total log-volume of all analysed exchanges.
The light area shows the volume of only the exchanges that are still in business as of end The blue inset plots schematically show the evolution of traded log-volume of the single exchanges that are listed next to the right plot axis. Note that the inset plots are ordered according to the dates of entry into service of the different exchanges. Hence, from these plots, the temporal origination of exchanges can be compared.
Data Source: [ 50 ]. The emergence of new Bitcoin exchanges significantly facilitated market entrance for the numerous primarily China-based investors to the Bitcoin market. Ultimately, the wave of new investments drove prices above the USD level for the first time. Figure 8 shows that the share in total traded volume of Chinese exchanges started to really break through during the second long bubble, in the phase of most intense growth. Before that, MtGox held the majority of the market share. The corresponding lost fraction had been replaced by trading volume on the uprising Chinese exchanges.
Volume share of the top 10 Bitcoin exchanges. Shown on the left scale is the percentage share of the Bitcoin volume traded on each exchange with respect to total traded volume.
Bitcoin price fluctuations are found to exhibit heavy tails following a power-law. Scaling exponents of. are estimated for multiple exchanges and time intervals. Estimates imply a finite second moment in the distribution of Bitcoin returns. We consider Bitcoin returns over various time intervals and from multiple digital exchanges, in order to investigate the existence of universal.
Notice the surge of the two main Chinese exchanges, OKCoin and Huobi towards the end of the second long bubble, as well as the constant decrease of the market share of until then dominating Japan-based exchange MtGox. As the bubble was developing at fast pace, on 2 October , the online drug market Silk Road , which enabled customers to buy illegal substances anonymously using Bitcoin, was shut down by the FBI and its alleged founder imprisoned [ 52 ].
Associated Data
This signalled to the Bitcoin community that the legal authorities had their eye on the cryptocurrency and intended to prevent any illegal activities related to it by all means. Silk Road was not the only, but by far the most popular Darknet drug market at the time. Therefore, its seizure had wide implications and hit the news headlines heavily. The closure of Silk Road symbolically set free Bitcoin as a proper investment for more cautious investors who, until then, were deterred by its illegal usage as drug money.
The subsequent build-up of a clean image for Bitcoin was also recognized by the US senate, which as a consequence announced an official hearing about the utility and future prospects of the digital currency [ 53 ]. The event is seen as a further beneficial factor contributing to the immense price surge seen in the second long bubble [ 54 ]. Figure 9 hints at the third contributing factor for the price growth during the bubble. The logarithmic hash rate power of the miners computing transaction blocks on the Blockchain is shown, as well as the logarithmic number of registered wallets. Notice that the hash rate grew at a much faster rate compared with the number of wallets over the period starting with the first long bubble.
The highest rate of increase was reached during the second half of , coinciding with the second long bubble. Growth of Bitcoin Blockchain network user base and increase in mining power. Superimposed to the Bitcoin price in log-scale left axis , the total hash rate and the number of registered wallets are shown right axis.
A second, more subtle acceleration of the hash rate can be seen around the nucleation of the third long bubble. The number of registered wallets is a measure for the size of the network of users operating directly on the Blockchain.
Scaling properties of extreme price fluctuations in Bitcoin markets
Hence, the hash rate increasing faster than the user base signals that on average miners enhanced their individual mining power during the period. This was most likely achieved through the usage of more efficient technical mining hardware. We conclude that, in addition to Bitcoin adoption effects in China and the image improvement of Bitcoin, we can see the effect of the ramp-up behaviour of hash power as an indication of increased mining sophistication. Concerning the improvement in hardware and mining technology, we note that the first long bubble played an important role as a precursor to the second one.
As the Bitcoin price increased for the first time to a fairly high level, miners were incentivized to invest in mining hardware.
At the relatively high price of a Bitcoin compared to the computational effort required to create it, it was profitable for them to enter the mining business. The increasing number of miners itself then triggered a feedback mechanism. It can be seen as a self-fulfilling, self-reinforcing bubble in the following sense: the larger the price, the larger the incentive to invest in hardware computing power; the more mining there is, the more activity there is on the Blockchain that attracts more and more users and buyers, the more the price increases.
Anticipating Cryptocurrency Prices Using Machine Learning
But, the faster growing variable should quickly become the hashing difficulty. The loop is closed when the incentive of miners to invest in hardware rises again. There are two major triggers for the crash that started in December and the following long drawdown period.
Firstly, the Chinese government suddenly prohibited financial institutions from using Bitcoin when the price of a Bitcoin was close to the peak of the bubble in December [ 55 ]. The announcement destabilized the currency and sent it into free fall. Secondly, during February , when there was still hope for the price to recover, MtGox suspended trading figure 8 and filed for bankruptcy protection from creditors after a major account robbery of at least missing or stolen Bitcoins of customers had been reported [ 56 ].
References
The demise of MtGox was perceived as a major setback for the Bitcoin world. Account holders of the exchange lost money that was not or hardly recoverable due to the difficulty of tracing it. These adverse circumstances set the currency off into a strong 1-year-long decline. The question remains whether these events led to exaggerated responses by the investor community.
The winners of the long Bitcoin price drawdown that took place from until mid were clearly the China-based Bitcoin exchanges. Within the 3 years following , Bitcoin volume formation on Chinese exchanges contributed to roughly a fold increase in global volume figure 7. It is this rising demand for Bitcoin from Chinese markets, originating during that time, which can be seen as a major precipitating factor for the nucleation of the third long-term bubble, whose start we identify around the beginning of This leaves open the question where this increased demand originated from?
We identify the devaluation of the Chinese Yuan as a main promoting factor for the rising interest in cryptocurrencies in China and thereby the formation of the third long bubble. This devaluation was motivated by the desire to raise the competitiveness of exporting firms. From there on, a continuous weakening of the Renminbi developed until January Development of Bitcoin in parallel to the Chinese Yuan.
A change of regime occurred in the Chinese currency simultaneously to the decline of Chinese exchange-traded Bitcoin volume b , right axis at the beginning of As a reaction to the depreciation of their currency from onward, Chinese market participants tried to transfer their money to what they perceived as safer stores of value, causing an outflow of capital from China [ 58 ].