2 About Bitcoin
2.1 Literature review
2.2 The functioning of the Bitcoin market
3 The Roll model and suggestions for improvement
3.1 The Roll estimate for the effective bid-ask spread
3.2 Estimation bias induced by positively autocorrelated expected returns
5.1 Discussion of Roll’s assumptions
5.2 Comparative analysis with other asset classes
Bitcoin is now in the centre of attention as growing transaction numbers show the rising popularity of the cryptocurrency. Media as well as bankers and financial institutions discuss Bitcoin’s potential for changing the traditional landscape of financial markets. Recently, in October 2016, the European Central Bank demanded stricter regulatory rules for cryptocurrencies such as Bitcoin because a growing adoption of Bitcoin as a currency would dilute the Central Bank’s control over the money supply in the Euro area (Сапера (2016)). This Bachelor Thesis will deal with the question of whether Bitcoin shows the tipical characteristics of a currency or rather an investment asset such as stocks or commodities. The answer to this question is highly important for assessing the disruptive potential of Bitcoin and its influence on major currencies.
Bitcoin since its inception in 2009 was designed as an alternative currency with different characteristics meant to innovate payment systems and financial markets. The Triennial Central Bank Survey conducted by the Bank for International Settlements showed that the average trading volume on the foreign exchange market was $5.1 trillion per day in April 2016. This makes the foreign exchange market the largest and most liquid market in the world. The starting point of this Bachelor Thesis is that by comparative analysis of liquidity in the Bitcoin market with the foreign exchange market and the stock market it is possible to develop a hypothesis for the classification of Bitcoin as a currency or an asset. That is because the degree of liquidity directly determines the efficiency of capital allocation and the investors’ access to trading opportunities. Liquidity and transaction costs are considered to "represent two sides of a coin" (Gomber and Schweickert, 2002, p. 3). The higher the transaction costs, the lower is liquidity in a market. That is why the focus of this analysis is laid upon the bid-ask spread in the Bitcoin market as it is considered a measure for transaction costs and thus, also liquidity. The bid-ask spread is the difference between the highest price sought and the lowest price offered for a security. It resembles transaction costs for the trader that arise due to inventory costs, order processing costs and information asymmetries (Zhang and Hodges, 2012, p. 1). Moreover, it can function as an indicator for liquidity in the Bitcoin market and can therefore be used to do a comparative analysis to the highly liquid foreign exchange market dominated by traditional currencies. The aim of this Bachelor Thesis is to investigate what drives the bid-ask spread. It is shown that Bitcoin as an asset is closer related to stocks than to other currencies such as the Euro or the US-Dollar.
The remainder of this Bachelor Thesis is structured as follows. Firstly, Bitcoin and its characteristics as a cryptocurrency are presented, literature on Bitcoin is reviewed and the functioning of the Bitcoin market is explained. Secondly, the estimation model proposed by Roll (1984) is presented and one possible extension of this model is discussed. Thirdly, the two models are applied to data on Bitcoin, the USD/EUR exchange rate and two stocks respectively.
2 About Bitcoin
Bitcoin is a cryptocurrency that was introduced in 2009 by Satoshi Nakamoto. A cryptocurrency is defined as "a digital currency produced by a public network, rather than any government, that uses cryptography to make sure payments are sent and received safely" (Cambridge University Press (2017)). That means that the Bitcoin system is a peer-to-peer network within which transactions are verified when the so called nodes in the network reach consensus on the validity of a transaction. When a transaction in the Bitcoin network is found valid, it is included and recorded in the blockchain. A full node is a computer where the whole blockchain, that is the whole transaction history, is saved. The higher the number of full nodes, the more secure the Bitcoin system. That is because a peer-to-peer system becomes more stable with a growing number of participants. Moreover, Satoshi Nakamoto solved the problem of double spending which was inherent to digital coins up to that point (Nakamoto, 2009, p. 2). That means that one Bitcoin can not be spent twice by the same user.
In addition to that, the Bitcoin network does not need centralized intermediaries such as banks. When Satoshi Nakamoto created the first block of the blockchain he (or she, as the true identity of Nakamoto remains unknown) included the headline of the Financial Times of that day in the software code, "Chancellor on brinks to second bailout for banks" (Sixt, 2017, p. 29). This shows that Bitcoin was developed in order to become an alternative payment system to the conventional financial system that was troubled by the repercussions of the financial crisis in 2008. The decentralized system creates trust without financial intermediaries because it is based on cryptographic proof and is not prone to fraud. Transactions made in the Bitcoin network are irreversible and manipulation of the blockchain is nearly impossible. Bitcoin is based on an open source software that is publicly available. Thus, no property rights can be claimed on the Bitcoin source code, a feature that underlines the decentralization of Bitcoin. The blockchain is a public distributed ledger for all transactions ever made in the Bitcoin network. One can trace back all the transactions to the first transaction made by Satoshi Nakomoto in 2010. Since then, the number of transactions grew steadily until it reached the threshold of 280,000 transactions per day in November 2016 with a total Bitcoin market capitalization of about $12 billion (numbers taken from coindesk.com). The growth in transaction numbers can be seen in the following figure.
illustration not visible in this excerpt
Still, Bitcoin has a few features that slow down its adoption as a currency. On the one hand, most people in developed countries may not see the necessity or the advantages in using Bitcoin because electronic payment systems already have become quick and efficient. On the other hand, Bitcoin transforms the internet into a monetary union (Sixt, 2017, p. 77) and especially for international transactions a fast execution to zero transaction costs in the Bitcoin network is beneficial. In addition to that, the decentrality and high transparency of the Bitcoin system attract a lot of users.
2.1 Literature review
Considering the characteristics of Bitcoin mentioned above, one can see that Bitcoin is designed as an alternative currency meant to change the traditional system of financial markets. The blockchain technology makes Bitcoin a more transparent payment system since all valid transactions are published in the blockchain. Moreover, as a decentralized peer-to-peer system it is supposed to be independent from government decisions and financial institutions. Since Bitcoin is still a very young phenomenon, literature is not as elaborated as in other field of studies. Still, the classification of Bitcoin as a currency or an investment has already been in the focus of a few studies. Yermack (2014) was the first to investigate on the question whether Bitcoin is to be considered a currency. He argues that Bitcoin does not satisfy the criteria of a typical currency being a medium of exchange, a store of value anda unit of account. According to Yermack (2014) the main problem is the high volatility exhibited by the Bitcoin exchange rate because people who seek to use Bitcoin as a currency are then exposed to high risk. Furthermore, he shows that Bitcoin is not correlated to any of the major currencies (Yermack, 2014, p. 3). Therefore, the volatility risk cannot be hedged. Since Yermack’s analysis (2014) was based on data of 2013 it is worth to evaluate his findings based on current data from 2016. Daily percentage returns from the Bitcoin/USD exchange rate are not correlated to returns of the USD/EUR exchange rate in the observed time period from October 2014 until October 2016. A correlation very close to zero of -0.04 is found which confirms the results of Yermack (2014) who calculates a correlation of -0.05 from July 2010 until March 2014 [p.22]. As can be seen in the following graph, the volatility measured as the standard deviation of daily returns for the preceding 30-day window was usually higher for Bitcoin compared to the USD/EUR exchange rate in 2016. Still, it never exceeded 5% and in some periods, for example from March to May, the volatility value goes closely with the volatility value of the USD/EUR exchange rate.
Figure 2: Volatility of Bitcoin compared to USD/EUR in 2016
illustration not visible in this excerpt
Therefore, it can be said that the findings of Yermack (2014) do not entirely hold true anymore, since the Bitcoin market has matured over time. That is most probably due to the rising transaction numbers. As more participants enter the market and transaction numbers increase, also market depth increases and single high volume transactions have weaker price impact. That inturn reduces price volatility (Sixt, 2017, p. 107). The reduction of Bitcoin’s price volatility is also fostered by the growing adoption of Bitcoin as a medium of exchange through big companies such as Microsoft and Dell.
The rising number of transactions could suggest that Bitcoin is gaining momentum in becoming an alternative currency. However, researchers find that Bitcoin is rather used as a speculative investment. Baur et al. (2014) show that only 7% of all Bitcoin users use it as a currency. They also find a decreasing percentage number of users classified as ’currency users’ from 2011 to 2013. These results comply with the findings of Glaser et al. (2014) who researched on the relation between the volume traded at Bitcoin exchanges and the transaction volume within the Bitcoin network. The results indicate that there is little relation between rising transaction volumes on Bitcoin exchanges and the transaction volume in the Bitcoin network which suggests that new Bitcoin users rather keep their Bitcoins in the wallets at Bitcoin exchanges to use them for speculative purposes. Ling (2015) explains that there are two opposing views on the value of Bitcoin as a currency. On the one hand, it can be argued that Bitcoin’s value is derived from its functioning as a medium of exchange. On the other hand, it can be said that a currency must be valuable and its value is defined as the sum of discounted cash flows. Cheah and Fry (2015) follow the latter argument in their research and find that the fundamental value of Bitcoin is equal to zero because Bitcoin has no intrinsic value and is not pegged to commodities or to any of the major currencies. A currency, though, should represent value in order to function as a medium of exchange (p. 33). Since its introduction the price of Bitcoin, that is the exchange rate of Bitcoin to e.g. US-Dollar, saw periods of extreme price surges sometimes followed by a plunge of the Bitcoin price. Hencic and Gouriéroux (2015) analyze the price development in the Bitcoin market and find that it shows strong evidence for the existence of speculative bubbles. So all in all, research literature gives the impression that Bitcoin is rather a speculative investment than a currency. Therefore, it is investigated in this Bachelor Thesis whether the hypothesis that Bitcoin is a speculative investment asset holds true. The focus, though, is laid upon a different aspect than in recent literature, that is the liquidity in the Bitcoin market represented by the bid-ask spread.
2.2 The functioning of the Bitcoin market
As already mentioned the foreign exchange market is seen as the most liquid market in the world. Especially in the case of heavily traded currency pairs such as the US-Dollar and the Euro, transaction costs in the form of bid-ask spreads are on average one pip large, that is $0.0001, and almost negligible. In order for Bitcoin to become a significant alternative to conventional currencies, Bit- coin users must not be subject to substantially higher transaction costs or else it can be assumed that Bitcoin will not reach the status of an alternative currency. That is the reason why the bid-ask spread is in the centre of this Bachelor Thesis’ analysis. First, however, one needs to take a closer look on how the Bitcoin market is working, how the bid and ask prices are developed and what influences them. In addition to that, it is shown why the bid-ask spread is highly relevant for Bitcoin users, although Bitcoin seeks to be a decentralized and independent transaction network.
Drivers of the Bitcoin price
Bitcoins can be traded for conventional fiat money at several different exchanges. The exchange rate for Bitcoin can differ between the exchanges because it is influenced by exchange specific characteristics, such as the number of transactions executed and transaction costs. The exchange with the cheapest and easiest trading process attracts more customers and thus, the price is likely to rise at this specific exchange (Brandvold et al. (2015)). Therefore, indices that compute the volume weighted average of the different exchange prices are often used to determine the Bitcoin exchange rate. But what are the drivers of the Bitcoin price? Since Bitcoin has no intrinsic value it is difficult to define what Bitcoin is worth. One special characteristic of Bitcoin is that its supply is capped at 21 million Bitcoins. The difficulty of the mining process, that is the Bitcoin creation process, will increase until the number of new Bitcoins created asymptotically reaches 21 million. Due to the predetermined supply of Bitcoins, the price of Bitcoin is solely demand-driven. Therefore, the Bitcoin price is dependent upon its adoption and its use as a medium of exchange. Other positive price drivers are crises that regard the traditional monetary system, such as the cypriotic banking crisis in 2013, because when conventional money is short people use Bitcoin as an alternative medium of exchange (Sixt, 2017, p. 21). Negative price drivers are, for example, the implementation of regulatory rules by governments or frauds and security risks on Bitcoin exchanges.
The relevance of the bid-ask spread
In order to pay with Bitcoins and trade on Bitcoin exchanges, one needs to create a Bitcoin address to which Bitcoins can be send and then stored in the digital wallet that belongs to the Bitcoin address. There are two possible ways to receive Bitcoins in the first place. Firstly, Bitcoins can be mined, that is they can be created, if the computer is able to solve the cryptographic problem. A lot of computing power, IT knowledge and capital investment is needed though for mining Bitcoins. Therefore, it is assumed that the average user of Bitcoin who intends to use it as a currency, will not begin with mining Bitcoins. Instead, he or she will buy Bitcoins at one of the many Bitcoin exchanges. There, the Bitcoin user will have to bear transaction costs other than brokerage fees that appear as the bid-ask spread, the difference between the highest price sought and the lowest price offered. These transaction costs are directly linked to market liquidity and are by definition paid by the liquidity demander, whereas liquidity suppliers earn the bid-ask spread. Thus, it can be concluded that the higher the bid-ask spread, the higher the transaction costs for every trade executed and the lower the liquidity in the market. Yermack (2014) mentions that Bitcoin exchanges pose a big obstacle for Bitcoin’s wide adoption. He claims that Bitcoin exchanges are not liquid and show significant bid-ask spreads [p. 11]. Although the transfer of Bitcoins within the Bitcoin network happens at virtually zero costs, traders need to buy Bitcoins at exchanges due to the difficulty of mining Bitcoins. At Bitcoin exchanges, though, transaction costs are present and the bid-ask spread is meant to compensate for inventory holding costs. Therefore, the necessity to encounter transaction costs at Bitcoin exchanges that are relatively higher than for other currencies might hinder people from buying Bitcoin.
As mentioned above, the transaction volume per day in the Bitcoin market is constantly rising meaning that liquidity is provided. However, the amount is only a small fraction of the trading volume on foreign exchange markets. A comparison of the relative spread values of Bitcoin, the USD/EUR exchange rate, the Allianz stock and the Bayer stock shows that the Bitcoin market exhibits high transaction costs when compared to the foreign exchange market of US-Dollar to Euro. In the following table it can be seen that the average bid-ask spread of Bitcoin is about seven times as big as the average bid-ask spread of the USD/EUR exchange rate. The difference to the two stocks of Bayer and Allianz is not as big, though.
Table 1: Relative spread comparison in bps
illustration not visible in this excerpt
However, descriptive statistics do not yield sufficient information to conclude that Bitcoin shows
- ISBN (eBook)
- ISBN (Book)
- File size
- 1.2 MB
- Catalog Number
- Institution / College
- University of Tubingen – Fachbereich Wirtschaftswissenschaft