FACULTY OF ECONOMICS AND BUSINESS
Regulatory impact on the Bitcoin price
BSc Economics & Business
Author: R.L. de Jong Student Number: 10757503 Thesis Supervisor: G. Ciminelli Date of Publication: 14/07/2017
ABSTRACT
This paper investigates the effects of regulations on the Bitcoin price using an event study multiple regression analysis. For official authorities such as central banks and governments it is of interest to stabilize the Bitcoin price in order for it to be a global reserve currency. Hence study the effect their regulations have on the Bitcoins price. For the regulations we make use of statements or jurisdictions made by central banks or governments and divide them in two variables namely positive and negative regulations. We find that positive regulations do not affect the Bitcoins price but that negative regulations increase the Bitcoins price. This result can be explained by the decrease in supply which increase the price at negative events and the result shows that the Bitcoins price is affected by regulations. For the analysis we used data from September 2011 till June 2017 and the regulations are taken from governments and centrals banks worldwide.
TABLE OF CONTENTS
ABSTRACT ... 2
TABLE OF CONTENTS ... 3
SECTION 1: Introduction ... 4
SECTION 2: Theoretical Framework ... 7
SECTION 3: Literature Review ... 14
SECTION 4: Data Sample ... 17
SECTION 5: Methodology ... 23
SECTION 6: Results and discussion ... 26
SECTION 7: Conclusion ... 29
SECTION 1: Introduction
Online payment systems are evolving quickly. Virtual currencies have been around for decades with David Chaum being one of the first to describe an anonymous electronic payment system that could be useable in the real economy (Chaum, 1983). Chaum was an American cryptographer who started DigiCash Inc. in Amsterdam. His corporation was established in 1990 and filed for bankruptcy in 1998; it failed. Ten years later in 2008Nakamoto released a paper about a peer-to-peer cryptocurrency called Bitcoin ( Nakamoto, 2008). The key difference between fiat currency and cryptocurrency is that a cryptocurrency does not rely on a third party like a bank. This currency relies solely on cryptographic technology and the peer-to-peer network, which later will be discussed. The Bitcoin enables users to make online payments while staying anonymous and has lower transaction costs compared to fiat currency (Reid & Harrigan, 2013). Its anonymity makes it attractive for illegal uses and therefore Bitcoin has widely been used on online black markets. The case of the online anonymous marketplace Silk Road which was used to sell controlled substances and narcotics confirms that the Bitcoin is being used for illegal activities (Christin, 2013). A milestone occurred on the 3th of May 2017 when the price of one single Bitcoin has exceeded the price of the gold standard for the first time. The market capitalization of the Bitcoin has exceeded 40 billion USD and it is the most widely used cryptocurrency with an average daily transaction volume exceeding 300 million USD in the first quarter of 2017 (blockchain.info 2017). The number of Bitcoin wallet users has increased by nearly 200% from January 2016 till January 2017 to 11 million (blockchain.info 2017). Nevertheless, the price of the Bitcoin is highly volatile and makes it not reliable as to store value at this moment
(Yermack, 2013). The main driver of the Bitcoins success is the blockchain technology and has become interesting for financial institutions and central banks. This interest led the central banking system of the U.S., the Federal Reserve System, to develop their own centralized cryptocurrency called the FedCoin (Koning, 2016).
In March 2009, the Governor of the People’s Bank of China, Zhou Xiaochuan, underlined the presence of the Triffin dilemma and asked what kind of international reserve currency is necessary to secure global financial stability and facilitate world economic growth. Zhou stated that this international reserve currency should be anchored to a stable benchmark and issued according to a clear set of rules concerning the supply. The Bitcoin is a kind of
international or supranational currency, it can be of use as international reserves held by the monetary authorities for conducting international transactions, intervention in foreign currency market as well as maintaining confidence in the exchange rate (IMF, 2013).
One factor that disqualifies the Bitcoin is the fact that their prices have been highly unstable with volatility that is typically much higher than for national currency pairs and both prices and volatility appear to be unrelated to economic or financial factors, making them hard to hedge or forecast (Yermack, 2013). If regulations by central banks and governments are able to affect the Bitcoin price, then they can as well affect the volatility. This leads to the research question of this thesis. This thesis investigates whether regulations affect the Bitcoin price. Regulations are defined by statements or jurisdictions made by governments or central banks and that makes this research unique compared to existing literature on the topic of the Bitcoin. The relevance to research the Bitcoins price comes from the Triffin dilemma and Keynes International Clearing Union, which both proposed a supranational currency as global reserve currency.
The methodology used to answer the research question is an event study done on two types of regulations, namely positive and negative regulations, where positive regulations improve further integration of the Bitcoin and negative regulations hinder the further integration of the Bitcoin .The data used ranges from the 18th of September 2011 till the 1th of June 2017. The multiple regression model used is derived from the Barro (1979) model which was used to analyse the gold price. My model takes supply and demand of the Bitcoin plus the regulations in account to estimate the Bitcoin price formation.
The results show for the period of 7 years no significant impact of positive events but a significant positive impact of negative events on the Bitcoin price. This suggests that central banks and governments can affect the Bitcoin price with regulations to a certain extent.
This paper is structured as follows: in Section 2 the theoretical framework is set up. Firstly, the Bretton Woods system and Triffin dilemma are discussed to outline the relevance of this research. Then the possible solutions to the Triffin dilemma like the ‘bancor’ and Special Drawing Rights are explained and at last the Bitcoin network is technically reviewed. In Section 3 a literature review is provided on the research done so far on the Bitcoin concerning Bitcoin price drivers. Section 4 describes the dataset used. The methodology is explained In Section 5. explained , Section 6 presents and discusses the results. Then in Section 7
conclusions are drawn to answer the research question, and the limitations of the analysis and suggestions for further research are discussed.
SECTION 2: Theoretical Framework
In this paper I am interested in the store of value function of the Bitcoin. Cryptocurrencies as the Bitcoin are designed like a monetary system and therefore the field of monetary
economics gives a theoretical framework to analyse the functions of currencies. First, this section describes the history of monetary systems starting from the Bretton Woods system and explaining the Triffin dilemma. Next, I will describe the Special Drawing Rights (SDR’s) the International Monetary Fund (IMF) uses nowadays and the solution Keynes proposed to solve the Triffin dilemma. Then I will explain what Bitcoin is and how it operates. At last, I discuss how the Bitcoin can solve the Triffin dilemma in the way Keynes envisioned.
2.1 Bretton Woods system and the Triffin dilemma
When the Second World War ended, there was a need to rebuild national economies and shape the global economy. As a result, the Bretton Woods system was established in 1944. The main features of the system were the creation of the International Monetary Fund (IMF), the creation of the International Bank for Reconstruction and Development (IBRD) and fixed but adjustable exchange rates.
In 1945 the IMF was founded to help operate a system of fixed exchange rates, in which all currencies were pegged to the dollar, in turn pegged to gold. Experts then considered this necessary to encourage international trade (Feldstein, 1998). Since the foreign currencies were pegged to the US dollar and since the United States had a stable political situation, the US dollar was being used as the global reserve currency. As a result the United States had to keep adequate gold reserves and settle external accounts with gold bullion payments and receipts.
The Bretton Woods system had fundamental problems and the problem relevant for this paper is the Triffin dilemma. The Triffin dilemma describes the conflict of interests between short-term-domestic and long-term international monetary policies. Robert Triffin foresaw a strong economic growth after the Second World War which would increase the global demand for US dollars. In order to satisfy the demand for liquidity, the United States had to increase the amount of US dollars in circulation and as a result, the inflation or price
level in the United States rose. At the same time the higher demand for the US dollar caused the US dollar to appreciate and that made the United States’ domestic export less
competitive. That causes resulted in a trade deficit for the United States but helped the global economy. If on the other hand the US authorities had chosen not to provide more US dollars and focus on domestic monetary policy, that would have hindered international trade. The stability of the Bretton Woods system relied on the ability to exchange US dollars for gold. To keep the US dollar pegged to the gold standard, the gold stock thus had to increase as well (e.g., by gold-mining) but by the end of the 1950s, the increase in dollars had exceeded the increase in United States gold stock and the growth of gold stock was not able to finance the growth of the world real output and trade and the confidence in the US dollar lowered (Triffin, 1960).
One aspect of the Triffin plan to solve this liquidity problem advocates as highly desirable the elimination of national currencies as a means of international reserves
accumulation by central banks (Triffin, 1961). This plan was partly implemented and the new type of reserve asset, the Special Drawing Rights (SDR), was created. The SDR are discussed in the next Section.
2.2 Special Drawing Rights and the ‘Bancor’
The SDR is an international reserve asset, created by the IMF in 1969 to supplement its member countries’ official reserves. SDR are neither a currency or a claim to the IMF. SDR is a fiat obligation and potential claim on the freely usable currencies of IMF members. The value of the SDR is defined as a basket of currencies that play a major role in the global economy. For members of the IMF it is mandatory to accept SDRs when the IMF demands it. There are two relevant issues concerning the SDRs.
First is its acceptability. Ronald McKinnon (1979) argued that the SDR would be less acceptable as reserve asset than the US dollar or gold. This was due the fact SDRs could only be used for official international transactions and not as private international money, like the US dollar.
A second issue is that SDRs are baskets of national currencies. The balance of payment would still permanently be in imbalance and there is no mechanism that gives an incentive to retain a balance close to zero. The countries providing the reserve currency still be in deficit and due to this the Triffin dilemma remains unsolved. During the Second World War, Harry Dexter White of the United States Treasury Department and John Maynard Keynes of the British Treasury started on their own ideas to reform the global monetary system in de post-war period. These two lead negotiators at the Bretton Wood Conference in 1944, White for the United States and Keynes for the United Kingdom, both had a plan to develop an independent financial institution that would promote international trade and financial stability . The
Bretton Woods System is a compromise of the Keynes and White plans, although the White plans predominated (Bordo, 1993).
Both plans had three significant differences. First, White’s proposed IMF had less influence than Keynes’ International Clearing Union (ICU) and would not allocate resources freely on demand from members. Keynes stated that countries with a deficit would not have a strong position and therefore needed resources to lower their deficit (Boughton, 2002). Second, White’s IMF was a multilateral institution rather than Keynes’ ICU that would be dominated by the two ‘founder-states’ to exert their plans. Nevertheless, in Keynes’ plans other states could join if they would fulfil special conditions. At third, White’s IMF would lend national currencies where Keynes created the ‘bancor’. The ‘bancor’ is a supranational currency which the ICU would issue to regulate currency exchange. The ICU would take the role of supranational central bank and all participating central banks would agree on initial value of their own currencies in terms of bancor and the value of bancor in terms of gold (Keynes, 1943). The central banks would then keep an account with the ICU to settle balances with other countries’ central banks. Creditors and debtors both pay 1% interest on their average balance. The only way to avoid this charge was keeping the international balance close to zero. The goal of this was to encourage surplus countries to buy deficit countries’ export to make the system symmetric and promote international trade. Nowadays the IMF still comes closest to the proposed ICU.
2.3 The functioning of Bitcoin
Bitcoin is a peer-to-peer electronic cash system and allows payments to be sent directly from one party to another without going through a financial institution and this makes the system decentralized (Nakamoto, 2008). A peer is a computer connected to the network and the difference structure between a centralized and peer-to-peer network can be seen in see Figure 1.The peer-to-peer computer network enables each peer to offer their resources to the network. In this case the resources are computational power, disk storage and network bandwidth. All resources are directly available for all users of the network and there is no central coordination. Peers are consumers and suppliers in the network and the peers within the network are equally privileged. The collaborative peer network allows them then to combine resources and finish greater computations than those that can be finished individually. In this paper Bitcoin is used to describe the entire network, bitcoin is used to describe one digital coin in the Bitcoin network and a peer is called a node. At the 3th of May 2017, there were approximately 13,5 million users with a Bitcoin wallet (blockchain.info 2017).
Figure 1: Centralized Network vs P2P-Network
Source: gigitribe.com
One digital coin is defined as a chain of digital signatures, which are linked to all previous transactions made. The bitcoins are stored in so called Bitcoin wallets. Owning a BTC, or a part of it, merely means that the wallet has obtained a certain balance, just like a normal bank account. These wallets have a private key and public key for each account. The private key is secured so that only the owners address knows it and it can be used for a digital
signature to sign transactions. The public key then is used for the other users of the network to verify transactions. Transactions are irreversible which reduces the need for a central authority since the central authority would be the party that reverses the transaction.
Within the system it is a problem for users to verify whether a coin has not been spent already. Normally a central authority would solve this double-spending problem. The problem with an central authority is, that every transactions needs intermediation of a third party like a bank. This costs time and results in transaction fees and the system depends on the well-functioning of the third party. The Bitcoins solution to avoid double spending is that the earliest transaction counts; and to know which transaction was first, all transactions are made public and the majority of nodes in the system need to agree on this one history of the order based on the chronological order. Transactions are saved in a public ledger or blocks and have a timestamp to chronologically link them. This gives every user the same information and makes the system fully transparent. These linked blocks then form the blockchain which is the most important technology behind the Bitcoin.
A big part of the Bitcoin relies on bitcoin mining where bitcoins are the resources that are extracted from the digital environment through calculations. Mining involves the record-keeping of all transactions and solving to proof of work to create new bitcoins done by the nodes. Ordinary nodes store only the relevant part of the blockchain and full nodes store the whole blockchain. Running the Bitcoin software by owning a digital wallet is all it takes to become a node within the network. All nodes keep record of the blockchain and ensure it is unchangeable.
Secondly, nodes use special mining software (and hardware) to solve the proof of work for each block. A block has a proof of work that needs to be solved before the network will accept the block. A proof of work is a formula that will only accept a few solutions. Solving proof of work is a random process with very low probability and the solution can only be found by trial and error. The low probability and trial and error process, forces the miners to pool their resources and calculate blocks in pools. Each block then carries a certain amount of bitcoins which will be divided amongst the miners in the pool. The low probability makes it unpredictable which node will solve the proof of work. A second way for miners to earn BTC’s is by charging transactions fees. These fees have increased by 354.41% from January 1 2016 till January 1 2017 (bitinfocharts.com 2017).When more nodes enter the network to mine
bitcoins, the proof of work is the mechanism that ensures mining gets more difficult and it limits the rate at which blocks are generated to one every 10 minutes and after every 210,000 blocks mined, the reward of a block halves to stabilize the supply of bitcoins created. In May 2017,the reward of solving a block was worth 12.5 bitcoins and the total amount of bitcoins is limited to 21 million. This idea of a fixed money supply shows similarities with the proposal Friedman (2008) made in his book A Monetary History of the United States, 1867-1960. Friedman’s book attributed inflation to excess money supply created by the central bank. Friedman’s proposed to fixed to monetary supply based on knows financial and
macroeconomic factors. When all bitcoins are created – approximately 2140 - , the currency will work deflationary (see Figure 2) since from then on the amount of bitcoins in circulation can only decrease and therefore the price level can only decrease as well which lead to deflation.
Figure 2: Fixed Monetary Base
Source: bitcointalk.org
Currency exchanges make it possible for Bitcoin user to exchange their BTC’s for traditional currency but with a risk. From 2010 till 2013, 40 Bitcoin exchanges were
established and 18 of them have closed within this period with customer balances wiped out completely (Moore & Christin, 2013). This is a big implication for the practical use of the Bitcoin.
Other limitations involves legal issues. Countries such as China, Ecuador, Iceland, India, Russia Sweden, Thailand and Vietnam have placed a ban or restrictions on making use
of the Bitcoin (cryptocoinsnews.com 2015). This makes the Bitcoin less useful in comparison to traditional currencies. These restrictions are mostly done to protect the own traditional currency, to prevent illegal payments or avoid money-laundering
2.4 Price Stability
There are several factors that influence the decision of central banks to hold a currency as a reserve currency. One essential element is that the currency’s value must be ensured through macroeconomic and political stability (Carbaugh & Hedrick, 2009). Without confidence in the currency’s value, the ability to serve as a reserve currency is undermined. Bitcoins monetary policies are based on its system design and cannot be influenced. This creates a rules to battle time inconsistency of monetary policies (Kydland & Prescott, 1977). What also is crucial for a currency to be a reliable store of value is a negligible volatility (Yermack, 2013). The implication of volatility is that retailers have to adjust prices frequently which is a costly activity. Secondly, when a currency is held largely by financial institutions as a reserve, they need to be able to use it at any time and the value therefore to be stable to from them rather than constantly changing to set proper expectations. Compared to other currencies, the Bitcoin shows a high volatility and therefore I would like to see the impact regulations of central banks and governments have on the Bitcoin price and thus volatility.
SECTION 3: Literature Review
An increasing number of scientific research aims to describe the characteristics of the Bitcoin and which factors drive the Bitcoins price. This section gives an overview of relevant research done on the Bitcoin. Previous studies have only looked at the effects of news on the Bitcoin price. This research aims to only look at the impact of official statements and laws set by governments and central banks. All studies examined show that the Bitcoin has a large
volatility compared to either gold or fiat currencies and that their price is influenced by news .
Research from 2013 done by Kristoufek has shown that the price of the Bitcoin is not determined by financial theories such as cash-flows models, uncovered interest parity and purchasing power parity and hypothesized that only speculation drives the Bitcoins price. To test this he checked correlations between the Bitcoin price and searches on Google trends and Wikipedia. He found a positive correlation for both pairs and conclude that speculation drives the Bitcoin price. Glaser et al.(2014) looked at the trade volume of the Bitcoin to see if it is used as investment or medium of exchange. They conclude that most users treat the Bitcoin as investment rather than medium of exchange.
Looking at weekly data from 2010 till 2013 Brière, Oosterlinck and Szafarz (2013) showed that investments in the Bitcoin have exceptionally high average return and volatility and the Bitcoin offers diversification benefits because of low correlations with other classes of assets. To analyze it even further Van Wijk (2013) tested the effects of global macro-financial factors on the Bitcoin price. For his research he chose the Dow Jones Industrial Average Index, the oil price index and the euro-dollar exchange rate. His findings were that these factors have an significant impact on the Bitcoin price. Nevertheless, Van Wijk (2013) did not take Bitcoin demand and supply in account but only the global macro-financial factors. All findings mentioned above show that the Bitcoin has beneficial characteristics for investment portfolios.
To research price volatility, it is of interest to know if the Bitcoin can be described as a currency. Yermack researched if the Bitcoin is a real currency by looking at the volatility in 2013. In his study he compared the movement of Bitcoin to the London gold price and US dollar using 3 years of data excluding weekends since markets are closed then. The Bitcoins exchange rate volatility was 142% where other currencies their volatilities ranged between 7% and 12%. Yermack showed that the Bitcoin compared to other currencies and gold has an excessive volatility. Later research indicated that in the period from January 2015 till June 2015 there was less volatility than from December 2010 till June 2015, although it seems temporary (Bouoiyour and Selmi 2015). For these two periods, Bouoiyour and Selmi discovered that the Bitcoins volatility is more influenced by bad news than by good news. Another outcome of this study was that the Bitcoin-dollar exchange rate showed nearly zero correlation with major exchange rates or gold and the lack of correlation suggests the Bitcoins price is not influenced by macro-financial factors and these findings are supported by
Bouoiyour and Selmi (2015).
Ciaian (2015) analyzed daily data from 2009 till 2015 by applying time-series models and concluded that macro-financial factors are not driving the Bitcoin price and they do not exclude that investor speculation affects the Bitcoin price development. Ciaian (2015) used the Barro (1979) gold standard model to test the traditional determinants of currency price, supply and demand, on the Bitcoin the price. The main difference between gold and the Bitcoin is that the demand of Bitcoin is driven by its value in future exchange whereas golds demand is driven by the intrinsic value. The result was that Bitcoin can be explained by a standard model of currency price formation and this conclusion is in line with studies done by Buchholz et al. (2012) and Bouoiyour and Selmi (2015).
In contrast to these findings that macro-financial factors do not drive the Bitcoins price, there is research done in Haubo Dyhrberg (2015) to analyze what elements in the world economy the Bitcoin is sensitive to. To investigate how to classify the Bitcoin, the study focused primarily on several aspects of the price volatility by using the GARCH framework. The data from the 19th of July 2010 to the 22nd of May 2015 was obtained from Coindesk price
Index. The conclusion was that the Bitcoin does reacts significantly to federal fund rates as currencies do ad can be used for risk averse investors to anticipate on bad news. The study also suggests the Bitcoin has come closer to the dollar and gold in terms of volatility
compared to the research done by Yermack (2013). Since it is not regulated it will not react exactly like currencies on the market.
Ciaian (2015) however tested a second hypothesis to analyze the influence of new information on the Bitcoin price. To measure available information the number of views on Wikipedia was being used and there was a positive relationship between the Bitcoin price and the arrival of new information. The conclusion drawn is that global macro-financial
development, looking at the Dow Jones Index, exchange rate and oil price, only have effect in the short run. The results of Ciaian (2015) imply that research done by Van Wijk (2013) and Dyhrberg (2015) may be biased. Ciaian (2015) argues that this bias comes from the fact both researches did not take market forces of supply and demand in account. This shows the importance of analyzing different price drivers simultaneously.
SECTION 4: Data Sample
This analysis seeks the determine what the regulatory impact is on the Bitcoin volatility and uses the Bitcoin price to analyze this. For the impact of regulations, we make use of an event study and this implies that we need to know the specific dates of the events, namely
statements or jurisdictions made by governments or central banks. Furthermore the value of the Bitcoin is expressed in US dollars has to be retrieved and then the other variables to run the regression.
4.1 Bitcoin price data
The Bitcoin Price Index (BPI) shows the exchange rate between the bitcoin (BTC) and US dollar (USD). The BPI uses specific criteria for Bitcoin exchanges to be included in the BPI to assure quality. Here the volatility is calculated for the whole period. For my analysis I use daily data ranging from the 18th of September 2011 up to the 1th July 2017 including weekends since Bitcoins can be traded at any time. All data is available on
http://www.coindesk.com/price/.
4.2 Event dates
A summary of all governments and central banks that have or have not acted on regulating the Bitcoin has been made by Courtneidge and Clarence-Smith (2017). In this paper we need to make a distinction between positive and negative regulations. Positive regulations improve further integration and the use of the Bitcoin whereas negative regulations aim to restrict the use of the Bitcoin or impose taxes on Bitcoin related activities. In total there are 40 event dates used in this paper. Tables A and B below provide an overview of the events used.
Table A: Negative event dates
29th of July 2013 Bank of Thailand stated that Bitcoin activities
are illegal in Thailand.
29th of August 2013 Russia states that roubles are exclusive means of payment and therefore implying that Bitcoin is
illegal (Tolkachev, 2013)
3th of December 2013 The central bank of China issued an official
warning against the Bitcoins risks (People’s Bank of China, 2013)
11th of December 2013 Central bank of Cyprus issued an official warning against the Bitcoins risks (Famagusta Gazette,
2012).
13TH of December 2013 European Banking Authority issued a warning
against the Bitcoins risks (European banking Authority, 2013).
17th of December 2013 Denmark’s Financial Supervisory Authority stated that Bitcoin will not be regulated (Finanstilsynet,
2013).
19th of December 2013 Iceland prohibits foreign exchange with bitcoins (MORGUNBLAÐIÐ, 2013).
19th of December 2013 Germany stated that Bitcoin is a financial
instruments like foreign currencies (Jens Munzer, 2013)
24th of December 2013 India’s central banks issued an official warning
against the Bitcoins risks (Reserve Bank of India, 2013).
16th of January 2014 A Canadian official from Canada’s Department of Finance stated that Bitcoin is not considered as a
legal tender (George-Cosh, 2014).
25th of March 2014 International Revenue Service (IRS) treats Bitcoin
as property which makes it taxable ( International Revenue Service, 2014). 21th of April 2014 Estonian Tax Authority stated that Bitcoin capital
gains are subject to taxes (The Law Library of Congress, 2014)
26th of April 2014 Canada’s Revenue Agency stated that Bitcoin
users will have to pay taxes on digital transactions ( CBC News, 2013)
6th of May 2014 Bolivia’s central bank bans the Bitcoin.
19th of May 2014 The Dutch court declares Bitcoin is not money
(The Law Library of Congress, 2014). 27th of May 2014 Poland imposes taxes on profits from sales of
Bitcoin (Eric Calouro, 2014)
3th of June 2014 Bank Indonesia issued an official warning against
the Bitcoins risks (The Law Library of Congress, 2014).
24th of July 2014 Ecuador bans the Bitcoin and plans own digital
money.
1th of August 2014 The ministry of finance of the Russian Federation announces plans to make the Bitcoin illegal by
law.
4th of August 2014 The national bank of the Kyrgyz Republic
confirms Bitcoin is illegal under national law. 15th of September 2014 The central bank of Bangladesh stated that any
transaction through Bitcoin is a punishable offense.
3th of October 2014 The ministry of finance of the Russian Federation proposed monetary penalties for Bitcoin use and
promotion.
9th of February 2015 Bitcoin ban is up for public discussion in Russia.
14th of September 2016 Bank Negara Malaysia issued an official warnings
against the Bitcoins risks (The Law Library of Congress, 2016)
10th of March 2017 US Securities and Exchange Commission (SEC)
disapproves a Bitcoin ETF proposal ( Winklevoss Bitcoin Trust, 2017).
Table B: Positive event dates
15th of October 2012 First Bitcoin hedge fund announced (Knyazev,
2012)
1th of July 2013 Winklevoss Bitcoin Trust is planned to be taken
public and filed for at the SEC (Winklevoss Bitcoin Trust, 2013).
3th of August 2013 First Bitcoin Hedge Fund from Malta was
launched (Matonis, 2013).
9th of October 2013 Brazil enacted law to normalize mobile payment
systems such as the Bitcoin(Lei No. 12.865, 2013).
6th of December 2013 The Croatian National Bank concluded the
Bitcoin is not illegal in Croatia (coinspot.ru, 2013) 13th of March 2014 Singapore announced it will regulate virtual
currency intermediaries including Bitcoin exchanges.
11th of April 2014 People’s bank of China stated it will not ban the
Bitcoin (Zhou Xiaochuan, 2014). 25th of April 2016 Bitstamp Bitcoin Exchange has been granted a
license by the Luxembourg government to be fully regulated in the European Union. 3th of June 2016 Bank of Japan amended act which will make
Bitcoin exchange regulation possible (The Law Library of Congress, 2014).
14th of August 2016 Russia considers dropping penalties on Bitcoin
activities (coindesk, 2016)
3th of October 2016 Russia does not consider the Bitcoin to be illegal any longer ( Tolkachev, 2016)
7th of November 2016 Bank of Japans amendment for virtual currency
exchanges regulation for the 2th June 2017 is approved(The Law Library of Congress, 2016). 1th of April 2017 Japanese parliament recognized Bitcoin as
11th of April 2017 Russia announced plans to make Bitcoin transactions legal within the country.
24th of April 2017 SEC announces reconsideration of the Bitcoin ETF proposal (Bats BZX Exchange, 2017).
The multiple regression model uses the general price level, the size of the Bitcoin economy, the velocity of the Bitcoin and the amount of bitcoins in circulation as independent variables to explain the dependent variable, namely the Bitcoin price. In table C a summary of the used variables is provided.
First for the general price level, 𝑝𝑝, I use the Consumer Price Index (CPI) of the United States since Bitcoin is mostly expressed in US dollars and the United States have a high number of Bitcoin users. The CPI data ranging from the 18th of September 2011 till the 1th of June 2017 is retrieved from DataStream (2017).
Then to measure the size of the Bitcoin economy, 𝑔𝑔, I use the total number of Bitcoin transactions per day following Ciaian (2015). The number of transactions is recorded in the blockchain (blockchain.info, 2017)
For the velocity of the Bitcoin, 𝑣𝑣, I follow Matonis (2012) and use Bitcoin Days Destroyed index which for any given transaction takes the number of bitcoin in a transaction and multiplying it by the number of days it has been since those coins were last spent (blockchain.info, 2017) .
Then for the total amount of bitcoins in circulation, 𝑏𝑏, I use the historical number of total bitcoins which have been mined(quandl.com, 2017).
Table C: Summary of variables
d2 2,083 .0115218 .1067453 0 1 d1 2,083 .0072012 .0845738 0 1 stock 2,083 1.27e+07 2584055 7295700 1.64e+07 velocity 2,083 7727273 9024369 0 1.73e+08 size 2,083 108677.1 90112.85 4523 369098 cpi 2,083 1.486431 .8719228 -.1995174 3.868357 close 2,083 364.7683 364.1209 2.05 2476.3 date 0 Variable Obs Mean Std. Dev. Min Max
SECTION 5: Methodology
In order to see if regulations have an effect on the Bitcoin price, this paper uses an event study as statistical analysis. An event study looks at the behavior of the price at an event or announcement and checks for abnormal returns. The analysis is done using OLS regression.
5.1 Event study
In this paper we make use of an event study to estimate the effect of event on the dependent variable. For the normal returns we look at the days without events over the period of 7 years and is called the event window. The event window is defined as period for which we want to estimate the events effects for on the dependent variable. In this paper I will use the 7 years excluding event dates for the estimation window and I use 40 separate events which each equal to one day for the event window. The Bitcoin price is affected by several factors and therefore a short event window is used to isolate the impact of the event as much as possible. First the estimation window for the full 7 years will be estimated and dummy variables 𝐷𝐷1 and 𝐷𝐷2 will take a value of 0. Then the event windows will be estimated
for the given 40 event dates and then dummy variables 𝐷𝐷1 and 𝐷𝐷2 will take a value of 1. The
estimate of the event window will be tested against the estimates for the estimation window to see if the coefficient for the events are significantly different from zero.
5.2 Regression model
To analyze the effect of the events on the dependent variable we include the dummy variable. Previous literature by Ciaian (2015) showed that the following multiple regression model can be used for estimate the Bitcoins price:
Where 𝑃𝑃𝑡𝑡𝑏𝑏 denotes the Bitcoin price at time t, 𝑝𝑝𝑡𝑡the general price level, 𝑔𝑔𝑡𝑡the size of the
Bitcoin economy, 𝑣𝑣𝑡𝑡the velocity of bitcoins in circulation, 𝑏𝑏𝑡𝑡the total stock of bitcoins in circulation, 𝛽𝛽5 the impact of the a positive event with corresponding dummy variable 𝐷𝐷1, 𝛽𝛽6
the impact of the a negative event with corresponding dummy variable 𝐷𝐷2 and 𝜀𝜀𝑡𝑡as error
term. Additionally we expect 𝛽𝛽1and 𝛽𝛽2to be positive and 𝛽𝛽3and 𝛽𝛽4to be negative (Ciaian, 2015).
The error term follows the following assumptions: Assumption 1: E(𝜀𝜀𝑡𝑡)= 0 for all t=1,2, …, n
Assumption 2: Var(𝜀𝜀𝑡𝑡)= σ2
Assumption 3: 𝑃𝑃𝑡𝑡𝑏𝑏 and 𝜀𝜀𝑡𝑡are not correlated
Assumption 4: There is no heteroscedasticity of the error term
5.3. Hypotheses
The research question of this paper concerns regulatory impact on the Bitcoins price. The impact of regulation is measured by the independent dummy variables in the regression. Regulations are split up in positive and negative regulations and that leads us to two
hypotheses.
Hypothesis 1
Positive regulations can make the use of the Bitcoin more safe for users and increase usage and thus the expectation therefore is that positive events will affect the Bitcoins price positively since the risk of using it is lower.
‘’The hypothesis states that positive regulations affect the Bitcoin price positively.‘’ H0: 𝛽𝛽5 = 0
Hypothesis 2
Negative regulations discourage people to make use of the Bitcoin or even forbid people to make use of the Bitcoin and thus the expectation is the negative events affect the Bitcoins price negatively.
‘’The hypothesis states that negative events affect the Bitcoin price.’’ H0: 𝛽𝛽6 = 0
SECTION 6: Results and discussion
In this section the results of the empirical analysis will be presented and I comment on the results. Then the assumptions of the OLS regression will be tested.
6.1 Regressions
In table 1 the results of regression analyses are shown with and without dummy variables. Model (1) in table 1 is used to check the used regression without dummy variables and Model (4) in table 1 shows regression (1) with dummy variables.
Table 1: Regression analysis for Bitcoins closing price
The coefficient for the size of the Bitcoin economy is not significant which is in contrast with current research done by Ciaian (2015). A possible explanation here is that the variable size is equal to the variable stock times variable closing price. The correlation matrix in table 2 shows a correlation of -0.9085 between size and stock.
* p<0.05, ** p<0.01, *** p<0.001 t statistics in parentheses adj. R-sq 0.695 0.695 0.697 0.697 N 2083 2083 2083 2083 (-26.58) (-26.53) (-26.08) (-26.03) _cons -1619.5*** -1619.3*** -1591.8*** -1591.4*** (4.24) (4.24) d2 175.9*** 176.0*** (0.05) (0.11) d1 2.589 5.771 (25.49) (25.45) (24.93) (24.88) stock 0.000130*** 0.000130*** 0.000128*** 0.000128*** (5.54) (5.53) (5.58) (5.57) velocity 0.00000298*** 0.00000297*** 0.00000299*** 0.00000299*** (0.27) (0.27 ) (0.76) (0.76) size 0.0000352 0.0000354 0.0000989 0.0000995 (27.47) (27.43) (27.04) (26.99) cpi 202.3*** 202.3*** 199.2*** 199.2*** close close close close (1) (2) (3) (4)
Table 2: Correlation matrix of the regression
Then the estimated coefficient for positive events is positive but not significant where the estimated coefficient for negative events is positive and significant. Different levels of significance do not change these results. This is in line with the current research done that showed that the Bitcoin price is more influenced by bad news than by good news (Bouoiyour and Selmi 2015). Nevertheless that research concerned the volatility where this analysis focused solely on the price formation.
Next, the adjusted squared values range from 0.695 to 0.697. The relatively high R-squared confirms the research of Ciaian (2015) which showed that Bitcoin demand and supply are main factors in the Bitcoin price formation. Adding the dummy variables did not affect R-squared significantly even though the coefficient for negative events was significant. This suggests some variables may have been omitted.
To summarize, the results of regression (1) shown in table 1 show no significant impact of positive events on the Bitcoins price and a significant positive impact on the Bitcoins price of negative events. This leads to the acceptance of the null-hypothesis of hypothesis 1. For hypothesis 2, we observe from table 2 the significant positive value for the estimated coefficient of d2 (p<0.001) which implies a higher Bitcoin price on negative event days. Therefore null-hypothesis for hypothesis 2 is rejected and the negative events affect the Bitcoins price as expected.
_cons -0.8025 0.8534 -0.0148 -0.9831 0.0522 0.1076 1.0000 d2 -0.0974 0.1153 0.0050 -0.1172 0.0144 1.0000 d1 -0.0520 0.0392 -0.0412 -0.0520 1.0000 stock 0.7160 -0.9085 -0.0057 1.0000 velocity -0.0423 -0.1677 1.0000 size -0.5905 1.0000 cpi 1.0000 e(V) cpi size velocity stock d1 d2 _cons
6.2 Testing OLS assumptions
To critically look at the used regression, underlying assumptions need to be checked for. The assumptions ensure the unbiasedness, efficiency and consistency of the estimations. First a test for heteroscedasticity is run to see if the variability is equal across the observations in the regression. The Breusch-Pagan test is used as shown in table 3 where the
null-hypothesis expects no heteroscedasticity.
Table 3: Breusch-Pagan test for heteroskedasticity
Table 3 shows that the null-hypothesis of no heteroscedasticity is rejected (p<0.0000) and this is an implication for the OLS regression used. The presence of heteroscedasticity makes the OLS regression inefficient but nevertheless unbiased in the estimations.
Second, in table 3 the results for the skewness/kurtosis test for normal distribution of the residuals is shown.
Table 4: Skewness/Kurtosis tests for Normality
The null-hypothesis in table 4 states that the residuals are normally distributed and is as shown in table 4 rejected (p<0.0000). The residuals are not normally distributed and is another implication for the OLS regression.
Third, a RESET test is run to see if the regression model has omitted variables.
Prob > chi2 = 0.0000 chi2(1) = 641.12
Variables: fitted values of close Ho: Constant variance
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity
resid 2,083 0.0000 0.0000 . 0.0000 Variable Obs Pr(Skewness) Pr(Kurtosis) adj chi2(2) Prob>chi2 joint Skewness/Kurtosis tests for Normality
Table 6: Ramsey RESET test
The null-hypothesis in table 6 states that the regression has no omitted variables and is as shown in table 6 rejected (p<0.0000). Therefore there are omitted variables that drive the Bitcoins price which are not included in this regression. The research done by Van Wijk (2013) that global macro financial factors such as oil prices are significant Bitcoin price drivers and these factors could be an explanation for the omitted variables in my analysis.
At last, an autocorrelation test was to see if the variables are correlated to eachtother.
Table 7: Autocorrelation test
The results in table 7 shows that the model suffers from autocorrelation (p<0.000) and that shows that the variable current value is affected by its previous values.
SECTION 7: Conclusion
In this section a summary of this paper will be provided and limitations and suggestions for further research will be discussed.
This paper has run an empirical regression to analyze the effects of regulations on the Bitcoins price for a period of 7 years concerning 40 events. A distinction was made between positive and negative events to see in which direction the price can be affected by
governments and central banks. The theory provided outlines in which way the Bitcoin can serve as a global reserve currency and help governments and central banks by looking at arguments given by Keynes (1943) and Triffin (1961) who outlined the need for a
supranational reserve currency.
Prob > F = 0.0000 F(3, 2073) = 27.74 Ho: model has no omitted variables
Ramsey RESET test using powers of the fitted values of close
H0: no serial correlation
1 1575.491 1 0.0000 lags(p) chi2 df Prob > chi2 Breusch-Godfrey LM test for autocorrelation
Current research done on the Bitcoin shows that its price is affected by both good and bad news. However, current research has difficulties to find the Bitcoins price drivers and until now it is best described by the Barro (1979) gold standard model.
The regression containing dummy variables for either positive or negative events was used to test two hypotheses. First, the effects of positive events on the Bitcoins price were tested. The result was that positive events do not affect the Bitcoins price at 1%, 5% and 10% significance. Second, the effects of negative events on the Bitcoins price were tested. The result here was the negative events increase the Bitcoins price at 1%, 5% and 10%
significance. This shows the Bitcoins price is more affected by negative events than by positive event which is in line with current research (Bouoiyour and Selmi 2015).
Regulations do affect the Bitcoins price and this shows that governments and central banks have the ability to affect the Bitcoins price, however regulations are just one of the many factors that affect the Bitcoins price.
The model used showed a relatively high R-squared, which makes it useful in theory but it the following two assumptions for using an OLS regression were violated that affect the estimations. First, OLS assumes homoscedasticity. By using a Breusch-Pagan test shown in table 3 heteroscedasticity has been found and that implies that the standard error of the OLS estimator is biased. This can lead to a statistical error by not rejecting the null-hypothesis McCulloch, 1985). More advanced econometric models should be used to solve this problem, but that is beyond the scope of this paper. Second, the OLS assumes a normally distributed standard errors. The skewness/kurtosis test in table 4 shows that this assumption has been violated. This can be an issue for the validity of the model used (Lumley & Emerson, 2002). Nevertheless, further research needs to take this in account.
Then, the Ramsey RESET test was used to check for omitted variables. The results in table 5 show that there are omitted variables which are important factors. Leaving these variables out leads the model to wrongly estimating the effect of the used independent variables and therefore leads to a bias (Howland, 2006). This shows that further research has to analyze other factors that affect the Bitcoins price formation.
At last the Breusch-Godfrey test showed us the model suffered from autocorrelation. Further research can reduce this problem by looking at returns instead of price levels.
REFERENCES
Advarsel mod virtuelle valutaer (bitcoin m.fl.) [Warnings Against Digital Currencies (Bitcoins etc.)], FINANSTILSYNET (Dec. 17, 2013),
http://www.finanstilsynet.dk/da/Nyhedscenter/Pressemeddelelser/2013/Advarsel-mod-virtuelle-valutaer-bitcom-mfl-2013.aspx.
Artem Tolkachev & Kseniia Osipova(Aug. 29, 2013), Possibilities and Risks of Using Bitcoin in Russia, ATPLaw.ru, http://www.atplaw.ru/useful/articles/vozmozhnosti-i-riski-ispolzovaniya-bitkoin-v-rossii.html
Anderson, T. G., Bollerslev, T. (1998). Answering the skeptics: Yes, standard volatility models do provide accurate forecasts. International Economic Review, 39, 885-905.
A.F. Bariviera, M.J. Basgall, W. Hasperué & M. Naiouf. (2017) Some stylized facts of the Bitcoin market
Barro, R. J. (1979). Money and the price level under the gold standard. The Economic Journal,
89(353), 13-33.
Bats BZX Exchange. (2017). United States Securities and Exchange Commission form 10-K.
Retrieved from Securities and Exchange Commission website:
https://www.sec.gov/rules/other/2017/34-80511.pdf
Böhme, R., Christin, N., Edelman, B., & Moore, T. (2015). Bitcoin: Economics, Technology, and Governance. The Journal of Economic Perspectives, 29(2), 213-238. Retrieved from
http://www.jstor.org/stable/24292130
Bouoiyour, J., & Selmi, R. (2015). Bitcoin Price: Is it really that New Round of Volatility can be on way?.
Bordo, M. D. (1993). The Bretton Woods international monetary system: a historical overview. In A retrospective on the Bretton Woods system: Lessons for international monetary reform (pp. 3-108). University of Chicago Press.
Bowman, R. G. (1983). Understanding and conducting event studies. Journal of Business
Brière, M., Oosterlinck, K., & Szafarz, A. (2015). Virtual currency, tangible return: Portfolio diversification with bitcoin. Journal of Asset Management, 16(6), 365-373.
Boughton, J.M. (2002). Why White, Not Keynes? Inventing the Postwar International Monetary System. The Open Economy Macromodel: Past Present, and Future.
Buchholz, M., Delaney, J., Warren, J., & Parker, J. (2012). Bits and Bets, Information, Price Volatility, and Demand for Bitcoin. Economics, 312.
Carbaugh, R. J., & Hedrick, D. W. (2009). Will the dollar be dethroned as the main reserve currency. Global Economy Journal, 9(3), 1-14.
classes.dma.ucla.edu/Fall15/161/projects/chan_c/Desma%20161/Project2.html
Croatia Allowed to Use Bitcoin, COINSPOT.RU (Dec. 16, 2013), Retrieved from: http://coinspot.ru/news/xorvatiya-razreshila-ispolzovanie-bitkoina/
cryptocoinsnews.com/top-10-countries-bitcoin-banned/
Chaum, D. (1983). Blind signatures for untraceable payments. In Advances in cryptology (pp. 199-203). Springer US.
Christin, N. (2013, May). Traveling the Silk Road: A measurement analysis of a large
anonymous online marketplace. In Proceedings of the 22nd international conference on World Wide
Web (pp. 213-224). ACM.
Courtneidge, R. & Clarence-Smith, C. (2017). Bitcoin and Blockchain technology Update:
Research Paper
Cyprus Central Bank Warns About Risks in Use of Bitcoin, FAMAGUSTA GAZETTE (Dec. 11, 2013), http://fama gusta-gazette.com/cyprus-central-bank-warns-about-risks-in-use-of-bitcoin-p21692-69.htm.
George-Cosh, D.(2014) , Canada Says Bitcoin Isn’t Legal Tender, THE WALL STREET JOURNAL (Jan. 16, 2014),
http://blogs.wsj.com/canadarealtime/2014/01/16/canada-says-bitcoin-isnt-legal-tender/.
Calouro, E.(2014), Polish Tax Man Says Bitcoin Mining Profits Subject to VAT, Retrieved from: http://www.newsbtc.com/2014/05/27/polish-tax-man-says-bitcoin-mining-profits-subject-vat/
European Banking Authority(Dec. 13, 2013), EBA Warns Consumers on Virtual Currencies,
Ciaian, P., Rajcaniova, M., & Kancs, D. A. (2016). The economics of BitCoin price formation.
Applied Economics, 48(19), 1799-1815.
Datastream. (2017) Thomson Reuters Datastream. [Online]. Available at: Subscription Service (Accessed: June 2017)
Engle, R. (1982). Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation. Econometrica, 50(4), 987-1007.
Glaser, F., Zimmermann, K., Haferkorn, M., Weber, M. C., & Siering, M. (2014). Bitcoin-asset or currency? revealing users' hidden intentions.
Graham, F. (1944). Keynes vs. Hayek on a Commodity Reserve Currency. The Economic Journal, 54(215/216), 422-429.
Feldstein, M. (1998). Refocusing the IMF. FOREIGN AFFAIRS-NEW YORK-, 77, 20-33.
Fleming, J. (1971). The SDR: Some Problems and Possibilities (Les DTS: Quelques problèmes et possibilités) (Los DEG: Ciertos problemas y posibilidades). Staff Papers (International Monetary Fund),
18(1), 25-47.
Friedman, M., & Schwartz, A. J. (2008). A monetary history of the United States, 1867-1960. Princeton University Press.
Howland (2006). Omitted Variable Bias. Introductory Econometrics: Using Monte Carlo Simulation with Microsoft Excel. Cambridge University Press
Höftin stöðva viðskipti með Bitcoin [Controls Stop Trading in Bitcoin], MORGUNBLAÐIÐ (Dec. 19, 2013), http://www.mbl.is/vidskipti/frettir/2013/12/19/hoftin_stodva_vidskipti_med_bitcoin/; for a rough translation of this article and comments, see Saevarg, Bitcoin Trading Illegal in Iceland According to Icelandic Central Bank
Katsiampa, P. (2017). Volatility estimation for Bitcoin: A comparison of GARCH models.
Economics Letters.
Keynes JM (1943) Proposals for an international clearing union, British government publication, Cmd.6437, London (April). Reprinted in J. Keith Horsefield ed. (1969), volume III. Documents: 19–36
Knyazev, A. (2012), The emergence of electronic gold , Retrieved from:
International Monetary Fund. (2013). Sixth Edition of the IMF's Balance of Payments and International Retrieved from:
https://www.imf.org/en/About
Investment Position Manual (BPM6). Washington, DC: International Monetary Fund.
International Revenue Service. (2014). , IRS Virtual Currency Guidance : Virtual Currency Is
Treated as Property for U.S. Federal Tax Purposes; General Rules for Property Transactions Apply.
Retrieved from:
https://www.irs.gov/uac/newsroom/irs-virtual-currency-guidance
Jens Münzer, Bitcoins: Aufsichtliche Bewertung und Risiken für Nutzer [Bitcoins: Supervisory Evaluation and Risks for Users], BAFIN (Dec. 19, 2013),
http://www.bafin.de/SharedDocs/Veroeffentlichungen/DE/Fachartikel/ 2014/fa_bj_1401_bitcoins.html.
The Law Library of Congress. (2014)., Regulation of Bitcoin in Selected Jurisdictions Koning, J. P. (2016). Fedcoin: A Central Bank-issued Cryptocurrency. November, 15, 2016. Kristoufek, L. (2013). BitCoin meets Google Trends and Wikipedia: Quantifying the relationship between phenomena of the Internet era. Scientific reports, 3, 3415.
Lei No. 12.865, de 9 de Outubro de 2013 [Law No. 12,865 of October 9, 2013], http://www.receita.fazenda.gov. br/Legislacao/leis/2013/lei12865.htm (Braz.).
Lumley T, Emerson S. (2002) The Importance of the Normality Assumption in Large Public
Health Data Sets. Annual Review of Public Health. 23:151–69
Matonis, J.(2013), First Bitcoin Hedge Fund Launches From Malta, Retrieved from:
https://www.forbes.com/sites/jonmatonis/2013/03/08/first-bitcoin-hedge-fund-launches-from-malta/#5a6d066e3e1e
Matonis, J. (2012). “Top 10 BitCoin Statistics.” Forbes. Retrieved from:
www.forbes.com/sites/jonmatonis/2012/07/31/top-10-bitcoin-statistics
McCulloch, J. (1985). Miscellanea: On Heteros*edasticity. Econometrica, 53(2), 483-483. Retrieved from http://www.jstor.org/stable/1911250
McKinnon, R. I. (1979). Money in international exchange: the convertible currency system. Oxford University Press.
McWilliams, A., & Siegel, D. (1997). Event studies in management research: Theoretical and empirical issues. Academy of management journal, 40(3), 626-657.
Moggridge, D., & Howson, S. (1974). Keynes on Monetary Policy, 1910-1946. Oxford Economic Papers, 26(2), new series, 226-247.
Moore, T., & Christin, N. (2013, April). Beware the middleman: Empirical analysis of Bitcoin-exchange risk. In International Conference on Financial Cryptography and Data Security (pp. 25-33). Springer Berlin Heidelberg.
[Notice on Precautions Against the Risks of Bitcoins] (issued by the People’s Bank of China, the Ministry of Industry and Information Technology, China Banking Regulatory Commission, China
Securities Regulatory Commission, and China Insurance Regulatory Commission, Dec. 3, 2013) YIN FA, 2013, No. 289,
http://www.pbc.gov.cn/publish/goutongjiaoliu/524/2013/20131205153156832222251/20131205153 156832222 251_.html (China).
Ocampo, J. A. (2009). Why Should the Global Reserve System Be Reformed?. Finance & Bien
Commun, (2), 79-89.
Osband, K., & Villanueva, D. (1993). Independent Currency Authorities: An Analytic Primer. Staff Papers (International Monetary Fund), 40(1), 202-216.
Reid, F. and Harrigan, M. (2013). An analysis of anonymity in the bitcoin system. Security and privacy in social networks, pages 197–223.
Reserve Bank of India, RBI Cautions Users of Virtual Currencies Against Risks (Dec. 24, 2013), http://rbi.org.in/scripts/BS_PressReleaseDisplay.aspx?prid=30247
Revenue Canada Says BitCoins Aren’t Tax Exempt, CBC NEWS (Apr. 26, 2013),
http://www.cbc.ca/news/ business/revenue-canada-says-bitcoins-aren-t-tax-exempt-1.1395075 Triffin, R. (1960). Gold and the dollar crisis. New Haven: Yale Univen.
Triffin, R. (1961). After the Gold Exchange Standard? Weltwirtschaftliches Archiv, 87, 188-207. Retrieved from http://www.jstor.org/stable/40434872
Triffin, R. (1964). The evolution of the international monetary system: Historical reappraisal
and future perspectives. International Finance Section, Department of Economics, Princeton
University.
Triffin, R. (1978). Gold and the Dollar Crisis: Yesterday and Tomorrow. International Finance Section, Department of Economics, Princeton University.
Tsay, Ruey S. Analysis of financial time series. Vol. 543. John Wiley & Sons, 2005.
Winklevoss Bitcoin Trust. (2017). United States Securities and Exchange Commission form
10-K. Retrieved from Securities and Exchange Commission website:
https://www.sec.gov/rules/sro/batsbzx/2017/34-80206.pdf
Winklevoss Bitcoin Trust. (2013). United States Securities and Exchange Commission form
10-K. Retrieved from Securities and Exchange Commission website:
https://www.sec.gov/Archives/edgar/data/1579346/000119312513279830/0001193125-13-279830-index.htm
Yermack, D. (2013). Is Bitcoin a real currency? An economic appraisal (No. w19747). National Bureau of Economic Research.
https://www.statista.com/statistics/218493/paypals-total-active-registered-accounts-from-2010/
