UNIVERSITY OF AMSTERDAM
FACULTY OF ECONOMICS AND BUSINESS BSc Economics & Business Bachelor
Specialisation Finance & Economics
Bitcoin and the Effects of Regulation
An event study of key announcements on new cryptocurrency legislation
Author: Chris Landi Student number: 10707654
Thesis supervisor: Cenkhan Sahin Finish date: January 2016
Statement of Originality
This document is written by Student Chris Landi who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
Bitcoin has been given increasingly more attention on social media and other news outlets and world leaders are faced with dilemmas. The need to regulate has never been bigger with new cryptocurrencies sprawling up rapidly and new ways being found to trade in and use cryptocurrencies. This paper will closely look at two important regulatory announcements made by government agencies regarding Bitcoin and cryptocurrencies. Specifically, I will test the hypothesis whether regulation, be it good or bad, has any effect on the value of Bitcoin. For good regulation, I will test whether returns go up and for bad regulation will do the opposite. What I found was that there is a significant effect, however the effect of good regulation is far less evident than that of bad regulation, indicating that this bullish-market still suffers from large dips when faced with uncertainty.
1.Introduction
On January 11th 2018, the South Korean justice minister announced that virtual
currencies are a danger and a bill is being written to ban cryptocurrency exchanges. South Korea being the third largest market, the impact was felt on the price of Bitcoin as it dropped 14 percent in one afternoon on Bitstamp exchange. Later that day Chief press secretary Yoon Young-chan revealed that closing the exchanges was one of the options proposed but was not a final decision. South Korean government feared exchanges weren’t paying all their corporate taxes and certain practices such as margin trading and shorting creating higher demand were seen as dangerous for investors (Jung-a, Harris, 2018).
With cryptocurrencies becoming more popular and widespread every day, and a relatively small degree of regulation, regulators need to step in. Fear of a price bubble has EU authorities tighten their oversight on the cryptocurrency industry. Vice-President of the European Commission Valdis Dombrovskis, has urged major financial watchdogs to warn the average consumer of the risks of the high volatility of investing in cryptocurrencies. Investors run the risk of losing their entire investments, suffering from security breaches, market manipulation and fraud. Coinbase, a large exchange, announced that it is investigating its own staff for suspicion of insider trading, with Bitcoin Cash. The focus of regulation is more on investor protection than protecting the financial systems, as cryptocurrencies are still relatively small in overall volume. The introduction of futures contracts by both Cboe Global Markets and CME Group could change that in the future (Brundsen, Murphy, 2017).
Do announcements surrounding new legislation on cryptocurrencies and in particular Bitcoin, impact the value of Bitcoin, and if it does, are the effects positive or negative? This question is highly important for regulators and governments that despite seeing the value of blockchain technology and cryptocurrencies in general, also see the dangers in the early stages of this industry (Brito, Castillo, 2013). This trade-off between nurturing a new industry yet also protecting early users from fraud, speculative risk and countering money laundering and terrorism financing leaves regulators with difficult decisions to make that depend on the ability of a particular
government to counter these illegal activities and the level of crime happening in a certain country. Security issues regarding national matters like safeguarding the national currency from speculative attacks is also of concern for many smaller governments (Plassaras, 2013).
No research has been done yet on the effects of regulation announcements concerning cryptocurrencies on the value of cryptocurrencies. The main focus on Bitcoin and regulation have been on how to regulate it further such as Sotiropoulou
and Guégan (2017) and Kaplanov (2012). Others have studied what the main drivers
are of Bitcoin price, one of the drivers being publicity and internet searches through which regulation announcements can indirectly affect the price of Bitcoin (Kristoufek, 2015). However, it is crucial for governments to know in what way their announcements affect the value of cryptocurrencies if they want to safeguard the savings of their citizens. A single announcement by itself can have momentary impact on the value of Bitcoin even if no further action is taken to pass the law (Jung-a, Harris, 2018) let alone the longer lasting effects of a law going into effect. Therefore, governments that do not intend to destroy the savings of their citizens must be cautious with their public statements.
My hypothesis states that a good law, meaning a law that should make the use or trade of Bitcoin easier and/or safer therefore increasing demand, will have a positive impact on Bitcoin returns. For bad laws it’s the opposite, they should decrease demand and have a negative effect on the return. Alternatively, Bitcoin is simply not affected by regulation, meaning that regulators have nothing to be concerned about. It is difficult to distinguish between good and bad laws but as Finnis (2011) defined a good or just law is one that promotes the common good. Simply said, the common good are the basic goods for example the value of knowledge, of friendship or of autonomy. This concept of basic goods might be scientifically unsatisfying, since Finnis says that only intelligent and good people will grasp the concept of these goods. However, since Bitcoin in itself is a technology that gives privacy and autonomy from large financial institutions, it can be said that any law that wants to hinder that is a bad law.
To fully capture any abnormal returns after the announcements of both the Chinese and Japanese governments announcements, an event study will be
performed on the returns. Usually event studies are used to assess the impact of an economical event on the value of a firm. Given that investors are rational actors, the effect of a certain economic event is directly implemented into the price of common stock (MacKinlay, 1997). Comparably, Bitcoin can be seen as a form of security or share of the blockchain technology behind Bitcoin and demand dictates the price to partake in that technology. Bitcoin investors are rational investors, this can be seen from the recent hard fork of Bitcoin Cash announced on 20th July of 2017 where investors anticipated the hard fork into the price of Bitcoin. In around 12 hours the price skyrocketed up 20% after investors came to know about the hard fork. The price increase of about 500 USD corresponds to the initial price of Bitcoin Cash when it got first released. MacKinlay reckons that an event study can also be used for measuring the effect of a change in regulation or policy on the value of a firm, as was done by William Schwert (1981).
To performs an event study, key events must be determined. The first event will be Japans law that sees Bitcoin as legal tender. The second event will be the Chinese ICO ban. ICO’s still not being regulated, China is the biggest market to ban them. Next, I analyse the normal returns in regular times, even though one can say there are no regular times for Bitcoin. With both the Constant Mean Model and the Market Model, I compare the actual return around and on the event days, with the return as predicted by the two models and see if there exist any significant discrepancy.
What I found was that the Japanese law, which is a good law as it makes it legal to use Bitcoin to buy goods and services, had little impact on the value of Bitcoin. Returns were only weakly positively affected by the new law. The Chinese ICO ban had a much more drastic effect on the price, showing large significant drops in Bitcoin’s return on both the day of the event and the aftermath. This means that the still fragile cryptocurrency market is strongly affected by uncertainty on the future legality of cryptocurrencies, and any positive development is simply accepted but is not value increasing as no new investors are taking advantage of these new laws.
The paper is structured as follows: I start of with the current context of Bitcoin regulation, namely how regulators could do it and how some are doing it. I explain what ICO’s are and the current context on that area of cryptocurrencies and I make
clear what difficulties arise with event studies. Then I show the methodology and how I will approach answering my research question. After that I reveal my findings in detail and propose a reason for them. Lastly I conclude with a short summary and some further suggestions.
2.Literature Review
2.1.How to regulate Bitcoin?
There exist three aspects that governments can regulate according to
Stavropoulos and Guégan. That is the Bitcoin system itself, the uses of Bitcoin and finally, the members of the system (2017). Starting off with the system itself, the most difficult aspect to regulate. A lack of a central authority means regulators cannot hold one entity accountable for the ongoing affairs of Bitcoin. Instead Bitcoin functions on the work of the developers for the code and the miners who validate transactions and mine Bitcoins so to speak as compensation. These cannot be directly regulated, regulators can only give out warnings to users and inform them and hear their complaints as best as possible through informative websites.
Secondly, governments can regulate the uses of Bitcoin and in particular the illegal uses of Bitcoin. Two biggest concerns are money laundering and the financing of terrorism but as with regular currencies, Anti-Money Laundering/Combating the Financing of Terrorism (AML/CFT) frameworks already exist and regulators have started to adapt them to cryptocurrencies. The lack of accountability and main representation means AML/CFT must tackle other entities that handle large volumes of cryptocurrencies, such as the exchanges and wallet service providers. However, even with these AML/CFT regulations in place, it still possible for users to simply not use these exchanges and make one on one transactions, as Bitcoin was intended to do.
Lastly, regulators can take on the members of the system. One inherent property of Bitcoin and the block chain is the irreversibility of transactions. If users make a mistake and send Bitcoins to the wrong address, there is no way to get it back. Therefore, the only way to regulate this is through disclaimers and properly informing users. Apart from regular users, service providers also need some attention.
Regulators do this through licensing and there are two ways of doing it. Some countries like Germany and France are able to incorporate cryptocurrencies into their existing licensing requirements for payment and financial service providers. The state of New York on the other hand has created separate licencing requirements through the Bitlicense Act. It looks at the owners and officers, the amount of capital and reserves the provider holds, whether it adheres to the AML/CFT regulations and cybersecurity (Sotiropoulou, Guégan, 2017).
2.2.What are regulators doing?
The Law Library of Congress in their report (2014) came to the conclusion that foreign jurisdictions have four ways of regulating Bitcoin: (1) not regulating it in any way; (2) taxing earnings from Bitcoin but not regulating it any further; (3) banning or limiting the use of Bitcoin and (4) lastly accepting Bitcoin as a legitimate currency and regulating it appropriately. The first group is still the largest and most countries including the Netherlands, Belgium, and also the European Union seem to lag behind when it comes to effectively regulating Bitcoin. They maintain a hands-off approach and don’t see cryptocurrencies taking over regular currencies any time soon or threatening the financial system as a whole.
The second group have set out certain laws that dictate how Bitcoin should be taxed but lack laws that address the use, trade or mining of Bitcoin. These countries include the United Kingdom who are planning to make Bitcoin more transparent but as of yet do not have laws regarding the use of Bitcoin. The UK treats Bitcoin as a foreign currency and charges VAT on the sale of Bitcoin when exchanged for pound sterling. Regarding profits and losses on Bitcoin, the UK imposes a capital gain tax. Also, Spain, Finland and Norway see Bitcoin as capital property and tax it accordingly and similarly to the UK with a value-added tax (Global Legal Research Directorate Staff, 2014).
The third group and perhaps most intriguing for my research is the group that has banned Bitcoin altogether or heavily limited its use. These countries include Thailand, Iceland and China. The Bank of Thailand has gone for the bluntest approach and banned Bitcoin outright. Iceland sees foreign exchange trading with Bitcoin illegal under its Foreign Exchange Act where it is stated that Icelandic currency cannot leave
the country. Lastly China, where the Peoples Bank of China has prohibited banks and payment companies from dealing with Bitcoin. The institutions are not allowed to acquire or sell Bitcoins, give Bitcoin related services to customers or to trade Bitcoins for Chinese yuan or foreign currencies (Rabinovitch, 2013). All in all, this seriously hinders further development of the industry as a whole.
The fourth group is the group with the most pro-active approach, seeking to further implement cryptocurrencies and merging the two worlds together. They recognise cryptocurrencies as valid payment methods and regulate by utilising existing financial laws or developing new regulations. Countries such as Brazil that according to Law No. 12,865, want to facilitate mobile payment systems and the creation of cryptocurrencies. Japan further legalizes Bitcoin by accepting it as legal tender by updating a part of their banking act. Germany sees Bitcoin exchanges as any other financial services company and must adhere to regular financial regulation and reporting requirements. Similarly, Sweden does not have a separate regulatory framework but sees Bitcoin as a financial service with certain requirements regarding reporting.
In conclusion, there are four broad approaches to Bitcoin regulation. What is the preferable way? Lawmakers have been tempted to make laws to ban the use of Bitcoin to stop it being further used for criminal ends even though there is no evidence Bitcoin is mainly used for illegal activities. According to a study analysing how and where Bitcoins are being used for money laundering, only 0,61% of total Bitcoins were used for illicit ends during the year 2013 to 2016 (Fanusie & Robinson, 2018). A ban then would be a blunt measure for a problem that can be solved within the Bitcoin technology, the blockchain. The blockchain is comparable to a ledger, where every transaction is logged and recorded in the blockchain. It is irreversible and the basis on which Bitcoin functions. Even though the users are largely anonymous, their personal wallet address is enough for law enforcers to track certain individuals and force them to give up other individuals that are part of the paper trail to eventually find the main culprit. It is a matter of understanding the technology and employing law enforcers that are able to tackle this problem. In fact, one should consider the economic advantages Bitcoin brings such as economic growth and the many businesses and jobs it creates as a consequence before banning it altogether
(Kaplanov, 2012).
The steps towards further regulation on an international level must also be considered. For example, the IMF’s main goals are to prevent countries from artificially devaluating a currency for their own advantage and to prevent currencies from having unstable exchange rates. The main concern regarding Bitcoin is the latter, and the IMF must protect nations from a possible speculative attack coming from Bitcoin. One way for central banks to counteract a speculative attack is to directly intervene in the foreign exchange market where central banks buy up weak currency with stronger currency at a fixed rate to counteract the depreciation. This requires reserves of the stronger currency which can be supplied by the IMF in emergencies. However, when an attack is being undertaken with a currency that is strong but not a member of the IMF like Bitcoin, this becomes a problem. Central banks are left on their own and their only possibility without reserves is to raise interest rates, only further damaging their own economy. The only way to counter this is to incorporate Bitcoin into the IMF, despite its non-member status and lack of centralized representation. This could be done by altering the definition of separate currencies described in Articles of Agreement of the International Monetary Fund Article IV section 5 to include not only currencies of colonies and other such territories but also cryptocurrencies. This way, some of the quota or contribution payed by member states would have to be paid in Bitcoin creating a bitcoin fund under IMF control. A second option would be that the IMF would directly buy bitcoin from its users, giving the digital currency more legitimacy and becoming more of an established currency. This would be much more difficult since it would have to buy it from exchanges like any other investor. On top of that, no Bitcoin user would benefit from selling their investment since the increased legitimacy of the IMF holding Bitcoin could raise Bitcoins value making it more profitable to hold on (Plassaras, 2013).
2.3.ICO’s
After seeing how Bitcoin could potentially be regulated and already is being regulated, an aspect of cryptocurrencies that has virtually no regulation but very much needs it is Initial Coin Offerings. ICO’s are similar to IPO’s as in they are used as an easy way to raise money for a start-up or a new venture but instead of releasing
shares of company, crypto-coins are sold as a token or certificate of ownership interest in the enterprise. The crypto-coins can appreciate in value if the company is doing good just like a regular share. These newly issued coins can be purchased with other forms of cryptocurrency making the whole thing completely anonymous. ICO’s have amassed 2.3 billion USD in 2017, over 10 times more than the year before, and as of writing, 50 ICO’s are taking place every month (Cohan, 2017).
According to Chohan, ICO’s create several regulatory and other risks within the accountability domain as they at the very top of technological progress but are similar to those concerns raised in the beginning of Bitcoin. Even if there are no fraudulent intentions with a certain ICO, there is still a lack of protection for investors. Due to the high risk and speculative aspects of ICO’s, the downside potential is significant especially for investors that cannot lose much. More regulation is needed to transform ICO’s into a cheap but most importantly reliable way of raising capital. Safer environment means more institutional investors investing on more stable horizons, bigger volumes and over more different tokens. With more accountability, ICO’s could thrive in the near future according to the SEC.
Seven regulators in China, banned ICO’s and demanded a refund to all investors or risk being sued. Before the ban, China raised over 400 million USD from over 100 000 investors, representing a significant part of the overall market (Chohan, 2017). However, the reality shows that cryptocurrencies are hard to kill in China as Chinese investors keep buying Bitcoin and ICO’s in the private over the counter market. Chinese authorities are having a hard time to keep cryptocurrencies in check only weeks after announcing the shutdown of public exchanges. Experts feared for a sharp drop in demand for Bitcoin given the large share that China occupies in the industry. The renminbi share of over-the-counter Bitcoin trading have gone from 5% before the ban to 20% in about a month according to the National Committee of Experts on Internet Financial Security. Chinese are finding alternatives to more mainstream exchanges such as the fairly liquid OTC market and mobile messaging platforms. It’s still possible for Chinese investors to partake in ICO’s if they already own Bitcoin or Ethereal but many non-Chinese platforms have already ramped up customer identification or deleted Chinese versions of their websites. Overall, Bitcoin demand in China is rising because of the Chinese government requirements that
demand ICO investors to be paid back in Bitcoin by the respective ICO sponsor in order to avoid prosecution (Wildau, 2017).
Recently, on April 1st 2017, Japan recognized virtual currencies as a legal
payment method. They updated a portion of their Banking act to include virtual currencies (Japan Financial Services Agency, 2017). The Virtual Currency act as its being dubbed doesn’t recognize cryptocurrencies as an official currency, however, it does accept them as legal tender, a viable way for consumers to pay for their goods and services. That means, Bitcoin is still being treated like an asset with certain aspects shared with that of a legally recognized currency (Chohan, 2017).
2.4.Potential Difficulties
Before a regulatory event study can be performed, it should be known what difficulties can occur. Binder (1981) analysed the effectiveness of the use of stock returns for such an event study. He specified that to measure the effects of regulatory changes, it is not always clear when expectations and implicitly stock prices change. These regulatory processes have multiple announcements and parliamentary discussions before actual legislation can be passed. This gives most firms enough time to anticipate the oncoming changes and the ultimate effect might not be more gradual as investors adapt expectations. Relating this to Bitcoin, a highly volatile stock essentially, a brand new industry with little legislation, new announcements can have a drastic effect on the further existence of cryptocurrencies.
Secondly, Binder says that it is not clear whether the effects of legislation are positive or not, as some stocks might gain from the new laws and other might lose. Bitcoin being such a particular security, the price being mainly driven by demand for transaction, and publicity (Kristofer, 2015), it comes as no surprise that legislation that increases demand or in other words, makes Bitcoin more within reach of regular investors or makes transactions easier, will drive up the price and vice versa. Furthermore, any new legislation that is covered by media outlets will also affect the price due to the publicity.
3.Methodology
September 4th, where the People’s Bank of China outlawed ICO’s outright is the day
I will consider the most important but considering the events that followed this announcement such as the closing of the biggest exchange of China BTCC and possible increased speculation beforehand, the event window will span from August 25th to September 14th. As follows, I will use an estimation window spending over
nearly a year before the events, namely from August 30th 2016 to August 24th 2017
with 257 observations.
Secondly, the day where Japan officially recognized Bitcoin as legal tender on April 1st 2017 will also be used as an important event, this time to see the effects of a
positive law going into effect. The event window will span from March 23rd to April
12th and the estimation window will be from August 30th 2016 to March 22nd, with 146
observations.
Since Bitcoin can be traded day and night, all year long but exchanges where indices and stocks are traded are only online during the week, the daily weekend returns that Bitcoin earns will be omitted and instead I will consider one single return over the weekends. The S&P500 and Shanghai Stock Exchange (SSE) price index needed to perform this analysis was retrieved from the DataStream add-in for Excel and the BTC/USD exchange rate from cryptodatasets.com. Specifically, the BTC/USD exchange price comes from Bitfinex Bitcoin exchange. I chose both the S&P500 and SSE since I am analysing specific events in Asia about a virtual currency traded globally.
To measure any abnormal returns during the event window, I will first use the constant mean model as described by MacKinlay (1997). To calculate the expected return, I use the following formula,
!"#$,# = ("#$+ *"#$,# +,-(*"#$,#) = 01(*
"#$,#), 2 *"#$,# = 0
where !"#$,# is the price return of Bitcoin on day t, ("#$ stands for the mean daily return
on Bitcoin and it’s assumed that on average, the daily return equals the average return
plus some error term *"#$,# that has mean zero and variance of 01(*
"#$,#).
MacKinlay says that despite the simple nature of this model, it is a robust model that often gets similar results to more complex models. In general, the variance
of abnormal returns does not change drastically with more intricate models making it a good benchmark.
Even though the Constant Mean Model provides results relatively easy, for potentially more accurate results one must turn to more refined models such as the Market Model. The Market Model eliminates any variation in the price of Bitcoin that might be due to variations in the market, leading to a smaller variance and in theory, better results over the Constant Mean Model. This though depends on the R-squared of the model (MacKinlay, 1997). The Market Model relates the return from the market to the return on Bitcoin assuming returns as normally distributed. By regressing the daily market returns on the daily Bitcoin returns during the estimation window I discover the coefficients 4"#$ and 5"#$ in normal times.
!"#$,# = 4"#$+ 5"#$∗ !7,#+ *"#$,# (MacKinlay, 1997). +,-(*"#$,#) = 01(*
"#$,#), 2 *"#$,# = 0
Here again !"#$,# is the price return of Bitcoin on day t, !7,# stands for either the
S&P500 daily return or that of the Shanghai Stock Exchange. Here too we have an
error term *"#$,# that has mean zero and variance of 01(*
"#$,#). 4"#$ , 5"#$ and 01(*"#$,#),
are the parameters of the model. In theory, this model eliminates the possible variance caused by a systematic crash or boom in the markets but as stated before, Bitcoin’s inherent high volatility compared to regular markets the results of this model are rather similar to the constant mean model.
Next, I turn my attention the event window. The abnormal returns are measured by subtracting the expected return as measured with either
2 !"#$,# = 4"#$+ 5"#$∗ !7,#+ *"#$,# Or
2 !"#$,# = ("#$+ *"#$,#
from the actual return happening on that day Rbtc,t.
8!"#$,# = !"#$,#− (4^"#$+ 5^"#$∗ !7,#)
Or
8!"#$,# = !"#$,#− (("#$)
Now I have the abnormal return occurring during the event window and the event day, for which I have to test for significance using a t-test. The abnormal returns
are essentially the error terms since we are looking at the difference between actual and predicted returns from the model and therefore when I test for significance, I use the variance of the error term. Generally, the variance of the abnormal returns is calculated with the following formula,
01 8! "#$,# = 01(*"#$,#) + 1 <=[1 + !7,#− (7 1 071 ]
where !7,# stand for the period-t return on the market, (7 the average return and 071
the variance of the market returns. <= represents the number of observations in the estimation window, which means that with enough observations, the right terms go to zero so that we can ignore the right term altogether (MacKinlay, 1997).
4. Hypothesis
My hypothesis has two distinct parts. First of all, I want to analyse whether “Good” forms of regulation lead to an increase in the perceived value and demand, and therefore a strong positive return. Good laws can be for example laws that make cryptocurrencies easier and safer to use or to trade, give more privacy and autonomy to its users.
The other side of my hypotheses is that “bad” regulation will have a negative effect on the perceived value, and a strong negative return. Bad laws can be laws that should restrict the use or trade of cryptocurrencies either partly or fully, restrict the privacy of users or limit their autonomy. The alternative hypothesis then automatically is that any form of regulations be it good or bad, will have no effect on the perceived value of cryptocurrencies and therefore not lead to significant abnormal return.
5.Findings
Consider the Japans law going into effect in April first. This law can be seen as a good form of regulation, something that will benefit the use of cryptocurrencies. One would expect that these laws are seen as value increasing as Bitcoin can be used for more mundane things in Japan such as buying goods and services and not merely being a highly speculative asset.
Looking at the Bitcoin US dollar exchange rate in the constant mean model we see two significant spikes in abnormal returns 5 to 4 days before the event. -9.16%
the 5th day and 7.22% the 4th day (Table 1a), we get t-values of -2.26 and 2.86
respectively. The Bitcoin Chinese Yuan exchange rate gives even more significant values with spikes of -9.23% and 10.00% (Table 1b) and t-values of -2.71 and 2.93 respectively. For a one-sided test, the 5.86% abnormal return on the event day is also significant at 95% confidence (t-value: 1.72).
The market model as expected gives similar results. Again, event day 5 and -4 give similar spikes both on the S&P500 and SSE. Table 2a depicts the S&P500 and shows spikes of -9.23% and 7,14% with t-values of -2.88 and 2.23 respectively. When Bitcoin is regressed on the Shanghai Stock Exchange returns we see spikes of -9.49% and 10.05% with respective t-values -2.78 and 2.94. For the event day, again a one-sided test at 95% has a significant abnormal return of 5.88% (t-value: 1.72).
All of this indicates these spikes are indeed significant yet cannot be fully attributed to the Japanese Law going into effect. We may assume these effects can be attributed to investors incorporating the news early into their beliefs yet the fact that they are seemingly random going up and down is more likely down to some other reason, possibly a pump and dump or even something as simple as a news coverage on national tv in one of the major markets such as the US.
Now let’s look at the flipside and consider China’s ICO ban. Bad forms of regulation don’t need to be permanent in nature and it very well may be that China reverses the verdict in the future, however these last resort form of regulation usually is due to the incapacity of a government to properly regulate these markets and specifically punishing those that are doing wrong therefore resorting to an industry wide ban until better alternatives are found.
Both the BTC/USD exchange rate and the BTC/CNY rate show strong evidence that the ban has major downside effects on the value of Bitcoin. Compared to the USD, Bitcoins abnormal return sits at -9.25% (t-value: -2.26) on the day of the announcement, and compared to the Chinese Yuan, the dip is even larger, -15.83%
(t-value: -3.55). On top of that, on September 14th, the biggest exchange in China,
BTCC, closed its doors causing the second big dip of -16.66% (t-value: -4.08) compared to USD and -16.53% (t-value: -3.70) compared to the CNY (Table 1).
And again, the market model does not differ in results. Large dips happening compared to the markets on the event day itself. Bitcoin drops -9.27% (t-value: -2.26) against the S&P500 and 15.80% (t-value: -3.53) compared to the SSE. And too for the closing of the BTCC exchange the abnormal returns are -16.70% (t-value: -4.08) and -16.56% (t-value: -3.70) for both markets respectively.
The cumulative abnormal returns also paint an interesting picture. When looking at the “BTC/CNY CAR CM-Model Good Regulation” graph, it is clear that after the weekend of the announcement, Bitcoins abnormal returns have a steadily upwards trend, indicating a possible small effect of the new Japanese law taking effect on the value of Bitcoin. Same goes for the Chinese ban where we see that after the announcement, the abnormal returns take a nosedive due to the new ban itself and the aforementioned exchange closing down due to the new law.
The results illustrate an interesting phenomenon that can be explained by the famous research of Kahneman, Knetsch, and Thaler (1991) about anomalies. They state that utility is not perceived as an absolute value based on wealth or welfare, but instead, actors mainly perceive relative changes in their utility compared to some neutral reference point. On top of that, losses in utility are perceived as larger and more devastating than gains or improvements of the same size. This effect is called loss aversion and it leads to the fact that investors prefer not losing nor gaining instead of gaining but also having a significant chance of losing a lot of money. When the Chinese government announced their new policy, there was a strong incentive of the regulatory institutions to attack all illegal activity with whatever means necessary, and many feared losing their entire investment. This fear caused massive sell offs in China simply to safe guard their investment from possibly being compromised at a later date. Rather than running the possibility of losing their entire investment, even if Bitcoin could and eventually did retain its value, investors rather played it safe and massively cashed in their holdings. At the same time, when Japan announced their law accepting Bitcoin as legal tender, even though it benefitted users, Japanese citizens without holdings weren’t necessary more inclined to massively buy Bitcoin.
Lastly, the two models gave very similar results and the Market Model was not effective in giving more insight on the relative movement of the market compared to Bitcoin. All four regressions with the Market Model revealed an R-squared less than
1% indicating that the market trend had virtually no impact on the price of Bitcoin. This is likely due to the extreme seemingly random volatility of Bitcoin compared to the markets.
6.Conclusion
This analysis was conducted to determine if new legislation had any effects on the value of cryptocurrencies and in particular Bitcoin. The objective was to find out if and how it did in order to have a clear understanding what the consequences are of certain government policies.
To do this, I analysed two periods around the dates where influential governments announced new legislation concerning cryptocurrencies. I chose one event that should have a positive impact and one where the effects should be negative, that the Japanese update to the Banking act recognizing Bitcoin as legal tender and the other being the Chinese initial coin offering ban. I calculated returns in normal periods through the constant mean model and market model. With these normal values, I was able to calculate any abnormal return around the date of the announcements (MacKinlay, 1997).
What I found was that positive legislation had a somewhat weak effect on returns, showing significant abnormal returns at a 95% confidence one sided t-test on the event date. This effect only showed when looking at Bitcoin to Chinese Yuan exchange rate or the market model including the Shanghai Stock Exchange. Negative legislation had a much more dramatic effect. Bitcoin returns dropped heavily on the day of the event and subsequent days taking another big hit when the biggest Chinese Bitcoin exchange closed.
I attributed this to fact to loss aversion, a concept thought of by Kahneman, Knetsch, and Thaler (1991). Much like gambling, trading in Bitcoin can earn investors big returns in short periods but shocks to the bitcoin system can turn those gains around quickly. Loss aversion in few words means people like earning money but they hate losing money much more. This is exactly what we see in Bitcoin trading. A negative shock causing a drop in price causes much bigger sell offs than a positive shock attracts new investors or more investing by current users.
How can governments use this information? This depends, if governments believe the blockchain technology should be nurtured and further explored in the future (Ansip, 2017) then it is clear now that negative laws should be used with prudence and the focus should primarily be on making it safer and more accessible to regular civilians. The instability and fear is not contributing to a prosperous future and should be avoided at all cost. Yet, when governments cannot oversee the situation and need to act quickly, sometimes there is no other option than to ban the use of cryptocurrencies outright. The biggest issues are still is a lack of supranational law making and enforcing, which could be the next step in incorporating Bitcoin into everyday life. That would mean either great changes to existing national institution or the creation of new ones. Moreover, as cryptocurrencies might play a bigger and bigger role in everyday life, governments should employ the right people to oversee these developments and make sure it is guided safely to a stable future.
I suggest further research is needed to analyse the effects not only on Bitcoin but on other cryptocurrencies as every cryptocurrency has a different purpose and some might be less prone to changes due to legislation. Another suggestion is that as time goes by and more legislation is implemented, whether the instability and shocks still persist, how they affect prices and if eventually, more regulation means more stability and therefore making it a safer investment and currency which should be the eventual goal of regulation.
Table 1a : Constant Mean Model
BTC/USD
Good Regulation Bad Regulation
Days to Event AR CAR AR CAR
-6 -1,00859% -1,00859% -0,01599% -0,01599% -5 -9,16288%** -10,17147% 0,51086% 0,49488% -4 7,22422%** -2,94725% 3,91030% 4,40517% -3 -0,20086% -3,14811% -1,11260% 3,29258% -2 -0,67026% -3,81837% 2,59407% 5,88665% -1 -0,45026% -4,26863% 3,31795% 9,20460% 0 3,47787% -0,79077% -9,25335%** -0,04875% 1 3,45383% 2,66307% 3,34660% 3,29784% 2 -0,76655% 1,89652% 4,35065% 7,64849% 3 -0,86395% 1,03256% -0,30672% 7,34177% 4 4,08822% 5,12078% -7,40883% -0,06706% 5 0,03536% 5,15614% -1,52759% -1,59465% 6 -0,03080% 5,12535% -1,87038% -3,46504% 7 0,86112% 5,98647% -7,91894% -11,38398% 8 -1,05631% 4,93016% -16,65590%** -28,03987%
** = significant abnormal return at 97.5% confidence one-sided * = significant abnormal return at 95% confidence one-sided
Bitcoin to US dollar exchange rates come from cryptodatasets.com, prices are retrieved from Bitfinex exchange and collected on this website.
The US dollar to Chinese Renminbi and both the S&P500 and Shanghai Stock Exchange come from Datastream Add-in from Excel
Table 1b : Constant Mean Model
BTC/CNY
Good Regulation Bad Regulation
AR CAR AR CAR -1,09090% -1,09090% -0,14286% -0,14286% -9,23905%** -10,32996% -0,59224% -0,73510% 10,00118%** -0,32877% 3,17397% 2,43887% -0,14470% -0,47347% -1,37087% 1,06801% -0,70187% -1,17534% 2,47713% 3,54513% -0,53110% -1,70644% 2,68585% 6,23098% 3,40937% 1,70293% -15,82627%** -9,59529% 5,86228% 7,56521% 3,66013% -5,93516% -0,85464% 6,71057% 3,73342% -2,20174% -0,89283% 5,81774% -0,86031% -3,06205% 4,06528% 9,88301% -8,11120% -11,17325% -0,07310% 9,80991% -2,43221% -13,60545% 1,59298% 11,40289% -1,91684% -15,52230% 0,74224% 12,14513% -8,13592% -23,65822% -1,29262% 10,85251% -16,52493%** -40,18314% ** = significant abnormal return at 97.5% confidence one-sided
* = significant abnormal return at 95% confidence one-sided
Bitcoin to US dollar exchange rates come from cryptodatasets.com, prices are retrieved from Bitfinex exchange and collected on this website.
The US dollar to Chinese Renminbi and both the S&P500 and Shanghai Stock Exchange come from Datastream Add-in from Excel
Table 2a : Market Model
S&P500
Good Regulation Bad Regulation
Days to Event AR CAR AR CAR
-6 -1,08933% -1,08933% 0,01853% 0,01853% -5 -9,23270%** -10,32202% 0,51171% 0,53024% -4 7,14553%** -3,17649% 3,92125% 4,45149% -3 0,13838% -3,03812% -0,99453% 3,45696% -2 -0,64258% -3,68070% 2,74354% 6,20051% -1 -0,32913% -4,00983% 3,36126% 9,56177% 0 3,33676% -0,67307% -9,26633%** 0,29544% 1 3,34369% 2,67062% 3,11920% 3,41463% 2 -0,76544% 1,90518% 4,42651% 7,84114% 3 -1,04548% 0,85970% -0,32478% 7,51637% 4 4,15854% 5,01824% -7,46409% 0,05228% 5 -0,03360% 4,98464% -1,23278% -1,18050% 6 -0,02321% 4,96143% -1,78784% -2,96835% 7 0,76150% 5,72293% -7,91043% -10,87877% 8 -1,27346% 4,44947% 16,70014%** -27,57891%
-** = significant abnormal return at 97.5% confidence one-sided * = significant abnormal return at 95% confidence one-sided
Bitcoin to US dollar exchange rates come from cryptodatasets.com, prices are retrieved from Bitfinex exchange and collected on this website.
The US dollar to Chinese Renminbi and both the S&P500 and Shanghai Stock Exchange come from Datastream Add-in from Excel
Table 2 b : Market Model
Shanghai Stock Exchange
Good Regulation Bad Regulation
AR CAR AR CAR -1,12046% -1,12046% 0,00050% 0,00050% -9,49422%** -10,61469% -0,52034% -0,51984% 10,04916%** -0,56553% 3,17810% 2,65826% 0,05253% -0,51300% -1,37692% 1,28134% -0,53537% -1,04837% 2,46832% 3,74966% -0,11099% -1,15936% 2,69871% 6,44837% 3,26464% 2,10528% -15,79891%** -9,35054% 5,87867% 7,98395% 3,66905% -5,68149% -0,83825% 7,14570% 3,73378% -1,94771% -1,49939% 5,64631% -0,90897% -2,85668% 3,94328% 9,58959% -8,11442% -10,97110% -0,12886% 9,46073% -2,40820% -13,37930% 1,83034% 11,29107% -1,91156% -15,29086% 0,50576% 11,79683% -8,12715% -23,41801% -1,08207% 10,71476% -16,55672%** -39,97473%
** = significant abnormal return at 97.5% confidence one-sided * = significant abnormal return at 95% confidence one-sided
Bitcoin to US dollar exchange rates come from cryptodatasets.com, prices are retrieved from Bitfinex exchange and collected on this website.
The US dollar to Chinese Renminbi and both the S&P500 and Shanghai Stock Exchange come from Datastream Add-in from Excel
-0.15 -0.1 -0.05 0 0.05 0.1 0.15 3/ 23/ 17 3/ 24/ 17 3/ 25/ 17 3/ 26/ 17 3/ 27/ 17 3/ 28/ 17 3/ 29/ 17 3/ 30/ 17 3/ 31/ 17 4/ 1/ 17 4/ 2/ 17 4/ 3/ 17 4/ 4/ 17 4/ 5/ 17 4/ 6/ 17 4/ 7/ 17 4/ 8/ 17 4/ 9/ 17 4/ 10/ 17 4/ 11/ 17 4/ 12/ 17
BTC/CNY CAR CM-Model Good Regulation
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 8/ 25/ 17 8/ 26/ 17 8/ 27/ 17 8/ 28/ 17 8/ 29/ 17 8/ 30/ 17 8/ 31/ 17 9/ 1/ 17 9/ 2/ 17 9/ 3/ 17 9/ 4/ 17 9/ 5/ 17 9/ 6/ 17 9/ 7/ 17 9/ 8/ 17 9/ 9/ 17 9/ 10/ 17 9/ 11/ 17 9/ 12/ 17 9/ 13/ 17 9/ 14/ 17BTC/CNY CAR CM-model Bad Regulation
References
Ansip, A. (2017). Answer given by Vice-President Ansip on behalf of the
Commission. Retrieved from
http://www.europarl.europa.eu/sides/getAllAnswers.do?reference=E-2016-009012&language=EN
Binder, J. J. (1985). Measuring the effects of regulation with stock price data. The RAND Journal of Economics, 167-183.
Brito, J., & Castillo, A. (2013). Bitcoin: A primer for policymakers. Mercatus Center at George Mason University.
Brunsden, J., & Murphy, H. (2017). EU financial regulation chief warns of signs of Bitcoin bubble. Financial Times. Retrieved from https://www.ft.com/content/7d4cebd8-e59d-11e7-97e2-916d4fbac0da
Chohan, U. (2017). Initial Coin Offerings (ICOs): Risks, Regulation, and Accountability.
Fanusie, Y., & Robinson, T. (2018). Bitcoin Laundering: An Analysis of Illicit Flows into Digital Currency Services. Defenddemocracy.org. Retrieved 27 January 2018, from http://www.defenddemocracy.org/media-hit/yaya-j-fanusie-bitcoin-laundering/
Global Legal Research Directorate Staff. (2014). Regulation of Bitcoin in Selected Jurisdictions. The Law Library of Congress. Retrieved from http://www.loc.gov/law/help/Bitcoin-survey/regulation-of-Bitcoin.pdf
Japan Financial Services Agency. (2017). About the result of public comment on "Cabinet Order etc. (draft), etc. revising part of enforcement orders of banking act etc.". Financial Services Agency.
Finnis, J. (2011). Natural law and natural rights. Oxford University Press.
Jung-a, S. and Harris, B. (2018). Bitcoin tumbles as South Korea plans trading ban. The Financial Times. [online] Available at: https://www.ft.com/content/0d5ff7d4-f67d-11e7-88f7-5465a6ce1a00 [Accessed 17 Jan. 2018].
Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1991). Anomalies: The endowment effect, loss aversion, and status quo bias. The journal of economic
perspectives, 5(1), 193-206.
the case against its regulation. Loy. Consumer L. Rev., 25, 111.
Kristoufek, L. (2015). What are the main drivers of the Bitcoin price? Evidence from wavelet coherence analysis. PloS one, 10(4), e0123923.
MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of economic literature, 35(1), 13-39.
Plassaras, N. A. (2013). Regulating digital currencies: bringing Bitcoin within the reach of IMF. Chi. J. Int'l L., 14, 377.
Rabinovitch, S. (2013). China bans banks from Bitcoin transactions. Financial Times. Retrieved from https://www.ft.com/content/40b78d2e-5d87-11e3-95bd-00144feabdc0
Schwert, G. W. (1981). Using financial data to measure effects of regulation. The Journal of Law and Economics, 24(1), 121-158.
Sotiropoulou, A., & Guégan, D. (2017). Bitcoin and the challenges for financial regulation. Capital Markets Law Journal, 12(4), 466-479.
Wildau, G. (2017). Bitcoin proves hard to kill in China. The Financial Times. Retrieved from https://www.ft.com/content/d576e4e4-c374-11e7-a1d2-6786f39ef675
timestamp close USD CNY BTC/CNY BTC/CNY return SSE SSE Return 30/08/2016 579,49 6,6799 3870,93525 3218,71 31/08/2016 576,15 6,6783 3847,70255 -0,006001833 3230,0510,00352346 01/09/2016 572,73 6,6799 3825,77913 -0,005697794 3206,791 -0,0072011 02/09/2016 579,85 6,6801 3873,45599 0,012462 3211,0450,00132656 05/09/2016 611,5 6,6752 4081,8848 0,053809522 3216,0230,00155027 06/09/2016 615,23 6,6803 4109,92097 0,006868437 3235,5090,00605904 07/09/2016 619,75 6,6644 4130,2619 0,004949227 3236,7730,00039066 08/09/2016 631,73 6,6639 4209,78555 0,019253899 3240,9690,00129635 09/09/2016 626,25 6,6799 4183,28738 -0,006294423 3223,024 -0,0055369 12/09/2016 611,62 6,6799 4085,56044 -0,023361277 3163,43 -0,0184901 13/09/2016 614,23 6,6799 4102,99498 0,004267356 3164,9960,00049503 14/09/2016 613,88 6,6712 4095,31626 -0,001871492 3143,277 -0,0068623 15/09/2016 611,81 6,6712 4081,50687 -0,003371995 3143,277 0 16/09/2016 610,01 6,6712 4069,49871 -0,00294209 3143,277 0 19/09/2016 609,79 6,6708 4067,78713 -0,000420587 3167,6320,00774828 20/09/2016 600,14 6,6714 4003,774 -0,015736599 3164,439 -0,001008 21/09/2016 597,43 6,6713 3985,63476 -0,004530535 3167,4450,00094993 22/09/2016 597,08 6,6697 3982,34448 -0,000825536 3184,6660,00543687 23/09/2016 603,29 6,67 4023,9443 0,010446064 3175,877 -0,0027598 26/09/2016 609,14 6,6699 4062,90289 0,009681691 3119,874 -0,0176339 27/09/2016 605,53 6,6682 4037,79515 -0,006179754 3138,475 0,0059621 28/09/2016 603,76 6,6715 4027,98484 -0,00242962 3127,588 -0,0034689 29/09/2016 604,6 6,6692 4032,19832 0,001046052 3138,6580,00353947 30/09/2016 611,1 6,67 4076,037 0,010872154 3145,1670,00207382 03/10/2016 612,67 6,67 4086,5089 0,002569138 3145,167 0 04/10/2016 610,98 6,67 4075,2366 -0,002758418 3145,167 0 05/10/2016 614,09 6,67 4095,9803 0,005090183 3145,167 0 06/10/2016 613,51 6,67 4092,1117 -0,000944487 3145,167 0 07/10/2016 620,13 6,67 4136,2671 0,01079037 3145,167 0 10/10/2016 618,87 6,7027 4148,09995 0,002860756 3190,7220,01448413 11/10/2016 641,87 6,7149 4310,09286 0,039052317 3208,6680,00562443 12/10/2016 637,63 6,7137 4280,85653 -0,006783226 3201,592 -0,0022053 13/10/2016 637,01 6,7299 4287,0136 0,001438279 3204,5310,00091798 14/10/2016 643 6,7254 4324,4322 0,008728361 3207,1590,00082009 17/10/2016 639,79 6,7394 4311,80073 -0,002920955 3184,153 -0,0071733 18/10/2016 638,68 6,7392 4304,19226 -0,001764569 3228,7320,01400027 19/10/2016 631,77 6,7393 4257,68756 -0,010804512 3229,6810,00029392 20/10/2016 632,46 6,7397 4262,59066 0,001151588 3229,367 -9,722E-05 21/10/2016 636,73 6,7597 4304,10378 0,009738941 3236,3110,00215027 24/10/2016 654,65 6,7716 4433,02794 0,029953776 3275,2650,01203654 25/10/2016 659,52 6,7778 4470,09466 0,008361489 3279,1020,00117151 26/10/2016 678,7 6,7687 4593,91669 0,027700092 3262,725 -0,0049944 27/10/2016 688,67 6,7779 4667,73639 0,016069012 3258,559 -0,0012768 28/10/2016 693,47 6,7797 4701,51856 0,007237377 3250,156 -0,0025787 31/10/2016 701,02 6,7708 4746,46622 0,009560242 3246,25 -0,0012018 01/11/2016 734,6 6,7749 4976,84154 0,048536177 3269,240,00708202 02/11/2016 750,85 6,761 5076,49685 0,020023806 3248,57 -0,0063226 03/11/2016 692,51 6,761 4682,06011 -0,077698608 3276,060,00846218 04/11/2016 707,62 6,7591 4782,87434 0,021532024 3272,24 -0,001166 07/11/2016 705,55 6,7758 4780,66569 -0,000461783 3280,65 0,0025701 08/11/2016 711,99 6,7796 4827,0074 0,009693569 3295,920,00465457 09/11/2016 724,54 6,773 4907,30942 0,016635984 3275,47 -0,0062046 10/11/2016 714,87 6,7923 4855,6115 -0,010534881 3320,40,01371712 11/11/2016 716,22 6,8156 4881,46903 0,005325288 3346,310,00780328 14/11/2016 707,43 6,8408 4839,38714 -0,008620743 3361,35 0,0044945 15/11/2016 712,17 6,8522 4879,93127 0,008377947 3357,77 -0,001065 16/11/2016 744,98 6,8695 5117,64011 0,048711513 3355,71 -0,0006135 17/11/2016 740,67 6,87 5088,4029 -0,005713026 3359,30,00106982 18/11/2016 753,97 6,8912 5195,75806 0,021098008 3342,99 -0,0048552 21/11/2016 738,99 6,8961 5096,14894 -0,01917124 3369,53 0,007939 22/11/2016 749,85 6,888 5164,9668 0,013503895 3401,190,00939597 23/11/2016 742 6,8928 5114,4576 -0,009779192 3393,66 -0,0022139 24/11/2016 737,61 6,9184 5103,08102 -0,002224395 3394,320,00019448 25/11/2016 740,36 6,9151 5119,66344 0,00324949 3415,540,00625162 28/11/2016 731,52 6,9022 5049,09734 -0,013783346 3431,350,00462884 29/11/2016 731,05 6,8914 5037,95797 -0,002206211 3437,750,00186516 30/11/2016 739 6,887 5089,493 0,010229349 3403,22 -0,0100444 01/12/2016 755,36 6,8949 5208,13166 0,023310507 3427,640,00717556 02/12/2016 774,88 6,8826 5333,18909 0,024011955 3396,71 -0,0090237 05/12/2016 750,62 6,886 5168,76932 -0,03082954 3355,63 -0,0120941 06/12/2016 757,36 6,8762 5207,75883 0,007543287 3350,25 -0,0016033 07/12/2016 765,01 6,8863 5268,08836 0,011584548 3373,920,00706514 08/12/2016 765,01 6,8792 5262,65679 -0,001031033 3366,7 -0,0021399 09/12/2016 770,5 6,9005 5316,83525 0,010294887 3385,150,00548014 12/12/2016 777,99 6,9152 5379,95645 0,011871949 3301,41 -0,0247375 13/12/2016 775 6,9015 5348,6625 -0,005816766 3303,520,00063912 14/12/2016 774,49 6,9049 5347,776 -0,000165742 3288,25 -0,0046223 15/12/2016 777,43 6,9356 5391,94351 0,008259042 3264,22 -0,0073078 16/12/2016 784,17 6,9545 5453,51027 0,011418287 3269,750,00169413 19/12/2016 790,59 6,9452 5490,80567 0,006838788 3264,67 -0,0015536 20/12/2016 797,99 6,9509 5546,74869 0,010188491 3248,78 -0,0048673 21/12/2016 829,34 6,9513 5764,99114 0,039346014 3285,040,01116111 22/12/2016 859,2 6,9467 5968,60464 0,035318961 3287,220,00066361 23/12/2016 918,99 6,9496 6386,6129 0,070034504 3256,41 -0,0093727 26/12/2016 906,4 6,9496 6299,11744 -0,013699823 3269,440,00400134 27/12/2016 936,43 6,9497 6507,90757 0,033145934 3261,11 -0,0025478 28/12/2016 981,7 6,9548 6827,52716 0,049112497 3248,14 -0,0039772 29/12/2016 974,74 6,9545 6778,82933 -0,007132572 3241,67 -0,0019919 30/12/2016 959,26 6,9495 6666,37737 -0,016588699 3249,590,00244319 02/01/2017 1019,3 6,9495 7083,62535 0,062589913 3249,59 0 03/01/2017 1037,5 6,9557 7216,53875 0,018763471 3283,450,01041978 04/01/2017 1139,6 6,9485 7918,5106 0,09727265 3307,450,00730938 05/01/2017 1003,2 6,8817 6903,72144 -0,128154045 3314,390,00209829 06/01/2017 898,5 6,923 6220,3155 -0,098990949 3302,79 -0,0034999 09/01/2017 903 6,9348 6262,1244 0,006721347 3320,530,00537122 10/01/2017 905,76 6,924 6271,48224 0,001494355 3310,49 -0,0030236 11/01/2017 779,54 6,9266 5399,56176 -0,13902941 3284,37 -0,0078901 12/01/2017 804,58 6,9018 5553,05024 0,0284261 3266,04 -0,005581 13/01/2017 828,12 6,8992 5713,3655 0,028869766 3259,27 -0,0020728 16/01/2017 830,1 6,8976 5725,69776 0,002158492 3249,77 -0,0029148 17/01/2017 903,99 6,8728 6212,94247 0,085097875 3255,330,00171089 18/01/2017 887,46 6,8342 6065,07913 -0,023799245 3259,770,00136392 19/01/2017 900,29 6,8766 6190,93421 0,020750773 3247,48 -0,0037702 20/01/2017 895,74 6,8755 6158,66037 -0,005213081 3270,330,00703623 23/01/2017 908,52 6,8555 6228,35886 0,011317151 3284,57 0,0043543 24/01/2017 886,1 6,8575 6076,43075 -0,02439296 3290,620,00184195 25/01/2017 893,35 6,8798 6146,06933 0,011460442 3297,960,00223058 26/01/2017 915,12 6,878 6294,19536 0,024100937 3308,06 0,0030625 27/01/2017 916,7 6,878 6305,0626 0,00172655 3308,06 0 30/01/2017 917,35 6,878 6309,5333 0,000709065 3308,06 0 31/01/2017 966,19 6,878 6645,45482 0,053240312 3308,06 0 01/02/2017 983,73 6,878 6766,09494 0,018153779 3308,06 0 02/02/2017 1007 6,878 6926,146 0,023654865 3308,06 0
03/02/2017 1015,7 6,874 6981,9218 0,008052934 3288,17 -0,0060126 06/02/2017 1022,6 6,8643 7019,43318 0,005372644 3305,80,00536164 07/02/2017 1052,1 6,8808 7239,28968 0,031321119 3301,7 -0,0012402 08/02/2017 1048,8 6,8788 7214,48544 -0,003426336 3316,270,00441288 09/02/2017 984,97 6,8672 6763,98598 -0,06244374 3333,230,00511418 10/02/2017 992 6,883 6827,936 0,009454487 3347,390,00424813 13/02/2017 996,5 6,878 6853,927 0,003806568 3368,460,00629446 14/02/2017 1013,3 6,8674 6958,73642 0,015291879 3369,60,00033843 15/02/2017 1013,9 6,8677 6963,16103 0,000635835 3364,4 -0,0015432 SUMMARY OUTPUT 16/02/2017 1038,5 6,8586 7122,6561 0,022905555 3381,87 0,0051926 mean return 0,00480759 17/02/2017 1056,2 6,8704 7256,51648 0,018793604 3352,96 -0,0085485 Regression Statistics variance btc 0,00116354 20/02/2017 1091,2 6,8782 7505,49184 0,03431059 3392,65 0,0118373 Multiple R 0,07390506 21/02/2017 1129,6 6,8829 7774,92384 0,035897981 3406,650,00412657 R Square 0,00546196 22/02/2017 1125,5 6,878 7741,189 -0,004338929 3414,940,00243348 Adjusted R Square-0,0014446 23/02/2017 1189,8 6,8774 8182,73052 0,057037946 3404,59 -0,0030308 Standard Error 0,0341353 24/02/2017 1185,4 6,8723 8146,42442 -0,004436917 3406,790,00064619 Observations 146 27/02/2017 1195,5 6,8714 8214,7587 0,008388254 3380,82 -0,007623 28/02/2017 1189,1 6,8688 8167,69008 -0,005729763 3394,50,00404636 ANOVA 01/03/2017 1233,2 6,88 8484,416 0,038777906 3399,890,00158786 df SS MS F Significance F 02/03/2017 1258 6,8827 8658,4366 0,020510616 3382,21 -0,0052002 Regression 1 0,0009215 0,000921503 0,79084144 0,37532901 03/03/2017 1289,2 6,8989 8894,06188 0,027213375 3369,93 -0,0036308 Residual 144 0,16779153 0,001165219 06/03/2017 1279,2 6,893 8817,5256 -0,008605324 3386,210,00483096 Total 145 0,16871303 07/03/2017 1232,4 6,8983 8501,46492 -0,0358446 3395,20,00265489
08/03/2017 1150 6,907 7943,05 -0,065684553 3393,41 -0,0005272 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95,0% Upper 95,0%
09/03/2017 1190,4 6,91 8225,664 0,035580035 3368,32 -0,0073937 Intercept 0,00464369 0,00283106 1,640263768 0,10313297 -0,0009521 0,0102395 -0,0009521 0,0102395 10/03/2017 1115,4 6,9164 7714,55256 -0,062136192 3364,1 -0,0012529 X Variable 1 0,42086981 0,47326356 0,889292662 0,37532901 -0,5145711 1,35631077 -0,5145711 1,35631077 13/03/2017 1238,5 6,9092 8557,0442 0,109208102 3389,560,00756815 14/03/2017 1245 6,9144 8608,428 0,006004854 33920,00071986 15/03/2017 1256,2 6,9115 8682,2263 0,008572796 3394,590,00076356 16/03/2017 1168,6 6,8974 8060,30164 -0,071631934 3423,010,00837215 17/03/2017 1070,4 6,9085 7394,8584 -0,082558106 3389,95 -0,0096582 20/03/2017 1040,5 6,9042 7183,8201 -0,028538518 3403,970,00413575 21/03/2017 1115,9 6,897 7696,3623 0,071346748 3415,280,00332259 Market Model Constant mean return model
22/03/2017 1039,1 6,8869 7156,17779 -0,070186991 3398,11 -0,0050274 E [R] AR CAR AR t-test E[R] AR CAR AR t-test 23/03/2017 1032,7 6,8873 7112,51471 -0,006101453 3401,820,00109178 0,00510319 -0,0112046 -0,0112046 -0,32824204 0,00480759 -0,010909 -0,010909 -0,3198131 24/03/2017 942,13 6,8882 6489,57987 -0,087582925 3423,770,00645243 0,00735932 -0,0949422 -0,1061469 -2,78135061 0,00480759 -0,0923905 -0,1032996 -2,7085513 27/03/2017 1042,7 6,8762 7169,81374 0,104819401 3421,2 -0,0007506 0,00432777 0,10049163 -0,0056553 2,943920778 0,00480759 0,10001181 -0,0032877 2,93197987 28/03/2017 1044,7 6,8861 7193,90867 0,003360608 3406,5 -0,0042967 0,00283532 0,00052528 -0,00513 0,015388307 0,00480759 -0,001447 -0,0047347 -0,0424202 29/03/2017 1041,8 6,89 7178,002 -0,00221113 3394,35 -0,0035667 0,00314257 -0,0053537 -0,0104837 -0,15683762 0,00480759 -0,0070187 -0,0117534 -0,2057631 30/03/2017 1041,2 6,8905 7174,3886 -0,000503399 3361,79 -0,0095924 0,00060653 -0,0011099 -0,0115936 -0,0325157 0,00480759 -0,005311 -0,0170644 -0,1556987 31/03/2017 1081,5 6,8918 7453,4817 0,038901308 3374,660,00382832 0,00625491 0,03264639 0,02105279 0,956382122 0,00480759 0,03409372 0,0170293 0,99950294 03/04/2017 1150,1 6,8918 7926,25918 0,063430421 3374,66 0 0,00464369 0,05878673 0,07983952 1,722168075 0,00480759 0,05862283 0,07565213 1,71860664 04/04/2017 1145,8 6,8918 7896,62444 -0,003738805 3374,66 0 0,00464369 -0,0083825 0,07145702 -0,24556675 0,00480759 -0,0085464 0,06710574 -0,2505489 05/04/2017 1140,4 6,8959 7864,08436 -0,004120758 3424,610,01480149 0,01087319 -0,0149939 0,05646307 -0,4392505 0,00480759 -0,0089283 0,05817739 -0,2617464 06/04/2017 1191,5 6,9002 8221,5883 0,045460339 3435,870,00328797 0,0060275 0,03943284 0,09589591 1,155192388 0,00480759 0,04065275 0,09883014 1,19178969 07/04/2017 1196,6 6,8988 8255,10408 0,004076558 3441,760,00171427 0,00536517 -0,0012886 0,0946073 -0,03775024 0,00480759 -0,000731 0,09809911 -0,0214311 10/04/2017 1220,3 6,9051 8426,29353 0,020737407 3423,69 -0,0052502 0,00243403 0,01830338 0,11291067 0,536200758 0,00480759 0,01592982 0,11402893 0,46700392 11/04/2017 1235,6 6,903 8529,3468 0,012229964 3444,260,00600814 0,00717233 0,00505763 0,1179683 0,148164213 0,00480759 0,00742238 0,12145131 0,21759689 12/04/2017 1227,4 6,8927 8460,09998 -0,008118655 3428,37 -0,0046135 0,00270202 -0,0108207 0,10714763 -0,31699362 0,00480759 -0,0129262 0,10852506 -0,3789501 13/04/2017 1186,9 6,883 8169,4327 -0,034357428 3430,610,00065337 14/04/2017 1206,8 6,883 8306,4044 0,016766366 3399,31 -0,0091237 17/04/2017 1240 6,8874 8540,376 0,028167615 3374,24 -0,007375 18/04/2017 1265,4 6,8852 8712,53208 0,020157904 3347,48 -0,0079307 19/04/2017 1260,5 6,8869 8680,93745 -0,003626343 3320,23 -0,0081405 20/04/2017 1308,5 6,8839 9007,58315 0,037627929 3321,710,00044575 21/04/2017 1327 6,8831 9133,8737 0,01402047 3322,80,00032814 24/04/2017 1345 6,8862 9261,939 0,014020919 3277,11 -0,0137505 25/04/2017 1371,1 6,886 9441,3946 0,019375597 3282,380,00160812 26/04/2017 1400 6,8908 9647,12 0,021789726 3288,93 0,0019955 27/04/2017 1440,3 6,8923 9926,97969 0,029009662 3300,810,00361212 28/04/2017 1415,6 6,8974 9763,95944 -0,016421939 3303,440,00079677 01/05/2017 1533 6,8974 10573,7142 0,082933032 3303,44 0 02/05/2017 1558,5 6,8967 10748,507 0,016530875 3292 -0,0034631 03/05/2017 1619 6,8918 11157,8242 0,038081312 3283,19 -0,0026762 04/05/2017 1607,1 6,897 11084,1687 -0,006601242 3274,79 -0,0025585 05/05/2017 1545,1 6,9 10661,19 -0,038160616 3249,4 -0,0077532 08/05/2017 1703,5 6,9033 11759,7716 0,103044927 3223,88 -0,0078538 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 SSE CAR CM-Model Good Regulation
09/05/2017 1760 6,9055 12153,68 0,033496267 3225,880,00062037 10/05/2017 1796,9 6,9036 12405,0788 0,020684997 3196,82 -0,0090084 11/05/2017 1853,9 6,9035 12798,3987 0,031706353 3205,980,00286535 12/05/2017 1735 6,9029 11976,5315 -0,064216405 3229,130,00722088 15/05/2017 1772 6,8975 12222,37 0,020526686 3236,20,00218944 16/05/2017 1786,2 6,8897 12306,3821 0,006873637 3259,960,00734194 17/05/2017 1870 6,891 12886,17 0,047112779 3250,97 -0,0027577 18/05/2017 1941,5 6,8902 13377,3233 0,038114762 3235,97 -0,004614 19/05/2017 1966,5 6,8926 13554,2979 0,013229448 3236,48 0,0001576 22/05/2017 2087,3 6,8924 14386,5065 0,061398136 3220,9 -0,0048139 23/05/2017 2249,6 6,8897 15499,0691 0,077333757 3206,67 -0,004418 24/05/2017 2395,5 6,8908 16506,9114 0,065025988 3208,90,00069543 25/05/2017 2268,1 6,8682 15577,7644 -0,056288361 3254,730,01428215 26/05/2017 2125,9 6,861 14585,7999 -0,063678233 3257,060,00071588 29/05/2017 2207,4 6,861 15144,9714 0,038336704 3257,06 0 30/05/2017 2146,7 6,861 14728,5087 -0,027498414 3257,06 0 31/05/2017 2191,8 6,821 14950,2678 0,015056453 3264,540,00229655 01/06/2017 2312 6,8049 15732,9288 0,052350969 3249,44 -0,0046255 02/06/2017 2405,9 6,8162 16399,0956 0,042342198 3252,430,00092016 05/06/2017 2636,9 6,801 17933,5569 0,093569875 3237,81 -0,0044951 06/06/2017 2844,6 6,7967 19333,8928 0,078084673 3248,820,00340045 07/06/2017 2644 6,7944 17964,3936 -0,070834117 3288,760,01229369 08/06/2017 2781,5 6,7936 18896,3984 0,051880671 3299,290,00320181 09/06/2017 2809 6,7992 19098,9528 0,010719207 3307,72 0,0025551 12/06/2017 2569,6 6,7981 17468,3978 -0,085374055 3288,38 -0,0058469 13/06/2017 2677,1 6,7981 18199,1935 0,041835305 3302,880,00440947 14/06/2017 2394,3 6,7969 16273,8177 -0,105794569 3278,7 -0,0073209 15/06/2017 2377,5 6,7975 16161,0563 -0,006929008 3280,550,00056425 16/06/2017 2437,5 6,8138 16608,6375 0,027695049 3270,77 -0,0029812 19/06/2017 2582 6,817 17601,494 0,059779527 3293,010,00679962 20/06/2017 2714,5 6,8263 18529,9914 0,052751053 3288,44 -0,0013878 21/06/2017 2624,4 6,8291 17922,29 -0,032795553 3305,450,00517267 22/06/2017 2672,8 6,8309 18257,6295 0,01871075 3296,27 -0,0027772 23/06/2017 2674,9 6,8387 18292,8386 0,00192846 3307,220,00332194 26/06/2017 2393,6 6,8381 16367,6762 -0,105241319 3336,090,00872939 27/06/2017 2521,2 6,8112 17172,3974 0,049165274 3342,10,00180151 28/06/2017 2518,2 6,7996 17122,7527 -0,00289096 3323,24 -0,0056432 29/06/2017 2472,4 6,7797 16762,1303 -0,021061008 3338,810,00468519 30/06/2017 2420,6 6,7795 16410,4577 -0,020980184 3343,390,00137175 03/07/2017 2524 6,7904 17138,9696 0,044393149 3347,040,00109171 04/07/2017 2579,9 6,7999 17543,062 0,023577404 3333,29 -0,0041081 05/07/2017 2598,6 6,7991 17668,1413 0,007129841 3358,77 0,0076441 06/07/2017 2593,2 6,8033 17642,3176 -0,001461597 3364,320,00165239 07/07/2017 2479,3 6,7999 16858,9921 -0,044400374 3370,10,00171803 10/07/2017 2318,3 6,8017 15768,3811 -0,064690164 3364,53 -0,0016528 11/07/2017 2283,8 6,8007 15531,4387 -0,015026428 3354,52 -0,0029752 12/07/2017 2375,6 6,7875 16124,385 0,038177168 3348,75 -0,0017201 13/07/2017 2330,1 6,7814 15801,3401 -0,020034554 3370,350,00645017 14/07/2017 2206,5 6,7841 14969,1167 -0,052667906 3374,820,00132627 17/07/2017 2220 6,7699 15029,178 0,004012351 3326,8 -0,0142289 18/07/2017 2302,8 6,758 15562,3224 0,035473956 3338,380,00348082 19/07/2017 2253,4 6,7555 15222,8437 -0,021814141 3383,840,01361738 20/07/2017 2865,1 6,7684 19392,1428 0,273884382 3398,420,00430871 21/07/2017 2659 6,7679 17995,8461 -0,072003221 3391,18 -0,0021304 24/07/2017 2769,7 6,7506 18697,1368 0,038969589 3404,40,00389835 SUMMARY OUTPUT 25/07/2017 2560,9 6,7511 17288,892 -0,075318742 3397,18 -0,0021208 mean btc r 0,00880523 26/07/2017 2527,74 6,7548 17074,3782 -0,012407611 3401,350,00122749 Regression Statistics variance btc 0,00199156 27/07/2017 2664,6 6,7376 17953,009 0,051459022 3403,520,00063798 Multiple R 0,01027201 28/07/2017 2784,8 6,7428 18777,3494 0,045916564 3407,06 0,0010401 R Square 0,00010551 31/07/2017 2854,3 6,7289 19206,2993 0,022844003 3427,790,00608442 Adjusted R Square-0,0038156 01/08/2017 2731,3 6,7189 18351,3316 -0,044514963 3448,410,00601554 Standard Error0,04471192 02/08/2017 2702 6,7222 18163,3844 -0,010241609 3440,49 -0,0022967 Observations 257 03/08/2017 2790,3 6,7233 18760,024 0,032848481 3427,75 -0,003703 04/08/2017 2860 6,7179 19213,194 0,024156153 3416,35 -0,0033258 ANOVA 07/08/2017 3396 6,7183 22815,3468 0,187483289 3434,590,00533903 df SS MS F Significance F
08/08/2017 3415 6,7038 22893,477 0,003424458 3437,120,00073662 Regression 1 5,3795E-05 5,37952E-05 0,02690896 0,86982998 09/08/2017 3340,4 6,6793 22311,5337 -0,025419611 3430,45 -0,0019406 Residual 255 0,50978475 0,001999156
10/08/2017 3405 6,6601 22677,6405 0,016408858 3415,96 -0,0042239 Total 256 0,50983854 11/08/2017 3643,4 6,6669 24290,1835 0,071107175 3360,18 -0,0163292
14/08/2017 4320,8 6,6687 28814,119 0,186245423 3390,350,00897869 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95,0% Upper 95,0%
15/08/2017 4151,8 6,6778 27724,89 -0,037801917 3404,970,00431224 Intercept 0,00882579 0,00279187 3,161249044 0,00176085 0,00332773 0,01432384 0,00332773 0,01432384 16/08/2017 4386,4 6,695 29366,948 0,059226852 3399,91 -0,0014861 X Variable 1 -0,0791326 0,48239958 -0,16403951 0,86982998 -1,0291272 0,87086199 -1,0291272 0,87086199 17/08/2017 4263 6,6728 28446,1464 -0,031355032 3422,9290,00677047 18/08/2017 4090,1 6,6796 27320,232 -0,039580561 3423,22 8,5015E-05 21/08/2017 3998,2 6,67 26667,994 -0,023873808 3442,270,00556494 22/08/2017 4081,9 6,66 27185,454 0,019403784 3445,790,00102258 23/08/2017 4130,8 6,6636 27525,9989 0,012526731 3443,17 -0,0007603 Market Model Constant mean return model
24/08/2017 4322,1 6,6607 28788,2115 0,045855287 3426,18 -0,0049344 E [R] AR CAR AR t-test E[R] AR CAR AR t-test 25/08/2017 4351,5 6,6645 29000,5718 0,00737664 3489,140,01837615 0,00737164 5,0029E-06 5,0029E-06 0,000111892 0,00880523 -0,0014286 -0,0014286 -0,0320119 28/08/2017 4385,1 6,6325 29084,1758 0,00288284 3521,750,00934614 0,00808621 -0,0052034 -0,0051984 -0,11637535 0,00880523 -0,0059224 -0,007351 -0,1327091 29/08/2017 4587,1 6,5975 30263,3923 0,040544952 3524,50,00078086 0,008764 0,03178095 0,02658259 0,710793739 0,00880523 0,03173972 0,02438874 0,71122461 30/08/2017 4568 6,5926 30114,9968 -0,004903464 3522,72 -0,000505 0,00886575 -0,0137692 0,01281337 -0,30795408 0,00880523 -0,0137087 0,01068005 -0,3071848 31/08/2017 4718,3 6,5969 31126,1533 0,033576509 3519,716 -0,0008528 0,00889327 0,02468324 0,03749661 0,552050519 0,00880523 0,02477128 0,03545133 0,55507555 01/09/2017 4907,7 6,5685 32236,2275 0,035663712 3526,350,00188481 0,00867664 0,02698707 0,06448368 0,603576676 0,00880523 0,02685848 0,06230981 0,60184566 04/09/2017 4205 6,5204 27418,282 -0,149457484 3539,460,00371773 0,0085316 -0,1579891 -0,0935054 -3,53348899 0,00880523 -0,1582627 -0,0959529 -3,5463554 05/09/2017 4375 6,5516 28663,25 0,045406492 3544,370,00138722 0,00871602 0,03669048 -0,0568149 0,820597179 0,00880523 0,03660126 -0,0593516 0,82016214 06/09/2017 4595,8 6,5246 29985,7567 0,046139453 3545,450,00030471 0,00880168 0,03733778 -0,0194771 0,835074291 0,00880523 0,03733422 -0,0220174 0,83658636 07/09/2017 4613,7 6,5006 29991,8182 0,000202147 3524,57 -0,0058892 0,00929182 -0,0090897 -0,0285668 -0,20329418 0,00880523 -0,0086031 -0,0306205 -0,1927781 08/09/2017 4304 6,4645 27823,208 -0,072306727 3524,05 -0,0001475 0,00883746 -0,0811442 -0,109711 -1,81482231 0,00880523 -0,081112 -0,1117325 -1,817559 11/09/2017 4198,7 6,5238 27391,4791 -0,015516864 3535,66 0,0032945 0,00856509 -0,024082 -0,133793 -0,53860247 0,00880523 -0,0243221 -0,1360545 -0,5450102 12/09/2017 4149,4 6,5329 27107,6153 -0,010363215 3538,940,00092769 0,00875238 -0,0191156 -0,1529086 -0,42752792 0,00880523 -0,0191684 -0,155223 -0,4295271 13/09/2017 3849,7 6,5306 25140,8508 -0,072553945 3543,780,00136764 0,00871756 -0,0812715 -0,2341801 -1,81766984 0,00880523 -0,0813592 -0,2365822 -1,8230987 14/09/2017 3235,3 6,5551 21207,715 -0,156444021 3530,46 -0,0037587 0,00912323 -0,1655672 -0,3997473 -3,7029777 0,00880523 -0,1652493 -0,4018314 -3,7029099 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 SSE CAR CM-model Bad Regulation
