Impact of Regulations on ICO Performance

Figure 7: Sample development process

Source: own elaboration

Measuring abnormal returns

To measure abnormal returns, expected daily market returns were estimated through calculating of CRIX:

Observed daily returns for every i token were calculated using closing prices from the data gathered on coinmarketcap.com:

Then, abnormal returns were estimated using the market model:

After that, the abnormal returns were cumulated for the first 30 days and later the first 90 days:

Where CAR stands for cumulative abnormal returns. The CARs were tested for significance with t statistics:

Average turned out to be 0,8, average equals -0,3. The relatively low cumulated abnormal returns for the first month and negative return after three months may be caused by three factors.

Firstly, the chosen sample mostly consists of ICOs from 2017 which were large in quantity but volatile in quality. According to the statistics by Blockchain.info (2017), over 43% of all projects that completed an ICO eventually shut down due to poor business models, unjustified use of blockchain and lack of financing. Secondly, one can assume that the CRIX index, which represents the market return in the model is a particularly accurate tool to predict returns on cryptoassets.Its broad nature and solid statistical approach for choosing index participants leads to the fact that CRIX returns reflect the price dynamics of the whole cryptomarket and do not allow for outstanding abnormality of the observed returns. Finally, three months seems to be enough for the market to adapt to the new information. The focus of investors shifts to a different blockchain project, which leads to the stabilization of the returns of a particular token and eventually causes it the price to decline.

Table 5: Descriptive statistics of the CAR

Statistics	1-month CAR	3-month CAR
Mean	,7974	-,2979
Std. Error of Mean	,14867	,15805
Median	,7325	,2345
Mode	,62a	-3,51
Std. Deviation	1,50884	1,60405
Variance	2,277	2,573
Skewness	-,153	-,455
Std. Error of Skewness	,238	,238
Kurtosis	-,488	-,527
Std. Error of Kurtosis	,472	,472
Minimum	-2,62	-4,24
Maximum	3,67	2,63

Source: own elaboration, SPSS

Descriptive statistics clearly indicate issues with skewness and kurtosis. One can presume this is caused by a limited number of observations and with a larger sample the assumption of high skewness may not hold true. Nevertheless, the cumulative abnormal returns for three months seem to be mostly concentrated around positive values.

Figure 8: Descriptive statistics of the 1-month (left) and 3-month (right) CAR

Source: own elaboration, SPSS

Linear regression model

Independent variables

To address the goal of the research and estimate the impact of regulatory factors on short-term performance of ICOs, a linear regression model is proposed. A regression was applied in a similar research design by Seys (2016), so there is a proven use case to back up the choice of the method. Cumulative abnormal returns will be used as dependent variables. A total of seven regulatory, business, infrastructure and IT-related factors were chosen on the basis of the established theoretical background and the literature study on the approaches to ICO regulation:

The amount of received funds

The jurisdiction of the ICO

Listing on the top 5 largest exchanges

Availability of the source code

Access of American investors

Bounty campaign

Token granting a share of profits or participance in governance

All variables except for the total amount of collected funds are coded through binary dummy variables, resulting in a following model:

Table 6: Descriptive statistics of the dummy variables (for full statistics see Appendix 3)

	Jurisdiction	Top5_Ex	OpenSource	American	Bounty	Profit
Mean	,7573	,7282	,5049	,5340	,6699	,5922
Std. Error of Mean	,04245	,04405	,04951	,04939	,04656	,04866
Median	1,0000	1,0000	1,0000	1,0000	1,0000	1,0000
Mode	1,00	1,00	1,00	1,00	1,00	1,00
Std. Deviation	,43082	,44709	,50242	,50128	,47255	,49382
Variance	,186	,200	,252	,251	,223	,244

Source: own elaboration, SPSS

The first two key hypotheses of the model are as follows:

Hypothesis 1. Regulatory factors have significant impact on the ICO performance.

Hypothesis 1a. Regulatory factors have a larger impact on ICO performance.

The amount of received funds

The firstvariable X is the amount of accumulated funds during the ICO. There is evidence to support the idea that the more investors believe in the project on the earlier state, the higher the potential of the project to reach product launch and fulfill the promises made in the white paper roadmap. If the blockchain ecosystem is fully developed and released, the tokens will find its place in the project's business model and the demand will grow, thus possibly increasing the token price on the secondary market. For example, Ethereum platform remains one of the most efficient token crowd sales and the token price has been steadily increasing since the platform officially launched. Even though the company faced significant technical malfunction with their blockchain which resulted in a hard fork and emergence of Ethereum Classic.

Hypothesis 2

The amount of funds accumulated during the ICO has a positive relation to the short-term performance of ICO.

The amount of funds raised is measured in USD, which implies the problem of scale. A solution proposed is data transformation from USD to natural logarithms. This allows us to simplify the case, leave the sample percentage changes equivalent andmake the new set of data more suitable for regression.

Table 7: Descriptive statistics of the original value and logs of the total funds received

Statistics	USD	Logs
Mean	28394399,7767	16,3696
Std. Error of Mean	4124901,29230	0,13828
Median	15000000,0000	16,5236
Mode	3400000,00a	15,04a
Std. Deviation	41863175,93229	1,40341
Variance	1752525499138172,800	1,970
Skewness	3,421	-0,573
Std. Error of Skewness	0,238	0,238
Kurtosis	13,905	0,486
Std. Error of Kurtosis	0,472	0,472
Minimum	260000,00	12,47
Maximum	257000000,00	19,36

Source: own elaboration, SPSS

Figure 9: Histogram with normal curve for original values of the total funds received (left) and logs (right)

Source: own elaboration, SPSS

Jurisdiction

The second independent variableis Jurisdiction. According to the literature study on the approaches of ICO regulation there seems to be a clear division between countries with ICO governance initiatives and countries with either zero tolerance to cryptocurrencies or at least large skepticism and complete lack of action on the topic. Jursidiction is represented by a dummy-variable equals 1, if the project is registered in an `ICO friendly' country: USA, UK, Switzerland, Luxembourg, Lichtenstein, Singapore, Australia, Lithuania, Germany, Japan or Canada. The dummy equals 0, if otherwise.

It is important to underline that defining actual jurisdiction of a particular ICO seems to be a challenge, which has long been recognized by the online community of cryptoinvestors, and yet has not been properly addressed. There is a number of ways one can assume the project's jurisdiction, none of them giving a guaranteed answer, rather aiding with hints that need to be correctly interpreted. In the sample used for this research, the jurisdiction is determined by the location of the founders or the majority of the team, if it is not possible to retrieve the data from the white paper and the website's terms of use. The same approach was used by EY (2017) and JP Morgan (2017).

Hypothesis 3.

The `ICO friendly' location and jurisdiction has statistically significant relation to its short-term performance, as some countries have implemented restrictive regulatory frameworks for cryptocurrencies or token sales in particular.

Listing

The third independent variable is also a dummy and shows, whether the tokens were listed on all the top five cryptoexchanges or not. The dummy equals 1, if after the completion of an ICO the tokens appear on OKEx, Bitfitnex, UpBit, Binance and Huobiand 0, if otherwise. There is certain evidence that supports the idea that investors that are willing to participate in post-IPO token trading are more willing to use large platforms, irrespective of the transactional costs: fees, long registration time, etc. (Hileman & Rauchs, 2017).

Hypothesis 4

Listing on all the largest exchanges has statistically significant relation to ICO performance.

Open Source

The fourth independent variable is the availability of the source code. If any of the source code material used in the development of the blockchain ecosystem is available for the public in any depository, the dummy equals 1, if there is no access to the code, it equals 0. Nowadays most projects that seek funding through token sales tend to leave at least some of the innerworkings of their IT systems open for the online community. Some researchers believe that this happened due to the open source nature of Ethereum, which offered one on the first major alternatives for the Bitcoin blockchain and led a successful ICO to fund itself (Chod & Lyandres, 2018). The information on the availability of the code was gathered through the analysis of the sample projects' white papers and additionally checked on GitHub (GitHub.com, 2018).

Hypothesis 5

Availability of the source code materials has statistically significant relation to ICO performance.

Access of American investors

On the one hand, the regulatory initiatives introduced by SEC (2017) made it possible for the token crowd sales to enjoy additional media attention and get recognized on a wider scale, thus reinforcing the reputation of the leading country in terms of the cryptomarket development. On the other hand, with some tokens being now officially regarded as securities, new transactional obstacles can arise for the American citizens that are willing to invest in ICOs. Usually the policies regarding the allowance of American investors are provided in the white paper, if there is no information on that topic, the access is assumed to be granted. The dummy equals 1, if the white paper does not imply any limitations for American investors and equals 0, if it does.

Hypothesis 6

Current policies regarding ICOs and the participance of American investors in ICOs developed by SEC have statistically significant impact on the performance of ICOs worldwide.

Bounty campaign

Bounty campaigns are used to incentivize the online community to spread information about the upcoming ICOs. I have covered some of the core mechanics of bounty campaigns in the first chapter of the thesis. Most of the companies in the sample utilized some form of a bounty campaign. Previous quantitative research on ICOs did not find statistically significant effects of bounty campaigns(Adhami et al., 2018), however some authors claim they do play a major role when it comes to less experienced investors.

Hypothesis 7

The presence of a bounty campaign prior to the token sale has statistically significant impact on ICO performance.

Token function

According to the report by JP Morgan (2018), 67% of the projects that are developing DLT platforms or ecosystems issue tokens that can be used only as a means of payment or access to services. However, there are projects that offer the ability to participate in key decisions by voting or even share the future company's profits. On the basis of the assumption, that such authority tokens will be more valuable than a new faster alternative to Bitcoin, I decided to code the governance function of the token as a last dummy variable in the regression. The dummy equals 1, if the white paper mentions that the token grants participation in governance or a share of the profits, and equals 0, if otherwise.

Hypothesis 8

If the possession of the token allows to participate in the profit share or governance of the company, ithas statistically significant impact on the ICO performance.



Empirical results

ICO performance after 1 month

In this chapter I will present the two linear regressions that show the relations between cumulative abnormal returns on short-term ICO performance and the chosen factors. I will elaborate on the results and their validity, attempt to interpret the significant coefficients and conclude withchecking the hypotheses.

To start with, I will focus on the Pearson correlation coefficients for the first regression. We can observelittle to no correlation between most of the factors. However, this correlation matrix provides us with two interesting insights. Firstly, there is a weak negative correlation between the presence of a bounty campaign and the profit share (or governance) function of the token. One can assume this can happen because tokens that confer rights similar to ownership in this sample would not be sold for a cheaper price just for supporting the campaign on social media. Secondly, there seems to be a weak but evident correlation between the cumulative returns during the first month and the availability of the source code. This point will be further elaborated on during the interpretation of the regression. Additionally, Durbin-Watson test shows satisfactory values so one can assume no autocorrelation.

Table 7: Pearson correlation coefficients

Variable	CAR1	Funds_raised	Jurisdiction	Top5_Ex	OpenSource	American	Bounty	Profit
CAR1	1	0,335	0,121	-0,075	0,219	0,139	0,133	0,047
Funds_raised	0,335	1	0,054	0,011	0,198	0,167	0,19	0,105
Jurisdiction	0,121	0,054	1	0,01	-0,153	0,197	-0,012	0,175
Top5_Ex	-0,075	0,011	0,01	1	0,137	0,085	0,082	0,115
OpenSource	0,219	0,198	-0,153	0,137	1	0,009	0,172	0,008
American	0,139	0,167	0,197	0,085	0,009	1	0,131	-0,023
Bounty	0,133	0,19	-0,012	0,082	0,172	0,131	1	-0,246
Profit	0,047	0,105	0,175	0,115	0,008	-0,023	-0,246	1

Source: own elaboration, SPSS

The goal of the first regression was to find linkages between regulatory factors and abnormal performance of token prices during the first thirty days right after the ICO. Their impact can now also be compared with the influence of other non-regulatory factors. The regression has a low level of R Square, but significant p-value for the coefficient before the total amount of funds raisedand relatively significant coefficient before open source dummy. Other coefficients are not statistically significant.

Table 8: Regression for the 1-month performance

R	R Square	Adjusted R Square	Std. Error of the Estimate	Durbin-Watson
,416a	,173	,112	1,33720	1,827
Model	Coefficients	t	p-value
	B	Std. Error
(Constant)	-4,135	1,594	-2,594	,011
Funds_raised	,269	,100	2,692	,008
Jurisdiction	,391	,324	1,209	,230
Top5	-,376	,304	-1,239	,218
OpenSource	,535	,278	1,926	,057
American	,205	,276	,742	,460
Bounty	,173	,303	,573	,568
Profit	,074	,288	,258	,797

Source: own elaboration, SPSS

The independent variables in the regression seem to explain very little variations of the dependent variable. However, the two significant p-values indicate that, regardless of the high variability of the CARs during the first month, there is still an evident trend. This regression can be interpreted as ademonstration of the unpredictable nature of initial post-IPO returns: it seems to partly support the first and the fourth hypotheses, however the connection to the amount of raised finances and the availability of the source code definitely cannot explain much about the behavior of the abnormal returns.Obviously, the first hypothesis is also rejected: there seem to be no significant impact of the regulatory factors.

These outcomes are partly supported by literature: some empirical studies also found the availability of the source code to be significant for ICO success. (Adhami et al., 2017). At the same timeElendneret al. (2016), Pieters & Vivanco (2017) and Pilkington (2017) presumed that the development of regulatory frameworks will help incentivize institutional investors to fund ICO projects and, thus, create additional demand on the tokens. It seems that the returns on ICOs are too sporadic, volatile and inconsistent due to certain inefficiencies of the market, which has only started to emerge.

To sum up, the empirical results show that the data is suitable for proposed analysis, but there is no evidence of linkages between regulation and ICO performance during the first month.

ICO performance after 3 months

The second regression is devoted to assessing the impact of the chosen factors during a longer time periodof 90 days. This period was used in the research model on the basis of the assumption that after three months of the public token sale, the market will have enough time to adjust to the new information and will more efficiently reflect the impact of certain drivers. However, it isnow evident that the situation on the market is more complex. At least judging from the used sample and the chosen factors, ICO performance seems to be far less predictable that traditional financial instruments.

The new Pearson coefficients for the CAR3 independent variable show little to no correlation, except only for the availability of the code -- a connection which was present in the previous model as well, though slightly weaker.



Table 9:Pearson correlation coefficients (only with the new variable)

	CAR3
CAR3	1
Funds_raised	0,3
Jurisdiction	0,059
Top5	0,25
OpenSource	0,312
American	0,35
Bounty	0,226
Profit	0,004

Source: own elaboration, SPSS

Some of the issues identified in the first regression are present here as well, however the model on the whole seems to produce more significant results. To begin with, the level of R Square is still too low to conclude that the chosen factors can comprehensively explain the values of cumulated abnormal returns. Durbin-Watson test gives satisfactory results, so we can say that the outcomes of the regresssion were most likely not inferred with by autocorrelation.

At the same time yet again there appears to be a group of statistically significant coefficients. This time the most precise coefficient is the one before a regulatory factor. We can observe highly significant difference between the CARs of projects that restricted the access of American investors in the white paper and the CARs of projects that either did not mention the issue or expressed no limitations. Other significant factors once again include open policies on IT-developments and the amount of attracted capital. Surprisingly, the availability of tokens on the five largest cryptoexchanges also has a significant coefficient in this model.



Table 10: Regression for the 3-month performance

R	R Square	Adjusted R Square	Std. Error of the Estimate	Durbin-Watson
,546a	,298	,246	1,39951	2,019
Model	Coefficients	t	p-value
	B	Std. Error
(Constant)	-5,428	1,668	-3,254	,002
Funds_raised	,214	,105	2,041	,044
Jurisdiction	,115	,339	,339	,735
Top5	,669	,318	2,104	,038
OpenSource	,756	,291	2,600	,011
American	,906	,289	3,139	,002
Bounty	,326	,317	1,030	,306
Profit	-,046	,301	-,152	,880

Source: own elaboration, SPSS

The new insights gathered from this model lead to a few key conclusions. The market perception of different factors can indeed vary with the growth of the observational window. First of all, one can suggest that American investors may be more risk-averse and will prefer to buy tokens a few months after the actual ICO, probably in a hope for getting a discounted price. Therefore, the lack of interest from the US investors which might be caused by the restrictions some ICO projects choose to implement, reveals itself later on the way, decreasing the CARs some months after the ICO.

Secondly, a possible interpretation of the growth of the impact caused by not being listed on major cryptoexchanges is that the exchanges themselves might provide additional marketing support and attract attention to the token just by listing it on their platforms. Thus, projects that decide not to pursue the goal of being represented ubiquitously among the key exchanges, may endure decreased returns due to the lack of investor coverage. Interestingly enough, the bounty campaign and the token function turned out to be not significant on both of the event windows for this sample. Even though bounty campaigns are used by a majority of ICO projects and seemingly believed to be an efficient marketing tool.

I find the second model to be more insightful and logical. I presume that the gradual reveal of the full impact of the regulatory factors is something entirely plausible, given that most of them do not actually interfere with the behavior of the early adopters of blockchain-based services and the enthusiasts of the cryptocommunity. Most of the current restrictions can be mitigated if the ICO team or the investors have substantial experience and are well-educated in the field. This implies that the regulatory initiatives mostly affect the second and thirds waves of investors, who are less experienced and can actually refuse to pursue the idea of buying tokens if presented with an immediate negative regulatory response. Therefore, regulation initiatives can possibly have larger impact on the ICO performance in the long run.

Review of the hypotheses

Table 11:Hypotheses

Hypothesis	1-month performance	3-month performance
1. Regulatory factors have significant impact on the ICO performance.	declined	confirmed
1a. Regulatory factors have a larger impact on ICO performance	declined	confirmed (most significance in the model)
2. The amount of funds accumulated during the ICO has a positive relation to the short-term performance of ICO.	confirmed	declined
3. The `ICO friendly' location and jurisdiction has statistically significant relation to its short-term performance, as some countries have implemented restrictive regulatory frameworks for cryptocurrencies or token sales in particular.	declined	declined
4. Listing on all the largest exchanges has statistically significant relation to ICO performance.	declined	confirmed
5. Availability of the source code materials has statistically significant relation to ICO performance	confirmed (slightly more than 5% significance)	confirmed
6. Current policies regarding ICOs and the participance of American investors in ICOs developed by SEC have statistically significant impact on the performance of ICOs worldwide.	declined	confirmed
7. The presence of a bounty campaign prior to the token sale has statistically significant impact on ICO performance.	declined	declined
8. If the possession of the token allows to participate in the profit share or governance of the company, it has statistically significant impact on the ICO performance.	declined	declined

Source: own elaboration



Discussion

Practical implications

There is a number of implications for future policies that can be drawn from the work. First of all, the inevitable growth of the popularity of distributed ledger technology will lead to what is commonly referred to as `tokenization' of the economy. Therefore, adequate legislative frameworks that will allow distributed networks to fully operate are necessary in the future. As stated in the thesis, there is currently no real consensus on whether or not ICOs can be regarded as fully legal securities or they require the emergence of a completely new legal status.

However, it is clear that even current regulative initiatives can affect the public perception of cryptocurrencies, ICOs and distributed ledger technologies on the whole (Nian & Chuen, 2015). The common misbelief that token crowd sales are mostly for scam projects -- short-lived companies created for the sole purpose of attracting capital -- is not completely rooted in misunderstanding, but it is definitely fueled by the announcement of ICO bans and legal cases against cryptoprojects.

The presented analysis of the impact of different infrastructural determinants established a few core concepts than can be useful for future policymakers to help build an efficient and trustworthy market:

Limitations lead to limitations. The regression showed statistically significant positive relation between the possibility of American investor participance and abnormal returns. This leads to the assumption that with the leverage of American capital more ICOs will reach their goals and more creative blockchain solutions will be able to attract capital. Regulatory measures that lead to restrictions of participance seem a counter-productive.

Staying open minded. The only factor being statistically significant in both regressions except for the amount of received funding is the availability of the source code. By making their intellectual assets public, such companies are able to engage the community into their R&D and foster new makers, not imposers. Moreover, is seems that this openness can be a leverage for the trust of investors. By making all ICO projects going open source, future policymakers will be able to accelerate both innovation and capital.

Marketing for marketing's sake does not give results. One of the few factors that never turned out to be statistically significant is the presence of a bounty campaign prior to the ICO. Obviously, this may be a phenomenon caused by the sample or by these specific research methods. However, I believe that the only possibly productive area, where ICO governance should be applied, is online marketing. Policies against targeting blockchain-related ads have become common practice among social networks, however, there is a lot of room for legislative work and improvement.

Limitations and further discussion

The first key limitation of this research is the sample used to gather data for both the event study and the regression. Looking for data on cryptomarkets is a a challenge that has already been established and currently there are means of getting some vital information online, however searching for deep details or historical data may become an issue. In order to get enough data for a large-scale retrospective quantitative study on ICOs or cryptocurrencies, one would have to find a whole team of researchers willing to aid with the information mining.

The second key limitation of the research is the lack of a proper model to assess the expected returns. There seem to be no other reliable index rather than CRIX, however, industry-centered indices will undoubtedly help to diversify quantitative research on cryptocurrencies and tokens.

The main area for further research has to be the general determinants of ICO success, as the current academical work on the topic is very limited. Using more sophisticated models of the event study one could possible yield more statistically significant results than the presented work.

Conclusion

The goal of this research was to identify whether current regulatory initiatives help ICO markets to grow and develop with the help of empirical data. The key insights gained in the thesis suggest that current methods are statistically efficient under specific conditions. However, some of them are possibly hindering the growth of cryptoeconomics and do not support token crowd sales, as they are based on restrictions, not on support and legitimization.

It is also important to emphasize that the research has helped to map out the imperfections of the ICO market visible from both the empirical results and literature on the topic: limited amount of quantitative data, skepticism towards good news and exaggerated declines towards bad news, overreliance on media channels and lack of common frameworks for value assessment.

On the one hand, it is clear that the ICO market itself is yet not ready to the amount of capital it managed to attain and has weak infrastructure. On the other hand, this is the perfect time for the development of policies for future regulation, as in the current phase the market is only emerging and would most likely benefit from certain amount of non-restrictive regulation.



References

Adhami, S., Giudici, G., & Martinazzi, S. (2017). Why do businesses go crypto? An empirical analysis of Initial Coin Offerings.

Baron, J., O'Mahony, A., Manheim, D., & Dion-Schwarz, C. (2015). National Security Implications of Virtual Currency: Examining the Potential for Non-state Actor Deployment. RAND Corporation-NDRI Turpin, J. B. (2014). Bitcoin: The economic case for a global, virtual currency operating in an unexplored legal framework. Indiana Journal of Global Legal Studies, 21(1), 335-368.

Bцhme, R., Christin, N., Edelman, B., & Moore, T. (2015). Bitcoin: Economics, technology, and governance. Journal of Economic Perspectives, 29(2), 213-38.

Bushnell, J. B., Chong, H., & Mansur, E. T. (2009). Profiting from regulation: an event study of the EU carbon market (No. w15572). National Bureau of Economic Research.

Chad R, Dougherty (31 Dec 2008). "Vulnerability Note VU#836068 MD5 vulnerable to collision attacks". Vulnerability notes database. CERT Carnegie Mellon University Software Engineering Institute. Retrieved 3 February 2017.

Chod, J., & Lyandres, E. (2018). A Theory of ICOs: Diversification, Agency, and Information Asymmetry.

Chohan, U. W. (2017). Initial Coin Offerings (ICOs): Risks, Regulation, and Accountability. SSRN.

Cong, L. W., Li, Y., & Wang, N. (2018). Tokenomics: Dynamic Adoption and Valuation.

Conley, J. P. (2017). Blockchain and the Economics of Crypto-tokens and Initial Coin Offerings (No. 17-00008). Vanderbilt University Department of Economics.

Dibrova, A. (2016). Virtual currency: new step in monetary development. Procedia-Social and Behavioral Sciences, 229, 42-49.

Elendner, H., Trimborn, S., Ong, B., & Lee, T. M. (2016). The cross-section of crypto-currencies as financial assets: an overview (No. 2016-038). SFB 649 Discussion Paper.

Flьhmann, D., & Hsu, P. (2018). Switzerland: Fintech outlook 2018. International Financial Law Review, 20.

Gartner (2017). Top Trends in the Gartner Hype Cycle for Emerging Technologies, 2017

Harwick, C. (2016). Cryptocurrency and the Problem of Intermediation. The Independent Review, 20(4), 569-588.

Hileman, G., & Rauchs, M. (2017). Global cryptocurrency benchmarking study. Cambridge Centre for Alternative Finance.

Holub, M., & Johnson, J. (2017). Mapping Bitcoin's Influence on Academic Research.

Jackson, O. (2018). France's crypto derivatives rules could be EU template. International Financial Law Review.

Jackson, O. (2018). UK eyes AML rules for bitcoin regulation. International Financial Law Review.

Jackson, O. (2018). US or Swiss approach for EU crypto regulation?. International Financial Law Review.

Kaal, W. A. (2018). Initial Coin Offerings: the Top 25 Jurisdictions and Their Comparative Regulatory Responses.

Klima, V. (2005). Finding MD5 Collisions-a Toy For a Notebook. IACR Cryptology ePrint Archive, 2005, 75.

Kothari, S. P., & Warner, J. (2007). Econometrics of event studies. Handbook of empirical corporate finance, 1, 3-36.

Labbй, A., Crabb, J., & Lai, K. (2018). Waking up and cracking down. International Financial Law Review, 3.

Lai, K. (2018). PRIMER: ICO regulation in Asia. International Financial Law Review.

Lamdin, D. J. (2001). Implementing and interpreting event studies of regulatory changes. Journal of Economics and Business, 53(2-3), 171-183.

Lim, J. W. (2015). A Facilitative Model for Cryptocurrency Regulation in Singapore. In Handbook of Digital Currency (pp. 361-381).

Luther, W. J. (2016). Bitcoin and the future of digital payments. The Independent Review, 20(3), 397-404.

McKinney, R. E., Shao, L. P., Rosenlieb, D. C., & Shao, D. H. (2015). Counterfeiting in Cryptocurrency: An Emerging Problem. In Handbook of Digital Currency (pp. 173-187).

Mokhtarian, E., & Lindgren, A. (2018). Rise of the Crypto Hedge Fund: Operational Issues and Best Practices for an Emergent Investment Industry. Stan. JL Bus. & Fin., 23, 112.

Motsi-Omoijiade, I. D. (2017). Financial Intermediation in Cryptocurrency Markets-Regulation, Gaps and Bridges. In Handbook of Blockchain, Digital Finance, and Inclusion, Volume 1(pp. 207-223).

Nakamoto, S. (2008). Bitcoin: A peer-to-peer electronic cash system

Narayanan, A., Bonneau, J., Felten, E., Miller, A., & Goldfeder, S. (2016). Bitcoin and Cryptocurrency Technologies: A Comprehensive Introduction. Princeton University Press.

Neyer, G., & Geva, B. (2017). Blockchain and payment systems: What are the benefits and costs?. Journal of Payments Strategy & Systems, 11(3), 215-225.

Nian, L. P., & Chuen, D. L. K. (2015). A light touch of regulation for virtual currencies. In Handbook of Digital Currency (pp. 309-326).

Pasztor, J. (2018). Bitcoin Investing--An Ethical and Regulatory Quandary. Journal of Financial Service Professionals, 72(2).

Pieters, G., & Vivanco, S. (2017). Financial regulations and price inconsistencies across Bitcoin markets. Information Economics and Policy, 39, 1-14.

Pilkington, M. (2016). Blockchain technology: principles and applications. Research Handbook on Digital Transformations, edited by F. Xavier Olleros and Majlinda Zhegu

Pilkington, M. (2018). The Emerging ICO Landscape-Some Financial and Regulatory Standpoints.

Rogers, Z. (2018). Blockchain and the state: Vehicle or vice?. AQ - Australian Quarterly, (1), 3.

Ryan, R., & Donohue, M. (2017). Securities on Blockchain. Business Lawyer, 73(1).

Santa Monica United States.

Sehra, A., Cohen, R., & Arulchandran, V. (2018). On cryptocurrencies, digital assets and private money. Journal of Payments Strategy & Systems, 12(1)

Sehra, A., Smith, P., & Gomes, P. (2017). Economics of initial coin offerings. Allen & Overy.

Seys, J. (2016). The Evolution of Bitcoin Price Drivers: Moving Towards Stability? (Doctoral dissertation, UNIVERSITY OF GHENT).

Switzerland: Initial coin offerings. (2017). International Financial Law Review, 1.

Trimborn, S., & Hдrdle, W. K. (2016). CRIX an Index for blockchain based Currencies.

Wang, D., & Moini, H. (2012). Performance assessment of mergers and acquisitions: Evidence from Denmark. E-Leader Berlin, 1-15.

Wilson, S (2016). "Blockchain: Almost Everything You Read Is Wrong"

Xiaoyun Wang, Dengguo Feng , Xuejia Lai, Hongbo Yu: Collisions for Hash

Functions MD4, MD5, HAVAL-128 and RIPEMD, rump session, CRYPTO 2004,

Autonomous NEXT: Crypto Fund List (2017)

Bitcoin and other cryptocurrencies in tax. (2018). International Tax Review, 1.

CoinDesk: State of Blockchain (2018).

EY research: initial coin offerings (2017)

Fabric Ventures and TokenData: The State of the Token Market (2017)

J.P. Morgan Perspectives Decrypting Cryptocurrencies: Technology, Applications and Challenges(2018)

PwC, Blockchain: What's happening with the technology in financial services (2017)

Societe Generale: cryptocurrencies, bitcoin and blockchain -- an education piece on how they work (2018)

CoinDesk. (2018). Scrap the White Paper: How to Evaluate Tokens and Blockchains - CoinDesk. [online] Available at: https://www.coindesk.com/scrap-white-paper-evaluate-tokens-blockchains/ [Accessed 10 May 2018].

Constellation Research Inc. (2016). Blockchain: Almost Everything You Read Is Wrong. [online] Available at: https://www.constellationr.com/blog-news/blockchain-almost-everything-you-read-wrong [Accessed 30 Apr. 2018].

Techcrunch.com. (2017). ICO funding hit a record $800 million in Q2 2017. [online]. Available at: https://techcrunch.com/2017/09/07/ico-report-q2-coindesk/ [Accessed 1 Dec. 2017].



Appendixes

Appendix 1. List of cryptocurrencies used in CRIX

№	Name	Token
1	Bitcoin	btc
2	Ethereum	eth
3	Ripple	xrp
4	BitcoinCash	bch
5	EOS	eos
6	Litecoin	ltc
7	Cardano	ada
8	Stellar	xlm
9	TRON	trx
10	NEO	neo
11	Dash	dash
12	Monero	xmr
13	NEM	xem
14	VeChain	ven
15	Tether	usdt
16	Ethereum Classic	etc
17	Qtum	qtum
18	ICON	icx
19	Binance Coin	bnb
20	OmiseGO	omg

Source: http://crix.hu-berlin.de/

Appendix 2. List of projects and sources used for regression data

№	Project	Token	Website	Historical data	№	Project	Token	Website	Historical data
1	Octanox	OTX	https://octanox.org	https://coinmarketcap.com/currencies/octanox/	53	AppCoins	APPC	https://appcoins.io/	https://coinmarketcap.com/ru/currencies/appcoins/
2	SolarDao	SLR	http://www.solardao.me/	https://coinmarketcap.com/currencies/solarcoin/	54	Centra	CTR	https://centra.tech/	https://coinmarketcap.com/currencies/centra/
3	Exchange Union	XUC	https://www.exchangeunion.com	https://coinmarketcap.com/currencies/exchange-union/	55	COSS	COSS	https://coss.io/	https://coinmarketcap.com/currencies/coss/
4	Suretly	SUR	https://www.surco.in/	https://coinmarketcap.com/currencies/suretly/	56	Wagerr	WGR	https://www.wagerr.com	https://coinmarketcap.com/currencies/wagerr/
5	Sphre	XID	https://sphereidentity.com	https://coinmarketcap.com/currencies/sphre-air/	57	Lunyr	LUN	https://lunyr.com/	https://coinmarketcap.com/ru/currencies/lunyr/
6	Metal	METAL	https://www.metalpay.com/	https://coinmarketcap.com/currencies/metalcoin/	58	Covesting	COV	https://covesting.io/	https://coinmarketcap.com/ru/currencies/covesting/
7	Peerguess	GUESS	https://peerguess.com	https://coinmarketcap.com/ru/currencies/guess/	59	KickCoin	KICK	https://www.kickico.com/ru/	https://coinmarketcap.com/currencies/kickico/
8	Bowhead	AHT	https://tokens.bowheadhealth.com	https://coinmarketcap.com/currencies/bowhead/	60	Dmarket	DMT	https://dmarket.io/	https://coinmarketcap.com/ru/currencies/dmarket/
9	Ethos	ETHOS	https://www.ethos.io	https://coinmarketcap.com/currencies/ethos/	61	LATOKEN	LA	https://sale.latoken.com/	https://coinmarketcap.com/currencies/latoken/
10	Ethbits	ETBS	https://www.ethbits.com/	https://coinmarketcap.com/ru/currencies/ethbits/	62	Blocktix	TIX	https://blocktix.io	https://coinmarketcap.com/currencies/blocktix/
11	CreativeCoin	CREA	https://www.creativechain.org/project/	https://coinmarketcap.com/ru/currencies/creativecoin/	63	Presearch	PRE	https://www.presearch.io	https://coinmarketcap.com/currencies/presearch/
12	Starta	STA	https://startaico.com	https://coinmarketcap.com/currencies/starta/	64	Utrust	UTK	https://utrust.io/ico	https://coinmarketcap.com/ru/currencies/utrust/
13	MyBit	MYB	https://mybit.io	https://coinmarketcap.com/currencies/mybit-token/	65	TaaS	TAAS	https://taas.fund/	https://coinmarketcap.com/ru/currencies/taas/
14	Primalbase	PBT	https://primalbase.com	https://coinmarketcap.com/currencies/primalbase/	66	Aeternity	AE	http://www.aeternity.com	https://coinmarketcap.com/currencies/aeternity/
15	EncryptoTel	WAVES	http://ico.encryptotel.com/	https://coinmarketcap.com/ru/currencies/encryptotel/	67	Humaniq	HMQ	https://humaniq.com/	https://coinmarketcap.com/ru/currencies/humaniq/
16	Dcorp	DRP	https://www.dcorp.it	https://coinmarketcap.com/currencies/dcorp/	68	0x	ZRX	https://0xproject.com	https://coinmarketcap.com/currencies/0x/
17	ZrCoin	ZRC	https://zrcoin.io/	https://coinmarketcap.com/currencies/zrcoin/	69	Aragon	ANT	https://aragon.one/	https://coinmarketcap.com/ru/currencies/aragon/
18	DAO.casino	BET	https://dao.casino	https://coinmarketcap.com/currencies/dao-casino/	70	Monaco	MCO	https://mona.co/en/	https://coinmarketcap.com/currencies/monaco/
19	Sociall	SCL	https://socialmedia.market/	https://coinmarketcap.com/currencies/sociall/	71	Hive Project	HVN	https://www.hive-project.net	https://coinmarketcap.com/currencies/hive-project/
20	BlockPool	BPL	https://www.blockpool.io	https://coinmarketcap.com/currencies/blockpool/	72	Worldcore	WRC	https://worldcore.com/	https://coinmarketcap.com/ru/currencies/worldcore/
21	ArbitrageCT	ARCT	https://arbitragect.com/	https://coinmarketcap.com/currencies/arbitragect/	73	Universa	UTNP	https://www.universa.io/	https://coinmarketcap.com/currencies/universa/
22	Dent	DENT	https://www.dentcoin.com	https://coinmarketcap.com/currencies/dent/	74	Storj	STORJ	https://storj.io/tokensale	https://coinmarketcap.com/ru/currencies/storj/
23	Ecobit	ECOB	http://www.ecobit.io	https://coinmarketcap.com/currencies/ecobit/	75	Blackmoon	BMC	https://blackmoonplatform.com/	https://coinmarketcap.com/currencies/blackmoon/
24	Datum	DAT	https://datum.org/	https://coinmarketcap.com/ru/currencies/datum/	76	Quantstamp	QSP	https://quantstamp.com/	https://coinmarketcap.com/currencies/quantstamp/
25	Voise	VOISE	https://www.voise.com/	https://coinmarketcap.com/currencies/voisecom/	77	Raiden Network	RDN	https://raiden.network/	https://coinmarketcap.com/currencies/raiden-network-token/
26	Global Jobcoin	GJC	https://www.globaljobcoin.com/	https://coinmarketcap.com/ru/currencies/global-jobcoin/	78	CyberMiles	CMT	https://www.cybermiles.io/	https://coinmarketcap.com/currencies/cybermiles/
27	Eroscoin	ERO	https://eroscoin.org/	https://coinmarketcap.com/currencies/eroscoin/	79	Civic	CVC	https://www.civic.com	https://coinmarketcap.com/currencies/civic/
28	Paragon	PRG	https://paragoncoin.com/	https://coinmarketcap.com/currencies/paragon/	80	Basic Attention Token	BAT	https://basicattentiontoken.org/	https://coinmarketcap.com/currencies/basic-attention-token/
29	Polybius	PLBT	https://polybius.io/	https://coinmarketcap.com/currencies/polybius/	81	SingularityNET	AGI	https://singularitynet.io/	https://coinmarketcap.com/currencies/singularitynet/
30	Dovu	DOVU	https://dovu.io/	https://coinmarketcap.com/currencies/dovu/	82	monetha	MTH	https://www.monetha.io/	https://coinmarketcap.com/currencies/monetha/
31	SunContract	SNC	https://suncontract.org	https://coinmarketcap.com/currencies/suncontract/	83	TokenCard	TKN	https://tokencard.io/	https://coinmarketcap.com/ru/currencies/tokencard/
32	Spectiv	SIG	https://www.spectivvr.com/	https://coinmarketcap.com/currencies/signal-token/	84	Minexcoin	MNX	https://minexcoin.com	https://coinmarketcap.com/currencies/minexcoin/
33	investFeed	IFT	https://www.investfeed.com/home	https://coinmarketcap.com/currencies/investfeed/	85	Tierion	TNT	https://tierion.com	https://coinmarketcap.com/currencies/tierion/
34	CanYaCoin	CAN	https://canya.io/	https://coinmarketcap.com/ru/currencies/canyacoin/	86	SONM	SNM	https://sonm.com/	https://coinmarketcap.com/currencies/sonm/
35	Starbase	STAR	https://starbase.co/	https://coinmarketcap.com/currencies/starbase/	87	indaHash	IDH	https://indahash.com/	https://coinmarketcap.com/currencies/indahash/
36	Po.et	POE	https://po.et/	https://coinmarketcap.com/currencies/poet/	88	INS Ecosystem	INS	https://ins.world/ru	https://coinmarketcap.com/currencies/ins-ecosystem/
37	Populous	PPT	http://populous.co	https://coinmarketcap.com/currencies/populous/	89	ZenCash	ZEN	https://www.zenprotocol.com/	https://coinmarketcap.com/currencies/zencash/
38	Patientory	PTOY	https://www.patientory.com/	https://coinmarketcap.com/currencies/patientory/	90	MobileGo	MGO	https://mobilego.io/	https://coinmarketcap.com/currencies/mobilego/
39	Cashaa	CAS	https://cashaa.com/	https://coinmarketcap.com/ru/currencies/cashaa/	91	Quantum Resistant	QRL	https://theqrl.org/	https://coinmarketcap.com/ru/currencies/quantum-resistant-ledger/
40	Encrypgen	DNA	https://www.encrypgen.com	https://coinmarketcap.com/currencies/encrypgen/	92	AdEx	ADX	https://www.adex.network	https://coinmarketcap.com/currencies/adx-net/
41	BOScoin	BOS	https://boscoin.io/	https://coinmarketcap.com/ru/currencies/boscoin/	93	WAX	WAX	https://wax.io/	https://coinmarketcap.com/currencies/wax/
42	Gladius	GLA	https://gladius.io/ru	https://coinmarketcap.com/ru/currencies/gladius-token/	94	Pillar	PLR	https://pillarproject.io	https://coinmarketcap.com/currencies/pillar/
43	Gnosis	GNO	https://gnosis.pm/	https://coinmarketcap.com/ru/currencies/gnosis-gno/	95	Santiment	SAN	https://santiment.net	https://coinmarketcap.com/currencies/santiment/
44	Stox	STX	https://www.stox.com/	https://coinmarketcap.com/currencies/stox/	96	TenX	PAY	https://www.tenx.tech/	https://coinmarketcap.com/currencies/tenx/
45	Leverj	LEV	https://register.leverj.io/	https://coinmarketcap.com/ru/currencies/leverj/	97	QASH	QASH	https://liquid.plus/	https://coinmarketcap.com/currencies/qash/
46	Substratum	SUB	https://substratum.net	https://coinmarketcap.com/currencies/substratum/	98	COMSA	CMS	https://comsa.io	https://coinmarketcap.com/currencies/comsa-eth/
47	Dimcoin	DIM	https://www.dimcoin.io/	https://coinmarketcap.com/currencies/dimcoin/	99	Status	SNT	https://status.im/	https://coinmarketcap.com/currencies/status/
48	Cofound.it	CFI	https://cofound.it/en/	https://coinmarketcap.com/ru/currencies/cofound-it/	100	Bancor	BNT	https://europe.bancore.com/	https://coinmarketcap.com/currencies/bancor/
49	PeerPlays	PPY	http://www.peerplays.com/	https://coinmarketcap.com/ru/currencies/peerplays-ppy/	101	SIRIN LABS	SRN	https://sirinlabs.com/	https://coinmarketcap.com/currencies/sirin-labs-token/
50	Incent	INCNT	https://www.incentloyalty.com/	https://coinmarketcap.com/ru/currencies/incent/	102	Tezos	XTZ	https://tezos.com/	https://coinmarketcap.com/currencies/tezos/
51	district0x	DNT	https://district0x.io	https://coinmarketcap.com/currencies/district0x/	103	Filecoin	FIL	https://filecoin.io/	https://coinmarketcap.com/currencies/filecoin/
52	Maecenas	ART	https://www.maecenas.co/	https://coinmarketcap.com/currencies/maecenas/
53	AppCoins	APPC	https://appcoins.io/	https://coinmarketcap.com/ru/currencies/appcoins/

Source: own elaboration

Appendix 3. Full descriptive statistics on independent variables

	Jurisdiction	Top5	OpenSource	American	Bounty	Profit
Mean	,7573	,7282	,5049	,5340	,6699	,5922
Std. Error of Mean	,04245	,04405	,04951	,04939	,04656	,04866
Median	1,0000	1,0000	1,0000	1,0000	1,0000	1,0000
Std. Deviation	,43082	,44709	,50242	,50128	,47255	,49382
Variance	,186	,200	,252	,251	,223	,244
Skewness	-1,218	-1,041	-,020	-,138	-,733	-,381
Std. Error of Skewness	,238	,238	,238	,238	,238	,238
Kurtosis	-,527	-,935	-2,040	-2,021	-1,492	-1,892
Std. Error of Kurtosis	,472	,472	,472	,472	,472	,472

Source: own elaboration, SPSS

Размещено на Allbest.ru