Journal of Sustainable Economics and Management Studies (ECOMAN) VOLUME: 2, ISSUE: 2 DECEMBER 2021 ISSN: 2718-1057 e-ISSN: 2791-8084 Istanbul Gelisim University Press Journal of Sustainable Economics and Management Studies (ECOMAN) (International Peer–Reviewed Journal) VOL. 2 • ISSUE 2 • DECEMBER 2021 • ISSN: 2718-1065 & e-ISSN: 2791-8084 Owner on Behalf of Istanbul Gelisim University Prof. Dr. Bahri ŞAHİN Editorial Board Prof. Dr. Kenan AYDIN Prof. Dr. Nükhet DOĞAN Prof. Dr. Nimet Hülya TIRMANDIOĞLU TALU Prof. Dr. Mehmet Selçuk USLU Assoc. Prof. Dr. Onur ÖZDEMİR Editor Assoc. Prof. Dr. Onur ÖZDEMİR Assistant Editors Asst. Prof. Dr. Şükran KAHVECİ Asst. Prof. Dr. Demet ÖZCAN BİÇİCİ Publication Board Asst. Prof. Dr. Ahmet Esad YURTSEVER, Specialist Ahmet Şenol ARMAĞAN, Res. Asst. Bahri Mert DEMİR, Lib. Officer Muhammad SHAHJAHAN Cover Design Özgür KIYAK, Gönül AKBULUT Tansu KISACIK © Istanbul Gelisim University Press Certificate Number: 47416 All rights reserved. Correspondence: Istanbul Gelisim University, Cihangir Mh. Sehit J. K. Er Hakan Oner Sk. No: 1 34310 Avcilar / Istanbul / TURKEY Phone: +90 212 4227000 Fax: +90 212 4227401 E-mail: ecoman@gelisim.edu.tr Web: https://dergipark.org.tr/en/pub/ecoman The Journal of Sustainable Economics and Management Studies is an international peer–reviewed journal and published biannually. The opinions, thoughts, postulations or proposals within the articles are but reflections of the authors and do not, in any way, represent those of the Istanbul Gelisim University. ii Advisory Board Prof. Dr. K. Ali AKKEMİK Prof. Dr. Nükhet DOĞAN Prof. Dr. İbrahim Halil EKŞİ Prof. Dr. Arhan S. ERTAN Prof. Dr. Anton Abdulbasah KAMIL Prof. Dr. Doğan Nadi LEBLEBİCİ Prof. Dr. Fatih SARIOĞLU Prof. Dr. Nimet Hülya TIRMANDIOĞLU TALU Prof. Dr. Mehmet Selçuk USLU Prof. Dr. Ahmet Burçin YERELİ Assoc. Prof. Dr. Hakan BEKTAŞ Assoc. Prof. Dr. Nilay YAVUZ Asst. Prof. Dr. Andrew Adewale ALOLA Asst. Prof. Dr. Festus Victor BEKUN Asst. Prof. Dr. Melik ERTUĞRUL & Reviewers for this Issue Asst. Prof. Dr. Ayla ALKAN Dr. Melike ÇETİN iii CONTENTS Pages iii Advisory Board & Reviewers for this Issue v Contents ORIGINAL RESEARCH ARTICLE 69-81 Should We Expect Bitcoin Markets to Be Efficient? Sema DUBE, Hayati ŞAHİN BOOK REVIEW 82-83 Book Review-1: The End of Normal: The Great Crisis and the Future of Growth Bülent EŞİYOK v Journal of Sustainable Economics and Management Studies (ECOMAN) ISSN: 2718-1065 & e-ISSN: 2791-8084 Vol. 2, Issue 2, December 2021 Article Statistics 1 1 ? 1 All articles submitted to our journal are analyzed by plagiarism detection tools. vi Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. Should We Expect Bitcoin Markets to Be Efficient?* Sema DUBE**, Hayati ŞAHİN*** Abstract We investigate whether Bitcoin markets demonstrate month-of-the-year effects, and whether such anomalies are present across markets that differ in terms of fees, trading requirements, size as well as the extent of legal support in their host countries. We use monthly return data for the period of 2015-2018 for Bitfinex, Bitstamp and Okcoin and find that returns were similar across the markets suggesting lack of internal frictions, and that all three Bitcoin markets showed a positive effect in December and a negative effect in January, followed by a positive effect in February. One explanation for the anomalies in the Bitcoin markets could be spillovers from the seasonal anomalies in broader markets, such as those posited by tax-loss or portfolio rebalancing hypotheses, which could result in some investors selling equity in December and repurchasing it in January and parking the proceeds in Bitcoin in the interim. If related to tax considerations, this situation could change as various jurisdictions start to enforce tax regulations for cryptocurrencies. Keywords: Bitcoin, Efficient Markets Theory, Month-of-the-Year Effect, Exchange Location, Calendar Anomaly, Cryptocurrency. JEL Classification: E42, G11, G12, G14, G15. 1. Introduction In an efficient market, asset returns should not exhibit any consistent patterns. Factors such as model misspecification, information asymmetry, behavioral issues, trading costs, and tax effects could all lead to markets inefficiencies (Fama, 1970, 1991). Even with markets increasingly utilizing technology to reduce frictions, patterns in asset-returns that go against market-efficiency have persisted. For example significant positive returns are observed before holidays due to behavioral issues (e.g., Lakonishok & Smidt,1988); while negative returns in December and positive returns in January arise due to trades induced by tax considerations, windows dressing, or information asymmetries (e.g., Rozeff & Kinney,1976; Haugen & Lakonishok, 1987; Ritter, 1988; Klock & Bacon, 2014). The Halloween Effect has been arise because people taking vacations during summer. Prior studies show that calendar-time Original Research Article Received: 27.07.2022 Accepted: 21.03.2023 * This work is based on the M.A. Thesis of the second author. Thesis Title: “A cross-market comparison of the month of the year anomaly for bitcoin.” Thesis Advisor: Sema Dube, Associate Prof. of Finance, Yeditepe University, Graduate School of Social Sciences, 2019. ** Corresponding Author, Associate Prof. of Finance, Yeditepe University, Department of International Finance, Istanbul, TÜRKİYE. E-mail: sdube@yeditepe.edu.tr ORCID https://orcid.org/0000-0003-2092-7308 *** Yeditepe University, Graduate School of Social Sciences, Financial Economics Program, Istanbul, TÜRKİYE. E-mail: hayati.sahin@std.yeditepe.edu.tr, hayatisahin6790@gmail.com ORCID https://orcid.org/0000-0001-7292- 3355 69 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. market anomalies exist in traditional currency markets as well (e.g., Bollerslev & Domowitz, 1993; Akram et al., 2008; Aydoğan & Booth, 2010; Kaul & Sapp, 2009, Popovic & Durovic, 2014; Caporale & Plastun, 2017). With globalization, global trading facilitated by technology, liquidity and use of leveraged instruments assets have become more correlated, especially during downturns (Elkhamhi & Stefanova, 2015; Warren, 2015), implying issues in one market can increasingly impact other markets. Negative returns in traditional markets in December if due to tax effects, followed by positive returns in January, could also lead to corresponding movements in and out of Bitcoins which could result in an opposite effect for Bitcoin markets, especially if Bitcoins are not taxed. Evidence for calendar-time market anomalies in the cryptocurrency markets such as the day- of-the-week effect (e.g., Decourt et al., 2017; Kurihara & Fukushima, 2017; Durai & Paul, 2018; Eyuboglu, 2018; Mbanga, 2018; Aharon & Qadan, 2019; Baur et al., 2019; Caporale & Plastun, 2019; Fraz et al., 2019; Kaiser, 2019; Yaya & Ogbonna, 2019), or the month-of-the-year effect (e.g., Eyuboglu, 2018; Baur et al.,2019; Fraz et al., 2019; Kaiser, 2019) is in general scarce, more so for the latter owing to the fact that crytocurrency markets have only been established starting 2009. Eyuboglu (2018), applying dummy variable regressions, finds significant positive Bitcoin returns for the months of February, October and November and insignificant positive returns for January during the period of 2013-2017 based on volume-averaged prices of all reported prices on the website Coinmarketcap.com. Fraz et al. (2019) applying a similar methodology find May, October and November to bring consistently higher Bitcoin returns than the other months, for the same period, based on price data from the cryptocurrency market capitalization website. Kaiser (2019), using similar data, examines month-of-the-year effect for 10 different cryptocurrencies using several methods during 2013- 2017 and 2013-2018 and finds that Bitcoin returns are significantly negative in January while other cryptocurrencies do not show any consistent patterns. Inefficiencies could also arise in terms of differences across exchanges. Bitcoin is the first and the most popular cryptocurrency, trading on several exchanges around the world since 2009 (Nakamoto, 2008; Brandvold et al., 2015). As of September 5, 2019, there were more than 200 cryptocurrency exchanges and Bitcoin had reached a market capitalization of about $194 Billion (Crypto-Compare, 2019a). To account for exchange-to-exchange differences, we examine the month-of-the-year effect in the three largest Bitcoin/U.S.Dollar (USD) exchanges: Bitfinex, Bitstamp and Okcoin. Bitcoin markets are not completely frictionless and Briere et al. (2015) find prices change based on the exchange. There are fee-differentials across exchanges with Bitstamp having the highest fees and Okcoin the lowest. Unlike Bitfinex and Okcoin, Bitstamp does not allow margin trading. On the other hand, Okcoin is the only exchange that requires ID verification for trading fiat currencies for Bitcoin (Bitfinex Exchange Review, 2019). Bitstamp has about 14 active markets and trading pairs and offers various payment methods, including credit cards. Bitfinex has around 128 active trading pairs and accepts only wire transfers. Okcoin has about 32 active trading pairs, and until June 2019, no credit card payments were accepted (Huillet, 2019). Daily trading volumes are highly volatile, with sharp peaks, but overall Bitfinex is larger than Bitstamp, which in turn is significantly larger than Okcoin (Crypto-Compare, 2019b). The exchanges are also based in countries with different 70 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. attitudes towards cryptocurrencies. Luxembourg, the base country for the Bitstamp exchange, supports the use of cryptocurrency as a means for investment and financing activities; Hong Kong, the headquarter for Bitfinex, is classified as a “fence sitter”; while China, the base country for Okcoin, is a strong opponent of the cryptocurrency markets (Thomson Reuters Report, 2017; Law Library of Congress Report, 2018). Bitstamp, Bitfinex and Okcoin have been active since 2011, 2012 and 2013 respectively. Bitstamp had changed its home country to the UK, another country supportive of Bitcoin, for about three years starting 2013 but has since moved back to Luxembourg. Baur et al. (2019) is the only prior study that examines the month- of-the-year effect in several Bitcoin exchanges. It uses heat maps to identity consistent patterns across time and exchanges and do not find any consistent month-of-the-year effect anomaly during 2011-2017 in Bitcoin returns for any exchange. We find a strong month-of-the-year effect anomaly in the Bitcoin markets during the period of our study. The results from our analyses across exchanges are broadly similar, and we find only minor differences with respect to the month-of-the-year effect anomaly among the three Bitcoin exchanges considered. The remainder of the work is organized as follows: Section 2 describes the data and methodology; Section 3 presents the results of our analyses; and Section 4 presents the conclusions. 2. Data and Methodology We collect the monthly closing price data for Bitcoin/USD from the Bitfinex, Bitstamp and Okcoin exchanges for 2015-2018 from Financial Contents (markets.financialcontents.com) and Crypto Data Download (cryptodatadownload.com) websites. Figure 1 shows the closing prices of the Bitstamp, Bitfinex and Okcoin exchanges for the period 12/2014-12/2018, and shows some differences in price across exchanges. The average daily closing price for Bitcoin during the sample period is $3,230.07 for Okcoin, $3,164.69 for Bitfinex and $3,075.87 for Bitstamp. The maximum price achieved during the sample period is $19,943.30 for Okcoin while it is $19,187.50 for both Bitstamp and Bitfinex. Okcoin always has the highest closing price while the Bitstamp exchange has the lowest one. During our sample period, Bitcoin prices have higher volatility in the Okcoin exchange with a standard deviation of above $4,000, compared to the Bitstamp and Bitfinex exchanges where the standard deviation of prices is around $3,700. 71 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. Figure 1: Bitcoin/USD Prices - Bitstamp, Bitfinex & Okcoin Exchanges, 12/2014-12/2018 100.000,00 10.000,00 1.000,00 100,00 Okcoin Bitfinex Bitstamp Notes: Prices are on logarithmic scale. Price data is from markets.financialcontents.com and cryptodatadownload.com We examine whether the Bitcoin markets are efficient by searching for the presence of the month-of-the-year effect anomaly in Bitcoin prices and whether any potential anomaly present in the market exhibits different characteristics across the three different exchanges. To ensure the stationarity of the variables used in our analyses, we apply the Augmented Dickey-Fuller (1979, 1981) unit root test of stationarity. Our results, which are not reported here to save space, show that the natural logarithm of Bitcoin prices are integrated of order 1 indicating that the series are stationary in first differences. 72 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. Figure 2: Monthly Bitcoin Returns - Bitstamp, Bitfinex & Okcoin Exchanges, 2015 - 2018 0,8 0,6 0,4 0,2 0,0 -0,2 -0,4 -0,6 OKCOIN BITSTAMP BITFINEX Notes: Underlying price data is from markets.financialcontents.com and cryptodatadownload.com We use monthly Bitcoin returns for our analyses estimated by the following equation: 𝐵𝑡 𝑅𝑡 = 𝑙𝑛 ( ) 𝐵𝑡−1 ...(1) where Rt is the monthly Bitcoin return and Bt is the Bitcoin price for Month t. We take the closing price in the last trading day at the end of Month t as the Bitcoin price for Month t. As shown in Figure 2, Bitcoin monthly returns across the three different cryptocurrency exchanges during the same time frame are essentially identical. We examine the month-of-the year effect using: 𝑅𝑡 = 𝛽𝐽𝑎𝑛(𝐷𝐽𝑎𝑛) + 𝛽𝐹𝑒𝑏(𝐷𝐹𝑒𝑏)𝑡 + 𝛽𝑡 𝑀𝑎𝑟 (𝐷𝑀𝑎𝑟)𝑡 + 𝛽𝐴𝑝𝑟(𝐷𝐴𝑝𝑟) + 𝛽𝑀𝑎𝑦(𝐷𝑀𝑎𝑦)𝑡 𝑡 + 𝛽𝐽𝑢𝑛𝑒(𝐷𝐽𝑢𝑛𝑒) + 𝛽𝑡 𝐽𝑢𝑙𝑦(𝐷𝐽𝑢𝑙𝑦) + 𝛽𝐴𝑢𝑔(𝐷𝐴𝑢𝑔) + 𝛽𝑆𝑒𝑝(𝐷𝑆𝑒𝑝) + 𝛽𝑂𝑐𝑡 (𝐷 𝑡 𝑡 𝑡 𝑂𝑐𝑡 )𝑡 +𝛽𝑁𝑜𝑣(𝐷𝑁𝑜𝑣)𝑡 + 𝛽𝐷𝑒𝑐(𝐷𝐷𝑒𝑐)𝑡 + 𝑒𝑡 ...(2) where t is the month of the sample period; the dummy (Di)t, where i represents the calendar month is 1 if t is i, zero otherwise; Rt is the monthly Bitcoin return; βi represents the average 73 January-15 April-15 July-15 October-15 February-16 May-16 August-16 December-16 March-17 June-17 September-17 January-18 April-18 July-18 November-18 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. return for Month i; and et is the error term at month t. A nonzero coefficient for a calendar month dummy indicates a month-of-the-year anomaly for that month. The absence of the month-of-the-year effect anomaly requires the finding that none of the calendar month dummy variables in Eq. 2 is statistically significant. If any calendar month shows significance in Eq. 2, we further test whether the observed differences in average returns between a significant month and the remaining 11 months are statistically meaningful. For instance, if the dummy variable representing January is statistically significant in Eq. 2, we examine whether the difference between January returns those of other months are significantly different from zero, using: 𝑅𝑡 = 𝑐 + 𝛽𝐹𝑒𝑏(𝐷𝐹𝑒𝑏)𝑡 + 𝛽𝑀𝑎𝑟(𝐷𝑀𝑎𝑟)𝑡 + 𝛽𝐴𝑝𝑟(𝐷𝐴𝑝𝑟) + 𝛽𝑀𝑎𝑦(𝐷𝑀𝑎𝑦) + 𝛽𝐽𝑢𝑛𝑒(𝐷𝑡 𝑡 𝐽𝑢𝑛𝑒)𝑡 +𝛽𝐽𝑢𝑙𝑦(𝐷𝐽𝑢𝑙𝑦) + 𝛽𝐴𝑢𝑔(𝐷𝐴𝑢𝑔) + 𝛽𝑡 𝑡 𝑆𝑒𝑝( 𝐷𝑆𝑒𝑝) + 𝛽𝑡 𝑂𝑐𝑡 (𝐷𝑂𝑐𝑡)𝑡 + 𝛽𝑁𝑜𝑣(𝐷𝑁𝑜𝑣)𝑡 +𝛽𝐷𝑒𝑐(𝐷𝐷𝑒𝑐)𝑡 + 𝑒𝑡 ...(3) Here the constant term, c, in Eq. 3 represents the average return on January and the coefficients (βi) represent the difference in average returns between the month i and January. The remaining variables are as defined in Eq. 2. All regressions use HAC standard errors and covariance estimators based on Newey and West (1987) that adjust for autocorrelation and heteroscedasticity. We also report Wald F test statistics for all regressions, which is commonly used for regressions with multiple dummy variables (Davidson & MacKinnon, 2004; Stata FAQ). 3. Results Table 1 shows the results from regressions based on Eq. 2. It shows that the month-of-the-year effect anomaly exists for January, February and December in all of the three cryptocurrency exchanges. February and December have significant positive returns while January has significant negative returns for all markets. For the Bitfinex exchange only, October also has the month-of-the-year effect anomaly with significantly positive returns, with results being only marginally significant for Bitstamp and Okcoin. Marginally-significant negative returns were observed across the exchanges for March, followed by marginally-significant positive returns for April. 74 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. Table 1. The Month of the Year Effect Anomaly for Bitcoin Returns, 2015 - 2018 BITFINEX BITSTAMP OKCOIN βi βi βi DJan -0.207959** -0.211302** -0.218873** DFeb 0.139105*** 0.135190*** 0.142209*** DMar -0.148174* -0.147677* -0.137116* DApr 0.151960* 0.139664* 0.105865* DMay 0.088382 0.116887 0.166571 DJune 0.080189 0.069706 0.043079 DJuly 0.088925 0.084686 0.074553 DAug 0.030136 0.028065 0.030197 DSep -0.014070 -0.014459 -0.017527 DOct 0.196078** 0.189732* 0.195249* DNov 0.049232 0.058348 0.053093 DDec 0.188917*** 0.183765** 0.194400** R2 / Adjusted R2 0.313227 / 0.103400 0.296794 / 0.081900 0.302630 / 0.089600 Wald F-stat. 13.7706*** 11.9857*** 12.9691*** Notes: The table reports regression results for Eq. (2) using Bitcoin returns from the Bitfinex, Bitstamp and Okcoin exchanges during 2015-2018. Each Di represents a dummy variable for calendar month i with a value of 1 when the month is i and 0 otherwise. It includes the coefficient of each dummy variable. The monthly Bitcoin return data is the dependent variable estimated using the Bitcoin closing price in the last trading day of the month. There are 48 monthly observations. Bitcoin prices are obtained from the websites markets.financialcontents.com & cryptodatadownload.com. R-squared and Adjusted R- squared values are reported at the end of the table. Regressions use HAC estimators adjusted for autocorrelation and heteroscedasticity (Newey & West, 1987). Wald F test statistics are reported in the last row. *, **, *** denote significance at 10%, 5% and 1%, respectively Results for differences in returns between months which show month-of-the-year effect anomaly (i.e., January, February, October and December), and the remaining months, based on Eq. (3), are in Table 2. The table has 4 panels, Panels A, B, C and D, reporting January, February, October and December effects, respectively. For each panel we report only months with statistically significant coefficients to save space. Overall, the results show the existence of significant differences for these months relative to others, as well as the uniformity of the results across exchanges. For January, Panel A of Table 2 shows that January returns appear smaller than all other months; significantly so relative to February, April, July, September, October and December. The only inter-exchange differences appear in relative significance. For May, the differences are only marginally significant for Bitfinex and Bitstamp while being significant for Okcoin, and for June the results are significant for Bitfinex and Bitstamp while being only marginally significant for Okcoin. 75 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. Table 2. Significant Monthly Effect Anomalies for Bitcoin Returns, 2015-2018 BITFINEX BITSTAMP OKCOI N βi βi βi Panel A. January Effect DFeb 0.347064*** 0.346491*** (0.361081*** DApr 0.359919*** 0.350966** 0.324737*** DMay 0.296341* 0.328189* 0.385443** DJune 0.288148** 0.281008** 0.261952* DJuly 0.296884*** 0.295988*** 0.293425*** DSep 0.193889** 0.196843** 0.201345** DOct 0.404037*** 0.401034*** 0.414122*** DDec 0.396877*** 0.395067*** 0.413272*** R2 / Adjusted R2 0.313227 / 0.037600 0.296794 / 0.081900 0.302630 / 0.035200 Wald F-stat. 9.175000*** 8.231900*** 7.988800*** Panel B. February Effect DJan -0.347060*** -0.346491*** -0.361081*** DMar -0.287280*** -0.282867*** -0.279324*** DSep -0.153180** -0.149648** -0.159736*** R2 / Adjusted R2 0.313227 / 0.103380 0.296794 / 0.081925 0.302630 / 0.089545 Wald F-stat. 12.327270*** 11.985740*** 12.969140*** Panel C. October Effect DJan -0.404037*** -0.401034*** -0.414122*** DMar -0.344252*** -0.337409** -0.332365*** DSep -0.210147** -0.204191* -0.212776* R2 / Adjusted R2 0.313227 / 0.103380 0.296794 / 0.081925 0.302630 / 0.089545 Wald F-stat. 12.327270*** 11.985740*** 12.969140*** Panel D. December Effect DJan -0.396877*** -0.395067*** -0.413272*** DMar -0.337091*** -0.331442*** -0.331516*** DSep -0.202987** -0.198224** -0.211927** R2 / Adjusted R2 0.313227 / 0.103380 0.296794 / 0.081925 0.302630 / 0.089545 Wald F-stat. 12.327270*** 11.985740*** 12.969140*** Notes: The regression results for Eq. (3) for January, February, October and December anomalies are reported in Panels A, B, C and D, respectively. The dependent variable, the monthly Bitcoin returns, are the same as explained in Table 1. Each Di represents a dummy variable for calendar month i with a value of 1 when the month is i and 0 otherwise. For each panel, only the months with significant coefficients are shown for space sake. R-squared, Adjusted R-squared and Wald F test statistics are reported for each panel. Regressions use HAC estimators adjusted for autocorrelation and heteroscedasticity (Newey & West, 1987). *, **, *** denote significance at 10%, 5% and 1%, respectively. 76 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. For all three exchanges, February returns shown in Panel B of Table 2 are significantly higher than the January, March and September returns. As Panel C of Table 2 indicates, October returns are October returns are significantly higher than the January and March returns. The October effect appears only in Bitfinex in Table 1, and is marginally significant for Bitstamp and Okcoin. In Panel C of Table 2, October returns for Bitfinex are also significantly different from September returns, but only marginally so for Bitstamp and Okcoin. Finally, Panel D of Table 2 shows that the December effect is consistent across the three exchanges with December returns being significantly higher than the January, March and September returns. 4. Conclusions We examine if the month-of-the-year effect anomaly is present in the Bitcoin markets during the period of 2015-2018, and whether the nature of this anomaly changes based on the market. We apply autocorrelation and heteroscedasticity adjusted robust least squares regressions with dummy variables for Bitcoin/USD returns in three different Bitcoin exchanges, Bitfinex, Bitstamp and Okcoin which have different fees, trading requirements as well as levels of support for cryptocurrency in the countries in which the exchanges are based. Overall, our results support the findings of the studies claiming that the Bitcoin markets are not efficient because of various anomalies (e.g., Urquhart, 2016; Kurihara & Fukushima, 2017; Makarov & Schoar, 2018; Reynolds et al., 2018; Fraz et al., 2019), and in line with Kaiser (2019) who suggests a reverse January effect based on cryptocurrency market characteristics such as uncertainty related to market fundamentals and informational inefficiencies; and contradict prior findings of weak efficiency such as Nadarajah and Chu (2017) and Yaya and Ogbonna (2019). In contrast to Baur et al. (2019), we confirm that month-of-the-year effects hold when different exchanges are considered separately. One reason for our observed positive December effect and negative January effect, that are the opposite in direction to the effects seen in traditional markets, could be that more money may flow into Bitcoins in December and may flow back out in January. Bauer (2019) suggests, based on trading patterns through the day and after-hours, as well as for different days as well as months, that institutional investors are a significant component of Bitcoin trades, especially USD-Bitcoin or Euro-Bitcoin trades. We do not seek evidence for trading for tax-effects in trading volume patterns as it has been estimated that up to 80% of the volumes may be wash trading (BlockChain Transparency Institute, 2018). Given that the Halloween Effect, which has been hypothesized as arising due to people taking vacations in summer needing to pay for them and not wanting to be invested in the stock market out of a fear of sudden downturns (Jacobson & Zhang, 2018), we would not expect a significant positive return for Bitcoins over summer months either as Bitcoins are considered highly risky. Table 1 shows none of the results for summer months to be significantly different from zero, although March has 77 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. marginally significant negative returns across exchanges while April shows marginally- significant positive returns. We would note that Bitcoins, as well as other cryptocurrencies, are relatively new and even regulatory attitudes have yet to be finalized. The supply of Bitcoins is limited, and ultimately finite. Events such as forks and halving also impact prices. The volatility in prices, and the lack of a determinant of value except for comparison with the dollar, suggests potentially less of a role as currency and more as a virtual commodity. In this regard, the US Internal Revenue Service has been increasingly active in informing taxpayers of their reporting requirements with regard to their cryptocurrency trades (Green, 2019). If jurisdictions start to enforce taxation related to Bitcoins to a significant extent, it would be of interest to repeat the analyses and determine if the December-January effects observed for our sample period persist. The Bitcoin market is the largest cryptocurrency market, and the most well-known. Even so it is relatively small in that there are companies on stock exchanges that have higher valuations than the entire market capitalization of Bitcoin. It is also a relatively new market with greater underlying risks. The concept of a virtual commodity itself is relatively new, and it is not entirely clear how the use of Bitcoin as a full-fledged currency will evolve. From a broader perspective, if such a market is part of a larger set of markets, any inefficiencies that exist in the wider markets could lead to corresponding patterns in return, even if that market itself has fewer inherent sources of inefficiency as money can always flow from one market to others. REFERENCES AHARON, D.Y. and QADAN, M. (2019). Bitcoin and the Day-of-the-Week Effect. Finance Research Letters, 31 (C). AKRAM, Q.F., RIME, D. and SARNO, L. (2008). Arbitrage in the Foreign Exchange Market: Turning on the Microscope. Journal of International Economics, 76, 237–253. AYDOĞAN, K. and BOOTH, G. (2003). Calendar Anomalies in the Turkish Foreign Exchange Markets. Applied Financial Economics, 13(5), 353-360. BAUR, D.G., CAHILL, D., GODFREY, K. and LIU, Z. (2019). Bitcoin Time-of-Day, Day-of- Week and Month-of-Year Effects in Returns and Trading Volume. Finance Research Letters, 31, 78–92. BITFINEX EXCHANGE REVIEW (2019). Cointelligence Research and Analysis for the Crypto- Economy. Retrieved from: https://www.cointelligence.com/exchanges_list/bitfinex/ BLOCKCHAIN TRANSPARENCY INSTITUTE (2018). Market Surveillance Report – December 2018. Retrieved from https://www.bti.live/reports-december2018/ BRANDVOLD, M., MOLNAR, P., VAGSTAD, K. and VALSTAD, O.C.A. (2015). Price Discovery on Bitcoin Exchanges. Journal of International Financial Markets, Institutions and Money, 36, 18-35. 78 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. BRIERE, M., OOSTERLINCK, K. and SZAFARZ, A. (2015). Virtual Currency, Tangible Return: Portfolio Diversification with Bitcoin. Journal of Asset Management, 16(6), 365–373. BOLLERSLEV, T. and DOMOWITZ, I. (1993). Trading Patterns and Prices in the Interbank Foreign Exchange Market. Journal of Finance, 48, 1421–1443. CAPORALE, G.M. and PLASTUN, A. (2017). Calendar Anomalies in the Ukrainian Stock Market. Investment Management and Financial Innovations, 14(1), 104–114. CAPORALE, G.M. and PLASTUN, A. (2019). The Day of the Week Effect in the Cryptocurrency Market. Finance Research Letters (in press). CRYPTO-COMPARE (2019a). Coins List. Cryptocompare Website. Retrived from: https://www.cryptocompare.com/coins/list/USD/1 CRYPTO-COMPARE (2019b). Exchanges. Cryptocompare Website. Retrived from: https://www.cryptocompare.com/exchanges/#/overview DAVIDSON, R. and MACKINNON, J. (2004). Econometric Theory and Methods. Oxford University Press, 244: New York. DECOURT, R.F., CHOHAN, U.W. and PERUGINI, M.L. (2017). Bitcoin Returns and the Monday Effect. Horizontes Empresariales, 16, 4-14. DICKEY, D. and FULLER, W. (1979), Distribution of the Estimators for Autoregressive Time Series with a Unit Root. Journal of the American Statistical Association, 74, 427- 431. DICKEY, D. and FULLER, W. (1981). Likelihood Ratio Statistics for Autoregressive Time Series with a Unit Root. Econometrica, 49, 1057- 1072. DURAI, S.R.S. and PAUL, S. (2018). Calendar Anomaly and the Degree of Market Inefficiency of Bitcoin. Madras School of Economics Working Paper No. 168/2018: Chennai, India. ELKAMHI, R., and STEFANOVA, D. (2015). Dynamic Hedging and Extreme Asset Co- movements. The Review of Financial Studies, 28(3), 743-790. EYUBOGLU, K. (2018). Examining Day of the Week and Month of the Year Effects in Bitcoin and Litecoin Markets. Cankırı Karatekin University Journal of the Faculty of Economics and Administrative Sciences, 8(1), 165-183. FAMA, E.F. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. Journal of Finance, 25, 383-417. FAMA, E.F. (1991). Efficient Capital Markets: II. Journal of Finance, 46(5), 1575–1617. FRAZ, A., HASSAN, A. and CHUGHTAI, S. (2019). Seasonality in Bitcoin Market. Nice Research Journal, 12(1), 2013-2017. GREEN, R. (2019). Watch Out Cryptocurrency Owners, The IRS Is On The Hunt. Forbes. Retrieved from https://www.forbes.com/sites/greatspeculations/2019/07/31/watch-out- cryptocurrency-owners-the-irs-is-on-the-hunt/ 79 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. HAUGEN, R. and JORION, P. (1996). The January Effect Still Here After All These Years. Financial Analysts Journal, 52(1), 27-31. HAUGEN, R.A. and LAKONISHOK, J. (1987). The Incredible January Effect. Dow Jones-Irwin: Illinois. HUILLET, M. (2018). Okcoin, Payments Processer Simplex Partmer on Card-Enabled USD Deposits for International Users. Cointelegraph News, June 17. Retrieved from: https://cointelegraph.com/news/okcoin-payments-processer-simplex-partner-on-card- enabled-usd-deposits-for-intl-users JACOBSEN, B., and ZHANG, C. Y. (October 1, 2018). The Halloween Indicator, 'Sell in May and Go Away': Everywhere and All the Time. Available at SSRN: https://ssrn.com/abstract=2154873 or http://dx.doi.org/10.2139/ssrn.2154873 KAISER, L. (2019). Seasonality in Cryptocurrencies. Finance Research Letters (in press). KAUL, A. and SAPP, S. (2009). Trading Activity, Dealer Concentration and Foreign Exchange Market Quality. Journal of Banking and Finance, 33, 2122–2131. KLOCK, S. and BACON, F. (2014). The January Effect: A Test of Market Efficiency. ASBBS Annual Conference: Las Vegas, Nevada, USA, 21(1), 423-434. KURIHARA, Y. and FUKUSHIMA, A. (2017). The Market Efficiency of Bitcoin: a Weekly Anomaly. Journal of Applied Finance and Banking, 7(3), 57–64. LAKONISHOK, J. and SMIDT, S. (1988). Are Seasonal Anomalies Real? A Ninety-Year Perspective. Review of Financial Studies, 1, 403-425. LAW LIBRARY OF CONGRESS REPORT (2018). Regulation of Cryptocurrency Around the World, June. The Law Library of Congress, Global Legal Research Directorate: Washington, D.C.,USA. MAKAROV, I. and SCHOAR, A. (2018). Trading and Arbitrage in Cryptocurrency Markets. SSRN Working Paper No: 3171204: London, UK. MBANGA, C.L. (2018). The Day-of-the-Week Pattern of Price Clustering in Bitcoin. Applied Economics Letters, 26(10), 807-811. NADARAJAH, S. and CHU, J. (2017). On the Inefficiency of Bitcoin. Economic Letters. 150, 6– 9. NAKAMOTO, S. (2008). Bitcoin: A Peer-to-peer Electronic Cash System, Unpublished undated Working Paper. Retrieved from: http://bitcoin.org/bitcoin.pdf NEWEY, W. K. and WEST, K. D. (1987). A Simple, Positive Semi-definite, Heteroskedasticity and Autocorrelation Consistent Covariance Matrix. Econometrica, 55(3), 703–08. POPOVIC, S. and DUROVIC, A. (2014). Intraweek and Intraday Trade Anomalies: Evidence from Forex Market. Applied Economics. 46 (32), 3968–3979. 80 Sema Dube, Hayati Şahin, “Should We Expect Bitcoin Markets to Be Efficient”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 69-81. REYNOLDS, J., SOEGNER, L., WAGNER, M. and WIED, D. (2018). Deviations from Triangular Arbitrage Parity in Foreign Exchange and Bitcoin Markets. SSRN Working Paper No. 3148094; Lugano, Switzerland. RITTER, J. (1988).The Buying and Selling Behavior of Individual Investors at the Turn of the Year. Journal of Finance, 43 (3), 701–719. ROZEFF, M. S. and KINNEY, W. R. Jr. (1976). Capital Market Seasonality: The Case of Stock Returns. Journal of Financial Economics, 3, 379-402. STATA FAQ (undated). How to Run a Likelihood and Wald Test. UCLA Institute for Digital Research and Education. Retrieved from: https://stats.idre.ucla.edu/stata/faq/how-can-i- perform-the-likelihood-ratio-wald-and-lagrange-multiplier-score-test-in-stata/ THOMSON REUTERS REPORT (2017). A World of Cryptocurrencies. Thomson Reuters Blog. Retrieved From: https://blogs.thomsonreuters.com/answerson/wpcontent/uploads/sites/3/2017/10/Worl d-of-Cryptocurrencies-graphic.pdf URQUHART, A. (2016). The Inefficiency of Bitcoin. Economics Letters, 148, 80–82. WARREN, R. (2015). Your Portfolio has more risk than you think. Forbes. Retrieved from https://www.forbes.com/sites/randywarren/2015/11/04/your-portfolio-has-more-risk- than-you-think/ YAYA, O.S. and OGBONNA, E.A. (2019). Do We Experience Day-of-the-Week Effects in Returns and Volatility of Cryptocurrency?, MPRA Working Paper No. 91429: Ibadan, Nigeria. 81 Bülent Eşiyok, “Book Review-1: The End of Normal: The Great Crisis and the Future of Growth”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 82-83. BOOK REVIEW-1 Bülent EŞİYOK* The End of Normal: The Great Crisis and the Future of Growth Author: James K. GALBRAITH Published 2014 by Simon & Schuster, 291 Pages. ISBN: 9781451644937 In this book, the author suggests that it is time that we reconsider economic principles or the way we understand them as they were not helpful to either foresee the world financial crisis or predict subsequent low employment and economic growth. Economic principles that predicted a return to a sustained growth after a shock were perceived to be “normal”. In first section, the author challenges our understanding of economic growth dynamics all along, starting from post-war era to the time of financial crisis of 2008. For instance, the author contends that a simple two-factor Solow growth model inclusive of capital and labour ignores the contribution of, among others, cheap oil and stable wages to the post-war era growth while overstating the role of technology, Solow-residual. Growth was important because it was the solution for poverty and unemployment. The author argues that although oil-shocks and resource costs pressures in 1970s proved to be inimical to economic growth, arms sale to resource rich countries and oil purchase in return of Treasury bills created a delusion that economic growth could be sustained in the years to come, thanks to increasing personal, household and corporate borrowing. In the second section, four factors-named as “the four horsemen of the end of growth”- are cited that would challenge what was perceived to be normal. The first factor is the rising resource costs relative to the prices of goods and services; the second is the ineffective military power; the third is the labour-saving opportunities in production brought by the digital revolution; the fourth is non-adherence to law and ethics in the finance sector. * Assoc. Prof., Istanbul Gelisim University, Faculty of Economics, Administrative and Social Sciences, Department of Economics and Finance, Istanbul, TURKEY. E-mail: besiyok@gelisim.edu.tr ORCID https://orcid.org/0000- 0001-7469-5309 82 Bülent Eşiyok, “Book Review-1: The End of Normal: The Great Crisis and the Future of Growth”, Journal of Sustainable Economics and Management Studies, Vol. 2, Issue 2, December 2021, pp. 82-83. Given that we are unlikely to jump on to a high economic growth path as in the post-period due to the reasons aforementioned above, the author starts a debate in the third section as to what can be done in a world characterised by a slow-growth. Due to high fixed costs stemming from extensive energy use in keeping the army operational, the author suggests a provocative measure to shed air force branch of army except troop and cargo transport; after all, maintaining a large military power proved to be useless in securing oil supply. Some other measures concern tax, wage, credit and social insurance domains to create a slow growth system. Such a system should create –the author argues- “a qualitatively different form of capitalism: based on more decentralised economic units with relatively low fixed costs, relatively high use of labour compared with machinery resources, relatively low expected rates of return, but mutually supported by a framework of labour standards and social protection.” The last line of the book calls attention to the question of climate change. Creating and sustaining a slow-growth system requires concerted efforts not only at national but also international level; therefore, it remains to be seen whether such a system can be built in near future. 83