May 31, 2019
Dear Sir/Madam, Re: Bitwise Comment dated- 05/24/2019 for file number File No. SR-NYSEArca-2019-01 I want to take this opportunity to query some of the points made in the above comment by Bitwise Asset Management. - In your report, you set out to demonstrate and conclude that “As a digital commodity, bitcoin’s spot trading market should be among the most orderly and efficient in the world.” “Once you clear away the noise, the data suggest that that is simply true.” (Pg 85. https://www.sec.gov/comments/sr-nysearca-2019-01/srnysearca201901-5574233-185408.pdf) However, it is known that ‘an orderly market is any market in which the supply and demand are equal. The orderly market would thus be said to be in a state of equilibrium.’. This fact is not the case with Bitcoin. Bitcoin has the following characteristics, which make it just the opposite: the supply is inelastic, (which increases volatility) and the demand drivers are opaque at best (https://seekingalpha.com/article/4169186-depth-look-economics-bitcoin). Moreover, if you are referring to market efficiency, then you have not defined whether it is ‘weak, semi-strong or strong’ and justified why it is so to base your research findings. You have not demonstrated that this market is orderly and efficient through the research you have conducted. You have picked and chosen data to paint a picture, but it excludes many factors which I will discuss. - You have excluded Korea from your analysis citing Capital Controls (See footnote Pg 34. https://www.sec.gov/comments/sr-nysearca-2019-01/srnysearca201901-5574233-185408.pdf) but have included China (where most of the Bitcoin mining is still undertaken) and Hong Kong (the latter is a marketplace located in your top 10) where Capital Controls are in place. Is this because of the Kimchi Premium that was seen in Korea just before the crash at the end of 2017? Is it not essential to include All countries with Capital Controls in your analysis? Furthermore, your definition “Arbitrage is “the nearly simultaneous purchase and sale of securities or foreign exchange in different markets in order to profit from price discrepancies.”1 By this definition, bitcoin should be among the most arbitrage-able goods in the world. Subject to exchange-level constraints, such as fees, withdrawal issues or the perceived risk of default, you would expect bitcoin to trade at the same price across exchanges around the world, since it can be bought, transferred, and sold across exchanges with extremely limited frictions.” (Pg 2. https://www.sec.gov/comments/sr-nysearca-2019-01/srnysearca201901-5574233-185408.pdf). This would be then untrue if there are different prices for Bitcoin in different countries- such as Korea and what we saw in India during the peak at the end of 2017. Also, in a paper by the London School of Economics (LSE), it states: “One additional factor to consider is that some companies manage exchanges in several countries. For example, Coinbase has operations in Australia, Canada, Europe, Great Britain, and the US. The order book of each of these exchanges are separate, and customers from different countries can usually only trade cryptocurrencies on their local exchange and in their local currency. However, the exchanges that operate across regions might be able to arbitrage and potentially circumvent some capital controls. While we do not find a consistent impact of having overlapping exchanges on the arbitrage spreads of countries, it is possible that the existence of this channel adds noise to the estimation of capital controls.” (See footnote Pg 4. http://personal.lse.ac.uk/makarov1/index_files/CryptocurrencyMarkets.pdf). What about Binance, Bitfinex and bitFlyer who are all non-US based but in your top10? - You have also restricted your trading pairs which is unusual as this marketplace is not soley based upon the trading pairs you have chosen: “We took into consideration all trading pairs where BTC was the base currency (e.g. BTC/USD rather than EOS BTC) and where the quote currency was either a fiat currency (e.g. USD, EUR, JPY) or a stablecoin / (e.g. USDT47).” (Pg 15. https://www.sec.gov/comments/sr-nysearca-2019-01/srnysearca201901-5574233-185408.pdf) This fact is convenient as there do exist arbitrage opportunities between cryptocurrencies-to-cryptocurrencies. This point is an important one to analyse because although AML and KYC exist for most exchanges for fiat-to-cryptocurrency transactions. However, for cryptocurrency-to-cryptocurrency trading, you can sign up on most exchanges (including most listed in your top 10) with a name and an email address. AML and KYC are not required. Due to this, it makes the argument of Capital Control weak and the risk of Manipulation of Bitcoin via this method a real threat. It should also be noted that exchanges such as “…..Bitstamp and Kraken, which allow trading against multiple currencies independent of the location of the customer. While this discussion suggests that in the majority of exchanges customers cannot easily arbitrage across at currencies, the exchanges that operate across regions might be able to arbitrage across regions and potentially circumvent some capital controls.” (Pg 8. http://personal.lse.ac.uk/makarov1/index_files/CryptocurrencyMarkets.pdf) - You state “We acquired historical bitcoin trade data from Kaiko55 and Nomics56 for parts of our analyses that required a continuous historical data set.” (Pg 18 https://www.sec.gov/comments/sr-nysearca-2019-01/srnysearca201901-5574233-185408.pdf). However, you have not provided a longitudinal picture of all the exchanges and provided a very short snapshot between 4/28/19 - 5/5/19. Is there a reason for just sticking with these dates? Why have you not offered a more comprehensive picture from 2017 for ALL the exchanges? - You have also just focussed most of your analysis on six exchanges out of the ten, and I quote: “As a baseline, we’ll use the six exchanges that have a BitLicense from the New York State Department of Financial Services.59” (Pg. 21 https://www.sec.gov/comments/sr-nysearca-2019-01/srnysearca201901-5574233-185408.pdf). If I were analysing and providing an argument for ten exchanges, then I would explain all 10. For example, it is essential to analyse Binance and Kraken because the former has not registered with the U.S. Department of Treasury as an MSB, and the latter has failed to pursue a BitLicense. Accidental oversight or not both exchanges have had a recent negative press, and the CEO of Kraken has said “…..traders are primarily interested that an exchange works quickly and does not suffer downtime. They also want security and privacy, including a “minimal document requirement for on boarding”. However, Kraken suggest that traders simply do not care about many of the things that concern regulators, including regulatory approval, “being protected from making risky investments” and, most controversially, “being protected from market manipulation” (https://cryptocoinspy.com/kraken-exchange-says-traders-dont-care-about-market-manipulation). These comments should be of concern. - Finally, this comment is of particular interest “The last spike at the end of the graph also indicates there was a stronger preference to trade 9.9 - 10.0 BTC than any other trade size in the 6 - 10 BTC range. We decided to cut off the graph after 10.0BTC, not because no trades larger than 10.0 BTC ever happen, but because the vast majority of trade volume occurs in the 0-10 BTC range and it is visually helpful to focus on this range.” (Pg 21. https://www.sec.gov/comments/sr-nysearca-2019-01/srnysearca201901-5574233-185408.pdf) Yes, it may be "visually helpful.” , but as you state “not because no trades larger than 10.0 BTC ever happen” and they do, so the narrative of your report could come out looking very different if these were included for all exchanges. The report, in other words, is not complete due to this additional exclusion. Best Regards, Louise