Recently, CoinMarketCap (CMC), arguably the industry’s best-known cryptocurrency market data service, announced an initiative to provide “greater transparency, accountability, and disclosure from projects in the crypto space.” The move followed recent reports on fake volume data and wash trading among cryptocurrency exchanges that were published last month.
Now, all exchanges are required to provide mandatory application programming interface (API) data to CMC by June 2019. Those who fail to do so risk getting delisted from the platform. So, can this brand new scheme cleanse the market from untrustworthy data?
Fake volume is one of crypto market’s chief problems: two reports
Recently, a number of researches highlighted the problem of fake volume among crypto exchanges, suggesting that the majority of platforms claim to handle unrealistic amounts of transactions. As explained by Changpeng Zhao, the CEO of Binance, some exchanges alter their volume to get ranked higher on popular trackers like CMC, and hence get exposure and attract new clients.
The Tie: 90% of the volume is fake, 75% of crypto exchanges look suspicious
While the problem of fake volume isn’t particularly new to the crypto market, at least two recent reports have stirred up a new wave of discussion. First, on March 18, trading analytics platform the Tie reported that almost 90% of cryptocurrency exchanges’ reported trade volumes may be fake, and that as much as three-quarters of those platforms have suspicious volumes.
To conduct the research, the Tie took the reported trading volume for the last 30 days of the top 100 exchanges. They subsequently divided that data by the exchange’s website visits over 30 days estimated by SimilarWeb to determine the volume per visit.
Thus, to calculate the expected volume, the researchers used a weighted average of the trading volumes per website visit across Binance, Coinbase Pro, Poloniex, Gemini and Kraken — resulting in $591 — and multiplied this number by the web views. The Tie explained that it picked these exchanges “because of large usage among institutions, reputation within the market, and because their web viewership appeared consistent with their reported trading volumes.”
“In total we estimated that 87% of exchanges reported trading volume was potentially suspicious and that 75% of exchanges had some form of suspicious activity occurring on them,” the organization tweeted at the time, adding that it affects the larger picture:
13/— The TIE (@TheTIEIO) March 18, 2019
If each exchange averaged the volume per visit of CoinbasePro, Gemini, Poloniex, Binance, and Kraken, we would expect the real trading volume among the largest 100 exchanges to equal $2.1B per day. Currently that number is being reported as $15.9B. pic.twitter.com/jZzezJMmKk
Notably, on March 21, two exchanges featured in the research as having questionable figures — LBank and Bit-Z — dethroned Binance in terms of the adjusted trade volume on CMC. According to research presented by the Tie, LBank’s estimated reported volume per website visit amounts to $65,850.
However, the Tie admitted that its research had certain limitations: Specifically, the website views didn’t take into account API, mobile application trades and desktop client trades.
Because of that, the data could, in theory, just mean that either a much more significant than average portion of LBank users use the API, desktop or mobile clients, or that an LBank user trades over $65,000 per session on average. The Tie notes:
“There were limitations to this report including some of the aforementioned, but the point of the exercise was to show those exchanges that appear most suspicious and to start a greater conversation around wash trading, transaction mining, and liquidity.”
Bitwise: 95% of bitcoin trading volume on unregulated exchanges appears to be fake
On March 20, another substantial report on fake volume surfaced. Issued by cryptocurrency index fund provider Bitwise Asset Management, it argued that 95% of bitcoin trading volume on unregulated exchanges appears to be fake or noneconomic in nature.
Notably, Bitwise sourced its data from CMC, which it claims includes a large amount of this suspect data, “thereby giving a fundamentally mistaken impression” of the actual size of the bitcoin market.
Bitwise ultimately wrote that the real market for BTC is “significantly smaller, more orderly, and more regulated than commonly understood” — amounting in reality to $273 million instead of the $6 billion reported on CMC.
To prove its point, Bitwise first analyzed Coinbase Pro as an example of a regulated exchange to outline trustworthy trading patterns, including an “unequal and streaky” mix of red (sell orders) and green (buy orders) trades, whose distribution fluctuates considerably at any given time.
Further, Bitwise studied spread as a parameter, noting:
“It’s [the spread is] $0.01. At the time this screenshot was taken, bitcoin was trading at $3,419. That means bitcoin was trading at a 0.0003% spread, making it amongst the tightest quoted spread of any financial instrument in the world.”
Coinbase Pro reported around $27 million in daily traded volume of BTC at the time of Bitwise’s analysis — as compared with $480 million reported by Coinbene. The latter was used by the index fund provider to demonstrate the patterns typical of what it characterizes as “suspicious exchanges.”
Suspect signs included an unlikely perfect alternating pattern of green and red trades, as well as a lack of trades with round numbers or small values. On Coinbene, buy and sell orders also appear in timestamped pairs, with one compensating the other. Moreover, the spread on Coinbene at the time of Bitwise’s analysis was $34.74: “that compares to $0.01 on Coinbase Pro. It is surprising that an exchange claiming 18x more volume than Coinbase Pro would have a spread that is 3400x larger.”
Additionally, as per the Bitwise paper, suspect exchanges showed consistent volume throughout the day, while on regulated exchanges, volume corresponded to waking and sleeping hours.
CMC’s response: the DATA alliance
On March 25, Carylyne Chan, global head of marketing at CMC, told Bloomberg that concerns over fake volume “are valid,” which is why more information will be added to the website to help users make better decisions.
"For instance, if an exchange with low traffic has $300M volume and just 5 BTC in its wallet, users will be able to draw their own conclusions without the need for us to make arbitrary judgment calls on what is ’good’ or ’bad.’ We want to state that our philosophy is to provide as much information as possible to our users, so that they can form their own conclusions and interpretations — and not introduce our own bias into that mix."
On May 1, CMC announced that it will require all crypto exchanges to provide mandatory API data, which includes their live trading data and live order book data, as part of a new transparency initiative titled “the Data Accountability & Transparency Alliance” (DATA).
The alliance was originally announced in CMC’s sixth anniversary blog post. The company explained that it has to deal with regular requests to delist crypto exchanges based on unverifiable information — such as screenshots of chat logs and emails — which is why CMC chose to empower its users to make more informed decisions and “provide a means for projects to differentiate themselves through enhanced disclosures” instead of applying harsh censorship:
“We are paying close attention to the growing discourse surrounding ‘fake volumes’ of exchanges. This is not a trivial problem to solve, as seemingly innocuous decisions can carry unintended consequences. To add to the complexity, we need to be mindful of the numerous use cases for our data – what some deem to be ‘fake data’ is information in and of itself that can yield interesting analyses, and it is important not to throw the baby out with the bathwater.”
Indeed, CMC seems to aim for a softer approach after removing a number of South Korean exchanges from its platform “due to the extreme divergence in prices from the rest of the world and limited arbitrage opportunity” back in January 2018, when it caused a major drop in the market.
Stressing that the new condition will be compulsory, the tracker stressed that any exchange that does not provide the data will be not be included in the price and adjusted volume calculations on the site. The changes will come into effect on June 14, 2019, CoinMarketCap noted.
Specifically, the required data includes exchange hot/cold wallet addresses (“indicative numbers to enable users to determine solvency of selected exchange”), live market-pair trading status (“more granular trading data at the market-pair level for further analysis”), live wallet status (“summary status of all possible deposits and withdrawals across currencies”), and historical trade data (“all time-stamped historical trades for tracking, and in some cases, compliance”).
“Our stance is that we do not censor any information, but rather will present all the information to users so that they can make their own judgments and decisions on the data presented,” Chan told Cointelegraph. “This philosophy of providing all the information rather than making our own judgment calls or censorship/curation is the same for data submitted by DATA members.” The global head of marketing at CMC added:
“As with all API endpoints submitted to us from exchanges (of which we now have 257 on CoinMarketCap) we work closely to ensure that the endpoints are up and running effectively. This constitutes the reported volume information that is presented on the site. The adjusted volume metric excludes those exchanges with fee rebates or transaction mining, and with the new mandatory data requirements, those that do not provide their live trade and orderbook data.”
When asked whether players have enough time to gather and submit the required information, Chan replied that the data “should not be technically hard for exchanges to provide,” and that the 45-day notice should ensure that there is enough time for everyone to join. According to her, no exchange has explicitly declined to join DATA so far:
“We carefully evaluated the requirements so as to make sure they are reasonable and not unnecessarily onerous for the majority of exchanges to provide. In fact, about 150 exchanges already submit this data, and we are simply waiting for the other exchanges to come up to speed on these data points.”
At this point, DATA is comprised of 12 exchanges: Binance, Bittrex, OKEx, Huobi, Liquid, UpBit, IDEX, OceanEX, Gate.io, KuCoin, HitBTC and Bitfinex.
Michael Gan, CEO of KuCoin, told Cointelegraph that CMC approached them about one month ago:
“After they introduced the whole idea, we soon decided to join DATA, as one of the early members. We are still communicating with CMC in terms of all the data submission and it will be done before the deadline.”
Starry Liu, head of marketing at OceanEX, told Cointelegraph that they were the only Initial Launch Partner of DATA that exists for less than one year. According to Liu, OceanEX approached CMC earlier this year to discuss “the idea of setting up an alliance to improve transparency across the whole industry,'' and soon joined the initiative:
“We found out this is also CMC’s goal. They actively asked about our feedback and acted really fast during the preparation of DATA.”
Liu specified to Cointelegraph that OceanEX integrated its data to the DATA project in less than one month. The information is collected, but not monitored by CMC, she confirmed:
“We prepare and stream raw data from OceanEx to them and CMC is taking a role more of collecting and disclosing data to the public, instead of monitoring. We may have different roles in this program, but we share the same vision and the ultimate goal, that is to benefit the community and the users.”
The KuCoin CEO, however, told Cointelegraph that CMC “mentioned that they will have a team to check all the data submitted, ensuring its accuracy as the DATA project expected.”
A representative of Exmo, the United Kingdom-based exchange that has applied to join DATA but has not been added to the official roster yet, told Cointelegraph that, while the new alliance is “a considerable step to the formation of the sustainable market,” the industry needs more solutions for the deep-rooted problems such as fake volume. Maria Stankevich, head of business development and communications at the exchange, wrote in an email:
“The fact that the data from the trading platforms will be collected in real time will expand the possibilities for studying the market and analyzing it and improve the understanding of the characteristics of various exchanges — that's for sure. But this does not mean that the data provided by the websites will be completely objective.
“The same thing applies to the additional information about the exchanges, that is supposed to be submitted by the projects and exchanges by themselves. We don't doubt the fact that it will be really useful for users — as it will be gathered in one place. But in general, we see it as an extension of the functionality of the CMC itself, an increase in its competitiveness in comparison with the other similar data aggregators.”
Cointelegraph has reached out to more exchanges currently listed in the top-50 by adjusted volume on CMC that are not part of DATA at this point — including Kraken and Coinbase Pro — but has yet to hear back from them.
Matthew Hougan, the author of the aforementioned Bitwise report, told Cointelegraph that he “admires CoinMarketCap for staring down the barrel of systemic fake volume in the crypto market.”
“We've seen a number of really robust responses to the problem of fake data, including from OpenMarketCap, Messari and Nomics, and I love that there is a diversity of smart people looking at this problem and driving towards solutions,” Hougan said. However, the researcher also suggested that the data gathered by CMC might still be unsubstantiated in the end:
“Ultimately, given the limited nature of regulations, the globally distributed nature of crypto trading, and the perverse incentives to exaggerate volume, the best approach is going to be ‘trust but verify.’ But CoinMarketCap has taken a good first step and is putting some teeth into its reporting requirements, and I'll be interested to see what develops from their efforts.”
Meanwhile, CMC plants to extend its initiative even further in the future. “Collecting the data is just the first step,” the tracker wrote. “With a larger dataset, more analyses can be run, and enable the introduction of new, meaningful metrics.”