Recently, one prominent “unfair player” on the crypto exchange market — a Chinese exchange called ZB.com, ranked 7th on the global TOP list by Coin Market Cap (CMC,) grabbed the attention of the CER team.
ZB.com, a cryptocurrency exchange, first appeared in November 2017 as part of the wave of global crypto awareness. After the Chinese government banned ICOs and crypto exchanges, the exchange chose the Independent State of Samoa for its jurisdiction and settled its headquarter in Hong-Kong. ZB.com steadily holds in CMC TOP-10 exchanges by total 24h volume, in both “reported” and “adjusted” sections.
Key Findings: While analyzing ZB.com, we found definite patterns of unnatural and obviously artificial trade volume performance on 10 out of the top-20 most-traded exchange’s pairs. Furthermore, we discovered that 4 DASH pairs volume on this exchange totaled more than $288mln accounting for 24.58% of the exchange’s total 24h volume and for 80.56% of all DASH traded on all exchanges. All of this directly points to the fact that trade volume manipulations are taking place on ZB.com. Interestingly, the amount of traffic to the trading platform of the exchange reduced in a direct ratio to the increased trading volume during the analyzed period. In addition, ZB.com receives an immensely small amount of traffic compared to fair exchanges from the TOP 30 CMC list. Specifically, Kraken’s number of Unique Visitors is higher by 17.9 times than ZB.com. Finally, the exchange does not have a wide community on social media and messengers, nor does it have a significant subscriber base.
While looking at a number of exchanges and deciding which of them to investigate, we noticed an interesting difference in ZB.com trade volume reported on different crypto resources. For instance, as of September 18th, the exchange was ranked 7th in the CMC TOP exchange list by 24h reported trade volume with about $411mln (fig. 1) and CoinGecko ranked it #2 with $1235mln 24h trade volume (fig. 2).
The daily trade volume difference of 3 times is pretty big, but the reason is pretty simple: CMC only reports 2 out of 4 ZB.com markets. It provides the data for pairs with BTC and USDT as quoted currencies only and omits ZB (exchange’s native token) and QC (a token equivalent to CNY) markets. CoinGecko appeared to be the only resource reflecting ZB.com’s trade volume in full. According to CoinGecko (fig 3), ZB.com’s QC market alone accounts for 50% of total trade volume, the rest of markets share the remaining half (USDT – 22.4%, ZB – 16.3%, BTC – 11.3%).
It is interesting that CMC doesn’t show two-thirds of ZB.com’s trade volume, despite the fact that CMC does count considerable trade volume of trading pairs of other exchanges’ coins, such as FCoin. However, it’s even more interesting that the exchange doesn’t claim its higher trade volume to gain rank. As according to CoinGecko, it usually outpaces its peers, particularly Binance, by total trade volume. And the same would be the case had CMC reported all trading volume claimed by ZB.com.
Maybe ZB.com doesn’t claim higher trade volumes because by doing so it would show it has its fingers in the pie?
Let’s try to figure it out by digging deeper and analyzing the exchange’s most liquid pairs. Here are the Top-20 in Fig. 4.
For some reason, Bitcoin Cash trade volume is duplicated by CoinGecko via displaying it for BCH and BCC tickers. As 4 duplicate pairs (quoted in QC, ZB, BTC, and USDT) accounted for $61.85mln, we deducted that number from total volume and adjusted each pair’s volume share in the total (see fig 4; column ‘Volume % adjusted).
Impossibility to get historical trade data via API from many crypto exchanges is a common problem.
And ZB.com is not an exception, as its respective API request returns only 50 most recent transactions. Moreover, although its API documentation states to have “since” introduced parameters designed to fetch historical trades referencing to transaction ID, the function simply doesn’t work even if applied. Therefore, not being able to analyze the exchange’s past trades, we focused on visual charts in our investigation. So, let’s get down to business!
While analyzing charts of ZB.com’s 20 most traded pairs, we found that literally half of them display patterns of unnatural volume performance. There are abrupt volume jumps by thousands of percents that can be seen on daily charts and obvious recurrent volume patterns that are distinguishable on smaller time frames. Let’s examine them more closely.
DASH/USDT daily chart shows that this pair was not so popular on ZB.com before July 2018 (not counting a one day spike on June 24th). But for some unknown reason, trade volume rose from hundreds to hundreds of thousands of DASH per day.
Similarly, DASH/ZB daily volume rose tremendously in July as well. It jumped from just 2-5 coins to more than 200k DASH in a matter of days.
And its hourly chart demonstrates patterns of recurrent trading activity, starting and fading at about the same hours each day. Such performance is most likely attributed to trading bots that are turned on and off daily.
XRP/BTC is the most recent ZB.com’s “star” whose volume soared sharply just a week ago from 1-2mln to over 200mln XRP per day.
Moreover, after the skyrocketing jump, its trade volume remained almost perfectly stable through the last week. This can be seen in both daily (fig 8) and hourly charts (fig 9).
LTC/ZB pair experienced about two notable stages of daily trade volume rise: about 500 times jump (from 140 to 70k LTC on average) in the second half of June and then, about the sevenfold rise (to more than 500k LTC) at the end of July. Each time the volume remained at the gained levels.
And its hourly chart shows clear recurrent volume patterns suggesting artificial trading activity.
BTC/USDT hourly chart displays repetitive patterns as well.
As for DASH/BTC pair, its daily chart shows some periods of rises and falls of trading volume varying from 1 to over 1.1mln DASH traded per day. Such a volume performance without significant price movements is definitely unnatural for fair markets.
Its hourly chart also demonstrates unnatural trade volume patterns with signs of recurrent trading activity.
DASH/QC daily chart shows the immense volume rise from hundreds to hundreds of thousands of DASH traded daily in July.
And its hourly chart demonstrates that the trade volume rises and falls do not correlate with price moves, which is an unnatural situation in normal market conditions.
BCH/ZB gained its “popularity” on ZB.com in the middle of August. As its daily volume rose from 34 to over 100k BCH. Natural? We do not think so.
QTUM/BTC daily chart shows unnatural hundredfold volume rises and declines, lasting through the huge trading activity starting at the end of June.
And here is its hourly chart demonstrating obvious recurrent trade volume patterns proving that trading activity is rather artificial.
LTC/USDT trade volume jumped more than 40 times in one day at the end of August and remains on that level until today (September 18th).
And it’s not surprising to see trading activity that is definitely not made by regular traders on LTC/USDT hourly chart.
*By the way, we already proved that the same artificial trading patterns are a common thing on trans-mining fee markets. You can read more here.
For reference, the combined 24h trade volume of 10 pairs with the obviously unnatural volume patterns detected and depicted on charts above account for as much as 45.28% of the exchange’s total 24h volume.
While investigating ZB.com’s trading pairs, we noticed that 4 out of TOP-12 most liquid pairs were DASH pairs. Trade volume of DASH/USDT, DASH/ZB, DASH/BTC, and DASH/QC combined totaled more than $288mln and accounted for 24.58% of the exchange’s total volume
But the volume of those 4 pairs accounts for as much as 80.5% in total volume of DASH traded on all exchanges around the world (according to CoinGecko).
Isn’t it strange that more than 80% of all DASH markets’ volume was produced by one exchange?
It might not be, but considering the suspicious patterns of unnatural trading activity spotted in all 4 pairs, we can conclude that the DASH volume is most likely manipulated on the ZB.com exchange.
Through this analysis, we can draw conclusions about the approximate number of unique users trading on the platform: we considered unique visits and the percentage of refusals (Bounce Rate) to determine the number of unique users who have spent more than 3 seconds on the website during the period from June to August.
The formula is UU = UV * (1 – Bounce Rate)
According to SimilarWeb (Fig 23 above), during the period from June to August, the number of UV of ZB.com was 55,304, the Bounce Rate was 45%, which means that the number of UU of the ZB.com exchange for this period was 35,947.
Let’s compare the results of the analyzed TOP 7 exchange ($410M Daily Trade Volume) with Kraken, which is the TOP 14 Exchange ($154M Daily Trade Volume).
According to Fig. 24, Kraken’s number of UV is higher by 17.9 times than one of ZB.com. In turn, the number of Kraken’s UU is 613,514, which is 17 times higher. Also, it’s necessary to note that Kraken has lower Bounce Rate, Higher Average Visit Duration etc.
In addition, an extremely interesting fact is that the amount of traffic on ZB.com lowered almost in direct ratio to its trade volume increase in the main trading pairs analyzed in the previous section.
Eventually, the main type of traffic coming to ZB.com’s landing page is referral (Fig.25). And as in the case of the Bitforex Exchange, the main referral source is CoinMarketCap. It signifies that thanks to the high fake trade volume, the exchange is in the TOP 7 list on the most famous rating platform of the crypto industry. As a result, it receives a significant number of new quality visitors without spending much money on the paid ad and building the community.
In spite of the results of the traffic analysis, we decided to analyze the size and activity of ZB.com’s community in social networks and messengers.
The Twitter account of the exchange has only 2.9k followers comparing to 327K followers of Kraken. Besides, on average ZB.com’s posts has 1 retweet and 3 likes.
However, the exchange has a huge Telegram group with 99K members. However, despite the fact that there is a constant influx of deleted accounts and bots, this group has very low activity. During 24h on 26 September, there were only 58 messages. To compare, in HKN Official Group, which includes 4.6K members (21,5 times lower than one of ZB.com), there were 45 messages. In spite of the fact that the number of members in these groups differs considerably, the activity of ZB.com is only 1.3 times higher.
Also, the share of the exchange’s traffic from the email delivery is 1.44%, which means that the company doesn’t have a significant number of subscribers in its base.
While analyzing ZB.com we discovered that chats of half of the 20 most liquid pairs accounting for over 45% of the exchange’s total 24h trade volume show obvious signs of volume manipulations. These are abrupt volume rises by thousands of percents and patterns of recurrent trading activity which are unnatural for fair markets.
Besides, four DASH pairs’ trade volume totaling $288mln and accounting for more than 24.5% of exchange’s 24h volume also accounts for over 80% of total DASH volume traded on all exchanges according to CoinGecko. Moreover, all four pairs’ charts display patterns of unusual trading activity.
In addition, the ZB.com website (both landing page and trading platform) receives an extremely low amount of traffic when compared to other exchange from the TOP 30 CMC list. Kraken’s number of UV is higher by 17.9 times than that of ZB.com. In turn, the number of Kraken’s UU is 613,514, 17 times higher than ZB.com. From June to August, the amount of traffic to the trading platform of the exchange reduced in a direct ratio to the increased trading volume. Also, similarly to BitForex, ZB.com forges trading volumes to stay on the very top of the rating and draw the attention of crypto investors and new users. However, the support team of the exchange doesn’t attract the incoming visitors to stay. Therefore, ZB.com loses the majority of newcomers and the opportunity to build a loyal community of users. All of the information represented above suggest that there are most likely prominent trade volume manipulations carried out on the ZB.com exchange.
But again, the main question that remains unanswered is the following:
“What should the global crypto community do to eliminate the fraud of exchanges and make the industry mature?”
There is almost no information on ZB.com’s management in the web. The only person publicly linked to the exchange appeared to be its co-founder Jimmy Zhao. And his Linkedin profile states that he is a Founder and CEO of Zillion Bits AG company which technology is used by ZB.com and another exchange named EXX. Being curious about that, we decided to look into another exchange built on the same platform, especially as it was ranked 12th by reported 24h volume on CMC.
And the picture we have seen suggests that most of EXX.com claimed volume is an absolute fake. The following screenshots speak for themselves.
Apparently, Mr. Jimmy Zhao is really experienced in creating the effective technology for crypto exchanges, which allows them to make a living by faking trade volume.
Don’t hesitate to contact us via [email protected], if you have suggestions on how to make these reviews more interesting and effective.
Enter your email address to subscribe to Hacken Reseach and receive notifications of new posts by email.