Crypto Market Anomalies
January 2021

This is our monthly digest of the most curious things that happened in the digital asset space.

Published in

Coinmonks

6 min readFeb 9, 2021

Anomalies in OKEx trading time distributions

While many exchanges have recently started to delist privacy coins, OKEx’s Monero (XMR) and ZCash (ZEC) markets remain active with abnormal trading patterns. XMR’s time of trade distribution on OKEx noticeably falls out of the common flow of activity observed on Binance and Huobi.

Percent of executed XMR spot trades by seconds on OKEx, Binance, Huobi, 2020. Source: NTerminal

Both XMR and ZEC second-of-trade graphs show abnormal distributions which indicate that the most active trading happens every 5 seconds. It is important to note that the observed sharp swings in trade volume by second were not caused by shortages in liquidity for these markets.

Percent of executed ZEC spot trades by seconds on OKEx, Binance, Huobi, 2020. Source: NTerminal

Privacy coins grabbed significant attention due to their anonymous ownership capabilities and subsequent use in darknet markets. The distinct surges in the above graphs may show not only dominant trends in trading bots but also potential trade patterns related to criminal activity. As a result, exchanges such as OKEx may be used as platforms to obfuscate illegal transactions. Wash trading can also be a good explanation for such trading patterns.

SENSO Token — Fake Trading Activity on Top Exchanges

SENSO token demonstrates suspiciously high and stable trade volume on Bittrex, KuCoin, and Poloniex. Observed trading patterns are consistent with wash trading.

Hourly volume on Bittrex, SENSO token, December 2021. Source: NTerminal

One of the ways in which new projects can gain recognition and adoption from users is by rising in rank on sites like CoinMarketCap. Project teams will often solicit the services of market makers who create fake activity on exchanges in order to boost the ranking of their token.

The diagram below reveals this commonly used trick to fabricate trade volume. Usually this sort of activity appears on the venue when there is one entity acting on both sides of the market — buying lower and selling higher of the same asset back and forth with themselves.

Hourly volume on KuCoin by price change, SENSO token, January 2021. Source: NTerminal

This indicates the manipulative nature of the average transaction size across three exchanges listing SENSO token.

Average transaction size on Poloniex, Bittrex, KuCoin, SENSO token, January 2021. Source: NTerminal

Abnormal Trading Volumes on FTX

Our investigation team constantly monitors abnormal activity happening on multiple market venues. The bar chart below highlights the anomalous bitcoin volumes traded on FTX in periods of low volatility. Of note are the readily identifiable peaks of 1 min trading volume that don’t significantly affect the price itself (the difference between open and close price).

Volume distribution by 1 min price change, FTX, BTC-USD, Mar’20. Source: NTerminal

Abnormal bitcoin volumes on FTX greatly exceeded the distribution during even minute price shifts in March 2020. Such volume spikes which are not correlated with price movement often indicate an unknown entity recursively buying and selling the asset.

For comparison, here is the distribution of BTC-USDT volumes on Binance in the same time period (March 2020).

Volume distribution by 1 min price change, Binance, BTC-USD, Mar’20. Source: NTerminal

Scheduled trading activity dominates Huobi and OKEx

One of the market manipulation metrics we pay attention to is the time of the trade distribution. Both Huobi and OKEx show abnormal distributions for second-of-the-trade and minute-of-the-trade graphs.

Number of executed spot trades by minutes on Huobi, Jan’21. Source: NTerminal

On Huobi, the spot market graph displays the convergence of various assets at the same time periods (15, 30, 45 minutes).

Number of executed spot trades by seconds on OKEx, Jan’21. Source: NTerminal

The vast majority of trades on OKEx are executed during the first seconds of the minute (1–4 sec). This distinct surge on the graph can be evidence of the exchange either participating in wash trading or looking the other way to allow a few trading accounts to dominate their order books.

Anomalous trades on Gate.io

Spot market trades on Gate.io show signs of falsified numbers. Below are the distributions of leading, second, and third digits for the size of executed trades as compared to Benford’s law expected distributions. Evidence based on Benford’s law has been used by ACFE to discern naturally occurring statistical deviations from fraud.

First, second, third digit distribution of executed trade sizes on Gate.io, LTC spot market, Jan 2021. Source: NTerminal

Gate.io demonstrates a noticeable overuse of certain numbers, possibly indicating non-standard trading activity on the exchange. Recent order size distribution for various coins deviates from other markets and contradicts Benford’s law. It’s worth mentioning that these anomalous trading patterns have been observed across trading pairs listed on this exchange.