- Crypto Integrity found that up to 88% of reported volumes in February were artificial at some of the highest reported volume exchanges
- On some less liquid trading pairs, the group estimates up to 100% of the volume is fake
Crypto Integrity, a project aimed at fraud detection and forensics within the crypto market, released a report that analyzed real order book depth on select liquid trading pairs and found that up to 88% of reported volumes in February were artificial at some of the highest reported volume exchanges. On some less liquid trading pairs, the group estimates up to 100% of the volume is fake.
While echoing similar findings aggregated by CoVenture’s liquidity report, which found that two of the top five reported liquidity pairs have less than 1 percent of their actual reported volumes, Crypto Integrity has offered new raw data which aggregated the sum total of trades that happened materially higher than the best ask/lower than the best bid (i.e. out-of-spread trade) and trades that happened at prices materially lower than best ask/higher than the best bid (i.e. in-spread trades).
Top reported trading volume in February across crypto exchanges…
…Crypto Integrity adjustments for wash trading at leading reported volume exchanges
The forensic shop believes the three most likely mechanisms to drive wash trading include:
- in-spread trades without limit orders: Exchanges just flat-out report fake trades with no change in order books. As long as there are no limit orders in the order book, exchanges can engage in this practice without risk of someone hitting the order
- in-spread trades with short-lived limit orders: Not necessarily malicious activity as arb and high frequency traders can exhibit this behavior by hitting orders quickly, which can lead to the benefit of tighter spreads and liquidity. However, when a majority of volume comes from orders present in the book for milliseconds something is amiss.
- trades near bid-ask caused by short-lived limit orders: A more sophisticated wash trading mechanism is to give the appearance of more organic flow, by producing sell orders near the best bid and buy orders near the best ask. An example: a limit order of 0.5 BTC hit by a market order of 0.5 BTC after 50 milliseconds can be easily classified as a wash trade.
From the sample size, these exchanges exhibited the 3 clearest wash trading mechanisms in February according to CI
Finally, the report measured liquidity metrics across the sample of exchanges by analyzing Handy liquidity (cumulative volume in an order book at remote prices, more liquid assets have higher Handy liquidity), bid-ask-spread (best bid-ask spread, with more liquid assets having narrower spreads), and weighted bid-ask spreads for 10 BTC. Using Coinbase Pro as a benchmark proxy as an exchange that reports higher levels of volume, but actually offers the liquidity to boot, many of the top reported volume exchanges clearly aren’t providing the level of liquidity one would expect given respective reported volumes (+$1B per asset pair).
Some believe it’s only a matter of time before the bear market shakes out these bad-acting exchanges. Chintan Sheth, a quant researcher at SeedCX told The Block, “I expect more consolidation going on with exchanges, with hundreds of exchanges out there, and massive volume reductions across the board from 2018, it makes sense why they’ve resorted to wash trading. It’s only a matter of time where they can’t keep up with expenses and they fold over, eventually you’ll have just actual flow.”