Wash Trading: Who, What, Why, and What Should We Do About It?

Many in the NFT community have been clamoring for more competition in the NFT marketplace landscape. One of the reasons for this is concerns about the concentration of volume on one centralized platform (OpenSea). Another is purely due to economics - transaction fees are high right now and as in any market, competition will drive these fees down and ultimately create a better user experience for the end consumer. So naturally, anytime a new marketplace launches, it is met with much excitement and anticipation. Throw in a token airdrop and things get even more interesting.

The LooksRare Marketplace launched as a "community-first" NFT marketplace on January 10th, 2022 as the latest contestant in the race to steal market share from OpenSea. In an attempt to quickly bootstrap volume on the marketplace, they launched what is known as a “vampire attack” on OpenSea (another “community-first” marketplace, Infinity, tried something similar last year). If you aren’t familiar with a vampire attack, here’s a tldr - Because of the open nature of the blockchain, anyone can look at the on-chain transaction data and identify OpenSea transactions. LooksRare did this and decided to airdrop $LOOKS tokens to incentivize users to participate in their new marketplace.

“Anyone with a combined 3 ETH trading volume or more on OpenSea over a six month period (from Ethereum Block 12642194 on 16th Jun 2021 to Block 13812868 on 16th Dec 2021) is eligible to claim $LOOKS, once they list an ERC-721 or ERC-1155 NFT for sale on LooksRare. …we’re distributing 120,000,000 $LOOKS tokens (12% of total supply) to the NFT community.” - LooksRare docs

This was a brilliant way to get OpenSea users to list items on LooksRare, but due to the aggressive emissions schedule of the $LOOKS token trading rewards (more on this below), participants in the LooksRare marketplace have engaged in rampant “wash trading” (buying and selling between two wallets, both of which they control) to capture the daily trading rewards.

We pulled the LooksRare transaction data for Meebits, one of the “no-royalty” collections that have seen the majority of wash trading occur, for the week of 1/11/22 - 1/19/22 and you can see from our transaction data below that almost all of the volume was from wash trading.

P.S. If you’d like programmatic access to this data, fill out this form to learn more about our upcoming public API

The “Wash Trade”

To fully understand why wash trading is occurring on LooksRare, you have to understand the $LOOKS token trading rewards. We won’t go into the full details of the rewards program here (official docs) but to summarize, $LOOKS are being awarded to both buyers and sellers on the marketplace proportionally based on the percentage of overall platform sales they represent in a 24-hour period.

• Example
• User buys or sells 10 ETH worth of Meebits on Day 1
• There is 10,000 ETH total volume on LooksRare from eligible collections on Day 1
• So on Day 1, User would receive 2,866.5 LOOKS
• (10 / 10,000) * 2,866,500 = 2,866.5
  

Once you have $LOOKS tokens, you can either sell them or stake them. Staking them earns you staking rewards in the form of WETH that is accumulated by the 2% marketplace transaction fee. The transaction fees are calculated at the end of each 6,500 block period (~24 hours) and then distributed proportionally based on % of staked $LOOKS tokens held by each user, similar to trading rewards. So if you are a trader doing a lot of volume, this is great because if you are staking the $LOOKS you are earning from trading you can recover some of the fees you are paying on transactions.

So what is a “Wash Trade” and why are they comprising ~99% of the volume on LooksRare right now.

Let’s look at an example of Trader X. Let’s say Trader X (or, Wallet X - they likely used multiple wallets) bought and sold a single LOOT between their two wallets for 14,305.06 WETH ($44,522,815 at the time). Looking at the other transactions for that day for the four “no royalty” collections we pulled out for this analysis, we can estimate what their $LOOKS rewards would be and roughly how profitable this wash trade was.

• Overall sales volume = 113,000 WETH
  • 93,000 WETH + ~20,000 WETH for other collections (we didn’t pull all transactions for this analysis, so we’re approximating here).
• Trader X volume = 29,000 WETH
  • 1x buy + 1x sell at 14,500 WETH
• Trader X LOOK Rewards
  • 29000 / 130000 WETH = 0.25 (% trading volume)
  • 0.25 * 2.8M $LOOKS (Daily Trading Rewards) = 700k $LOOKS
  • $LOOKS price on 1/19 = $5
  • Trader X Rewards = 700,000 * $5 = $3.5m USD

Based on the above, this trader would have received roughly 1/4th of the $LOOKS trading rewards distributed that day (29,000 WETH / 130000 WETH = 0.25 * 2.8M $LOOKS). On 1/19/22, $LOOKS was worth roughly $5.00, so if this wallet turned around and sold all of their $LOOKS rewards immediately they would have realized a profit of somewhere between $3-3.5million.

It’s not a certainty that traders are turning around and selling their $LOOKS rewards, though. In fact, it’s more likely that wash traders are long-term $LOOKS HODLers who are staking their $LOOKS so that they can continue to capture transaction fee revenue even after the trading rewards have ended. If you are bullish on the potential of LooksRare to continue to capture market share from OpenSea in the long-term, this would be the smart move. We also think that wash trading will be less rampant on LooksRare after the daily trading rewards are halved in February.

What are we doing about it?

At CryptoSlam, we strive to provide ultimate transparency for NFT data across all blockchains. So when we had several community members start reaching out to us about unusual sales numbers, we knew we had to take action to address the issue.

Cleaning up our data

We’re not 100% finished yet but we have made a lot of progress cleaning out the wash sales from our data. So far we’ve done the following:

• Wrote an algorithm to detect transactions that we believe to be wash sales. We’ll be flagging these and/or filtering them out of certain places on CryptoSlam.io*
  

• Cleaned up transaction history in our database to remove wash sales from our NFT Collection Rankings by Sales Volume
  

• Added an element in our UI to indicate if a sale is likely a wash sale in the Activity tab for most Collections.

*we think that we are catching ~99% of wash sales as of right now, but we’ve found a few cases where we’ve missed what appear to be wash sales after some manual review. We’ll continue working to improve this algorithm as time goes on.

These are just the first steps we have taken to solve the immediate problem. We have a lot of exciting new things our team is working on to help continue to solve the wash trading problem in a more sophisticated and accurate way. We’ll be building these wash trading features into all of the new tools (👀) we’re working hard on for collectors and investors to help provide them with the best NFT data (fyi, we’re hiring!).

The reality is, there will be more marketplaces that launch with incentive structures similar to the $LOOKS reward program to help bootstrap volume and there will inevitably be bad (savvy…?) actors who take advantage of this. Wash trading isn’t a problem that is likely to disappear, we just need to handle it appropriately and in a transparent way. We’ll continue to do our best to do this.