Introduction

Hadeswap is a newly launched Automated Market Maker (AMM) NFT marketplace on Solana to improve the NFT liquidity and its trading experience. It let users trade their NFTs using liquidity pools, the same as the swaps occurring on decentralized exchanges. Hadeswap is a platform with 0% fees from the start. The royalties are also set to 0% by default, however, users have the possibility to pay them if they wish to do so. Just like any other AMM, it allows users to collect SOL fees as people swap in and out of their pool.

The non-fungible tokens (NFTs) gained massive popularity in 2021 and their sales saw explosive growth before the current bear market. However, the global financial slowdown has not slowed the sales of NFTs in 2022 and new projects continued to develop and release their projects. While there are numerous factors affecting the price and sales of a collection, this new technology is prone to be abused and bad actors may try to manipulate the sales and price of these digital collectibles to make them appear more valuable through a process called wash trading.

A wash trade is a transaction where the parties on both sides of the trade are the same person, or two people colluding to manipulate the value and liquidity of an asset. In the case of NFT trading, the owner of an NFT sells it to a new wallet address, which it also controls, to make the NFT appear more valuable. As many NFT marketplaces allow users to trade their NFTs by only connecting their wallet, and even charging no fees, wash trading becomes cheaper to execute, and its possibility increases.

Although it is challenging to detect wash traders in general, with the transparency of blockchains it is possible to track suspicious transactions and exclude them from the total sales. With that being said, the NFT sales activity on the Hadeswap and the amount of wash trading that took place on this platform have been investigated in this analysis. The assumptions and the methodology used to detect wash trading have been summarized in the following section.

On a side note, the Chainalysis NFT crime report, the CoinDesk’s article on wash trading, and the NFT Wash Trading on Ethereum by Pinehearst are great resources to dive deeper into the matter.

Methodology

Two types of wash trades, including the self-financed and recursive wash trades, were considered in this analysis. First, all of the sales were selected by filtering the Hadeswap marketplace. Then, the tx_id of wash trades was determined, using the method explained in the next two sections. Next, all of the sales were filtered with and without these transaction IDs to obtain washed and legit sales, respectively. Finally, these two were either joined or unionized to be able to compare them simultaneously. besides that, the NFT collections were labeled using the dim_nft_metadata and dim_labels tables.

> NFT buyers were considered the “wash traders”. > > More details could be found in the SQL code.

Conclusion

In conclusion, the analysis of NFT wash trading on Hadeswap has demonstrated that a negligible portion of its daily sales was operated by wash traders. While its sales percentage has been nearly 10% daily, the washed volume has been less than 1% which is a positive sign. The number of wash trades progressively increased in the first few days after the launch of the platform, then decreased to an acceptable amount. However, it has started to increase again in recent days. Among the multiple addresses that were involved in the wash trading on the Hadeswap, only a few of them generated a considerable volume. Some of these addresses washed even more than 100 NFTs, however, were not able to wash high volumes. From the collection's perspective, the majority of known washed collections were small projects with a low NFT price floor, which indicates that smaller collections were more susceptible to wash trading in general. The small scale of wash trading that was conducted on these collections might not even be utilized by bots. One might conclude that these small collections had a higher chance of being wash traded compared to other well-known or larger collections. Last but not least, the comparison of wash trading on different NFT marketplaces has demonstrated that while Magic Eden has had a high number of wash trading taking into account its scale of transactions, it has not had the highest percentage of wash trades compared to its total NFT sales. On the other hand, CoralCube and Solanart have registered a considerable volume of wash trading conducted by numerous wash traders during the second and third weeks of October, resulting in a considerable percentage of total NFT sales on these platforms being wash trades. OpenSea has also been among the top marketplaces with the highest amount of wash trading during multiple days in the measured period. Overall, high volatility in the wash trading percentage of each marketplace has been observed in the past month, indicating that each of these marketplaces has been vulnerable against wash trading. As an example, it is worth mentioning that, while Hadeswap has had a higher percentage of wash traders and washed NFTs on its platform, the total percentage of number and volume of washed trades have been higher on Magic Eden.

Marketplaces Comparison

Legit vs Washed Sales Overview on Hadeswap

Daily Legit vs Washed Sales on Hadeswap

Wash Traders on Hadeswap

Washed Collections on Hadeswap

Self-financed Wash Trades

Self-financed wash trades (as pointed out by Chainalysis) happen when address A provides funds for address B, then address B buys an NFT from address A and returns the fund. To detect these wash trades, the fact_nft_sales and fact_transfers tables were joined with the following conditions:

sales.purchaser = transfers.tx_to AND
sales.seller = transfers.tx_from AND
sales.sales_amount <= transfers.amount AND
sales.block_timestamp > transfers.block_timestamp

Recursive Wash Trades

Recursive wash trades (no official naming) happen when address A sells an NFT to address B, which sells back the NFT to address A. To detect these wash trades, the fact_nft_sales table was self-joined to monitor two consecutive sales with the following conditions:

sales_first.mint = sales_second.mint AND
sales_first.seller = sales_second.purchaser AND
sales_first.purchaser = sales_second.seller AND
sales_first.block_timestamp < sales_second.block_timestamp

Insights

• Although the 0% fee structure on Hadeswap provides a good foundation for wash trading, only around 6% of the total sales have been under the assumed conditions for wash trading;
• Even so, less than 1% of the sales volume has been washed since the launch of the marketplace;
• Interestingly, around 11% of the total buyers were among the wash traders on the platform;
• The daily averages of legit transactions have been much higher compared to the washed ones;
• Expectedly, the average price of NFTs that were washed has been considerably lower than the average NFT prices on the platform;
• While the daily share of the wash trades, wash trades and washed NFTs has varied between 5 to 15 percent, the daily share of washed sales volume has never gone higher than 1%;
• The number of wash trades progressively increased in the first few days, then decreased to an acceptable amount. However, it has started to increase again in recent days;
• Multiple addresses were involved in the wash trading on the Hadeswap marketplace, however, only a few of them washed a considerable volume on the platform;
• Among the wash traders, there were addresses that washed more than 100 NFTs. Nevertheless, the washed volume by these traders has not been high;
• Besides the many unidentified collections that were washed on the platform, the majority of known collections were small projects with a low NFT price floor. This indicates that smaller collections are more susceptible to malicious behaviors;
• The small number of sellers, or wash trading counterparts, has also demonstrated that most of these trades were operated on a small scale and not even by utilizing bots;
• These small collections have also had a higher percentage of wash trades among all of their sales activity compared to other well-known or larger collections;
• The comparison of wash trading on different marketplaces has demonstrated that while Magic Eden has had a high amount of wash trading considering its scale of transactions, it has not had the highest percentage of wash trades compared to its total sales;
• CoralCube and Solanart have registered a considerable volume of wash trading by numerous wash traders during the second and third weeks of October;
• OpenSea has also been among the top marketplaces with the highest amount of wash trading on multiple days in the measured period;