Wash trading - general explanation

As the first epoch is on the verge of conclusion, and prior to @chaoslabs proposing a rewards distribution, I would like to address a significant concern related to wash trading.

What constitutes wash trading in the opinion of @chaoslabs?

I understand that in dydxv4 there is a very small pool of traders and market makers, which consequently means that situations where we encounter the same maker-taker pair during matching are likely to occur quite frequently. I am not asking you to disclose the full methodology for identifying wash traders; rather, I seek a general understanding of the process.

For instance, there are a pair of accounts with the following characteristics:

  1. An event where account1 is the maker and Account2 is the taker occurs more than 1000 times, and vice versa, also more than 1000 times.
  2. These accounts have similar sources of funds and are connected to the same addresses in EVM networks.
    Can these accounts be called wash traders?

You can begin to imagine what’s going on… Oh boy. There are a few big players here who are laughing all the way to the bank. No prizes for guessing who they are.

It is also very easy to create a script where funds are send to new addresses all the time causing (1) to be not true for those addresses, while still being operated by the same individual.

Ahead of the publication of our incentive distribution report for the first trading season, we are set to present the community with an in-depth wash trading analysis, which will also delve into the specific methodologies and criteria we employed to scrutinize trading patterns. We aim to foster transparency and provide a better understanding of how we differentiate between legitimate trading activities and manipulative practices.

Regarding the aspect of wash trading, the criteria for identifying such activities are multifaceted. Wash trading typically occurs when trades are artificially created to give an impression of heightened market activity. This can be either through addresses that are controlled by a single entity or through a group of addresses colluding to generate deceptive trades. Key indicators include repetitive, non-economic trading patterns, and a lack of genuine change in ownership.

In your specific scenario, If the addresses are either controlled by the same entity or if the addresses are colluding to earn rewards then it is most likely wash trading. However, if the majority of the 1000 trades are consistently one-sided (such as A consistently selling to B) and if these trades are conducted between addresses that are independently controlled and not in collusion, the likelihood of these transactions being classified as wash trading diminishes. It’s important to consider the broader context of these trades, including the market conditions and the economic rationale behind these transactions, to make a more informed determination. Our analysis framework is designed to detect and scrutinize such nuances, ensuring a comprehensive evaluation of trading activities.