Introduction

This dashboard includes a deep dive into ThorChain’s Ethereum bots. In the first part, I define what a ‘bot’ is, by investigating the swap count per minute for all ETH users. Having defined what a ‘bot' is, I move on to the main part of this dashboard, the analysis, where I look into which pools bots are interacting with the most. I will look into the daily and the total USD volume, the total and daily swap count per pool - and the total and daily bot count per pool. Lastly, I will do some digging outside of Flipside to find out whether bots interact with synths or just non-synths, and finally, look into the profits made by some of the analysed ETH bots. Enjoy!

In order to tackle this question, we first need to establish what a ‘bot’ is. The pie chart below shows the number of users who have swapped a certain number of times per minute - we can see that majority of ETH users swaps 1-5 times per minute, which could be a bot activity but could also be human behaviour. Initially, I thought to define a ‘bot’ as a user who swaps a minimum of 50 times per minute, however having written this query, it’s clear to me that there are not many bots that execute such a high number of trades per minute. Due to these results, I thought it would be best to define a ‘bot’ as a user who swaps more than 5 times a minute - hopefully this definition will filter out the majority of humans from the rest of this analysis!

The next step is to look at the behaviour of those bots - what are they swapping, which pools are they interacting with the most?

Key points:

• The most popular pool, when we look at the count of swaps executed by these ETH bots, is the BNB-BUSD pool, which has over 5k swaps since the beginning of this year. The second pool with the biggest number of swaps is the ETH-ETH pool, which means that these bots must also be interacting with synths and not just non-synths.
• The most USD volume has been swapped by ETH bots in the ETH-ETH pool, which has a total volume of over 100M, what a huge amount! In a later section of this dashboard, we will look into how this number translates into the profits made by these bots.
• The BNB-BUSD pool has also quite a big USD volume, with almost 53M since the beginning of this year - it makes sense for users to be swapping in/out of stable BUSD, and making profits as BNB goes up in price!
• The daily count of bots shows us that there were days where there were a lot more bots than on other days - which could be caused by general speculation in the market or more orders being fulfilled and hence more bots coming into play.
  • The total count of bots also agrees with the USD volume, we can see that most bots have been swapping in the BNB-BUSD pool and again, the ETH-ETH pool. We can also see quite a few bots that interacted with the ETH-USDC pool, which probably made them some profits by selling ETH when it was going up in price for another stable, this time USDC.

We can also look at this (rather messy) scatter chart, which shows us the daily count of ETH bot (remembering that these are the users who make at least 5 transactions per minute) coloured by the USD swap volume. We can correlate these two aspects - as we can see that with more bots, more USD is being traded. The next step is to look at how this USD volume translates in terms of the profits, or even possible loss, made by these ETH bots…

Now that we know what these ETH bots are swapping the most and we have seen the huge USD volumes, what about the profits they made? By looking at some of the ETH bots, we can see that most of them manage to make a profit, with the biggest profit being just over $8k. I am quite surprised at these results - given the swap volumes, I would expect these profits to be a lot higher. However, we have to remember that the bots might be taking minimal profits - i.e. on ETH’s move from $1200 to $1210 etc, which explains to some extent why these profits might not be as high as what I initially expected.

Conclusion

This dashboard has looked deeply into the behaviour of ETH bots found on ThorChain. I have firstly investigated what a fair definition of a bot is, because this is where every analysis starts - with a good, coherent definition. If I decided to change the definition of a ‘bot’ to have a higher or lower number of swaps per minute, all of the results presented above would have been completely different. After deciding to proceed with a definition of a ‘bot’ being a user with more than 5 transactions per minute, we have found out what pools they are interacting with the most, by looking at the bot count, the swap count and the USD volume in each pool. By doing some manual investigation outside of Flipside (i.e by looking at the bot addresses manually), I have also found out that they do swap to/from synths, but mostly, bots tend to be swapping for non-synth assets. I have also analysed how much profit some of these ETH bots have made, and we have seen that it is not as much as me - or you, would think. Surprisingly, we have also seen two bots that made a loss!