- Following the DAO decision on May 20, 2021, the ETH swap fee was increased from 0.10% to 0.15%.
- We have now collected over a month of data, but which has proven difficult to analyze due to the high level of noise in market activity over this period.
- Herein it is proposed that the fee be reduced back to 0.1%, while we work with data scientists and community members to extract meaning from the existing dataset.
- In the future, the process of increasing the fee, and reducing it again can be repeated to continue to test the conclusions of the developing theory.
Since the launch of v2.1, the subject of pool fees has been a polarizing issue. There are compelling arguments to make for both reducing and increasing the fee and in both cases, the foundational theory for making these decisions is virtually non-existent. Members of the BancorDAO have attracted attention to this issue, and are beginning to propose experiments that would allow future decisions to be influenced by the data gathered during these changes. This is not a simple task.
Running these experiments is complicated by the fact that we only control one variable, and the significance of its effect on trade volumes has yet to be demonstrated. To separate the signal from the noise, data should be collected over substantive periods, and then returned to some baseline level. Repeating the process several times adds confidence to the conclusions of the analysis, and contributes to the developing theory.
I submit that the time interval associated with the increased ETH pool fee has been long enough that a return to 0.10% is probably overdue. Resetting this variable, and continuing to collect data is part of the process; this proposal does not make any judgement as to whether 0.10% or 0.15% is preferable. We simply don’t know yet, and moving the fee back to where it was after this testing period is complete is an implicit requirement of the investigation. The question is when should the fee be moved back?
For some community members, the difficulty of extracting meaning from the available data might be mysterious, while others with an earned erudition in data science will likely see the challenge as the beast it really is. Regardless of where on this spectrum you find yourself, I am ambitious that the value of repeated trials is probably well-understood. I am actively speaking with potential on- and off-chain data analysts to help with the interpretation of the results gathered during these fee experiments. After these efforts yield fruit, future changes to the ETH pool fee can be more hypothesis-driven, and likely be performed on smaller time intervals.
To DECREASE the pool fee on ETH from 0.15% to 0.10%, vote FOR
To leave the pool fee on ETH UNCHANGED at 0.15%, vote AGAINST