Historical Fee Change Data Dump

Historical Fee Change Data Dump

TLDR; We are currently experimenting with changes to the fee rate on many different pools. This data repository can be used to observe the historical average daily volume and daily fees generated in the time period of each fee.

Two folders within Fee Changes :

  1. Avg Daily Vol-Fees at Each Rate:
  • Historical fee changes and the average daily volume and fees at each rate
  • Great overview for historical fee changes
  1. Avg 30d VolFees:
  • Average daily volume and fees for the most recent 30 days (above) vs the previous 30 days (below - i.e. 60 to 30 days ago).
  • Most useful for pools currently being experimented with


Since the launch of v2.1, we have had over 200 pools added to Bancor of which about 160 are whitelisted. During this time, we have implemented approximately 90 fee changes focussed around 40 pools.

Statistically, the fee changes we have implemented in days gone by do not correlate with volume changes indicating that there is room to increase the fee rate on many pools without detracting from LP yields. We seek to change this by correlating the fee changes with statistically significant volume changes and identify market tolerance levels.


Example from Avg Daily Vol-Fees at Each Rate

Interpreting x-axis:
0 # Index position
21 days # Duration that fee was at this rate
2020-12-20 # Start time of fee change
WBTC # Pool of interest

Left y-axis: Average Daily Volume (blue)
Right y-axis: Average Daily Fees (red) set at 1% of volume

Note that this is the daily average for the duration that the fee was at this rate. While this is an excellent benchmark it can lose the ability for a meaningful comparison if the durations are very different (hence the section below).

Example from Avg 30d VolFees

This dataset compares the average daily volume and fees for the most recent 30 days (right) with the previous 30 days (i.e. the time frame from 60 days ago to 30 days ago) with the centre date listed in the x-axis and title.

In this example, the recent experiments with the fee rate on the USDC pool show that even though the average daily volume has decreased in the recent 30 days, the average daily fees have increased almost 100%.

Csv files:

Each folder contains the raw data in a csv file for ease of comparison.

Data Notes

This dataset will be continually updated every day or two. This is currently a manual process from Dune Analytics as it is contingent on marrying the token addresses, pool addresses and bancor converter addresses.

Feel free to use this data for other governance proposals.

This is still a work in progress so if you have some suggestions for formatting or specific queries/data you would like to see let me know here.

Note that the pool naming convention hasn’t been perfect in the past so might require searching (e.g. AAVEBNT not BNTAAVE, and PSP not PSPBNT)

Many thanks to @tfns and @glenn for feedback.


Damn, this is great stuff.

Thank you so much @NIX !!!