A number of factors have converged over the past nine months that have significantly impacted protocol fee revenue:
Sharp increase in Q1 and Q2 due to high gas prices and increased volume of open orderbook (limit order) trades
Sharp decrease in Q3 due to lower gas prices (in particular, the explosion of Flashbots and other MEV schemes) and increased competition for open orderbook liquidity
Going forward as L2s start to gain traction and in the absence of some compelling event, the downward trend is likely to continue.
Hypothesis
If protocol fees are eliminated (temporarily, in this case), open orderbook volume will increase. Additionally, swap volume could increase as well due to user behavior/loyalty.
The data gathered from this experiment can be used to inform a tokenomics redesign, which is top-of-mind for protocol stakeholders.
Note: We could also reduce the fee temporarily rather than eliminating it, but that would provide a weaker signal and not be as useful an experiment.
Other Factors to Consider
Staking performance will be negatively impacted; however, it is already being impacted by the factors described above.
Voting Options
Yes (temporarily eliminate protocol fees on open orderbook trades)
Adding a note to say that this is currently a temperature check and that we will do some more work on defining what the experiment will measure, etc., prior to moving forward with a formal proposal.
I’m in favor of this proposal. I’m curious to see, prior to removing the protocol fee, if the team has any ideas on their vision for the tokenomics redesign. I’m sure once live data is collected it can influence certain decisions, but the last time this discussion came up on the “Future Proofing the Protocol Fees” thread, I feel like the conversation ended without any real ideas.
I wish I came into this conversation with my own ideas lol. Don’t want to be a backseat driver - I believe “we” (community + team) will figure it out.
Thanks @nikita for proposing this. I will vote 1. Yes as well.
This experiment will hopefully equip us with the necessary data points to shift towards a more sustainable model, or at the very least validating (and addressing) the fact that the current one is anachronistic vs how product/market fit evolved.
I would focus in particular on two data points during the experiment:
open orderbook volume → expected to increase. When available, one could even break it down by application.
% of limit orders filled / # limit orders created → expected to increase. This latter metric is just a heuristics for the impact pf have on the likelihood of limit order execution. We ran an analysis internally in February that showed that ~25% of limit orders created in Matcha are eventually executed.
For both metrics, we will have to settle for correlation with a simple pre/post analysis, but I think it’ll be enough signal. Obviously, both metrics depend heavily on market and gas conditions, which hopefully don’t change drastically.
After the vote passed and executed in effect yesterday, the protocol engineer updated the nativeOrderFeatures contract destination and it needs to be added in data pipeline to reflect correct count of fills.
the Success Fill Rate charts in dashboard means the (1 - revert rate) on chain, i.e. the success rate of settlement after the orders are matched/taken and signed by both taker and maker. We gonna try to add another % of matching order success rate (i.e. % how many orders are picked up by takers from the open orderbook pools). Presumable assumption is without protocol fees, this rate can go higher as (adjusted) price is more competitive.
added the fill rate chart for v3 limit orders in the dashboard. So far we have seen a similar level of fill rate compared to before vote (with protocol fees):
