Thanks for taking the time to make your proposal to reduce the fee levels on the Vega network from the conservative launch levels and make them more palatable for traders, and apologies for the delay while we discussed the data internally and sought feedback from traders in the community.
We think reducing the network wide fees is a great idea, and that now would be a good time to bring the network fees down from the conservative levels set at launch to something more in line with other similar platforms and markets.
We do have a few comments on the specifics of your proposal and an alternative set of proposed changes based on how the mechanisms in the protocol work and the data we have observed on-chain, which I’ve set out below.
Infrastructure Fee (market.fee.factors.infrastructureFee)
This is probably the most critical part of the fee as it economically secures the Vega network and bridge against attacks, and needs to be sufficient to attract validators to the chain as the network grows to incorporate more nodes. Although these fees are supplemented initially by VEGA emissions, we think that
0.02% is slightly too low at this stage of the protocol’s life and would propose setting the infrastructure fee to
0.03% instead, for now to ensure the network can remain safe and robust.
Maximum Liquidity Fee (market.liquidity.maximumLiquidityFeeFactorLevel)
This network parameter is actually not intended to be used to directly control fees on markets. It exists to set an absolute maximum, taking into account all potential scenarios such as extremely low liquidity or trading volume, where market makers may need to set much higher fees in some markets to make it worthwhile to support a market remaining open.
The protocol already incentivises market makers to lower fees when conditions allow it, which has already happened once since launch across all 3 active markets. LP fees at the time of writing are half what they were when you made the initial proposal. If this mechanism doesn’t work correctly in the longer term, the protocol itself may need some changes. However, we believe from informal community feedback that it’s likely the LP fees will reduce further over time via this mechanism, as market makers learn how the protocol works and become confident in their technical integrations and risk management.
Forcing fees lower via this method (which it is not designed for) is more likely to lead to LPs reducing their engagement with the network than the intended outcome. We recommend leaving this parameter unchanged for now, and monitoring how the protocol performs over the coming weeks and months.
Maker fee (market.fee.factors.makerFee)
The maker fee on Vega fulfils a similar role to the LP fee, in that it incentivises people to provide liquidity. The difference is that the LP requires a longer term commitment while the maker fee is earned by anyone posting an order that is matched. It is therefore preferable to incentivise LPs more than makers, in order to maintain liquidity at all times so that liquidations will be successful and traders are always able to exit positions when they want to. For this reason, we recommend a larger reduction in the maker fee to
0.005%, as market makers that support Vega markets can also earn the LP fee by making a commitment on chain.
These revised levels we suggest (summarised below) would enable total fees on Vega to come down to
0.05% or lower assuming LPs reduce their fee bids as they are incentivised to. This would be much closer to or in line with the fees on other DEXs and exchanges (and lower than the total in your initial proposal), while still enabling the network to remain secure and allowing the LP fee market mechanics to operate. These could of course be brought down further in future if trading volume increases sufficiently.
What do you think about this revised proposal? It would require two on-chain votes, and we’d be happy to announce our recommendations regarding these changes to the community and assist with the proposal transactions if required.