The recent submission of the Lofty proposal points to a serious weakness in the proposal process. There are serious questions about whether it is technically, or even legally feasible to implement it. For reference, many concerns were raised in the associated temperature check and include, but are not limited to, the possible lack of reliable real-time price oracles, the possible KYC requirements/complications, the resulting inability to liquidate tokens in a real-time decentralized manner, legal and structural questions regarding tradeability, etc… and liquidity questions.
Allowing proposals that are not carefully assessed before they are submitted for a vote risks overwhelming Algofi governors with a large number of proposals that cannot be either technically or legally be implemented, or present grave risks to the protocol that have been intentionally hidden or lost in a flood of information. The feasibility of some categories of proposals like changing an emission schedule can be assessed in an instant. Other proposals like adding a new asset to the lending market should require an extremely careful assessment from the Algofi devs and others with respect to the above concerns and others.
Obviously we don’t want a totally gated proposal process, but there are some settings that could be adjusted to make the proposal process more selective.
Minimum veBANK to propose (currently set to 100k) basically acts as a filter on spam proposals, given that we have >50M veBANK already, seems potentially wise to bump that to something more like 500k (we should all think about a reasonable number), this would limit those able to actually propose, but there are enough users at that level that it wouldn’t be total deal breaker.
The other parameter is proposal-delay, basically 1 account can only create a proposal every so often, right now the delay is set to 1 week. Maybe slightly higher would be wise.
Last is the quorum value, initially this was set at 5M, but it seems given the quick adoption across the protocol that a slightly higher number would be smart 10-15M?
I think these would be a good first step (as early as we are) and if bogus/invalid proposals persist we could take a more interventionist approach.
Totally agree. Many questions regarding the viability of this proposal we’re raised and not adaquately addressed. At a minimum maybe Algofi devs should weigh in on any proposal that makes it on chain. Similar to how the Foundation makes their position known for each proposal. It’s encouraging to see DragonFi making their position known and why. Maybe there should be a soft vote in the governance forum along with veBank.
given the global voting power is ~63M veBANK and by looking at the holders list on algoexplorer i def think that the numbers you mentioned make totaly sense.
since a proposal is voted on for 14 days the delay to launch a new proposal might make sense to be 14 days too?
on the first proposals over 20M veBANK has already voted on with ~13 days left + starting Oct 28 BANK emissions should start so more BANK will be released into circulation that could be turned into veBANK. thats why i feel like 15M quroum is preferabble over 10M
Thanks for responding. The suggestions regarding increasing the minimum veBANK required to propose, increasing the timeout period for proposers and increasing the required quorum for proposals to pass are all steps in the right direction, but in my opinion they are insufficient.
It is very easy and inexpensive to acquire the proposal minimum of 100k veBANK or even the possible change to 500k. BANK is quite affordable on the DEX. The proposer can lock for 4 years and increase the voting power to those minimums by four times making the effective minimums 25k and 125k BANK which can be acquired outright for few thousand USD. The proposer can even receive delegated votes from others they are coordinating with. In the worst case, malign or perhaps just ineffective, but well intentioned, proposers can put the protocol at risk for just a few thousand or less USD. If a flawed proposal created a lending market for an asset that was easily exploitable then millions could be at risk. We want to defend against a situation where malign parties can risk a few thousand and exploit the community for millions.
Defending against these scenarios doesn’t mean the proposal process would have to be totally gated. First, proposals could be slotted into various categories like “Protocol Parameter Changes” and “Add a New Asset to Algofi Lend”. Protocol parameter changes I would assume are generally technical feasible, but if it is beyond changing parameters that are currently designed to be tuneable it would be very helpful to have you or other devs weigh in on the technical feasibility, risks and benefits of the change. If it was something designed to be tuneable like emissions there wouldn’t be any further assessment necessary. Proposals to add a new asset to the lending market should be treated much more carefully. One way to improve this process might be as simple as building a checklist for new assets with line items assessing it on things like - liquidity, technical feasibility in general, technical effort required (e.g. less than 2 weeks or a heck of lot more), KYC implications, regulatory/legal concerns and including who reviewed each item. Adding some structure to the process at minimum would help.
Paul’s response on the Defly proposal lays out three things that should be present when considering adding a new asset. I think it fits the checklist that you are talking about. You bring up a good point about the relatively low costs of buying Bank. A couple thousand or even tens of thousands is relatively small compared to 100+ M.
Good point the proposed borrow factor and collateral factors should be required be present in all proposals to add assets to the lending market in addition to items Paul mentioned and the other ones I mentioned. We need some structure/process added to prosposals for new assets.