About the circuit breaker, some suggestion from Fernando from Balancer:
Built in circuit breakers are coming in v2
Since PieDAO manages the pool we can monitor for funky stuff and stop trading as soon as we understand one asset is failing, in this regard the real-time Risk statistics explored by the discussion with Gauntlet Network would be helpful.
For now, there can be a DAO signed tx prepared to be broadcast with a huge gas price similar to Maker’s emergency shutdown.
This is fine because it just halts trade until you guys decide what to do
Will discuss with balancer team once I have a more concrete proposal.
Exposure to possible collapse of a single asset does not outweigh the generated revenue from being a liquidity provider in my opinion. So I think its fine for now.
Some risk can be prevented and some can’t. Having protection from the risk of an asset failing while still being able to generate fees from providing liquidity can be a selling point for the cash++ product. I don’t see how, other from being fun the matroesjka doll argument is applicable.
In light of the Maker situation with the unfortunate liquidation auctions I think we should think about how we phrase these products and communicate the possible risks they bring. The exposure is controlled price wise but the risk of failure is multiplied.
In terms of actionable steps I will continue development of the pool as it is now. And think about a proposal for circuit breakers and discuss it with the Balancer team.
It’s true the pool can be paused but most of these attacks can be executed within a single transaction making it impossible for the DAO to respond.
EDIT: Found out about circuit breakers in Balancer v2 so we’ll migrate to v2 when the time comes.
Great point, feel free to draft a risk report document that should be transparently linked in the frontend. From my side, I don’t think anybody here is trying to miscommunicate anything, would be great to have that by launch.
Generally speaking I don’t think there is a way to completely remove risk, the probability of one of those assets failing before there is a circuit breaker seems thin while possible. This is stuff for a Term of Service.
Chaosnet ethos makes sense to me as long everybody involved know their risk.
I think it’s too much of a hassle for the end user to buy up to 3 individual insurances. It would be awesome if you could include that automatically in the BTC++ Pie. Haven’t found any service that is currently offering insurance on synthetic BTC. Maybe indirectly with options?
I like the idea of embedded insurance, it’s a bit of a complicated product tho, from an operational perspective I think it might work like this.
There is a iBTC++ Pie (i == insurance) which only holds the insurance positions following the weights, when people are minting/buying BTC++ there is a checkbox saying something like:
Add insurance to this 0.4 BTC++ position.
Basically you would receive a BTC++ + iBTC++, assuming liquidity is enough to insure large positions they could be put together in a single token. Thoughs?
With regards to possible risks regarding collapse of one of the underlying assets.
Currently Balancer does not support circuit breakers but they plan to support them in V2 which is expected to launch in 2021.
After some discussion with the core contributors I propose the following intermediate solutions in this order:
We implement a bot which watches the prices of the underlying assets on decentralised exchanges and immediately halt trading if one of those assets has a fast decline in price.
We offer a Pie which includes insurance through options like Alessio suggested and possibly let the DAO provide the liquidity needed to offer them. These options would insure the underlying value of the pool in BTC.
We work together with the balancer team to speed up the development of the circuit breaker without breaking compatibility. And migrate the pool to a complete Balancer V2 implementation when that launches.
I feel comfortable this would sufficiently cover the risks associated.
I think I have a way to implement insurance without exposing us to any other protocols. We deploy a second pool with trading disabled called iBTC then we support minting two different versions, BTC++ and iBTC++ . BTC++ is just the straight uninsured pool with trading and higher risk. iBTC++ is 1% iBTC and 99% BTC++ . During normal times, redemption of iBTC++ only gives you back the assets from BTC++. During a black swan, an insurance redeemable flag is set, disabling new minting on iBTC++ but changing redemptions to include the BTC++ along with a proportional share of the iBTC pool.
The DAO could potentially participate in iBTC to give it a value boost at the beginning as @mickdegraaf suggested. Then we could redeem the DAO’s portion when iBTC exceeds something like 40% of the outstanding BTC++.