Hey everyone, reminder that we have our AMA with Reflexer today @ 1pm EST!
Matt of UniWhales.io (WILL NEVER ASK FOR FUNDS)
I'm curious about their "Governance Minimization" - Many of us think that one of the reasons Indexed.finance has done well is due to their light governance. What specific advantages does there governance have over Maker and Reserve?
Any others? Drop 'em in
Can you please explain how collateral auction mechanism and debt auctions will work? Specifically, we have seen multiple troubles with similar mechanisms ( for example at maker dao). What are the protection methods in case of major volatility?
Let's start with some intros -- Stefan, can you tell us about crypto bio/background and how you ended up working on Reflexer?
Then, can you give a quick update on what Reflexer is, plus how it's going today? We'll then dive right into questions
I worked for around 3 years in crypto doing various things from scaling solutions to financial instruments. I then ended up as a Binance X grantee when I did most of my stablecoin research and found out about the initial design that DAI was supposed to have years ago.
In March last year when DAI broke way above the peg I coded a demo of the initial design and pinged Ameen about it. We didn't know each other before, I just knew that he wanted to build something called Metacoin.
We decided to work together and bring the initial Maker vision to life and that's how Reflexer started.
I was annoyed at Maker going gov maxi, published the metacoin paper, then was lucky enough to have Stefan show up with the skills to build everything and hire me 🙂
Reflexer is a project where we seek to build non pegged, stable assets. Think of how EUR floats against the dollar.
We are building an asset like that called RAI. RAI is backed by ETH and managed by an on-chain, PID controller.
Cool. Anything else you’d like to add before we open things up @sionescu?
Just that I'm excited to finally build a stable asset that's not tied to fiat 😊
Alright folks, have at it.
There is a lot to unpack with the mechanism @sionescu, and I'd venture to guess many of the folks in this audience don't have an understanding of it. Perhaps you could walk us through some of the core parts in more detail?
It seems like MakerDAO was design inspo, so feel free to weave in comparisions when appropriate. I think that'd be helpful since most folks understand it quite well
Currently we're using an auction mechanism similar to Compound's: the protocol offers a fixed discount in collateral auctions and you can bid with RAI and buy ETH at that discount.
The collateral auction/liquidation mechanism also has another difference compared to Maker: you can attach an insurance contract to your CDP/position and in case you get liquidated, that insurance contract can first try to save your position. In case it can't save it, you get liquidated.
As for debt auctions, currently the logic is very similar to Maker's although we have the option to add a staking pool that also protects the system. The staking pool would get diluted in case the protocol is underwater.
William Phan | Hex Capital
if governance is allowed to turn off the PID controller during unstable situations, why is there a need for a PID controller? How does the system return to stability with no control scheme in place?
The protocol has more flexibility in setting its own monetary policy.
For example, if the market price of RAI is above the moving peg, we need a way to bring it down. The problem is that no one might want to mint RAI and sell it on the market (think of Black Thursday).
In this case, the protocol will devalue RAI. Basically, the moving peg will start to go down and each unit of RAI inside the protocol becomes cheaper than before.
This allows people who have open CDPs to leverage even more because 1 unit of RAI is becoming cheaper compared to 1 unit of ETH.
Moreover, RAI holders expect that, because the peg is moving down, the market price will most probably follow at some point. This compels holders to sell now instead of later when the market price already depreciated.
This highly depends on RAI liquidity and on the controller parameters we set. That being said, we want to get to a point where the RAI market price and the moving peg are close to each other almost all the time (think of DAI) and the moving peg varies by less than 10% per year.
Yes we plan to talk with Curve about this and see what we can build together.
We will seek to integrate with other protocols ranging from money markets to options/futures platforms and show other teams how they can leverage the repricing mechanism behind RAI. We wrote a couple of examples and focused on money markets and options:
Moving the "peg" up is actually what helps contract the supply or at least compel people to go long on RAI.
When the moving peg goes up, 1 unit of RAI will become more expensive compared to 1 unit of ETH. RAI holders will also expect that the market price will follow the moving peg at some point, meaning that if you want to make profit you should long RAI before the rest of the market.
This is a self-fulfilling prophecy because the fear of someone else going long will make you go long.
For ETH we're currently using Chainlink, for RAI a 16 hour TWAP from Uniswap v2.
Oracles will, for better or worse, still be governed in the long run and we're looking for solutions to make them more resilient.
30 minute warning 🚨
Do you expect the #1 use case of SAFE users (for the audience, that means user who generate RAI with their ETH) to be leverage?
Additionally: in reading your docs, it seems like a key assumption is that economic agents of Reflexer will be shorting Rai, using derivative products built on top, etc. -- are you doing anything to incentivize these L2 applications/primitives to get built?
The first integration we want to do is with something like Opyn. We'll post a bounty in the next week.
Additionally we plan to talk with other options protocols that have recently come out and present the case of RAI where you can take a put or a call depending on your expectations of how the peg will move.
L2 is still tricky because there's no clear winner so initially we'll focus more on existing L1 protocols.
If the RAI moving peg goes down and if you levered with RAI, you expect that the market price will also follow at some point. This means that you'll be able to buy RAI at a cheaper price from the market and close your position.
Another detail is that while the peg is going down, your collateralization ratio goes up (assuming the ETH price doesn't tank) because each unit of debt you took gets cheaper compared to ETH
Hmm interesting -- in what conditions would the Rai peg go down while the ETH price remains stable?
A surge in demand for RAI which would make its market price go up and thus RAI would start to be repriced down inside the protocol.
Got it, thanks Stefan. 15 minute warning folks! Get your questions in please.
Maker's collateral auctions are good for price discovery (no oracle needed) but bad for capital efficiency and composability (can't use flash loans for example), at least until dutch auction liquidations. how does Reflexer handle this?
This is why I've been thinking of ways to add insurance in the protocol and potentially a staking pool that takes the initial risk in case the protocol is underwater but also receives part of the protocol's cashflow.