RE: Will Steem developers fix/upgrade the Steem Dollar? (Multicollateral Steem)
It is my opinion that for an asset that has the objective of being a stable pair for another one (in this case USD) using a volatile one as collateral (steem) it has to have a backing of at least 2 to 1.
The changes introduced in HF20 had the aim of preventing SBD to be overpriced by removing the mechanism that prevented the printing of more SBD when the open market was overvaluing it.
We all know what the end result was...during the bear market SBD holders had the incentive to convert their asset to steem and sell it on the open market to prevent or mitigate losses and in the process increased the effective inflation of steem to 17% creating more selling pressure for the underlying token further eroding the peg.
In it's current form SBD is just a source of instability for steem and it should be ditched or fixed.
The idea of having multiple tokens as collateral is good in theory only if the backing assets are not correlated. Being that the crypto market pretty much moves in the same direction regardless of which coin you are looking at I don't believe that it would make much of a difference.
The problem with paying interest is that it has to come from somewhere. Banks are able to do that because they charge a fee on their loans. DAI charges a stability fee if you want to close your CDP. In essence MAKER acts as a bank that loans the value of one USD to all the parties that open CDPs.
In a way the steem blockchain charges a fee to anyone that converts their SBD if you aqcuire it above it's intended value or if the debt ratio exceeds 10% since you would not be getting back the value of one dollar if you do it (assuming that you buy it when it's priced at one USD).
A conversion under those conditions in a stable or bull market is beneficial to steem holders as it would be equivalent to burning steem. In light of the fact that the market is prone to downsings and in a long term downtrend it backfires (remember the 17% inflation we had in the last year) I am in favor of using a different pegging mechanism.
What I propose is the following:
- Have a 2 to 1 backing.
- Remove the virtual supply and instead create a hard coded smart contract where 10% of the inflation is used to create sbd.
- Cap the issuance of sbd to 5% of the supply (this would ensure a 2 to 1 backing).
- Allow anyone to create sbd by sending steem to the smart contract in a similar way as how bitshares and Maker do it with BitUSD and DAI.
- Pay interest but only if the external price of sbd is above a certain threshold (this could be a dinamic mechanism that uses the witness price feed to estimate the rate to be paid).
EDIT: with HF21 the SPS already uses 10% of the inflation to create SBD so I don't see a need to pay content creators with it as this could potentially lead to an emission of steem dollars above the intended 10% of the supply.
My argument is that Steem being the exclusive backing for Steem Dollars is why the peg couldn't hold. It was based entirely on the assumption that Steem itself could hold value without anything backing Steem. Maker works because there is multi collateral which is set to back it so it's not like Maker Dai is going to be backed exclusively by Maker tokens.
And this is basic derivatives. You don't really need 2:1 to hold a peg. You just need a better underlying asset. For example if you used actual USD to back BitUSD then it would hold the peg. Because BitUSD was backed by BTS is like trying to build a house on a shaky foundation. So my suggestion is build the house on as stable of a foundation as you can (which isn't really going to be USD elusively) and if you want to see how to do it right just look at the Libra whitepaper which explains exactly the right way.
tldr
My argument is not in favor of sticking with SBD 1.0. My argument is in favor of scrapping SBD 1.0 and replacing it with SBD 2.0. SBD 1.0 has failed and needs to be replaced or upgraded. In my opinion the better way of doing it is in my post here: https://steemit.com/steem/@dana-edwards/fix-upgrade-the-steem-dollar-toward-a-proposal
Back SBD 2.0 with stable SMTs. Stable SMTs can be created relatively easily and even Tether could launch an SMT and we'd have one. The logic behind the pseudo code would be for every 1 SBD 2.0 token created, X amount in Y percentage of stable SMTs must be locked as collateral to be redeemed in a process which destroys the SBD 2.0 created.
Reference
https://www.openlibra.io/
DAI is not backed by multiple assets...yet. The upgrade that will allow this is scheduled to go live on November 18th so that is not the reason it has held the peg. In addition to being over-collateralized the stability fee and the DAI savings rate play a role in maintaining the peg.
In the end these financial instruments can only hold the peg if the underlying asset(s) have enough stability and value to back them up. As I mentioned before I am still not convinced that multicollateral backing is enough to achive this. In my opinion you need a basket of assets that are not positively correlated in addition to an adjustable interest rate that can influence the supply and demand for the stable coin.
All of your points are valid and I agree. This is why we need more discussion.
Some ideas you present are not bad. The idea of letting people generate SBD by sending Steem to the smart contract is good. The problem, why limit it to Steem? What backs Steem then? If Steem were backed by SMTs and not just generated out of thin air then this might work but when Steem is being generated out of thin air of course it doesn't work.
Interest has to be paid for the simple reason that if you are pegging to something unstable like the USD then you are already losing wealth for anyone who holds their life savings in that peg. If there is no interest then no one will be willing to choose this over Dai or over a lot of other better things like gold. I do know you need a plan to generate the interest and this can be generated and the revenue from that generation shared as interest.