Automate alUSD debt cap increases! 🚀 🤖

whalegoddess

Discord discussion begins here -- https://discord.com/channels/738480789464940634/826152864837140492/849038734061338624

Alunara

If there ever comes a timeboxed increase of debt I would like to suggest an exponential increase instead of a linear increase. It doesn't really make sense to grow something linearly in a space that grows exponential. If for example the protocol debt has grown by 10x the linear amount doesn't help that much. As far as I know previous cap raises were also something in the order of 10%, 15% or 20% more. The increase percentage doesn't have to be big, could be a measly 2% per increase. For example, if the total debt at the moment is 300 million and you want to increase it with 1 million per week now, I would suggest to phrase it as a 0.33% per week increase instead.

In addition to a timebox I would also suggest to only increase the cap if the total debt is getting close to the cap, maybe like 10%. This way you don't needlessly increase the cap when there's still plenty of debt available and it doesn't balloon unnecessarily out of proportion.

jerre

In general, it is an interesting idea to have an automatic debt ceiling increase to lighten the workload for the team at this moment. However, I also believe that we should be very cautious.

As Butler ATTS [Colonel Yunt] stated in the discord (sorry, not able to tag him here):

yea depegging risk is the main concern here
this is the main reason we're capping ETH at 2k at the start of alETH
although we have liquid alUSD/stable markets, there is a point at which the demand for alUSD and sell pressure may force users to transmute, which is something we want to avoid now and in the future
the best way to remove this barrier is to create buy pressure for alUSD and other alAssets. Since we don't have liquidations as an incentive for that demand the best way to do that is additional DeFi integrations. Until we have those we're going to have to live with a gradually increasing debt ceiling and exercise caution in this regard

Furthermore, I believe that there might be other/better solutions when this can get managed by the DAO. In that case, I assume the community will be able to increase the debt cap whenever they want without the 'main team' needing to do all the work.

ButlerAndTheThirdStringers

jerre Yes, this sums up my concerns and I suspect it's the reason not everyone voted to unanimously increase the debt cap by 20% last time. In general I think this is too dangerous to automate. If the peg becomes weak because we continue to flood the market with alUSD or other alAssets that forces people to use the Transmuter which results in a decrease to the yield of borrowers.

A slow measured approach here with semi-frequent votes and discussions on whether we should raise the cap is safer and will make our job of creating a sustainable Alchemix easier. For now I'm strongly against the idea, but if in the future we are able to create a ton of DeFi integrations for alAssets that result in more buy pressure perhaps we can revisit the idea then.

whalegoddess

Some additional considerations I have:

When the DAO is live, we will still want to reduce the number of votes/code reviews/churn ops. I don't believe the problem statement (redundant work is required to raise debt cap) goes away with a DAO.
@Alunara By keeping the limit at linear and choosing the right slope for the line, we are ensuring we will only undermint alUSD, never overmint. We can also use an exponential function with a low % per week (per @Alunara)
Couldn't agree more with caution @jerre. I believe my numbers are modest choices.

yuyoman

Maybe we can automate this task with a formula. We can include for example when alUSD available is below 1M and with certain amount of deposits (DAI) and certain amount in the transmuter (DAI) (I mean a ratio > x) then we can expand by x% the minting or a fix amount of million alUSD. The developers can create this recipe and we can then change it when we see it is necessary.

What I am trying to say is we can find a formula that fits what the protocol wants and automate this task

scoopy

whalegoddess from your perspective it seems obvious to increase the debt cap whenever we encroach upon it, but we are closely monitoring the health of alUSD and measuring the effects of actions and market conditions to make data driven decisions regarding the parameters we can change. For example, we will soon phase out the alUSD-only pool. There is 40m alUSD in there. What effect will that have on the peg? We cannot know for sure, so taking cautious baby steps is our preference over potentially shocking the market.

Perhaps in the future when alUSD is ubiquitous in defi and has a long history of holding its peg, then we can adjust the debt cap algorithmically.

whalegoddess

scoopy Thanks for responding. I can appreciate the work required to monitor the health of the peg especially during potentially destabilizing events.

Mixitup88

All of your discussion on the consequences of the different plans of increasing the debt is well and good but it seems to me that all of you are overlooking one thing, market demand, When the market comes to your door and demands that you accommodate the demand of the market and you are unable to meet that demand then the market moves on to the next door (your competitor). At this time it appears that you do not have competition but tomorrow you will and the customers that you turned away because of self-imposed restrictions will do business with your competitor or may even become your competitor. This brings me to my case wherein I opened an account a month ago and found your platform to be a good idea, so I decided today to double my account and after having deposited I attempted to borrow but was denied due to the restriction on new alUSD this upset me because I had plans for the funds and I am not willing to wait and I was not informed about the restrictions before I went to the trouble and expense of making the deposit, therefore, I have decided that will wait over the weekend and if I am unable to borrow against my deposit Monday I will then withdraw my deposit and move on, never to return and the few people who follow my experience in the crypto market will be advised to avoid Alchimex. If you do not meet "market demand" your business dies.

virtualevil

Mixitup88

Valid opinion and if this was a mature and battle tested idea I'd agree, but this magic, and magic is dangerous when not wielded by an adept.

Reputations and trust are built over years and destroyed minutes.

Mixitup88

Whalegoddess, and virtualevil, I appreciate your response and agree but I give you fair notice I am 65 years old worked in construction all my life, I can see in my mind how any type of construct from a dog house to a skyscraper goes together with all the details but concerning computers, the internet and all things in the virtual world my mind is blank, but I intend to be involved in the virtual world for the rest of my days because I am physically unable to build anymore but this virtual world frustrates me and makes me a grumpy old man and proud of it so I will be an irritant at times. You can take my comments personally if you want but you shouldn't one day soon I'll die and go away and you'll miss me.