This publication proposes revising the Debt Interest Rate from 2% to 4% on Prisma Finance.
Abstract
We propose increasing the Debt Interest Rate parameter from 2% to 4%. mkUSD borrowing appears too cheap compared to current stable borrow rates causing redemptions and a mkUSD price hovering over $0.994.
Motivation
Current Prisma APRs largely outweigh the loss incurred from depositing one sided into Curve pools.
Borrowing remains cheap compared to other protocols
For reasons already discussed in other forum threads, I don’t think it makes sense to raise the interest rate to 4% right now.
Without belaboring the point that has already been made a few times elsewhere, a few quick points on this:
(1) I think we should be trying other methods to get the peg back to a good place other than raising interest rates, this is the bluntest measure Prisma DAO has and if we go to 4% now there is on space to raise interest rates anymore. We should not do this IMO until we have exhausted other attempts (like some of my suggestions in the peg thread) that may help without an interest rate change, particularly around controlling the mint cap.
(2) I think comparing the interest rate to the other protocols in the post is misleading because all of those models are different than Prisma’s which is a liquity fork, so as a result the other models listed do not have redemption risk (meaning users are willing to pay more for other reasons). At 4% Prisma is arguably not competitive in the market place and will likely drive away significant TVL, who wants to borrow for 4% + risk getting redeemed when you can go to aave and pay on average 5% (even with occasional large spikes) with no redemption risk?
(3) I don’t think we should be raising interest rates at all right now without trying other measures first, but if we are going to raise rates, why not start with 3% instead of 4%? That at least leaves some space for another future rate raise if needed instead of another 100% rate raise, it doesn’t make sense to me why we would go immediately to 4%, at that point we will be out of additional options in regards to future rate raises.
Given all this, I have to be against this proposal for now.
None of the solutions you proposed are possible to implement and we have no idea if they would even work. We however know for a fact that higher interest rate is likely to reduce borrowing and it is the fastest most effective tool we have right now.
Michael from Curve made this immediate observation “borrowing is too cheap here”. We can sit around and wait for more redemptions or we can take action right now.
Lowering the mint caps in order to temporarily put a freeze on minting is not possible to implement? Why not?
I fully expect my arguments to go into the wind here and the rate raise to happen, I also expect it will only temporarily solve the issue and we will be right back at this place within a week or two with the same problem if we don’t do anything in regards to the mint caps because ultimately the sell pressure will not relent even with the rate raise after a temporary fix. We need better long term solutions and I think that the solution is more likely related to adjusting the mint caps than anything to do with rates. The only major difference in a few more weeks after this rate raise is going to be the loss of TVL.
Also, again, why not at least go to 3% and incrementally test what is needed instead of going straight to 4%?
Didn’t you say @Sidn3yGottlieb you wouldn’t support another rate raise after the last one? You just think we have no other options at this point?
We’d be burning a ton of emissions going to minting and turning users away from the protocol by lowering mint caps.
Also, again, why not at least go to 3% and incrementally test what is needed instead of going straight to 4%?
Because 3% won’t make a difference when most protocols have borrowing at 5%+. We can also lower it back down if mkUSD stabilizes.
I won’t vote in favor of this and let holders decide but it’s the right decision for the time being and the general consensus is that it’s mostly likely to have a rapid impact on the peg.
We need better long term solutions and I think that the solution
We have some, they just aren’t ready yet. In the meantime, APRs are 20-30%+ across the protocol so we have to do what we can.
I get the points and appreciate the thoughtful response but I fear no one is taking the points about needing to contain the mint pressure seriously enough. Ultimately the rate raise cannot fix this for more than a week or two if we don’t also stem mint pressure (and really stemming mint pressure may be enough even without a rate raise in combination with redemptions!).
Regardless I have made my points, and I guess we will see what happens after yet another rate raise.
I’m one of the largest customer on the platform and here’s my request: raise the IR!
Why? Because even with 2x CR backing, my position was close to the front of the redemption queue. The root of this problem is the low peg as a result of low IR.
Raise IR, most borrowers including myself don’t care for the speedbump when we’re raking in double-digit APR. It’s easy to debase a high peg anyway.
People would be happy to pay over 4% on USD if it meant their 3.5% yields x10… it’s a speculative bet that price appreciation of ETH yields outperforms USD IR
Enacting policies that mitigates redemption takes priority over lower IR, because redemption eats into TVB much more than IR, which is barely noticeable even at 4%
Right now, I don’t see IR at even at 5% would affect TVL because:
The high PRISMA emission we are having
The market rates are higher than ours
People are lined up to get PRISMA to play the flywheel, it’s a highly speculative phase of the launch and the primary thing to monitor is the mkUSD peg leading to redemptions.
Raising the IR on borrows is the most sustainable response as far as mkUSD peg goes.
let’s say (generously) that the mkUSD peg is off by 0.5%.
we know that redemptions are happening very frequently even at a 1%+ effective fee (!!).
that means oracle deviation is accounting for the other >= 0.5% … i.e. it is MORE than half of the problem
Possible Solution
increasing the redemption fee *floor is actually a powerful move to combat this
a heightened redemption fee will kick in on top of the floor (proportional to the size of the redemption/mint against total tvl) as soon as the first redemption or mint occurs.
during times of high oracle vs. spot price volatility this fee will throttle redemptions, giving time for the oracle to catch up.
this mechanism charges a high fee to bots who try to play the oracle arb, which is again, the majority of the issue at play
Misc
if some folks think raising the redemption fee floor is a step too far, keep in mind: we can also adjust the minuteDecayFactor, which will help the redemption (and mint) fee recover to the floor level quicker to make up for any big spike.
the chart below should highlight the concentration of redemptions around specific blocks where oracles are stale.
note the redemptions all happening in small clusters of blocks. meaning the first bot doesn’t take it all. there is usually more room for a 2nd, 3rd or even more bots to capture a profit behind him. These later bots are the ones who will be charged a higher fee, and potentially dissuaded by the high redemption fee.
FWIW - I agree with you @Tetranode that dealing with the redemption problem is the most important thing, and as another large borrower on prisma I don’t have a problem paying the interest, my concern is that raising the interest rate is not actually going to fix the redemption problem (we will just be left with high interest rates and the same problem soon enough, which could also lead to follow on effects of prisma becoming far less competitive in the marketplace when the high emissions period ends in less than 2 weeks).
So rather than apply this temporary band-aid (and i think governance wise it likely will be applied in the short term which is fine) I would rather see us actually attacking the root of the problem and fixing this more for the long term.
I think there are other problems at play that could actually help fix the issue though regarding redemptions and they have more to do with controlling the mint cap (and especially what kind of minting happens right after a redemption) and making sure the redemption fee is set correctly according to the info from @wavey 's excellent post.
Checked with Charlie and we’d like to bundle this with a 1% minimum redemption fee on all collaterals to prevent redemptions due to oracle redemptions following @wavey 's post
quick follow-up thoughts on the fee floor increase topic:
redemptions are among the best mechanisms for fixing a bad peg, as it’s literally pulling mkUSD out of dex pools and burning it
therefore it would be misguided to think this will have a positive effect on the peg
rather, the correct argument for implementing a fee floor increase is to throttle the volume of “artificial” redemptions that are caused primarily from oracle lag
my suggestion to the next Liquity fork would be to add a fee amplifier. this would give operators more control to adjust how intensely fees amplify. in combination with decay factor, this could be a nice way to tune different collaterals based on the quality of their oracles; creating a cool-down period after each redemption to wait on an oracle update.
