FIP - 71 - 100% full collateralization though fraxlend + seniorage

Ok, I did not get this from you original post, people are required to keep their CDPs open, even if they do not want to? That is very risky for users, because the protocol can unilateral decide to increase the interest rate and they get REKT. I do not think we will attract a lot of borrower this way, even if we start at an interest rate of 0%.

When the CR is below 100%, it means we are NOT solvent, that is kind of how it is defined. POL is calculated as collateral in the CR calculation, so that can not make it more solvent. The way we currently prevent a bank run is by using locked liquidity, this locked liquidity can absorb sell pressure and can not be used during a bank run. We have enough locked liquidity now to avoid a run on the bank.

Locked liquidity is far more efficient in avoiding a bank run, because it effectively removes liquidity. The locked CDP’s you propose, actually increase the ammunition for a potential attack by lending out FRAX, putting the peg at risk.

1 Like

Where did you get that from? Of course not.

Again, from scratch:
Let’s start from now,
now, with the liquidity that the AMO owns, we are effectively solvent.


Idk what you mean with locked liquidity, locked FXS? FXS supply expands during recollateralization, those newly printed FXS get sold for collateral, if liquidity is thin that is not a PRO for FXS’s price but, anyways:

-Fraxlend goes live → Lending happens
-People will supply collateral in excess to the credit extracted (FRAX) from the Money Market to meet Collateral Requirements. (Regular overcollateralized lending.)
-the Collateral Requirements would be dynamic in the sense that the protocol would request as more collateralization as there is protocol liability on the Seniorage side, at the moment of opening new CDPs.
-UCR and OCR balance must be met in $ terms, ofc this must happen gradually @ first, but practically having a “REAL” CR of over 100% definitely will allow us to achieve it, we are just getting more and more secure as lending grows.
-Holding an oversupply of collateral into the CDP positions (vs credit lent out in FRAX) can justify an interest rate hike up to the tot value of that collateral -1$ in an absolute worst case scenario. (this is already how overcollateralized stablecoins work)
-People can obviously redeem at anytime but if IR are hiked paying is not an option.
You cannot go “I refuse to pay so I colse” so the protocol is able to extract as much value as needed
-If people redeem FRAX’s supply contracts as it should be.

As you can understand, covering Seniorage’s liabilities with lending would only make us MORE collateralized because, basically, you can assume IR would be paid.
Max pain scenario you can hike as much as there is overcollateralization happening in fraxlend.

The inflection point can happen when OCR fails to be balanced with UCR, thats the top of an expansion phase, IR are low but then liquidity exits from Fraxlend bcs collateral requirements are too high and demand for leverage fails to keep up with seniorage debt.

You see, low /high IR are an incentive/disincentive to borrow in the long run.
Lower/higher Collateral requirements @ CDP stipulation are an incentive/disincentive to borrow immediately.
This acts as an anti run mechanism:

As liqudity gets out of the MM, Seniorage prints FXS to recollateralize and collateral requirements drop as IR hikes while FXS price dumps aka borrow now

This doesn’t even take into acount what I wrote about creating a Collateral Reserve from IR (you just count it ouside the CR of the Seniorage as we are doing now with that extra liquidity that puts us over 100%CR) by selling FXS on a TWAP in fraxswap to acquire collateral during the expansion phase and while FXS price is surging due to Seniorage requirements for minting as we decollateralize.

If I’d be able to edit something about my OT would be the part about FXS price floor, which isn’t accurate.
It’s more of an FXS selling pressure dampening, king of like a forward guidance sort of thing.

2 Likes

That is super cool, taking the good parts to what went bad in other protocols and use it to upgrade an existing one will create a new performing and more stable asset

Keep up with the good work! :grin:

@Denett Let’s go straight into the absolute worst case scenario Ser, shall we?
At the end of the day tha market is gonn seek those max pain points so, it’s really in our interest to think about it.


(This is a little schematic I made of how the FRAX’s Seniorage system works now, it can be helpful)

Lets start from expansion:
-Bull phase, life is good
-FRAX supply is expanding thanks to a demand increase
-cosequently Seniorage gradually decollateralizes, meaning that it slowly requires less collateral and more FXS to mint new FRAX, but FXS’s supply contracts at the same time because it’s being burnt to create new FRAX.
-This supply decrease because of the burn/demand increase because of decollateralization creates a mismatch that drives the FXS price higher.

What happens in the Money Market at the same time?
-IR go down, as they do for all other Money Markets, our advantage though would be that at the same time FXS’s price increases thanks to Seniorage still making up very profitable.
-Because FXS is increasing in price we start to gradually sell it with a TWAP sell order on Fraxswap, the more decollateralized we get, the higher the FXS’s price, the more we sell for collateral. (sell high)
-A positive market momentum incentivizes demand for leverage, like I said, the protocol would be trying to balance out OCR with UCR, so Collateral Requirements to open new CDPs would gradually increase.

Enter the inflection point
There will eventually come a time where Seniorage is too undercollateralized and demand for lending simply can’t keep up mainly because of two reasons:

  • maybe a sharp drop in collateral value (blow off top), causing people to not want to leverage anymore.
  • collateral requirements simply become too high and act as a disincentive to lending regardless of low IR
    (as we previously denoted: Collateral reuqirements impact lending in the present, IR do so in the future)

So then, at this point we enter the contraction phase:
-Seniorage starts recollateralizing, it also utilizes that excess Collateral Reserve that we built by selling FXS denominated Interest Rates on the way up to do it faster.

  • IR in the Money Market Hike
  • Minting of new FXS, usually sold to recollateralize, can be dampened by our FXS’s reserves increasing during the IR Hike in this phase.
    (As the FXS’s price dumps we get paid a lot of FXSs as IR increase in this phase)
  • If the Collateral reserve is enough to meet the recollateralization demand, we HODL the excess FXS obtained through IR hikes in this phase (buying low)

At the bottom of this phase another inflection point needs to happen for the system to go back to balance, if this doesn’t happen we suffer a run and we go to 0.

Let’s see how the incentives get aligned at this point:

  • As OCR fails to keep up with UCR, Seniorage is recollateralizing as fast as our Collatreal Reserves created during the expansion phase + FXS’s price action can take.
  • at this point people are getting fudded about the peg, the worse the unbalance in OCR/UCR the worse the fud.
    To go back to euqilibrium we need essentially two things in this phase:
  • More collateral
  • FRAX’s supply contraction (meaning less relative collateral needs)

As people run to fraxlend to close CDPs they need to buy FRAX to close their positions, this helps us with the peg.
If they close CDPs that contract FRAX’s supply, this also helps.
IR can be hiked to OCR value -1$ causing us to raise a lot of money to recollateralize + FRAX’s supply contraction.
Then…
The Collateral Requirements gradually drop in this phase,
also a very important thing to understand is that when borrowing new FRAX from the Money Market the user can effectively SHORT the peg.
I put up collateral, borrow FRAX, sell it and if it goes to 0 so does my debt.
The more the FUD, The better.

Low Collateral Requirements to open new positions + an increase in the demand to short the peg should successfully incentivize new lending demand and this closes the circle.
:exploding_head:

1 Like

OK so,
unfortunately the forum doesn’t allow me to edit the OT or the Title anymore.
I apologize if presenting it as an FIP may have been a bit rushed.
Certainly the ideas can benefit from a thorough process of discussion, expacially because the concept of backing Seniorage’s liabilities with the overcollateralization from a Money Market is quite new and I never heard anyone talk about it before what happened with UST during May 10, plus this might represnt a big structural change in terms of FRAX’s economic model.

Please understand that the rushing wasn’t meant for any reasons other than that:

  • We had a partnership with Luna and that, together with the “algo” label, were looking to be enough for the market to get completely FUDded over FRAX/FXS.
  • With what happened, I tought that, to use Sam’s own words, we would need to “redifine what being an algorithmic stablecoin means”
  • I saw the team, plus the whole community working together to come up with solutions on how to achieve better security gaurantees during this crisis.
  • There was already another proposal up that IMHO was motivated by the best intent but not going in the right direction (100% CR and wUSDC/ PSM clone + IR) and I thought that MAYBE some of those discussions might also have been priced in by the market.

For context:
I got a nice feedback from both FRAX’s, CRV’s and CVX’s communities when I first posted this on Twitter.
Then Sam liked it and followed me, said it was good suggestions!
The day after Travis, the FRAX’s CTO did the same, so that ended up giving me the confidence to put togeter the proposal in what I felt was the time of need.

Even so, I sent the draft of the proposal to both Sam and Trevis and asked if they could review it before punlishing it to be sure I wasn’t being inappropriate.
Sam didn’t reply but I was told and understand that he recieves probably hundreds of DMs everyday.
Travis told me that, to him, it looked like the proposal was very though out, to go ahed and post it if he didn’t reply!
I asked another admin to poke his shoulder but he also told me to go ahead and post, so I finally did.

In regards of the grant:

I know grants aren’t an uncommond form of recognition, mostly used to reward the work of programmers that can add value to various DeFi protocols in different forms.
I’ve seen performance fees paid to strategists (also a nice suggestion from @Sparrow )
, grants paid for mareketing purposes on 1min commercials on Youtube shows and also bounties paid to outright avoid exploitation.

Our space values inclusion and decentralization,
I’m not a coder but, as you can judge by my proposal, I am able to conceptualize and propose an economic design.
Now, personally, I think that is also worthy to be considerated as added value to the protocol/community, as long as it’s worthy of being adopted.

Of course if the idea ends up not being judged as valid by the community / team, the proposal can be dropped and it ends up there.
On the other hand IMO, if my idea ends up being adopted by the protocol, that would clearly indicate it as a valuable contribution and I don’t see why asking for some recognition would be a wrong thing to do.

In conclusion:

I would say to keep this thread up for the discussion of the implications of merging/overlapping the undercollateralized design of an Algo Stable to the overcollateralized design of a Money Market Stablecoin.
Eventually, I will post another proposal when Fraxlend comes out to determine whether or not the community thinks its appropriate to integrate any of this into its mechanics and in FRAX’s protocol economic design, obviously postponing the grant argument up until then.

1 Like

Solvency can be calculated by subtracting the liabilities from the assets. If this is positive you are solvent, if it is negative you are insolvent. In the case of Frax, the liabilities are the outstanding Frax supply and the assets are the collateral. A CR of 89% means that we have $89 of collateral for every $100 of outstanding FRAX. So by definition this means that Frax is insolvent as long as the CR<100%. This does not mean we will lose peg, because for that to happen more that 89% of the FRAX needs to be redeemed.
We have a lot of concentrated liquidity (FRAX-USDC on Univ3 for example) that is locked for up to 3 years. The FRAX in the pair can not be redeemed, the USDC in this pair will buy FRAX and lock it up in case of a sell off. So the actual FRAX that can be redeemed is much lower than outstanding supply, meaning a bank run can not deplete all the the collateral and we can not unpeg.
The locked liquidity is working as de facto collateral, so practically we are overcollateralized, but in reality we are not.

That locked liquidity is a crucial part of the design is not well understood. Many Frax forks have removed the locked liquidity and got REKT.

Lending out FRAX via the Lending AMO or via Fraxlend does not make Frax more solvent, because for every Frax lend out we get exactly 1 FRAX worth of CDP. No matter how overcollateralized the CDP’s are, they are still just worth the amount borrowed. So when lending out FRAX, the assets increase with the same amount as the liabilities leaving us with the same (in)solvency.

Most people think that DAI is overcollateralized, but that is a common misconception. All CDP’s are overcollateralized, but DAI itself is just 100% collateralized. For every DAI lend out they own 1 DAI worth of CDP.

Since lending out FRAX does not help with solvency or preventing bank run (it makes it worse, because people can easily short FRAX), I do not understand the problem you are solving.

1 Like

Max pain is the ultimate stress test:

  • A big FXS/FRAX whale suddenly starts to sell of their FRAX and FXS and starts a FUD campaign.
  • FXS price drops
  • People panic and start selling more FRAX and FXS
  • FRAX unpegs and FXS drops further
  • If we run out of collateral we need to sell FXS
  • More FUD, more panic until only those that can not sell because they are locked are still holding FRAX/FXS.

In the current situation this scenario can not play out, because there just is not enough free FRAX in circulation to unpeg it (thanks to locked liquidity).

If we add a lot of lending as you propose, this might play out different, because then people can borrow FRAX and sell that as well. If we lend out enough the peg could break, causing more FUD and more panic.
We would then increase rates to force repayments, but that is a slow process. After some time the peg will likely restore, but most damage to our reputation has then already been done.
The attackers that start this bank run could profit via shorting FXS, so even in the case of a repeg they still profit.

1 Like

@Denett Saying that DAI is ovrecollateralized is a misconception? :thinking: :sweat_smile:
You seem to, completely, disregard the fact that lending is a profitable business Ser.

1st) Interest Rates can be hiked as fast as overnight up to 99% the $ value of the excess collateral present in the CDPs, if needed.
Did you wittness what happened to DOLA recently?
Pretty much that.

This is valid as long as Interest Rates are not fixed.

2nd) He also explains quite well why Shorters aren’t a problem at all.

“A Soros attacker with unlimited capital should be every stablecoin protocol’s wet dream, not their nightmare.
Shorters are the protocol’s customers. A large shorter is a large customer.”

Those are all profits that come through the lending market that help you back Seniorage, plus buy pressure for FRAX when CDPs are closed, plus FRAX’s supply contraction (needed in the contraction phase.)

This litterally solves the security problem that Seniorage has @ scale.
Seniorage, impacting FXS’s price, litterally solves the adoption/unprofitability problem that overcollateralized stables have.
Those are the problems.

Honestly, Sam gets it:


Seniorage’s liabilities (FXS, and kept in place as it is) backed with credit/debt from lending.
That is because IR make lending profitable up to the extent the Money Market is overcollateralized.

I know It surely isn’t the esaiest thing to understand, I’m gonna work on some schematics to make it more visually immediate.

The idea is cool, however,
It looks very unfavorable to borrowers, as they will have to take great risks and most likely high IR.
As soon as the bank run happens, those CDP owners will absorb the $FXS sell pressure from the open market. IR is determined by when the bank run happens. What if I open a CDP today and it happens tomorrow?
I don’t think there is enough interest for this lending product.

The risk of bank run is high.
It’s good we can ask some borrowers to take the risk so that we can be 100% solvent.
Borrowers: but why do I have to take such big risk?

@messi Ty for your questions Ser.
For CDP owners if FXS experienced sell pressure it might be percieved as a discount (bcs IR would be paid in FXS or insta-swapped for FXS by the protocol, depending on the phase of the contraction/expansion cycle) but that would not be really accurate because IR would ultimately be priced proportionally to the $ value of the oversupply of collateral in the tot CDP positions.

The fact is that Variable Interest Rates already work like this in all other Money Markets like Inverse or MakerDAO.
They can be hiked to protocol needs, very quickly, to the purpose of contracting the supply of the stablecoin (credit to the users).

Our advantage vs the competition would be in two things:

  • As we approach the inflexion point at the bottom of the contraction phase (FUD/BEAR/RUN),
    besides demand to close CDPs is buy pressure on FARX, Seniorage recollateralizes from IR profits, Collateral Reserves built during expansion and eventually FXS selling.
    Because Seniorage is recollateralizing, Collateral Requirements become lower and lower plus, from the borrower’s perspective, buying FXS for future IR is cheaper:
    Your are incentivized to open a short position on FRAX with us vs the competition because we offer the best deal, plus we actually are the ones providing the credit in FRAX.

  • Thanks to Seniorage we stay more profitable than the competition during expansion, as we decollateralize FXS’s price is growing and our Collateral is put to work in AMOs.
    These profits can really flow in two directions: yeld or lending subsidies (aka growth/adoption vs security/resilience)
    This would give us a nice advantage vs competitors in the lending space IMO.
    Lots of levers to pull in this sense too!
    :pray:

Thanks for the response.
I have no problem understanding the importance of lending market for FRAX, in profitability as well as helping stabilize the $FXS market.
But I have concerns about the magnitude it can help during a bank run event.
One party exerts a big force in one direction, you’ll need a force as big in the opposite direction to keep it balanced.

I like your analogy, I’d say we can think of it this way:

Debt vs Credit

Protocol Debt = Emissions are protocol’s debt, in the sense that people have a credit vs the protocol that they can redeem through Seniorage. For collateral.
So the protocol has a debt, users have credit.

Protocol Credit = Lending acts exactly in the opposite way, creating a debt for the users, that sooner or later they will have to pay it back to the protocol.
So the protocol gets credit, users contract debt.

That is exactly fitting in the analogy you just made because “debt” users are “short” vs the peg, “credit” users are long the peg!

Protocol doesn’t own the CDP, users do.
I’m not sure what you mean by protocol gets credit in Lending. Protocol only gets some interest, not the debt, nor the collateral.

Meaning that the protocol is the creditor and users contract debt.
In Seniorage the user credits its collateral to mint FRAX and the protocol gets a debt that will need to be repaid when redemtion will happen.

CDPs are owned by the users, IR are a stream of protif that happens proportionally to collateral oversupply, the protocol has a claim on that excess collateral that can be exercised through IRs or liquidation.
If you borrow 100$ (120$ collateral) you will never be able to repay 100$ and get your collateral back. You will have to pay IR on top.
That’s always been the case.
That’s the profits that borrowing generates and thay can go as high as 19.99$ in this case
(to force your repayment).

This seems to be the main point of our disagreement. So I will do a small example.
I am buying a house worth $100k, I get a $50k mortgage from Bank A at a floating interest rate and put in $50k of my own.
After a while, bank A needs some cash so it wants to sell my mortgage to Bank B.
How much is Bank B willing to pay for this mortgage? It would pay around $50k for it, because that is the value of this mortgage.
If Bank B has just this mortgage and also borrowed $50k to pay for it, its balance would look like:
Assets: $50k mortgage
Liabilities : $50k borrowed
The collateralization ratio of bank B is 100%
Note, that the balance would look the same if the house I bought was worth $1M instead of $100k and I put in more money of my own.
Conclusion: My mortgage is overcollateralized, Bank B is 100% collateralized.
Same for Maker DAI: CDP’s are overcollateralized, Maker DAI is 100% collateralized.

If executed correctly it can be profitable for sure(avoid bad debt!), I like it and we are already doing it and we will expand further with Fraxlend, but it can not turn an undercollateralized stablecoin into an overcollateralized stablecoin. And I understood that that is your claim (because it is in the title). The reason is that borrowers can leave whenever they like (overnight) and then the protocol is just as undercollateralize as before.
If that is not your claim and you are just saying that lending can make us money, then I do not understand why you want these special rules with tying the LTV with the CR or something, this sounds just dangerous. The LTV should be set such that the loan is safe from bad debt. Avoiding bad debt is the number one goal in lending.

This part is not clear “up to 99% the $ value of the excess collateral present in the CDPs”. Interest rates are usually specified in percentages.
You should not be allowed to set an insane rate overnight, because then you can just rug the borrowers and nobody would want to touch your protocol. You could set it to for example 100% per year and then you earn 0.27% per day. That is high enough that most borrowers would want to repay their loan quickly. If they on average repay in a week, you will earn 1.9%. That is a lot, but not enough to get you collateralized and you just killed your borrowing market.

So unless DOLA can rug its borrowers overnight, it would unpeg in case of an attack, just that it can not be kept unpegged forever, because the attacker would run out of funds.

That’s the key point. around 50k, but never 50k, I can gaurantee.

This recent argument, always from Nour, allows me to explain why a model like mine would be fine in a worst case scenario.
I’ll make a very simple example that can also, hopefully, illustrate the mechanics of variable Interest Rates in a cryptocurrency’s Money Market:

He’s argument is, basically:
Inverse can contract the supply of DOLA at will, by raising rates and forcing repayment or liquidations, when it wants, proportionally the demand drop for DOLA.

Aka:

  • User/Market borrows 100$ of DOLA
  • Inverse gets 120$ of collateral (for instance, 20% overcollateralization or OCR)

If the demand for DOLA goes to 0, supply gets contracted to 0. How? IR.
If the market asks to close the CDPs say, by redeeming 1$ at a time they will never be able to pay 1$ to have 1.20$ back.
It’s gonna be 1.01$ to have your 1.2$ back. (in your 50K$ analogy, yes ALMOST 50k$)
This because lending is a profitable business, those are IR that need to be paid.
Next $? it’s gonna be 1.02, then 1.03 and so on.
If demand keeps dropping and the peg is endangered, the Money Market keep raising rates, because it wants its stablecoins back, you either pay and give them back to us or we raise ABOVE the 20$, you are liquidated and :wave:

As demand for collateral redemption (or CDP closure) goes up, IR go up proportionally, they go hand in hand so it doesn’t matter if you want to get back 1$ at a time or 100$ alltogether.

For these reasons when demand for DOLA goes to 0, and all its circulating supply is removed from the market Inverse will not be left with 0.
It will be left with profits from the IR paid by the market for borrowing.
How much? 20$, no more than that.

Ofc who redeems/closes first pays almost as much as he borrowed ( almost 50k)
But in a complete bank run scenario the last ones who redeem pay A LOT.

So there you have a Money Market who suffers a run and it’s left with profits

If we think about the same scenario for FRAX and we don’t account for locked liquidity (thanks btw, I was not aware of how that worked)
If we have a CR of say… 80%, then we are undercollateralized by 20% (UCR)

Aka. 100$ of emissions and if people will come to redeem them all we’ll have a 20$ debt.

My suggestion is we keep both the undercollateralized AND the overcollateralized designs running at the same time so if the bank run happens we will be left with no profits, but no debt also!
+20-20= equilibrium

(I actually think that Overcollateralization + Undercollateralization will result in Overcollateralization * Undercollateralization, in the sense that the models are complementary in a way that will allow scalability like never before, besides the fact that they can withstand a bankrun)

Somehow you omit the factor time. Interest rates are paid per block and are stated per year. The current borrowing rate at DOLA right now is 3.86% per year and the max interest rate is 58% per year. At the max rate it takes 8 days to earn one percent and 5 months to earn 20%. Borrowers can and will pay back a lot sooner and remove all their collateral in case we hit the max rate, so I do not see how you think that you can earn at least 20%.
You are giving arguments to show that all DOLA is recallable and that is true, but there are no guarantees on how much interest is earned on the way out.

This is a valid point you bring up, definitely worth discussing and actually one that I was thinking through these days…

The key point here would be that, for security reasons, if we assume a single borrower for all the Money Market’s TVL (which isn’t going to be a real world scenario AS LONG AS we discuss this and disincentivize this kind of attack) If he’d want to close all the position at once and basically cause a bank run on the lending side of the protocol, we should be able to instantly charge the correct IR amount which would be amounting to 100% OCR.

If we assume just one borrower, borrowing 100$ of FRAX and placing 120$ of collateral,
IRs should have the highest rate of responsiveness to what the demand for redemption/closure is.
(In this case demand is binary so: -1: you have demand -0: you don’t)
If the market has 0 demand for your stables IRs should be hiked to max out the protocol’s potential for profitability because that also means that the protocol’s giving itself the highest chances of succesfully contracting the supply of the stable to meet demand. (as Nour keeps suggesting https://twitter.com/NourHaridy/status/1527267954068447232?s=20&t=W9ffDk-pvHRM9jqcZ0FA5A)
So, in this very extreme case, 0 (aka you don’t have any demand) we instantly charge all the IRs we can on the closing position. Which will be those 20$.
If the position is redeemed by half, say 50$ we charge just 10$ and so on…

This could be considered like an “algorithmic” way to equilibrate IRs.
Consider that this would be the most extreme of the possible scenarios and would mostly affect “whale” activity, disincentivizing attacks in this sense and promoting a healthy usage, divestification and protectiong for smaller CDPs in Fraxlend.

This is because the closure of any normal sized position would hardly affect the totality of the borowing demand/ Fraxlend’s TVL, so any regular user shouldn’t experience any sharp increase or decrease in IRs based on the closure of his position alone.

This would gaurantee solvency during the contraction phase, plus if you factor in:

  • IR subsidies financed with organic profits generated by our AMO’s strategies.
  • Dynamic Collateral Requirements to open new CDPs, inversely correlated with Seniorage recollateralization (Seniorage recollateralizes → less collateral is required to open new CDPs: LTV goes up) acting as a potent incentive
  • Higher rate of protifitability, even from IRs, compared to the competition because of the strategy/exploitation of the correlation between FXS’s price action and Seniorage’s phase’s (contraction/expansion).

I believe we could still achieve a very high level of competitivness for our CDPs!

Governance being able to exercize that claim on OCR against Borrowers, with IRs or liquidations is exactly the “Govvernance vs Users” conflict of interest that is typical of Money Markets.
The lending side should alwys be able to force a complete supply contraction in case borrowing demand drops to zero by maxing our IRs.

Eventually inflection from expansion to contraction will happen when D.Collateral Requirements will become too high and will end up disincentivizing borrowing so OCR will fail to keep up with UCR:
UCR>OCR
:thinking:
If maxing out IR from the get go , when this hapens, is percieved as too much or too high as a fee, we could make some simple calculations on what is the exact velocity we think we want to have in a worst case scenario (IMO velocity/severity of IR hikes should be calibrated proportionally to borrowing demand drop and recollateralization in the Seniorage), then come up with what exactly is the max extractable value in IR according to OCR and that velocity/severity.
If, for example, we want to be more mellow and we would eventually end up collating 50% of that OCR (worst case, complete flush), lets just decide it beforehand and then use that 50% as the reference to try to balance it with UCR. :bulb:

  • Then, during contraction, when the borrowing demand drops OCR shrinks accordingly to how much value we extract from it with IRs, plus ever CDP closed is buying pressure and supply contraction for FRAX, that help us to move us to equilibrium.
  • If Seniorage sees the UCR>OCR (or UCR > max extractable value from OCR) demand for expansion should drop too and then the undercollateralized side should start the recollateralization process
  • As Seniorage recollateralizes (even using our collteral reserve accumulated selling IRs fees for collateral during expansion) D.Collateral Requirements drop in Fraxlend, accordingly.
  • At the end of the day is a long vs shorts kind of game, if we think borrowers are short the peg (if FRAX depegs and goes to 0 their debt goes to 0), as Collateral Requirements drop to open a new short position vs the peg, borrowing demand should resume in this phase.