[Working group] - Redesign NDOL mint and burn algo

[ericjuta]

Two aims,

• keep collateral pools in balance for trader borrowing capital efficiency, help maintain stability
• have a defacto framework to progressively onboard new collateral pools and assign a target weighting.

Factors to take into account :-

• Average mint collateral price for pool and how far current price is
• Predetermined assigned target weighting per collateral to get collateral AUM to reach as a ratio of total minted ndol
• How much traders are profiting from the collateral pools reducing the pool amounts vs minted amount
• Historical Price volatility of the collateral asset on multiple timeframes

We want to :-

• penalise minting too much of an overweight asset
• penalise burning for an underweight asset
• reduce collateral debt ratio if number go down and people want to sell the dump
• "Heal pools" via minting or direct deposits if they're getting hammered by profitable traders
  (Can tweak more for min profit factors)
• Refrain from using bond treasury funds to burn ndol to reduce collateral debt as a last resort
• Incentivise minting an underweight asset with less fees to bond for necc and lock collateral in

Also making it clear,
IL is combated by staking NDOL for nNECC via 1% fee on bond purchases.
The algorithm change is in addition to this as well as NECC/NDOL LP bonds.

Lemme know thoughts taking these into account, we should formulate an algorithm together ❤

[ericjuta]

[fullmetalengineer]

Hey Eric, I'm Tac (TacticalMinivan) from the discord.

reduce collateral debt ratio if number go down and people want to sell the dump

You're sure you want dynamic debt ratios? Are there any other protocols you know of that use that approach?

[coryondrejka]

+1 to this. Seems like step 0 is to crib off of others' learnings.
In terms of penalizing underweight pools, what about simply using incentives a la Osmosis, Trisolaris, etc, have the nNECC emission get divided among underweight pools?

[resdegen]

Revenues from the protocol might be used to create an LP NDOL - [Stablecoin] on Trisolaris or Rose. That will help the peg.

It is a loop:

• People would gladly mint NDOL because they know they can swap it at any time regardless of utilization rates
• Therefore collateral increases
• Which increases revenues and reduces utilization rates
• With more revenues you even add more liquidity to the LP and more people are interested in having NDOL since they can stake it for nNecc, which accrues value from revenues

Repeat.

Extra: trying to partner with Trisolaris or Rose to provide a high farming APY for the NDOL - [Stablecoin] pair. Increased revenues by Necc Protocol incentives the usage of NDOL and therefore brings volume to Trisolaris or Rose. Besides, revenues from the protocol are used to provide liquidity to the LP, increasing TVL for the DEX.

[Added] Alternative: Use protocol revenues to market buy ETH / BTC / NEAR and mint more NDOL, reducing utilization rates

[case-flatline]

ideally all NDOL activity happens directly through the protocol - this is actually a key part of the rebalancing process

[resdegen]

Yes, ideally. After these loops are done an collateral is increased and utilization rates are decreased, people could directly swap on the DEX using the LP or burn NDOL using the protocol. Now we could incentivize the burning mechanism to use the protocol instead of the DEX

[jommi9]

Rather do Wannaswap than tri as the fee is 0.2 vs 0.3%

[case-flatline]

@rej156 it might be useful to try and get in touch with X, the lead dev of GMX, and ask him how different GLP rebalance strategies performed in the early days

He's a good dude, would definitely be down to chat

@xdev_10 is his TG, I can always reach out as well

[BAXYCode]

for what it's worth, I sent him a message on TG requesting his counsel for the protocol

[leaf22]

Initial thoughts:

• Significantly increasing cap of Target Adjusted Fee for overweight assets should achieve our penalty goal.
  Traders currently seem unaffected by minting fees. Unsure if higher fees would have a significant impact, but the worst-case appears to be more stable fee income. We should be aggressive with this initially.

• I would like to add the average Utilization Rate of collateral pools over a specific timeframe as another factor to take into account. With price volatility concerns in mind, the impact could be a percentage change based of the original target weight, which was assessed according to these risks. This should lead to better overall utilization and higher fee income.

• How much traders are profiting from the collateral pools reducing the pool amounts vs minted amount
  I'm unsure whether this is something we should initially take into account. Assuming no underlying gamification issues, I do not see the reason to adjust based on realized profits. Direct Deposits/Active rebalancing however seem necessary to counteract pools potentially getting depleted.

• Refrain from using bond treasury funds to burn ndol to reduce collateral debt as a last resort
  Could you elaborate a bit on why you view this as a last resort measure?
  At the end of the day, NECC holders would be the only ones benefitting from a vastly overcollateralized protocol (more collateral than can possibly be redeemed).

With the assumption that NDOL redemption values go up naturally and the protocol will eventually become overcollateralized, burning part of the treasury does not result in money being lost. If you will, we are just transferring it to settle a currently outstanding debt.
I think of it as a conditional burn where we agree to transfer a corresponding amount of funds back to the treasury once a certain level of overcollateralization is reached.

[ericjuta]

chad reply, agree with all you've said.
wrt burning ndol from the bond treasury, I think it's more that in other ohm forks, they allocate treasury reserves to yield and repay the debt.
In our case, it would be burning the ndol to reduce collateral pool debt and incurring a treasury debt initially but yes repaying it from protocol fees later on
Not an issue, just being aware of it is fine, treasury reserves are there to be used.

[0xInoculum]

Do you have a technical writeup on the mint|burn algo with equations or can you point me to the specific source code for it?

[ericjuta]

https://docs.necc.io/ndol
https://docs.necc.io/rebalance-mechanisms

[ericjuta]

bit of a naive mint and burn algo, lemme know what you think

[ericjuta]

for minting, the diff above target weight is a multiplier factor

[0xInoculum]

No worries. I think we can get this sorted out. Just need to follow the logic before I can contribute anything of value here.

[ericjuta]

oh this is what I propose as a new one, not current btw
links above are current in previous comment

[0xInoculum]

Oh. Got it. I'll try to catch up.

[NotNarNar]

Probably off topic, but I think the biggest reason ETH is overweight is because ETH is the asset people bridge to Aurora, and it’s the gas token.

They land on the app and it displays mint ndol with ETH. I doubt they even know that you can mint with more than one asset.

Adding smart contract risk is probably not good, but maybe the protocol could swap the ETH to underweight assets automatically before minting.

Or somehow make it more obvious that there are discounts for minting with other assets.

People are lazy and a lot of them probably don’t know which dexs there are or how to find them since this ecosystem is so new. I would guess they mainly want to bridge over and buy necc and not think about swapping ETH first and looking up other apps to do that.

Maybe even something in the ui to show them the best deal, or default to the best deal in the drop down, and then link to a dex for swapping?

Also another thing I’m aware of is how other CEXs incentives holding onto bad trades, or balancing the trades by paying people to hodl their longs when the market is down through rates. Not sure if that would be useful. Sorry to derail your algo talks.

[ericjuta]

Yeah you're talking about a ZAP in smart contract that splits it up via AMM swaps
Had a PoC couple of months back on that 😊

Good shout bout the UI part, yup someone mentioned that showing the info for average redemption ratio should be on the Mint tab too

Was thinking on your last point maybe that we could mint a nft to represent your position and you stake that for fee rebates or something, who knows

[case-flatline]

++ from me on ui, I’d take it even further and completely remove anything regarding calculations for NDOL minting, and only show:

1. How much NDOL will be minted
2. Avg (maybe a weekly avg?) redemption ratios for each asset
3. Fee and whether asset is under/overweight, with suggestions for each
4. Warning explaining that they are providing assets as a liquidity provider, and their assets will be borrowed by derivatives traders

[ASAP-Beans]

Paraphrasing what has already been discussed in discord but recording it for the tape.

I agree with sentiment on focusing on increasing AUM to replenish peg over time via fee accrual. Paying out of treasury reserve isn't a complete no-go but not first resort in my opinion.

Work flow should be;

• Implementation of new mint and burn algo, most recent proposal seems fair to incentivise correct weightings (this first restores investor confidence)
• Tweaks to necc platform fees, so the house captures additional revenue on previous 'win-only' trades for traders (I can see there are already discussions on how this would be done).
• Implementation of additional assets to increase AUM, we should choose low hanging fruit from a easy of implementation POV as well as strategical in terms of assets increasingly favoured by traders on Aurora or cross-chain (we implement this only after the previous two as we have the investor confidence of new algo from the new participants who are joining because we added support for 'x' token as well as already having tweaked the numbers prior to the addition of new AUM).

The con of this workflow is it's completely unknown how long it may take to restore peg via the more organic route, we know this protocol needed some bull before bear and I'm not sure there is an easy long-term fix without waiting for those assets numbers to increase favourably but in the meantime it does make sense to focus our attention on re-organising some of the game theory to lessen its impact as our increased fees over time slowly bridge the gap.

[cryptonesta]

I think we are drinking too much hopium in regards to the fee structure saving us at this point. I agree the overall fees must be higher, but it's important to keep in mind that the USD 700.000 or so that has so far been gathered through fees are fees received from exploiters maxing the ETH utilization. After the exploiters took off, ETH utilization has been around 5-6% and fees have been nothing to write home about.

Our number 1 priority should be to once again increase the AUM. For that to happen we need to

(1) Make it attractive to invest. The peg is a HUGE turn off at this point in time.
(2) Make sure we have a functional product. An exploit can happen the best of us, but we have to make sure it does not happen again. If there is a hole somewhere, then its better to stop the trading.

Thinking a more aggresive fee structure will save us at this current AUM is not realistic imo.

[ASAP-Beans]

I am in agreement I don't expect the fee structure to make up for large portion of the ndol de-peg in the short term and neither am I against using the treasury to plug this gap.

However a direct deposit to fix the peg will just hasten additional exits and leave us back to where we were before which whilst having some benefit isn't the most capital efficient hence why I wouldn't have it as option #1.

I believe someone in the discord mentioned a 'QE-style' utilisation of the treasury funds with perhaps a planned disbursement period of X% contributed towards peg restablisation. Why I think this method is important is the expectation re-anchoring without the stop-gap solution of using a large quantity of treasury funds for exits for n participants but no long-term fix.

The cold hard reality is the redemption value isn't improving until the prices of the underlying assets are so I believe we should at least signal to re-anchor expectations by implementing the above fixes and perhaps contribute funds from the treasury on a disbursed basis which also helps close the gap whilst not inciting a mass AUM exit.