[NIP-70] Plan to manage Notional V3 reserves and conduct periodic NOTE burns

This proposal lays out a high-level framework for an ongoing process to manage exchange rate risks of Notional reserves and execute regular, weekly NOTE buybacks funded by protocol revenues.

NOTE buy and burn

We propose that governance commit to dedicating 20% percentage of the protocol’s revenues to NOTE buybacks.

Every quarter, this percentage of fees across all tokens on all deployments would be converted to ETH and transferred to the Mainnet treasury manager. This ETH would then be used to fund regular, weekly purchases of NOTE such that all the ETH is used by the end of the quarter. Weekly ETH sales would be split between use as incentives for sNOTE holders and the buy and burn per a percentage agreed upon by governance.

This structure would imply that the protocol’s buyback rate would be somewhat backward-looking. For example, the buyback rate for April - June would be determined by the revenues earned in January - March.

Minimizing gas costs

Converting all protocol fees to ETH and transferring them to a single contract on Mainnet will require a significant number of transactions and gas costs. By executing this operation only once a quarter, the protocol can keep this cost to a minimum.

We would also propose an additional caveat of a minimum trade size on Mainnet. If the amount of reserves in a certain token to be converted into ETH is less than $1,000 for example, we would propose to ignore these reserves and leave them in that token.

Managing exchange rate risk of reserves

Notional accrues fees across a variety of different tokens, but its expenses are primarily denominated in USD, with a minority of expenses denominated in ETH. With this in mind, we believe that reserves should be periodically converted to USDC to match the protocol’s expenses with the caveat that we want to maintain some positive price exposure to ETH.

Specifically, this is how we believe accrued reserves should be handled by token.

ETH - retain any accrued ETH reserves.
USDC - retain any accrued USDC reserves.
LSDs / LRTs - convert accrued reserves to ETH.
Stables ex USDC - convert accrued reserves to USDC.
All other tokens - convert accrued reserves to USDC.

Implementation Details

The next step will be to convert Notional V3 reserves to ETH and USDC per the above methodology and implement the recurring buybacks using the current percentage split between sNOTE and reinvestments and NOTE burns (75% burn, 25% sNOTE reinvestment).

The team will target implementation of this proposal such that processes are online to begin conducting buybacks per the agreed pace by the start of Q3 this year.

For V3 revenues accrued between Jan and March of 2024, they will be used to conduct buybacks during either the remainder of Q2 or for a minimum of 4 weeks, depending on when these buybacks begin.

How do we handle the following scenario

90k in Stables , 10k in non Stables across 11 tokens and the value is <$1K.
50% allocated to buy and burn.
Assuming $1k minimum trade size and due to minimum trade size we can’t touch 11 tokens

1)We still use 50% revenue and all 50k comes from stables for BnB

2)We use 50% of 90k i.e 45k for BnB

In case of Option2. The amount that was supposed to be used but wasn’t used how will thaat be handled.

  1. Ignore and move on
  2. Use in future transactions

LSDs are less risky and yield bearing when compared with LRTs. Shouldn’t we be holding LSDs and only converting LRTs.
So ideally it should be USDC , ETH , 1 biggest LSD in terms of $$ in our reserves

Adding this in case it helps understand the scale of burn in different scenarios i.e revenue and price.

The ones highlighted in yellow are the scenarios that I think are most likely to happen.Also, the values are # Note tokens burnt.

Yearly estimates can be derived after multiplying by 12

  1. I think that we would do option 2. We wouldn’t seek to “make up” buy and burn revenue from tokens that have generated larger fees. If we do that, I am pretty indifferent to whether we just ignore those revenues forever or use the min future transactions. If the trade size is under the threshold, by definition it’s a small amount of money that doesn’t make much of a difference. I would lean toward choosing whichever option is cheaper and easier to implement technically.

  2. We could hold LSDs instead of ETH, but I think just holding ETH is simpler, less risky, and more impactful for Notional. All things equal, it’s better to have ETH deposits than LSD deposits because there are more strategies that ETH can be used for. And in practice, we’ve seen that ETH lending on Notional tends to have better yield than lending LSDs anyway due to the higher utility of ETH.

This information was very useful to me, thank you.

Edited main post to add implementation details section