Centrifuge LP Rewards: Ideas for Improvements

Centrifuge Governance Proposals
1

In the last couple of weeks we’ve put out a call for the community discuss how to improve the way LPs are rewarded. We’ve had the first discussion on our discord channel #wg-cfg-rewards and I wanted to present the first set of ideas that have been discussed for broader feedback.

Contributors: @Tjure07, @Ash, @lucasvo, @zaatar, @omegafattyasses

What’s the goal of our LP rewards?

Centrifuge is a decentralized network that requires stakeholders to hold CFG and use it to participate in the network by voting, staking & performing other work. To seed this community of token holders the protocol mints rewards to early users to give them tokens for them to use and become long term incentivized token holders.

Below are a few updates to the CFG Rewards program that the community has discussed in an effort to overhaul the rewards in the coming weeks. We’d love to hear what you think about these ideas and would like to have an open discussion before starting to poll for these proposals.

Note: the numbers below are using a price of $1.5 per CFG, the 30 day TWAP is currently $1.60.

Discussed changes

The rewards working group discussed 3 topics around the Centrifuge rewards program:

  • External protocol rewards for users who interact with Centrifuge indirectly for example through providing liquidity in an Aave Market that has Centrifuge collateral
  • Different reward rates for TIN/DROP: TIN token holders have a significantly more involved role in the network, they take first loss of pools and have to perform a lot of work underwriting.
  • Lockup incentive: The goal of the program is to reward long term involvement in the network. Should we encourage users who intend to hold CFG and stay involved longer.

External Protocol Rewards: Aave

With the integration of Aave being close to launch we should start discussing how the community wants to award users of Tinlake Pools via external protocols. Given how different all DeFi protocols are it’s best to look at them case by case. Below we are proposing a reward framework for Aave markets that accept Tinlake LP tokens as collateral.

The Aave protocol allows issuers such as RWAMarket to create a pool of pools. Tinlake LP tokens (TIN or DROP) can be used to borrow a stablecoin from the market giving the issuers of the Tinlake pool additional liquidity to borrow from.

Aave has been very successful at launching new markets (e.g. the Polygon and Avalanche Markets) with a liquidity rewards program. We believe it’s a good idea to reward early liquidity providers in the market. Having significant liquidity early on is important in ensuring the protocol functions well from the start.

How the Aave Protocol Distributes Rewards

The Aave protocol distributes rewards very differently from how the Tinlake rewards are built. Aave rewards are fixed and split proportionally between all liquidity providers in a market. This means if the liquidity in the market doubles, the rewards per $ in provided liquidity will get cut in half. Rewards are allocated over a timeframe and are distributed on an ongoing basis. This means when planning Aave rewards the community needs to decide on a timeframe upfront, make an assumption of how big the market will grow and allocate rewards accordingly.

For the initial launch of RWA markets in Aave, we propose to decide on an allocation for 3mo. That will give the community time to evaluate how the program is going and adjust them based on these learnings.

Aave Rewards Launch Proposal

Target Market Size

An important aspect of the rewards is the target size. We believe a significant amount of liquidity for Centrifuge pools can come from Aave. Looking at capital demand in the next three months we believe the market can and should reasonably grow to USDC 50M in size.

For that reason the below proposal for rewards models them out at $50M in market size.

Target Reward

Looking at early rewards programs a target reward rate that would be an equivalent of 10% return in rewards seems competitive and would adequately reward early Aave users to provide liquidity in the new market.

When we model this at different sizes the rewards return looks as follows (based on a price of $1.50 per CFG):

952×664 45.5 KB

Proposed parameters for the Aave Centrifuge Markets launch

Mint CFG 833,333 for the first 90 days of the protocol launch and deposit these into the Aave Rewards Distributor Contract.
Distribute 9250 CFG per day as rewards across all AAVE Centrifuge markets.

With this proposal we would see an additional issuance of 0.76% of the current total supply for Aave Protocol rewards.

External Protocol Rewards: MakerDAO

While MakerDAO’s strategy for RWAs is still under discussions within MakerDAO, discussing a framework for CFG rewards is tough. We currently don’t know what direction the community wants to go and how arrangers fit into this. With this uncertainty it’s probably better to hold off and decide on a framework once this is more clear. In the long run we should address this topic and also make sure to refer back to the conversation started in the forum by Paper.

Originator Rewards, TIN & DROP Rewards

Currently AO rewards are 1/4th of TIN & DROP rewards which means they are still rather high and it often leads to AOs receiving rewards both for providing liquidity in the junior tranche as well as for the loans they originate. Rewards for investment in the junior tranche should outweigh rewards for asset origination, therefore further reduction of AO rewards could be considered. An increase in TIN rewards means that rewards for AOs that provide their own TIN are not changing significantly and instead we encourage additional junior tranche investment by AOs as well.

1518×292 39.3 KB

Lockup Incentive

While the network is still at an early stage we want to incentivize long term holding of CFG. This could be done by giving users an option to lockup their rewards in exchange for a higher reward rate. Due to technical complexity the implementation of this will take some time.

The current implementation for claiming rewards can be modified to add a second extrinsic that allows a user to claim locked tokens instead of unlocked ones. When using this extrinsic users would get an extra bonus on top but can only receive tokens after a minimum time delay. To keep the lockup mechanism as simple as possible the claims module should only offer two options: unlocked rewards or locked rewards. The option to vary rewards depending on time will push the technical complexity beyond what can quickly be implemented in the near future.

Due to the timeline for this, we do not propose to agree on any parameters as of now and wait until the technical work is closer to completion before doing so. But if the community supports this proposal development of this can probably be accelerated.

6 Likes
2

Thanks @lucasvo for the proposal, below are my thoughts on this, would appreciate input from other participants as well to make sure the right decision is made here and DAO democracy is utilized to the fullest.

Lockup proposal - in my opinion, a lockup is very much needed, but I am not sure it needs to be incentivized further with rewards. There are some existing examples such as the “cooldown” mechanism used by Maple Finance (here) that prevents a withdrawal for X time. I believe this is the correct structure for less-liquid assets such as those on Tinlake, and will give some certainty over the funds locked in the pool and the available liquidity for AOs. Something like 90 days may be a good start, or perhaps 30 days for DROP and 90-180 for TIN.

External Protocols - in my opinion, this needs to be a uniform proposal for all external protocols, if there is one. Not sure why AAVE would be getting rewards but MakerDAO would not be, and if they are not ready, I would say that waiting and then implementing a uniform rewards process is the right way to go. I also believe 10% rewards for external protocols is a bit rich, looking at the current APYs on AAVE, with the addition of DROP APY, this will be a fair amount more than what I believe has been the average there.

AO Rewards Decrease - while some decrease may be in order, in my opinion, this is a very steep cut, a ~90% reduction from current levels, which have just been reduced by 20% in September. TVL has not yet grown substantially, and the price has actually dropped since the last rewards reduction, and this doesn’t make sense to me at this stage. Perhaps another 20% decrease, in line with the previous actions, is warranted, but 90% is steep. With the current ETH gas fees, AOs are spending $500-$700 per each loan to lock/finance and then additional to repay, and this decrease would really hurt.

TIN Rewards Increase - I believe this is a far too steep of an increase here as well, perhaps a 20% increase may make sense to incentivize new participants.

I am not sure if putting the TIN onus on the AO is scalable. Certainly, some skin in the game should be required, but in my view, the beauty of the protocol is the ability to attract market participants, spread the risk and get rewarded for it. Surely, TIN deserves a higher rewards rate than DROP, but I think the proposed changes are very drastic. I would love to see some cadence and a bit more certainty around these changes, we have had 20% decreases every few months, this may be OK, however, this seems a rather large change and not appropriate for this time.

1 Like
3

Thanks for your comments @prankstr25

uote=“prankstr25, post:2, topic:3237”]
I believe this is the correct structure for less-liquid assets such as those on Tinlake, and will give some certainty over the funds locked in the pool and the available liquidity for AOs. Something like 90 days may be a good start, or perhaps 30 days for DROP and 90-180 for TIN
[/quote]

I think there’s a small misunderstanding: tocken lockups for the LP tokens are super interesting and something that would be very interesting for the protocol though it’s not what’s discussed here. This is about locking up the CFG rewards that users receive.

When we launch pools on Centrifuge Chain we will probably be looking into lockups as one of the features that will follow soon thereafter.

As the Aave Markets with Centrifuge assets are a more novel thing than the Aave V2 market on mainnet it’s probably a better idea to compare this to the Aave Polygon market and others. As mentioned though, this is an allocation for rewards for the first three months and with the target of growing this market significantly more rewards will go down quite a bit from there.

On your overall statement on issuer vs TIN rewards, in most pools the issuers actually hold a large part of the TIN and thus this change will not affect these pools adversely. Additionally it actually creates an incentive that is good for the investors and the safety of these pools: it incentivizes the issuers to provide more of the junior capital which means they have more of their own value at stake.

4

Thank you for the clarification @lucasvo and appreciate the color on the community call.

I think to sum up our concerns, with so many moving targets, a gradual approach to changes would be much more preferable, at least that will give everyone a chance to prepare and adjust. I believe you are on board with that, from what I head in the call.

Would be great to hear feedback from others in the community as well :slight_smile:

1 Like
5

I think the framework for the AO, TIN, DROP rewards is right, but the discrepancy is too large.

For example: TIN:AO Reward is 50:1, junior capital is valuable, but I question whether it’s valuable to that extent.

A possible solution is to simply further reduce both the TIN and DROP rewards to be closer to / less extreme in difference from the AO Reward.

This could something like this on Reward Yield basis:

AO = ~2%
TIN = ~30%
DROP = ~20%

But, I think it may make sense to look into the way this works in the real world or perhaps have someone else way in with more experience on how to incentivize such yield returns.

I’d be most curious to understand the logic of this framework, or why the numbers that have been, why they have been chosen.

Thanks for the continue work and transparency on this

1 Like
6

Aave Rewards Launch Proposal

I’m really excited that the Aave protocol allows issuers such as RWAMarket to create a pool of pools. Something that has not been possible on Centrifuge before and will make it a lot easier for issuers to finance real world assets.

Having significant liquidity early on is important in ensuring the protocol functions well from the start.

One part of convincing users to place their liquidity into the RWA market on Aave & Centrifuge is high liquidity rewards at the beginning.

1 Like
7

Tin Reward increase - Not sure I really understand the motivation behind the proposed
redistribution of relative rewards between AO / TIN / DROP providers. I feel scarcity should drive the reward focus, namely what is the relative difficulty in attracting these different groups to the platform. My perception is that TIN providers already earn a very good APY return on their investment (particularly relative to the apparent risk) & boosting this with additional CFG rewards seems over-kill. If there is difficulty in attracting TIN investors then one approach to address this would be to reduce the TIN investment threshold from 50K to say 10K to make it more accessible, as at the moment I would speculate that TIN investing (particularly those that are not also AO) is limited to people with investment portfolios > $1M.

Lock-up of CFG rewards - Personally not keen on this idea. Short duration locks are pointless and longer ones will become a new barrier to entry. As there are currently huge numbers of CFG unlocking for for early investors and the CFG team it would feel petty to lock rewards for small investors.