In this formula, the variable reward rate is updated via governance according to TVL growth.
The proposed adjusted formula for LP rewards is:
Reward Rate = Total Rewards / TVL
In this formula, the variable total rewards is updated via governance. Total rewards would now represent a global daily reward, therefore the rate adjusts depending on TVL and would go down gradually as TVL grew. The global daily reward would be set to 30,500 CFG, which is approximately how many rewards were earned system-wide today.
Has the last proposal been implemented? How will this effect the different rates for TIN, DROP, etc, voted on in the last proposal?
The way I understand it, this would dissientivize investors from adding significant investments as it would lower their own rewards. Please correct me if I’m misunderstanding this.
This is not related to adjusting the individual rates of DROP, TIN, but more so creating a streamlined method to adjust reward rate with TVL. The rate changes that will happen with TVL rising are changes that would typically be made anyway using the previous structure, but now we will not need frequent governance proposals to make it happen.
Maybe to add a bit more context on how the reward mechanism works; the way rewards are calculated, we can’t specify a % value tied to token supply for rewards. A value still has to be set by governance and updated on a semi frequent basis.
If we simply took the total CFG awarded per day today (this is displayed on CFG Rewards | Tinlake | Centrifuge as “Rewards earned today”) the value would be CFG 30,500. I think this is probably the most reasonable number.
Thanks @lucasvo this makes sense. How frequently do you anticipate governance updates to daily rewards going forward? It would be nice to have triggers that will cause a rewards rate asjustment, so that we can anticipate and project numbers going forward. For example, TVL=100mm, 150mm, etc, once the trigger is reached, it would cause a rewards adjustment vote.
Also, can someone please clarify how the inflation is calculated, is it based on circulating token supply, total supply or another number?
Thank you to everyone for clarifying and answering my questions.
I think most people define inflation as the % of the total supply issued.
The idea of restructuring the rewards is that they would become independent of TVL. I don’t think defining this would make sense as this is what we want to avoid. However anyone can propose a change to the rewards whenever they see necessary.
I see, so the current level of CFG 30,500 per day would stay static for the foreseeable future? I thought there would be some periodic adjustments to bring it in line with the new total supply issued.
For example, as of today, I am seeing a total of 430,011,123 CFG issued, so the daily reward based on a 3% inflation would be 35,343 CFG. Lets say a year from now, that total supply would increase by 3% to 442,911,456 CFG, so the daily reward would go up by 1,000 CFG roughly. Not a big change I suppose, but wanted to understand the mechanics.
If the goal is to make it fairly static, would it make sense to set the daily rewards to the 3% goal as of now, which is 35,343 CFG/day?
That would increase the rewards beyond what they are today, personally I don’t think that’s necessary. You’re also not counting the rewards for other markets like the RWA Market into total rewards as they are right now.
As I mentioned before, I think rewards should be reviewed by the community periodically or when things change significantly. This can always be discussed on the governance call for example.
The initial proposal was light on the implementation details and I wanted to follow up with more details on what the implementation could look like.
The essence of the proposed reward model is:
Reward Rate = Daily Global Rewards / TVL
This however is a bit simplified as it does not take into account that one dollar in the junior tranche count differently than one dollar in the senior tranche for example. In the most recent proposal token holders voted for the following different rates:
Type
Reward
TIN Reward
CFG 0.00072
DROP Reward
CFG 0.00048
AO Reward
CFG 0.00004
To implement the global rate the reward mechanism takes these as points and multiplies each $ invested/originated in the different categories by this factor. These are example numbers. Rewards in practice will depend on actual TVL and will be different.
Name
Amount
Factor
Points
Junior Token Value
$1,500,000
0.72
1,080,000
Senior Token Value
$8,500,000
0.48
4,080,000
Issuer Origination Total
$8,000,000
0.04
320,000
Total Points
5,480,000
This means the formula for rewards can be described as follows:
Reward per point = Global Rewards / Total Points
Example:
Rewards per Day
CFG 30,500
(set by governance)
Reward per Point
CFG 0.0056
Rewards per Day / Total Points
Effectively for the above example this would lead to the following reward rate:
Type
Reward Rate per $
Rewards per $ in Junior
CFG 0.0040
Reward per Point * Junior Factor
Rewards per $ in Senior
CFG 0.0027
Reward per Point * Senior Factor
Rewards per $ in loans
CFG 0.0002
Reward per Point * Issuer Factor
Should the on chain poll pass, the reward mechanism will be updated to the above specification.
Thanks @lucasvo, in this scenario, to figure out the points, we need to take into account the TVL across all pools, correct? If so, is there an easy way to look up the current Tin, Drop amounts across all pools?
Part of the implementation would be adding this calculation to the subgraph which will allow everyone to read out the current state and who gets what points. Until then your best bet is to model it based on the data already in the subgraph today (i.e. adding up all tranche values and calculating it manually).
