[PIP-10] Should PICKLE rewards be evenly distributed across the stablecoin pools?

Authors

@Larry_Cucumber @BigBrainBriner @0xpenguin (the core devs)

UPDATE: Proposal is now open for voting here.

Intro

As suggested by Robert Leshner from Compound Finance here, the allocation of PICKLE rewards to the various stablecoin pools should be revisited, in light of DAI returning closer to peg.

The current percentage allocation of PICKLE rewards are as follows:

  • Pickle Power (PICKLE/ETH LPs) - 75%
  • psCRV (from staking sCRV in pJar 0) - 4%
  • pUNIDAI (from staking DAI/ETH in pJar 0.69a) - 1%
  • pUNUISDC (from staking USDC/ETH in pJar 0.69b) - 9.45%
  • pUNIUSDT (from staking USDT/ETH in pJar 0.69C) - 10.5%

Issue

The original reward allocations was predicated on the fact that DAI was often wildly off-peg (sometimes fluctuating to up to $1.05). With the recent changes to Maker’s collateralization ratio requirements, DAI has now largely returned to peg.

Proposed Solution

The reasons for a skewed allocation of PICKLE rewards is no longer present, and so the rewards should be adjusted accordingly.

We propose allocating an equal share of PICKLE rewards to each of the non-PICKLE/ETH farming pools (i.e. the stablecoin pools). There is currently no valid reason why one of the stablecoin pools should be favoured over another.

If this proposal is accepted, each of the non-PICKLE/ETH farming pools will be allocated 6.25% of PICKLE rewards, for a total of 25% across the stablecoin pools.

4 Likes

Equal distribution between the stablecoin pools sounds sensible.
The other option is incentivising the Bitcoin jars in PIP-9 (although it might be better/safer to add rewards to that pool later, as suggested in PIP-9).

Sounds good to me …

100% agree. Let’s make it

Actually on second thought, I think the 25% allocations to the pUNIDAI, pUNUISDC, and pUNIUSDT pools are too high. It should probably be revised downwards overall.

We are effectively overpaying for the TVL (since any additional Pickle in these pools is a bonus over simply pooling at Uniswap), and giving these pools with no skin-in-the-game too much Pickle with which they can dump (without becoming invested in what Pickle is about, and holding and staking Pickle in the PIP-8 staking pool).

The fair amount (APY-wise) at which these poolers would be indifferent towards staking at pickle.finance, should be equal to the performance fee we take from them (e.g. if we take a 4.5% performance fee, then the fair amount would be to award that pool with an equivalent ~5% APY in pickle rewards (probably just a bit higher to account for fluctuating pickle prices).

Summary: 25% of all pickle emissions is currently paid to pools without skin-in-the-game.
In reality, to attract any of the TVL in those UNI pJars, we just needed to provide more pickle APY to those pJars than we take in performance fee (~4.5%).
We could probably cut emissions to those pools by 10% and still keep all the TVL (just throwing out 10% as number, could be lower or higher).

Some rough back-of-the-envelop numbers…
If we take a 4.5% performance fee, then a ~10% Pickle bonus APY is more than enough to incentivize new TVL to join. Right now, some of those jars are getting around %30 APY (at pickle price of $19).

So we could probably slash the 25% of all emissions to the pJars in half down to 12.5% and still keep the Pickle bonus rewards APY on those pJars around 15%, and thus keeping and attracting the TVL we already have.

4 Likes

Strongly agree. Pickle’s continued emission is a great asset to keep incentivizing behaviors. But that does not mean we should be overpaying. Reducing PICKLE rewards to people with no skin in the game makes sense. And I am going to repeat myself here, but we should seriously start thinking about ways to incentivize long term presence in the jars and in the farms through dynamic withdrawal fees, vesting or proportional rewards based on lock-ups.

1 Like

This is a very good point.
We need more maths here

I strongly disagree with this notion. (The one I quoted, the parent allocation change in itself is perfectly fine by me.)
We need actors inside of the DAO who actually use the product and not just suicide pool boys to keep a balance. The rewards for the PJars try to give ownership to the people who actually use the protocol and is not to be considered in a “APY only way”. If people want to dump let them dump, but at least give them the option.
Although this original Idea of liquidity farming got forgotten over the summer and is no longer true by itself it should still be considered IMO.

If you are scared of the dumping that’s happening with rewards from pjar rewards, the inflation that is to be allocated to the pjars is about 1.25% over the next week or so and will decline further into the future.
I think the allocation of 25% to the pjars is perfectly fine and see no reason to change.

1 Like

This is not a suggestion to take the pJar rewards away, rather, a suggestion to balance the rewards in a way where we are attracting TVL (“those who actually use the product”) without overpaying, and avoiding attracting opportunistic/temporary liquidity.

I disagree with the idea “if people want to dump, let them dump”. Keeping the price stable (and not allowing it to fall into a downward spiral like Curve) will also promote confidence in the reward pJar users receive, further incentivizing them to consider staking their Pickle, rather than auto-sell.

Think of this way. Imagine paying every Uber user a $10 reward to use the service. You are subsidizing the end-user to use the service, but you are overpaying. None of those users will stay and a simple $2 subsidy should suffice if the goal was to simply attract end-users.

However, with the -%60 crash and halving, it now appears the rewards on pJars are closer to a fair and sensible level.

I tend to agree with @0xBoxer on the point of the allocations to suicide pool vs non-suicide pool. In a vacuum where all we had to do was offer a better return than Curve’s sUSD pool (for example), then yes, we would just have to provide is a farming yield that covers the Jar fees. However, there’s other yield aggregators now in this space we have to stay competitive with (see https://yaxis.io/farms for example).

Your point about “overpaying” for TVL is a valid one. I’d totally be welcome to a discussion of that topic on another thread :slight_smile:. However, this topic was intended to be limited to a discussion on whether the farming rewards across the stablecoin pools should be equalized. It seems that people agree on that point, so we’ll be moving this to a Snapshot proposal shortly.

2 Likes

Against, DAI makes up a fraction of the TVL with rewards still outpacing usage.

I like the idea of revisiting emission distribution but I’m not sold on the proposal to split evenly between all types of stable coins. It seems like using a blunt instrument where a scalpel might be the better tool. Absent the need for incentivized distribution, I would think we could just pair distribution to the size of the pool on uniswap? If we use the current pool sizes, distribution would be:
75% powerpool
6.25% to sCRV (fixed at 1/4 of the 25%)
5.175% to ETH / DAI
7.07% to ETH / USDC
6.49% to ETH / USDT

Just an idea! I like the idea of these being dynamic and algorithmically calculated on a semi-regular basis if possible.

Guess I missed the boat on this one, huh

I think I agree with what you’re saying here. But the proposed 6.25% across the board isn’t actually too far from those numbers. We can re-visit this next week but I don’t think it’s going to make that big of a difference in the short term.

Pickle Distribution PIP-10

PIP-10 intends to bring the distribution of PICKLE token to parity on all farms.
Below is a quick overview of the current, and propsed PICKLE rewards.

Current PICKLE Distribution
| Pool     | Value  | APY    | PICKLE APY | Allocation |
| ---      | ---    | ---    | ---        | ---        |
| DAI/ETH  | 3.55MM | 26.95% | 17.24%     | 1%         |
| USDC/ETH | 30.8MM | 21.88% | 15.5%      | 9.45%      |
| USDT/ETH | 48.6MM | 23.56% | 10.61%     | 10%        |

Proposed PICKLE Distribution
| Pool     | Value  | APY    | PICKLE APY | Allocation |
| ---      | ---    | ---    | ---        | ---        |
| DAI/ETH  | 3.55MM | 26.95% | 107.75%    | 6.25%      |
| USDC/ETH | 30.8MM | 21.88% | 10.25%     | 6.25%      |
| USDT/ETH | 48.6MM | 23.56% | 6.63%      | 6.25%      |

Assuming we do not factor in the halving, or fluctation in performance of UNI farming we can use linear scaling to see how this change will impact the APY of PICKLE on other pools.
DAI/ETH LP is currently the highest APY yield in LP tokens, and also the highest APY in PICKLE.
This makes it the most attractive jar to enter even today at 1% rewards.
The DAI/ETH LP jar has at best, 10% and at worst 6.25% of TVL compared to USDx/ETH LP.
Why? Is it because the rewards are not enough?
The rewards are already the highest yet the attraction is not there.

The proposal to stabalize the APY across the pools because the peg has returned will greatly impact PICKLE users in terms of “double dipping rewards” favoring a small minority (< 5%) at the expense of the many (measured in TVL).

It would be more prudent to ladder up the DAI rewards with increasing LP, perhaps doubling per week or in a staggered way to the 6.25%, but at current this change is too drastic.

2 Likes

One thing that is not being talked about is that the current TVL of the DAI jar is severely suppressed due to the low rewards. For reference, these are the values for $ locked in the jars:

ETH/DAI Jar: $3,552,466.73
ETH/USDC Jar: $30,769,744.47
ETH/USDT Jar: $48,491,472.38

This is likely caused by the skewed farming rewards we had prior to Maker’s 101% CR change.

When the difference in value locked is not an order of magnitude, it would be more reasonable to compare. I also think it is likely that the DAI farming APY will eventually fall towards what it is in the other jars as more liquidity flows into the system.

In other words: I don’t think we are considering that the ETH/DAI jar could potentially 10x in size if we equalize rewards, which is kind of what we’re trying to do here.

1 Like

Actually on second thought, I think the 25% allocations to the pUNIDAI , pUNUISDC , and pUNIUSDT pools are too high. It should probably be revised downwards overall.

What I think is that we should keep or even increase those number, but also increase the proportion share that PICKLE holders take from the jars.

I’m not sure what exactly it is that you are suggesting here. Care to elaborate?

This is off-topic as it is concerned with the 75/25 split between the ETH/PICKLE farm and the other farms. This thread is specifically about the distribution within the 25%. I would recommend starting another thread if you want to talk about the 75/25 split.

This is important, I hope it passes - obviously it can be revisited and one day there may be an opportunity to increase it on a particular pjar if it brings more TVL.