The DAO’s FLX Incentive Campaign: 2023 Goals and Roadmap

General Discussion
1

The DAO’s FLX Incentive Campaign: 2023 Goals and Roadmap

Foreword

First, as always I am thankful for this community. We are an eclectic bunch and include everyone from literal cypherpunk-anarchists and citizens of hyper-inflationary economies to degreed economics researchers and GenZ memelords. Hell, this is likely the first forum post after having been graced by Vitamin B himself.

As different as we are, we all share one important vision - if you are a part of this community it is because you believe this is the current best-chance at creating a sovereign stable money. At the DAO we will continue to do what we believe is in the best interest of this goal. We hope you stay with us for this next part of this journey. I believe we are positioned to grow the protocol and the technology in strides this year!

Let’s have a frank discussion about FLX incentives…

Problem Statement

  • I believe the FLX staking incentive is far too expensive for the protection it provides, and it is time to path towards its elimination
  • I believe the FLX incentives experiment has introduced problems that are contributing to a perpetually negative redemption rate

FLX Staking

The purpose of FLX staking was to create a pseudo-insurance plan for the protocol. During certain bad debt events, stakers would be slashed and this slashed value would be used to backstop the protocol’s balance sheet. As compensation for this risk, stakers historically have received FLX token as a reward. This makes sense.

Or, it did make sense. Given the size of the protocol’s current collateralization, the size of the protocol’s debt outstanding, and the value within the staking contract, I do not believe it makes sense anymore.

The current size of the staking contract is ~$3.2M. As it is a Uni v2 contract, this value is equally split between ETH and FLX. During a slashing event, 1/3 of the contract’s LP is taken. So, there’s ~$0.5M or so of ETH at current values, plus whatever the market would now value the FLX in the LP (probably not highly). optimistically, the protocol could expect at most ~$1M of value from the staking contract.

So, what are we paying for this insurance product? Well, stakers split 87.5 FLX a day or about 320k FLX per year. At the currently market value of $10ish per FLX, that’s $320K USD per year for protection of (maybe) $1M. That is insane, and it looks more insane when you consider that the protocol is running higher than 300% collateralized as a protocol.

Should stakers even want to provide liquidity at these prices?

Proposal

  • Reduce FLX staking rewards FROM 87.5 FLX per day TO 0 FLX per day over the next three months.
    • Reduce from 87.5 FLX per day to 65 FLX per day in February
    • Reduce from 65 FLX per day to 40 FLX per day in March
    • Reduce from 40 FLX per day to 20 FLX per day in April
    • Eliminate FLX Reward in May

FLX Incentives

The purpose of FLX incentives for RAI activities was two-fold:

  • Reward users for helping to grow and/or stabilize the protocol
  • Put governance tokens in the hands of the protocol’s users

Failure is a strong word, so I won’t use it, but the current model has not “succeeded”.

It is a commonly held belief (both by myself and quite a few community members that are much more intelligent) that RAI/DAI, especially those incentives attached to uniswap v3 RAI/DAI are a major contributing factor in perpetuating negative redemption rates. FLX rewards are routinely dumped by mercenary capital which does nothing to strengthen the protocol or give community members a larger voice.

Our current iteration of Curve is perhaps just as non-constructive, though we have plans to address this in the future. Regardless, the pool is relatively small, and we don’t effectively engage in the bribery wars for this to be an effective use of the DAO’s resources.

Frankly, RAI/[USD] liquidity is a “nice-to-have”, but it is non-essential to the protocol’s oracle architecture. The protocol relies on Uni v2 ETH/RAI for its feeds, and this is the only essential pool in which we must encourage liquidity.

There are currently first-pass discussions in the community about transitioning this to a uni v3 RAI/ETH pool, but that is, again, outside the scope of this post.

What is pertinent to this post is that we are, again, spending way too much resources on way too few ineffective and biased liquidity dollars. Therefore, I propose the following.

Proposal

  • Reduce Uni V3 RAI/DAI LP + Minter rewards FROM 40 FLX per day TO 0 FLX per day
    • Reduce from 40 FLX per day to 10 FLX per day in February
    • Reduce from 10 FLX per day to 0 FLX per day in March
  • Reduce Curve Finance rewards FROM 20 FLX per day to 0 FLX per day
    • Reduce rewards to 0 FLX per day in February
  • Increase Uni V2 RAI/ETH LP FROM 20 FLX per day to 70 FLX per day
    • Increase from 20 FLX per day to 70 FLX per day in February
  • Add mint requirement to Uni V2 RAI/ETH LP

Summary

  • Staking not worth it for the protocol (and probably stakers)
  • USD-related RAI incentives are wonky and not worth it for the protocol

695×126 8.93 KB

3 Likes
2

A more calculated approach might be to not increase incentives for RAI/ETH at all and see what happens to market price and liquidity during the period of scale down then revisit increasing incentives afterwards.

Fiddling with too many moving parts (especially with i rate change) seems too chaotic. If we focus on scaling down first, analyse and reflect what happens, it might be easier to take next steps.

3

Regarding FLX/ETH staking rewards, reduction may be fine. However, can it be zero at the terminal? Is there enough (other) motivation for the stakers with zero rewards?

4

Strongly support reducing or removing RAI/DAI LP and Curve LP.

It’s been a kind of cheaty way to make RAI appear more stable than it really is since the day it began. Time to take the training wheels off.

1 Like
5

I’m not sure if the mint requirement is necessary for the RAI/ETH pool incentive. LPs are basically affected by the impermanent loss. The mint requirement may bring the liquidity down to ~1/4 of current level.

1 Like
6

Hi everyone,

I’m Brian from Gamma Strategies, which is an active liquidity manager & market maker on Uniswap v3.

I am new to the Reflexer community, and was inspired by Ameen.ETH’s tweet here (https://twitter.com/ameensol/status/1618639412103110656?s=20&t=iYED0KM6zSq9CcMm1rf2SQ) to look into the liquidity situation for RAI/DAI.

To summarize, I believe the main issues are:

  1. Liquidity isn’t being rebalanced appropriately resulting in a consistently negative redemption rate, and out of range liquidity

  2. As a result, FLX staking incentives are incentivizing suboptimal (wider than needed or out of range) liquidity as it requires users to manually rebalance ranges on Mainnet, which is quite expensive to do.

  3. Incentivizing full-range Uni V2 liquidity will result in a lowered efficacy of FLX expenditures given the lack of capital efficiency. The incentives will not go as far in reducing price impact on trades leading to very unstable pricing and further depegs from redemptionPrice

Proposed Solution:

Incentivize an actively managed position that concentrates liquidity around the redemptionPrice (found here: RedemptionPriceSnap | Address 0x07210B8871073228626AB79c296d9b22238f63cE | Etherscan)

We ran a backtest on this strategy with the following:

Assumptions:

  • Liquidity Provided: $200,000 (50/50 RAI/DAI)
  • Liquidity Range: +/- 1.4% of redemptionPrice
  • Rebalance Triggers: 11.97% (distance from currentPrice to upper/lower which will trigger a rebalance)
  • Days backtested: 299

Results:

  • Impermanent Gain / (Loss) : 5.03% gain over holding
  • Final Value: $201,045
  • Value if held 50/50: $191,419
  • Rebalances: 47 (so approximately 1 rebalance every 6 days)

LP Position vs. Holding
image

Main Benefits:

  1. Concentrating liquidity in a +/- 1.4% range is much more capital efficient than full range liquidity → Around 14,000% more capital efficient than a full range position.
  2. It is orders of magnitude cheaper to incentivize a tight concentrated range vs. a full-range position
  3. Liquidity would be concentrated on the redemptionPrice which is where it should trade, thus lowering the impermanent loss for LPs. No incurring losses by following off peg prices.
  4. The concentration of liquidity aligns with the goal of reducing the amount of FLX emissions because you don’t need as much liquidity as a full-range position

Happy to get the community’s thoughts here! And if the community is interested, we can do a more thorough proposal via another thread. Thank you!

7

Strategy Simulation - Ranges & Rebalances

image
image710×289 40.6 KB

8

Thanks for the simulation. The strategy looks interesting. However, from my point of view, the discussion of the full range (RAI-ETH) incentive is focusing on something like “fully collateralized AMM oracle”.

1 Like
9

Gotcha! Yeah totally understand. However, I think that RAI/DAI given the tight correlation is the cheapest liquidity you can incentivize. Probably the Curve liquidity and FLX staking is a bit excessive. I think the right move could be a majority going to ETH/RAI and a minority going to RAI/DAI. Given the tight range you can get on RAI/DAI, you’ll get the most liquidity per incentives with it. And if you can focus the liquidity on the redemption price, it may solve the issue of the perpetually negative redemption rate.

10

Thanks @0x-kingfish for kicking this off!

RAI/ETH uni v2 mint requirement & rewards

I think I agree with @deif and @kyoronut and maybe we can leave the current RAI/ETH pool as is—both without the mint req and at only 20 FLX/day. If the mint req was added, liquidity would be thinner as kyo mentioned, and that might justify a higher FLX payout. Also if the subsidy is lower, the influence on the rate from not having the mint requirement is also lower, so it isn’t quite as bad IMO. I would still be open to adding a mint requirement in the future if controller bias is a more urgent issue than having liquidity in the uni v2 pool for the RAI/ETH price oracle.

Summary:

  • leave RAI/ETH uni v2 rewards at 20 FLX/day
  • leave the mint requirement off (for now)

FLX staking

On this one I agree with @0x-kingfish and disagree with @kyoronut - I would like to see the FLX staking rewards go to 0 and the FLX/ETH insurance pool disabled for now as well. My original reasons for wanting this were:

  • I hate the UX of ETH/FLX staking, the locked rewards escrow is annoying too
  • communication about how the staking auctions works also leaves much to be desired, most RAI users don’t seem to know for example, that only 30% of the staked ETH/FLX is auctioned off during a slashing event
  • because RAI hasn’t had any bad debt, we don’t actually know if anyone is running the staking auction keepers, so there is a chance that bids are low and the insurance pool doesn’t actually do anything
  • ETH/FLX stakers can’t vote in governance
  • increased liquidity for ETH/FLX is actually inversely correlated with RAI governance security IMO, especially bc staked LPs who are likely to be long-term aligned can’t vote, and yet an attacker can buy out their FLX and use it to vote…

However, in the process of writing this post, I realized that we actually fucked up ETH/FLX staking even worse than I had originally thought when we updated the popDebtDelay to 28 days. The problem is that the thawing period for ETH/FLX stakers is still 21 days, so in the event of a looming failed liquidation cascade, the ETH/FLX stakers could bounce in the first week and leave the system unprotected.

So I think we should completely remove ETH/FLX staking rewards and the entire staking/slashing/insurance pool flow immediately. We can potentially add it back later if we better think through the UX, security, and if the community demands it.

Summary:

  • reduce FLX staking rewards to 0 in February
  • eliminate the stake slashing flow from the RAI system (for now)
2 Likes
11

How does the backstop mechanism look like without the staking?

12

Before being all-in to L1 uni v2 RAI/ETH, how about having one-month Arbitrum bootstrap campaign? (+later on Optimism when the Bedrock edition is live)

13

@kyoronut , without staking the backstop will be just debt auctions (like DAI).

1 Like
14

@ameen @kyoronut

image
image701×285 7.59 KB

Let’s go with this then.

15

What to do with 40 (= 70 - 30) tokens?