[Proposal] Incentivize a tfUSDC/USDC liquidity pool
Summary
Divert 18% of tfBUSD and 18% of tfTUSD Farm rewards to a low fee tfUSDC/USDC liquidity pool. This would redirect approximately 4,368,000 TRU a year. Balancer V2 is the prefered option for this pool.
Motivation
The current liquid exit system is complex, unreliable and encourages bankruns. I propose a liquidity pool as a better alternative. See Thoughts on the current lender experience.
Projected impacts
The liquidity pool should grow into the high millions, providing a deep secondary market and ensuring people can always sell their tfUSDC.
Based on the current pool, I project a roughly 0.15% reduction in APY from USDC dilution. Roughly 0.3% of APY is being redirected to the liquidity pool, roughly half of which is USDC and half of which is tfUSDC.
Increased usability of the liquidity pool should increase fees from trading, which counteracts the APY reduction.
For further consideration
Hal suggested using tfBUSD/tfTUSD incentives instead. This would gently encourage movement to the tfUSDC pool and require roughly 18% of those incentives to give similar emissions. I am generally supportive of encouraging lenders to migrate to one pool as it also helps improve liquidity and allows us to potentially split pools in more interesting ways.
If someone has arguments for an alternative to Balancer for the pool, please give them.
Yes, we should incentivize a tfUSDC/USDC Liquidity Pool
No, we should not incentivize a tfUSDC/USDC Liquidity Pool
0voters
The pool should be incentivized by diverting 10% of tfUSDC Farm Incentives.
The pool should be incentivized by diverting 18% of tfTUSD and tfBUSD Farm Incentives.
That was @tylerw suggestion. Sushiswap and Uniswap tend to be better for pairs that with higher volatility. tfUSDC/USDC is a fairly predictable trading pair, so he suggested a low-fee Balancer pool structured similar to the wstETH/WETH Balancer Pool that would gradually increment up. He also noted that Uni V3 is fairly difficult to set up liquidity mining on.
In the summary, I switched âdiverting 10% of tfUSDCâ to âdiverting 18% tfBUSDC and 18% tfTUSDâ due to polling results.
Softened wording around Balancer. My reasoning is that the domain experts implementing this may find good reasons to switch during development, so I want to give them some flexibility on implementation. The on-chain vote will cover specifics.
Increased tfUSDC/USDC liquidity allows for better entry/exit abilities. Swapping USDC for tfUSDC
and visa-versa can be cheaper than depositing/withdrawing directly. Itâs useful to be able to swap tfUSDC for USDC during liquidity crunches like the one happening now.
I like Balancer and its ability for an address to change swapping fees. We could grant the timelock this ability to allow the DAO to adjust the swap fees.
Iâm in favor of diverting rewards from the tfUSDC pool. While itâs more efficient to incentivize users to use a single stablecoin and pool, doing so opens up regulatory and centralization risks.
gm gm gm. Yes, the innovation of the withdrawal fees is an important mechanism that I actually really love.
I beleive that a tight spread like the dai/usdc 1bp pool would be perfect, with a sliding upwards price. so either univ3 or curve v2 are my faves.
First of all, thank you for putting in the effort and making the proposal. It sounds to me like a great idea. At the same time, as I was going through the proposal, a few questions occurred to me.
I understand that the proposal applies only to the USDC pool. What I am a bit concerned about it is that it may come across as unfair to the users who deposited into other pools. What if instead, it could solve the liquidity problems in all of the pools? I am aware that it may be better to take smaller steps to test the waters with one pool first, but I would like to know the OPâs and yes-votersâ rationale behind such a decision. Maybe thereâs something that Iâm missing here.
Using AMM to provide users with a pool exit option implies that some of the poolâs assets wonât be available for withdrawal, as some of the tfUSDC is gonna be used to provide liquidity on AMM. So it definitely wonât allow everyone to withdraw. To me, it may help only to a limited extent. Is it possible in any way for us to calculate how exactly the liquidity situation is expected to improve after the change?
Also, if we want to incentivize users to put USDC along with tfUSDC in an AMM pool, why not incentivize holding tfUSDC directly? USDC/tfUSDC pair liquidity providers are probably aware that they will provide exit liquidity for tfUSDC holders and hence their exposure to tfUSDC is going to be very heavy. They need to be comfortable with tfUSDC exposure anyway. If so, why hold tfUSDC/USDC and not just tfUSDC and directly allow other tfUSDC holders to withdraw their USDC? I sort of feel like incentivizing holding tfUSDC directly may turn out to be a more efficient way of providing liquidity to the pool.
Not sure if we should be seeking short-term minor liquidity reliefs that could make pool participation less profitable and could cause more people to leave, possibly resulting in more liquidity problems in the future. Maybe the go-to solution should rather be revisiting the loan disbursement strategy and establishing a policy that would prevent such liquidity crunches in the future?
I see an interesting opportunity in bootstrapping a larger liquidity pool for tfUSDC on AMM. Exit liquidity is one thing, but enough selling pressure might create a chance to acquire a fair batch of tfUSDC at discount. In theory, such a mechanism should work - but if so, shouldnât it drive organic growth of this AMM pool, without incentives? So far it hasnât turned out to be the case. So if the mechanism doesnât seem to be as efficient as expected, maybe we should reevaluate the idea of growing the AMM pool artificially?
Maybe there is a way we could improve how the TrueFi pool itself handles liquidity? There is a mechanism of liquid exit penalty, which is designed to disincentive leaving the pool at low liquidity. Maybe we could think up a mirroring mechanism that would allow the lender to enter the pool at a discount when the liquidity is low? Just thinking out loud here. This could bring some of the arbitrage opportunities to the pool natively, without the need of incentivizing holding USDC/tfUSDC, instead of just tfUSDC.
Iâm not an opponent of the proposal, but Iâd love to have it really thought through before going live with it.
If something doesnât make sense, please point it out.
Iâm open to all kinds of feedback.
I really did view this as âtesting the watersâ. I have no issue expanding to other pools if this is successful. tfUSDT at a minimum would be worth adding.
Well I donât expect everyone to withdraw. I just want to improve availability without punishing holders in the event of a default. What should happen is that the price of tfUSDC will free float in the liquidity pool. So it will drop to say 1.03 and that will be the equilibrium between buys and sells, like bonds. The only way withdraws would be impossible is if tfUSDC goes to 0.
Buying tfUSDC only improves liquidity through liquid exiting, See my other post on why liquid exiting is a problem.
4, 5, 6. Yes, this is a plan to jumpstart a long term liquidity market. AMMs have an issue where low liquidity is self-perpetuating. People donât trade on it because the AMM is too small, so people donât LP because the fees are too low. I have had thoughts on how the pools could profit from buying USDC at a discount and actually make this very profitable for holders, but want to do one step at a time.
I have considered adjusting liquid exit and the mirror idea is interesting, but
If someone gets early information on a significant default, it wonât matter much. That person is going to pull out and leave everyone else with amplified losses.
The best pricing signals are from the market, not formulas. To get good pricing data, we need a good secondary market. A high volume AMM could actually give us the on-chain data we need to develop more sophisticated pool management.
