[RFC] Yield on VST Enabled by Vesta Saving Module

For edits, please see the end of this post for the list of edits.

Motivation

  • Since introducing popular collaterals types, Vesta has consistently run into liquidity crunch where VST would depeg downward as people sell the VST immediately after producing it from a vault.
  • This was often alleviated by a market downturn (May 2022 and early November 2022) as falling prices force people to repay loans and close their vault positions. However, this is not a very sustainable solution as we cannot control when market downturns happen and a downturn is technically against the interest of the protocol. The protocol succeeds based on people collateralizing more crypto assets.
  • On Nov 25, Vesta introduced Vesta Reference Rate, which shifted the downward peg defence mechanism from redemption to a dynamic interest rate that reacts to the peg. Vesta now has a source of revenue that could be used to facilitate the growth of the project in a sustainable manner.

Proposed Solution - Vesta Saving Module

Summary

The Saving Module is a place where users can lock up VST for a period of time to earn an interest rate. The locked up VST will be used to backstop liquidation (similar to the current stability pools, but is second to order to the current stability pools), further securing the system against long-tail event. Such an arrangement would effectively make VST the most secure crypto-backed stablecoin in the whole market. The interest rate revenue is funded by the interest that is paid by users on their minted (i.e. borrowed) VST based on the Vesta Reference Rate model.

Overview

  • Enter anytime - users will be able to enter into the Saving Module at anytime and the locked up amount will receive the interest rate emission. Users will not be able to claim the locked amount until the lock period expires.
  • Deposit timelock - the Saving Module will have a max time lock of 90 days, and users will be able to select any number of lock up time between 0 to 90.
    • We are utilizing the veCRV model for the time-lock, providing people with multiple choices on the time-lock.
    • The longer someone locks up for, the more reward he or she will get on a linear scale. For example, if someone locks up for 90 days, he would get all the reward. On the other hand, if someone locks for 30 days, he would get 1/3 as much reward as the person who locks for 90 days.
    • The upper limit time lock is determined by duration of historical peg cycles.
    • Vesta will provide the option of auto re-locking so they don’t have to come back every month to re-lock.
  • Funded by Vesta Reference Rate - the incentive will be fully funded by Vesta Reference Rate. Please see this link for more information on the Vesta Reference Rate. A portion of VRR will be dedicated toward the Saving Module.
  • Emission tuning based on peg - since the Saving Module is likely to attract capital away from LP and potentially even attract external capital to buy VST, we plan to tune the emission according to the peg of the stablecoin. When the peg approaches $1, VRR decreases, the Saving Module reward rate would decrease, and vice versa.
  • General stability pool - the locked up capital in the Saving Module will be used to backstop liquidation if any stability pool is empty. Please note that the asset-specific stability pools will still have priority treatment when it comes to liquidation. Only when a asset-specific stability pool runs out of capital will the capital in the Saving Module be used.
    • Liquidation by the Saving Module will work the same way as stability pool with liquidation bonus. When a position is underwater and the asset-specific stability pool cannot cover it, the general stability pool will be used to liquidate and the liquidated collateral would be eligible to be claimed as the collateral is distributed to every in the pool on a pro-rata basis.

Implementation Plan

Initially, we plan to implement both yield cap and capacity cap. Effectively, the emission will be calculated by min{highest possible emission enabled by VRR × VRR’s allocation toward the Saving Module, amount of emission needed to sustain 10% for all capital within the Saving Module}.

Yield Cap

We plan to start the Saving Module with a 10% yield cap. The yield cap is in place to minimize APR dilution especially for those who enter into the Saving Module early.

Yield cap’s selection criteria is mainly influenced by the relative cost of minting. The Saving Module’s yield should not exceed cost of mint by too much or it would encourage arbitrage.

Capacity Cap

We plan to start the capacity of the Saving Module at 1m at adjust accordingly. The capacity cap is in place so the team can monitor where is the capital flowing from.

Capacity cap is mostly in place to let the team monitor initial capital flow. Over the long run the amount in the Saving Module should correlate with the amount of VST needed to bring VST back to peg.

VRR Allocation Cap

VRR’s allocation toward the Saving Module will also be adjusted. VRR will not be fully allocated toward the Saving Module as the rest will still go to the emergency reserve to be used in case of severe situations, such as severe VST depeg. We plan to initiate that parameter at 70%.

Monitoring

  • Overtime, we’ll be monitoring for two data points:
    • The peg: how much of the Saving Module were from people swapping in from other assets? how much of the Saving Module’s capital are from internal deposit places such as the liquidity pool and the stability pools?
      • If majority of capital comes from internal depositing pools + new VST minted, then the structure is not ideal and not achieving the goal of fixing the peg.
    • The equilibrium rate that the users accepts for the lock up time required.
      • 10% is an interest rate that we will target for the start, but overtime people may find it high or low.

Alternative Approaches

Incentivizing LP directly

Incentivizing LP directly would attract capital to enter LP directly but it is a rather “non-targeted effort” as people could enter LP with VST, which would not help with the peg. Current incentivization schemes such as gauges are also facilitators of farm and dump as many gauge programs (such as veBAL gauge) do not support extended vesting on reward. We plan on tackling the part of expanding LP with our new staking program once that becomes available. By then, the Saving Module and LP incentives could be live at the same time as these two items are not mutually exclusive.

Setting Up a USDC or FRAX Single Pool to Buy VST

Vesta could take the approach of asking people to deposit USDC or FRAX directly, which will then be swapped to VST by Vesta to support the peg. This will directly push the price of VST back up but there’s a critical issue here: we cannot guarantee that at the time of redemption (end of the time lock) we can swap the bought VST back to USDC at a level that we bought the VST. This is extremely likely if VST supply is constantly growing and that the lock time does not guarantee a time of decreasing demand.

Timeline

We plan to iteratively scale this product, with the initial version being a simple one for the purpose of monitoring behaviors.

Further iteration of the Saving Module will see the boost mechanism being implemented along with a supposedly new staking program and the capital being mobilized by Vesta. For the initial version, Vesta will not utilize this capital for any means. Any change to this policy will go through governance.

We look to enter the implementation stage immediately after this proposal passes. We expect this project to take less than a month to implement.

Edits

2022/12/30

  • Included auto-lock feature
  • updated to included the use of Safety Vault’s capital toward backstopping liquidation - effectively having safety vault as a general stability pool that liquidates after the single stability pool runs out. Such an arrangement would effectively make VST the most secure crypto-backed stablecoin in the whole market.
  • Added another parameter: VRR Allocation Cap
  • updated the target yield to a lower level of 7%.

2023/01/17

  • Changes locking system to look more like vote-escrowed CRV, where users will have the option to lock in CRV to earn more. After consulting with Risk DAO, we realized it’s better to utilize the veCRV model so the protocol may gauge for how long people would like to lock up for. In the new design, users will have the ability to lock in VST for either 30, 60, or 90 days. We did not go with a continuous lock time as discrete time is better in terms of implementation.
  • Changes yield cap to 10% to be more in line with current market condition.

2023/01/18

  • Changing the name of the feature to Vesta Saving Module to prevent confusion with the word vault.
  • Changing locking system to be continuous. People now have the option to select any number of days they want to lock for.
1 Like

I like this idea a lot, and I think it will greatly help to strengthen the peg as VST is locked into the vault and VRR interest flows into it.

Question: Is the time-lock for the whole vault or is it per-user-deposit? For example, is it going to be a vault that lasts 90 days and users can deposit at any time in that 90-day period unless it is full, or is it going to be users locking up their VST for 90 days regardless of when they deposit?

This part also concerns me a little:

Vesta will have the capabilities to use the VST in the vault to fund the protocol’s development and expansion. For instance, Vesta will use the funds in the Safety Vault to increase liquidity in the stability pool thus ensuring more efficient liquidations of collaterals.

This seems a little risky to me, especially when we’re dealing with locked deposits. What would users do if the DAO decides to allocate SV funds in a way they don’t like? Is there a way to pre-announce the ways in which SV funds could be used before a vault time period, and to limit the DAO to only using the funds in that manner during the lockup?

Even in the stability pool example given above, I’m concerned that there’s a significant risk that users won’t be getting back 100% of their VST deposit at the end of the locked term. Stability pool funds are basically used as a liquidation backstop during downturns which is great for people looking to score cheap collateral, but it’s also quite possible that the downturn continues and stability poolers end up with collateral worth significantly less than their original VST deposit. Is there a way to guarantee that SV depositors will receive 100% of their deposit back at the end of the lockup period if funds are utilized in this manner?

I know that there’s supposed to be no reward without a little risk, but there’s already a substantial amount of risk in just using DeFi with the kind of collateral Vesta is built on. I’m hesitant to create an attractive yield product with somewhat obscure risks that new users may not fully comprehend, so I’m mostly looking for ways in which to improve user safety and awareness.

2 Likes

Thanks for reading into the proposal!

The timelock will be applied on a per-user basis. A vault that lasts 60 days (we just changed the time requirement after digging into the historical chart a bit more) as a whole will not do much for fixing the monetary supply as after each period we’ll potentially see a sell order of VST.

The part where Vesta will take VST from the vault toward liquidating in the stability pool will make use of of an internal dex trader that we’ve built. It would first check the profitability of the trade then execute only if the liquidation trade is expected to be profitable. As a result the risk is close to 0. We’ll also outline this feature in more depth.

1 Like

Sounds promising. I agree with that the safet vault makes more sense than the other 2 options.

I’m not sure how I feel about the 90 day lock, because it might feel too long for some people that might otherwise decide to hold VST. I understand the rationale behind it and there may be enough people who find it worthwhile. I think this proposal is a great first step to be possibly tweaked later.

One idea is to have a tiered system, where more of the VRR goes to longer lockers. This could make it tempting for people who are not sure if they are able to lock for 90 to still hold VST, and they might hold for 90 days anyway by redepositing in lower lock time vault.

Another idea which I prefer as it makes it more tempting to lock in my opinion, is to offer the possibility of removing your VST earlier at the expense of the interest. That interest could then be fully distributed to continued lockers, or spread between them and the protocol.

How do you feel about 60 days? I guess they changed it after making the OP, see Mikey’s reply to me.

Another idea which I prefer as it makes it more tempting to lock in my opinion, is to offer the possibility of removing your VST earlier at the expense of the interest. That interest could then be fully distributed to continued lockers, or spread between them and the protocol.

Isn’t this just removing the time lock entirely? That sounds like it would be the same as having no lock at all. I think the lock-in is necessary because it actually removes that VST from circulation for a time, which is the main goal of the vault.

Hey guys, I’ve just updated the post. Deleting the cap on interest to be more of a soft cap, and changed the time lock from 90 days to 40 days. Let me know what you guys think. @Maswasnos @Mofe

1 Like

40 days seems like a reasonable amount of time. It should insulate VST from short term fluctuations while still allowing users to reevaluate their positions on a regular basis.

Do you have any ideas on how to encourage liquidity from external sources rather than from pre-existing VST pools? Maybe the SV could have some small token incentives for depositing in USDC or FRAX with an internal contract DEX trade to convert it to VST for the vault deposit.

Transaction-based incentives would be very game-able. Using what you mentioned as an example, I could start with VST and just swap to USDC directly before joining the pool, capturing the incentives.

1 Like

That’s very true. Maybe the initial rollout doesn’t need additional incentives and we can just see how it performs in the first two months or so.

1 Like

Great feature in my opinion. Just a question, does it make sense that liquidity is on curve? a 80/20 approach like balancer would not make more sense?

Regards!

Hey welcome to our community. I believe a 80/20 pool is more of an option for the governance token. For a stablecoin it’s more optimal to use these tight-spread stablecoin pools.

1 Like

This is a great proposal to strenghten the peg and give utility to VST.

I would ideally remove the timelock period. I believe it does not provide true value as the yield reward should be enough to imcentivize people to stake in the safety vault , and might even deter some people to use the vault by fear of locking their capital.

1 Like

I think the 40-day lock is necessary mainly because Vesta wants to use that liquidity for other purposes, such as reinforcing stability pools. If you could withdraw at any time, it’s possible that everyone in the pool could choose to withdraw at once while funds were being used elsewhere and not everyone would be able to get their funds out. With a time-lock, it’s possible to have funds on-hand for withdrawals when terms mature while still using those funds elsewhere in the protocol during the term deposit.

It’s mostly a safety feature so that funds can be handled more reliably. Vesta could do a non-time-locked vault but I don’t believe it would be safe to use those funds anywhere else.

I personally don’t think 40 days is all that long anyway; many protocols such as Frax offer configurable time locks and users often choose terms much longer than 40 days.

I understand. Indeed if Vesta wants to use the liquidity then a timelock is necessary. To be honest I am also not a fan of Vesta using that liquidity for the same reason that you stated before. If Vesta wants to go forward with that to optimize the utility of those locked VST then it should be crystal clear what they will be used for and what are the risks involved. For me, you should get yield simply in exchange for the risk of holding VST as you allow people to borrow and you take the risk of depeg.

Or potentially if there is added risk due to the use of those funds, then additional yield could be added in exchange for those risks.

1 Like

Hey thanks for your response. The original intention of having the lockup is for monetary reason - a lock up would ensure that capital is staying in the system for longer and decreases the probability of capital flight especially in a depegging scenario. I’ve just updated the post to include how we plan to use the capital as a new general stability pool. So now there’s actually a brand new use case for the locked capital. Liquidation could actually be profitable in this case so it’s technically added benefit to safety vault as well.

1 Like

Hey guys, I’ve just updated the post once again. This update is fairly substantial, and you can see a summary of the update at the bottom of the first post. I’d love to see what you guys think. @Maswasnos @Mofe

1 Like

Thanks for the update!

  1. 7% seems like a reasonable starting point for yield. The other basic parameters like the 40-day lock and $1m cap sound fair as well.

  2. I like the basic idea of using this as a universal stability pool of sorts, but I’m wondering if there are statistics on how often this may have been utilized in the past. Do the stability pools need to be backstopped frequently? Just trying to get an idea for how often the capital in the vault may be used.

Thanks for your speedy feedback Maswasnos! Historically no stability pools have been emptied before so such a general stability pool actually would have never been used. However I expect some capital from existing stability pools to move over so the outlook may change.

1 Like

Hi @mikey and all,

I welcome the efforts undertaken by team to help stabilize the peg, especially now that introduction of VRR has made depeg even more harmful to Vault users.

I do, however, believe that Safety Vault is very unattractive product for almost everyone in DeFi. Reasons:

  • Interest rate of 7% is just too low given that we failed to attract sufficient liquidity even with much higher APR on Liquidity mining program
  • Locking capital even for newly updated term of 40 days is also hard pass for most of us.
  • Funding Safety Vault with VRR means that it will offer higher APR in case of lower peg which is extremely undesirable as it means you have probably lost on the value of locked capital
  • In the event of peg stability around 1.00 VRR revenue will be extremely low and thus Vault will offer extremely low APR which will frustrate locked users
  • Locking period increases probabilities of dumps once period expiries thus bringing additional uncertainty to VST as stablecoin

I believe that right solutions are:

  • incentivise liquidity directly with other stablecoin partners that are willing to offer LP partnerships (Liquity has been doing this already for some time. Also, Sperax is offering shared liquidity with it’s stablecoin)
  • add VST as collateral to lending protocols (some rumours where mentioned in discord)

Hi guys, what is going to be my incentive to deposit VST in the safety vault if the APY is capped at 7% while my minted VRR on VST can cost me much more than that? I would not put any cap APY on the safety vault and would increase its max capacity to 30% of the VST TVL. 1M might no mean much of the difference to aim the equilibrium PEG. The safety vault could be earning 100/30XavgVRRX70%. Therefore 30% on safety vault could really be enough fire power to drain some VST offered from the market, sustaining its Peg, and at the some time contributing to reduce the VRR and make it quite more sustainable and self feeding.