• AIP
  • [AIP-21] ALCX Bond Program

The Problem:

Alchemix emits approx 16k ALCX weekly towards liquidity providers across our pools. At current prices, this amounts to roughly $5.5m per week, or $20m per month. This is an astronomical figure for temporary liquidity, and presents itself as a possible bottleneck to growing Alchemix. Solving this is critical for protocol sustainability.

The Solution:

In order to address this problem, we are proposing an Alchemix bond program in partnership with Olympus. We will offer 1600 ALCX weekly available for purchase via bonds, which is roughly 10% of the current weekly farming rewards. We will initially target our three flagship liquidity pools: ALCX/ETH SLP, alUSD3CRV LP, and alETH Saddle LP.

With the accumulated LP tokens from the sales of ALCX bonds, we will also participate in farming with the protocol owned LP tokens. In addition to earning ALCX in each of these pools, we have the opportunity to earn extra governance tokens from other projects. ALCX/ETH SLP also earns SUSHI and alUSD3CRV LP also earns CRV and CVX via convex. If we start incentivising bonds for the d4 Saddle LP (alUSD, FEI, FRAX, LUSD) and the mStable alUSD feeder pool, we will also earn TRIBE, FXS, LQTY, and MTA.

We anticipate that some might say that the DAO is stealing from the farmers if it enters the farming pools. The counter to that is that many in the community have expressed that emissions are too high, and if the schedule is unchanged, it would be detrimental to the project in the long term. However, we are wary of changing the emissions schedule because when other projects tried this, it didn’t work very well, hurting the token more than helping it. This program will result in the DAO progressively diluting farmers, which would accelerate the disinflationary schedule of ALCX. This would make ALCX scarcer on the market, reduce the need for large farming budgets, and make Alchemix and its various markets much more sustainable.

The farmed ALCX would just return to the DAO balance sheet and can be used to fund other things or even be returned to the bond program. The other tokens, likewise, would be added to the DAO balance sheet. For CRV and CVX, we would accrue these tokens and use them to boost our pool in Curve and the 3CRV rewards would be added back into the alUSD3CRV pool to boost liquidity even further. For the other governance tokens, they can become a proxy for Alchemix meta-governance, collateral assets, grant money, emergency rainy-day fund, token value accrual, or deployed in DeFi to earn extra yield.

Implementation:

In essence, Olympus is offering it’s bonding program to Alchemix. This would include having the UI built for this program, helping us set everything up, and continuing to maintain the parameters for the program. In exchange for their support and expertise, Olympus would take a 3.3% fee on all ALCX bonds sold. Olympus will use the ALCX they earned from their fees as extra backing for OHM, which would help to reduce the circulating supply even further.

This chart tracks the average discount of OHM bonds relative to the market. They vary greatly but average around 6-7%. If we target 1,600 ALCX bonds weekly, 52.8 ALCX will be paid to Olympus DAO, and with a 6-7% average discount, Alchemix DAO would convert approximately 90% of the purchased ALCX bonds into LP positions, which would be added as permanent liquidity.

The Olympus bond contracts have dynamic prices for the bonds to ensure that there is constant buy-pressure. They achieve this through the discount rate. If the mean discount of ALCX bonds is relatively low (under 7%), we would likely look to increase the amount of ALCX available for bonds, and vice versa, if the mean discount is too high (over 10%), we would look to decrease the amount of available bonds. If this program proves to be successful, approximately 10% of a year's worth of emissions at current prices would purchase the protocol $15-20m in liquidity. This amount would be highly reflexive relative to the ALCX price, and could deviate significantly in either direction. In the optimistic scenario, ALCX price appreciates and two years from now, Alchemix will control over 100m of LP tokens, and earn millions annually from them.

These ALCX bonds will have a one week vesting period. This is to ensure that discount buyers can’t immediately flip their positions because they typically will be buying at a discount. The ALCX bonds will continually vest, so you will be able to do partial claims before the vesting period is over if you want to sell, farm, or LP with them.

Conclusion:

The ALCX Olympus bond program will initially take approximately 10% of the farming emissions to purchase the LP tokens essential to our protocol. This will create sustainable, permanent liquidity for the protocol. It will accelerate the disinflationary schedule of ALCX and provide additional revenue to the DAO in a variety of tokens, which can greatly enhance our ecosystem in myriad ways. While this will ultimately lower the farming rewards, we view this as beneficial for the long term sustainability of Alchemix.

    scoopy This is awesome! Alchemix owning it's liquidity seems like a great way to scale and guarantee a floor level of liquidity. Even better if it reduces sell pressure on ALCX and increases CVX holdings.

    One suggestion on implementation:

    • (1) In addition to accepting LP positions with these bonds, accept DAI and ETH directly.
    • (2) Take half of the ETH or DAI obtained and purchase alETH or alUSD directly from the protocol to boost yield
    • (3) Then finally create the LP positions using the 50/50 alETH/ETH or alUSD/DAI and proceed as specified in this proposal, staking and earning for the treasury.

    This would have additional positive effects in that it would boost yield every week and create new liquidity positions, instead of just buying existing ones off the market as this proposal suggests 🙂

    Love this proposal and the collaboration with Olympus, who obviously have been very successful with their bonding programme. To me owning part of the liquidity will become more important over time as we add more assets to Alchemix and as the protocol grows in TVL. The current inflationary rewards scheme has been very successful in attracting liquidity but at the cost of having mercenary capital, dumping, and price depression of the ALCX token. So I'm all for this proposal and to start selling LP bonds.

    My only question would be: are we being ambitious enough with the ALCX allocation? This bonding programme can clearly be a game changer, and while it's good to start with a relatively low allocation (at 1600 ALCX a week) to test the waters, wouldn't it be better to increase the allocation for bonds / to ramp it up over time? Especially because we are now discussing 3 different LP bonds (ALCX/ETH SLP, alUSD3CRV LP, and alETH Saddle LP), so the total amount of liquidity acquired per week will be relatively low in the beginning.

      0xFelix we wanted to start conervatively to ensure that the discount wouldn't be too high. If it does prove effective, I'd be more than open to increasing the total amount allocated to the bond program

      0xFelix I second this, and I also wouldn't mind increasing the lock up period here. I think Alchemix supporters that intend on sticking around long term and participating in the DAO don't intend on selling anyway and would likely participate regardless, and it would prevent the game of people picking up the bond and dumping a week later for profit if the price hadn't depreciated. If possible maybe it can be a lever for the DAO to experiment with, since the priority should obviously be to sell out the bond each week

      Comparing these two scenarios, only looking at ALCX/ETH for simplicity:

      1. As proposed, 10% of issuance is set aside to offer bonds, where the treasury purchases ALCX/ETH LP tokens with ALCX, paying a premium (which, to the buyer, is a discount).
      2. 5% of ALCX issuance is market sold for ETH, and it is deposited into the SLP with 5% of ALCX issuance with the treasury retaining the SLP tokens.

      Can we enumerate all the differences between the two? Does the bonding discount help Alchemix at all, or just offer an arb that people can use to drain the treasury? Scenario 1 results in Alchemix owning existing liquidity rather than adding new liquidity to pools, which may or may not be an advantage. Also, the 3.3% Olympus fee may be a net positive in that integrations bring value.

      As a Senior Product Manager, I've always thought ALCX and OHM were made for each other as a product market fit. Awesome to see this, I'd vote for it if I wasn't a 9,9 degen.

      Such a big fan of this, seems like a galaxy brain move that provides flexibility, another revenue stream for ALCX holders, and a solution to the excessive emissions, all while not altering the original emissions schedule!

      Absolutely F'ing love this. Giggabrain long term move to boost and control liquidity for the protocol.

      "If this program proves to be successful, approximately 10% of a year's worth of emissions at current prices would purchase the protocol $15-20m in liquidity. This amount would be highly reflexive relative to the ALCX price, and could deviate significantly in either direction. In the optimistic scenario, ALCX price appreciates and two years from now, Alchemix will control over 100m of LP tokens, and earn millions annually from them."

      Care to explain more about this part?
      What is the pessimistic/conservative scenario?
      Which risks were identified?

        I love the overall concept: the protocol paying to acquire permanent liquidity that it owns, rather than temporarily renting the liquidity at a very high rate.

        That said, the discount rate being targeted of 10% over 5 days is waaaaay too high. A 10% discount makes sense as a target for Olympus with their OHM bonds, given single-staking OHM has had yields in the 7,000-20,000% APY range. A 1.25% discount for 5 days would already be greater than the 5-day yield on the ETH/ALCX pool (currently showing a 90% APR). When you then factor in that the ALCX purchased with the bond can be claimed as it vests over the 5 days and single-staked for 69% APR, the deal gets even better.

        I recommend revising the proposal to target a much lower discount, maybe 1.5% or 2.0% at the very highest. I understand that most users will see that as being too small if it's being presented the way that Olympus does with their bonds, because they aren't doing the math to realize that's really a 109.5-146% APR. Maybe if that's explicitly spelled out in the UI, we can make it work. Heck, maybe even start with the insane 10% (or slightly less insane 5%) discount rate to start and get attention, but then quickly bring the target down to a number more in line with the existing rates the DAO is paying for liquidity.

          Thank you for bringing up this proposal. I think it is a wonderful symbiosis between Alchemix and Olympus to foster a partnership of this kind.

          A brief word about bonds and discounts. The bonding mechanism operates as a secondary market in which the discount is a function of bond capacity and bond demand, where bonders try to beat other investment strategies like staking. So in your case, the amount of ALCX used to sell bonds and the amount of liquidity you target decides how much bond capacity there is within a certain timeframe. Given a certain bond capacity, the bonders on the secondary market decide at which discount they purchase your bonds. The discount is not configured by anyone. Olympus does not dictate to you at what discount you should sell. The discount is merely a function of supply and demand, which is also why bonds work without any price oracle. If the discounts trend higher, there is less demand for the available supply aka bond capacity. If the discounts trend lower, there is more demand for the available supply. Bonders will purchase bonds at the discount they think is competitive for a given "payment" asset, which is ALCX in your case. For e.g. OHM the 5 day staking ROI is about 6% right now, which is why the bonds sell around this range at this point. For e.g. ALCX the discount will try to beat whatever staking or other investment strategy there is, considering supply and demand on the secondary market of your bond programs.

          We will be super happy to help everyone figure out how bonds work and we will also be glad to help you configure capacities of your bond programs, because these things tend to be a bit tricky given dynamic market conditions. For these reasons I think it is super smart to work together with people who built the bonding mechanism and operated bond programs successfully in the past.

          jamesk
          From my understanding, instead of emitting ALCX to people who simply stake ALCX in the single sided staking pool, we would be providing bonds which offer discounted ALCX over a vesting period. It would be incentivized to purchase these bonds if the discount rate deems it worthy and as users pay in DAI/ETH + LP tokens, this would help the protocol permanently lock in LP tokens that earn fees + DAI/ETH which could be used in a multitude of ways (purchase alAssets.)
          The pessimistic scenario + risk in this case would be if people who purchase bonds simply do so in order to dump the ALCX tokens back into the market. However, if staking ALCX had positive incentives, then this could lead to a scenario where users purchase bonds to attain ALCX at market discount then stake ALCX. This is like a (3,3) scenario in that users benefit + protocol benefits. This is a more sustainable way of emission and could really grant the protocol a strong moat of permanent liquidity. Correct me If I am wrong in my understanding but the real risk IMO is not doing something like this.

          delitzer

          The discounts aren't set in stone with olympus bonds. There is a dynamic price where there is a function of time between bond sales that determines the price. The longer the time between purchases, the lower the bond price becomes. The discount rate is purely determined by supply and demand. We can change the vesting period or try to fine tune the ALCX available to hit a certain discount %, but such fine tuning will require the program to be up and for us to have some data to go on.

          100% in support of this proposal.

          Question...
          Is this proposal in addition to (the good chance) ALCX being added to Tokemak?
          Or as contingency, should ALCX not receive the required votes?

          To say it's neither would imo, feel a little short sighted considering the implications of having a reactor would mean for the team and community. If nothing else, it would surely simplify the above proposal...

          Regardless of the Tokemak reactor vote at the end of this month, ALCX will get a reactor eventually, of that I'm very confident. The reactor model, imo resolves everything this proposal is concerned with and to a degree that is far simpler to understand:

          'A single ALCX pool with IL protection, that facilitates all ALCX based pools.'

          Jmo, but I think this should be considered a variable within the above proposal.

            Gravity I would say that we can definitely use a TOKE reactor + a bonding programme at the same time since they are very complimentary. Especially at the start the TOKE reactor will only be good for assets traded against ETH/USDC - so this would most certainly support the ALCX/ETH trading pair. But the beauty of this bonding proposal is not so much for trading ETH/ALCX but for the alToken/Token pools. Owning liquidity in those pools is in my opinion way more important since these pools will make or break Alchemix as a product. On top of that, I think that for alToken/Token pools we should own the liquidity and we shouldn't have to rely on a LD in Tokemak to solve this for us. (And I'm saying this as a big fan of Tokemak.)

            Absolutely love this proposal. A real Win-win-win

            • A win for newcomers -> they can buy ALCX at a lower price
            • A win for ALCX holders -> less incentives required for liquidity providers, so more value for your token
            • Another big win for ALCX holders -> additional stream of revenue: both liquidity farming and trading fees (eternally!)
            • A win for ohm and interprotocol cooperation

            There is only one thing I am a bit concerned about. This bond program gives traders arbitrage opportunities. They can buy ALCX at a discount and sell it for profit leading to a downward pressure on the price. You can argument that the same counts for Ohm, but I believe that's not 100% true. The Ohm community has a very strong staking mentality (3,3), also boosted by the enormous APY's. Furthermore, although the ohm protocol sells bonds with a discount, it is till making a lot of money on every sale as it is selling something that is backed by a value of "a few dai at max" (1 Ohm) for multiple hundreds of Dai. (currently)

            Since these mechanisms are not at play at ALCX, the bond program might put some downward pressure on the price of ALCX - also leading to impermanent loss in the LP positions.

            Maybe good staking incentives might mitigate this downward pressure in the future? (e.g. fee sharing, boosted APY on loan payback, dao/NFT functionalities,...)

            Interested in hearing your thoughts about this.