Note: See this post for the latest proposal

I’d like to make a proposal to adjust emissions. The ambition of this AIP is to better position Alchemix to take advantage of the Tokemak ecosystem that is incrementally being built out in DeFi.

As a precursor to this proposal, it’s a good idea to review AIP-22, where @scoopy made the initial proposal for an ALCX Token Reactor, as well as outlined a possible future roadmap. This proposal aims to follow up on some of those ideas for the future. I have also detailed some of the broader context, background and issues towards the bottom of this proposal.

tl;dr

This AIP consists of two proposals, each of which are aimed at accruing additional TOKE. The additional TOKE accrued would be used to boost the Alchemix stake in the ALCX Token Reactor, thereby unlocking additional TOKE emissions and boosting the pre-existing flywheel effect.

The two fully formed proposals are:

Commence TOKE bonding with up to 7.5% of total ALCX emissions.
Deploy 72k ALCX to the ALCX Token Reactor ($21.6M @ $300/ALCX).

Proposals A and B can be approved independently of each other.

A quick note about the utility of TOKE
TOKE is used to direct liquidity in the Tokemak ecosystem. Alchemix will use TOKE to ensure sufficient liquidity for its ALCX Token Reactor. TOKE would also be needed by Alchemix in the event that the project decided to add reactors for alAssets. The theoretical addition of any alAsset reactors would result in competition for the Alchemix TOKE supply (LD votes), and could result in Alchemix needing to resort to bribes in order to ensure sufficient liquidity across all of its tokens in future. In the event that Alchemix has an excess supply of TOKE (i.e. can direct more liquidity than needed), it could opt to receive bribes for its excess votes (via VoteMak).

Proposal A)

Accept TOKE as an additional token type for ALCX bonds.

As an initial suggestion, I would propose to start with 2.5% of total emissions, increasing to 7.5% over a period of 4 weeks (i.e. 2.5% per fortnight). This would be drawn from existing ALCX/ETH SLP emissions (25% -> 17.5% over the same timeframe). I’m sure there will be feedback on the specific allocation and timeframe, so hopefully that can be tuned through discussion on this proposal.

This proposal will enable Alchemix to establish a new source for TOKE, in addition to the existing tALCX LP and TOKE LD rewards. At current emissions rate (_16k ALCX/week), a fully subscribed bond should yield somewhere around 1,600, 3,200 and 4,800 TOKE/week at 2.5%, 5% and 7.5% respectively (after factoring in 10% for discounts and Olympus Pro cut). For reference, the calculation used is: 16000 * 0.8 * [.025, .05, .075] * 0.9 * (ALCX/TOKE ratio).

Once the allocation reaches 7.5% of total emissions, the TOKE received from bonding would equal or exceed the current TOKE harvested from the ALCX Token Reactor. The TOKE sourced from these bonds would then be directed back towards the ALCX Token Reactor in order to boost the existing rewards on both sides of the reactor.

For context to the incremental march towards 7.5% of total emissions, I would ideally like to propose a greater allocation to TOKE/ALCX bonding. As a reference point, CVX/ALCX bonds currently receive an allocation of 20% of total emissions. However CVX is a proven strategy and TOKE remains speculative. In addition, the ALCX/ETH SLP emissions were only recently reduced by 37.5% as part of AIP-31 (from 40% of total emissions to 25%). This proposal would reduce it by a further 30% over 4 weeks. Intuitively it makes sense to make more gradual adjustments over time in order to avoid shocking the LPs too much. Violent adjustments to the pool may lead to dumping of ALCX, which would be counter-intuitive for the Alchemix effort to harvest more TOKE since the ALCX price is a key governor in the TOKE rewards calculation. Additionally, it is preferable if exits from the ALCX/ETH pool occur via the ALCX/ETH Bonds, (currently 10% of total emissions), so it makes sense to avoid unnecessarily overwhelming/saturating that bond (if at all possible).

Benefits:

Alchemix can source additional TOKE rewards, to result in a 2x of harvested rewards each week. These additional rewards would be deployed to Tokemak for a boosted flywheel effect.
Alchemix reduces the allocation towards rented liquidity and instead invests in a longer-term liquidity solution.

Risks:

Potential for emissions shock to ALCX/ETH SLP, which could potentially lead to some dumping of ALCX. - Negative price movements on ALCX also impact the rate of TOKE rewards, so this may offset some gains in the short term.
Holding additional TOKE increases exposure to Tokemak, which could backfire in the event that any of Tokemak’s execution risks are realised. However, the ALCX emissions that are proposed for this bond are only being spent on rented liquidity, so any additional risk is arguably muted.

Proposal B)

Stake 20%[1] of treasury (36k ALCX) in the ALCX Token Reactor.

[1] Was originally 40% (72k ALCX), reduced to 20% after initial feedback from participants.

As a precursor to describing this proposal, I fully expect that the 40% proposed allocation will be contentious and subject to debate. I acknowledge that Alchemix should generally be conservative with its management of treasury funds and should not pursue unnecessarily risky strategies. Community consensus will be necessary to determine where exactly this proposal falls on that spectrum, and I’d be surprised if the final version of this proposal remains at 40%.

The Alchemix Treasury has approximately 180,000 ALCX. These tokens are currently unproductive, but could be used to boost TOKE rewards by deploying a portion of them to the ALCX Token Reactor.

At the current ALCX price (~$300 USD), this would result in the ALCX LP deposits increasing by about $21.6M. I estimate that this would deliver a next-cycle boost of approximately 200 TOKE (~+30%). See below for notes on the difficulties of modelling the results.

It should be noted that a side effect of this increase is that ALCX holders would implicitly bear the brunt of this adjustment. My modelling shows that there would be a 10% hit to TOKE emissions to the non-Alchemix LPs in the ALCX Token Reactor, whilst other Token Reactors would see an approximate 3% reduction in their LP TOKE rewards. This may result in some of the ALCX LPs withdrawing from the ALCX Token Reactor, or migrating over to the tALCX Staking Pool for a relatively greater APR. Either outcome would result in Alchemix collecting relatively more of the ALCX Token Reactor TOKE rewards.

As background information, the move by Alchemix to stake treasury assets in Tokemak would elevate it to Stage 5 in the Tokemak DAO Evolution.

Benefits:

Accumulate additional TOKE rewards by deploying an otherwise unproductive asset.

Risks:

Implicit Smart Contract risk by deploying to a third-party project.
ALCX staked in the ALCX Token Reactor is locked for (worst case) one Tokemak cycle (currently ₂₄ hours, to be extended to 1 week).

Other options

For completeness, I'd highlight that there are always other options for accruing TOKE. A good example is AIP-20-A from @Alunara .

The case for considering further strategies for accumulating TOKE directly is that the ALCX Token Reactor is currently quite unbalanced (i.e. overweight on the LP side). This is mostly due to the success of the tALCX Staking Pool, and means that any additional volume directed to the LD side provides relatively more TOKE rewards than the same value directed towards the LP side of the reactor. For reference, Alchemix currently holds ₄₇% on the LP side and ₂₅% on the LD side. Boosting the LD vote also is ideal for Alchemix because it straddles the ALCX Token Reactor and would collect additional rewards on both sides.

If there is consideration of whether a replay of AIP-20-A could be an attractive strategy, it should be noted that the Tokemak reactor cycle duration is currently 1 day, but is slated to increase to 1 week in the future. This represents the maximum possible lockup time for staked ALCX, which may be relevant in future in the event that leveraged positions in Abracadabra need to be reinforced.

Background

The ALCX Token Reactor has been live for ALCX staking for the past 3 months. Alchemix has been participating in this reactor initially as a Liquidity Director (LD), by first staking TOKE and then later voting for the ALCX Token Reactor when LD voting commenced in late November. Alchemix is also indirectly participating in the ALCX Reactor via tALCX staking, whereby it accepts tALCX deposits into a staking pool in return for ALCX emissions.

This staking arrangement has been successful so far and enables Alchemix to collect TOKE rewards from both LP and LD positions in the ALCX Token Reactor. To date Alchemix has harvested approximately 40k TOKE.

Risks

One important factor to consider is that Tokemak has not fully launched yet, so there remains a number of risks to the protocol. This includes execution risk, Smart Contract risk, and the fact that the protocol has not yet been proven in use - so it’s unknown how effective it will be at addressing the broader liquidity problem. This should be a factor when considering this AIP.

Relating to the Tokemak roadmap, the project has outlined the remaining steps in a recent medium post.

Context

The term “TOKE Wars” has been thrown around a lot lately, and VoteMak has recently been acquired by [REDACTED]. It’s anticipated by many that protocols will need to resort to bribing for TOKE LD votes, so it arguably makes sense to invest efforts in acquiring additional TOKE reserves as soon as possible.

In terms of TOKE holders, Alchemix was until very recently 4th on the list of tTOKE holders, but has recently slipped a couple of places. There are notable entrants into the list of holders, including Olympus. There are also some major holders of unstaked TOKE.

Tokemak Pair Reactors

Broadly speaking, the addition of new Pair Reactors has a negative impact on the TOKE rewards allocated to the ALCX Token Reactor via implicit dilution in TOKE LD votes.

Tokemak Pair Reactors have recently come online, and FRAX and Spell are two notable entrants there. The protocols behind these tokens are behemoths and they have the capacity to suddenly allocate significant resources to their Tokemak positions, leading to a sudden change in the competitive landscape. Frax are already notable - they have both the FXS Token Reactor and the FRAX Pair Reactor. The FXS Token Reactor has the largest TVL by a significant margin and is therefore attracting the greatest number of TOKE rewards by a significant margin. See also a recent Snapshot vote proposing to add a FRAX Pair Reactor, as well as this post from Tokemak hinting at further FRAX integrations and bonding of TOKE.

ALCX Token Reactor Performance

The ALCX Token Reactor has been active in some form or another for approximately 3 months now. ALCX Reactor TVL is currently at about $80M, but swings wildly due to recent volatility in the ALCX price. TOKE LD voting for the reactor currently sits at almost $20M, and has been a bit more stable (also because price moments in TOKE are shared across all LDs).

Since launch, Alchemix has harvested approximately once per week, collecting about 3,400 TOKE per week on average. The reactor is however constrained in terms of TOKE LD voting. The tALCX staking program has been wildly successful and has resulted in _130k tALCX staked, accounting for almost 50% of the TVL in the LP side of the reactor. Dollar for dollar, additional investments in the TOKE (LD) side of the reactor currently give the best return for both Alchemix and all other tALCX holders.

Competitively, the addition of each new Pair Reactor is likely to negatively impact the Alchemix LP position in the ALCX Token Reactor, because it drains away LD votes and therefore reduces TOKE emissions to the LP side of the reactor.

In summary, Alchemix has performed well so far, but it’s anticipated that other projects will start to direct greater attention towards Tokemak and that this will therefore increase the competition for TOKE. The addition of new Pair Reactors will also negatively impact the Alchemix LP yields. It is therefore sensible to consider reinforcing the Alchemix LD position by accumulating additional TOKE.

Notes on Tokemak Reward Distributions

Tokemak have described their allocation of TOKE emissions in their article The Balancing Act. Effectively, the system is inherently reflexive, and any interaction with the Token Reactors has a ripple effect that impacts LPs and LDs across the protocol. Participants in the protocol are incentivised to pursue the greatest returns, which should implicitly deliver an optimised allocation of liquidity.

Variables that can impact the allocation of TOKE emissions include:

Number of Token Reactors and Pair Reactors.
Number of staked LP tokens in each Token Reactor.
Unit price of each staked token.
Number of LD votes attracted by each Token Reactor.

The highly reflexive nature of the protocol makes it increasingly difficult to model the behaviour beyond the next cycle (currently 24 hours in duration).

**SCOOPY MOD ABUSE EDIT**

final parameters for the AIP, as posted below by dix are as follows:

OK, so after some additional discussion in the Discord, here is an adjusted proposal:

Alchemix to stake 22% of current treasury in the ALCX Token Reactor (69.420k ALCX).
New TOKE Bond @ 2.5% of weekly emissions.
Boost tALCX emissions from 10% to 15%.
wETH/DAI bonds to remain as-is @ 5% each.

Emissions are to be drawn from:
5 points ALCX/ETH SLP emissions (25% -> 20%)
2 points CVX bonds (20% -> 17.5%)

All other emissions to remain unchanged.

Notes:
ALCX LP on Tokemak
When I prepared this proposal, the ALCX in the treasury was 180k ALCX. It's now over 300k ALCX due to a delay in harvesting ALCX for supply of the bond contracts, so 20% => 60k ALCX. It's so close to 69.42k that I've bumped it up to that, so we can sail on the power of memes. We've had some back and forth discussions about whether it's responsible to allocate this much to Tokemak. I'm in favour of this allocation and I would argue that it's easy to reduce this amount via an AIP if the situation changes.

tALCX Emissions
tALCX emission adjustments have not been discussed yet, but a deep dive with @ov3rkoalafied and @barree in the Discord has indicated that it probably makes sense anyway. The justification for increasing tALCX emissions is that we can see there is a decent chance of capturing some low hanging fruit (tALCX holders who do not stake in Alchemix). These holders attract over 50% of the TOKE emissions to the LP in the ALCX reactor, and it delivers the most benefit to Alchemix to attract them across. Additionally, boosting the tALCX emissions by 50% (to 15% total) rewards ALCX/tALCX holders, and does not incur the same dumping effect that cyclical bonding would.

TOKE Bond
Also a comment on the TOKE bond @ 2.5% - this might seem low, but it delivers some immediate benefit in TOKE collections. I see it as a positive that the DAO can choose to readjust the emissions allocation afterwards in response to feedback/observations about the performance.