How to deploy the spartan bucket capital

While we wait for the outcome of the vote on the Enzyme finance proposal (here), Marh mentioned in the discord that he would like to deploy the asset management in three phases:

1) deploy a small % of capital over predefined strategies

2) deploy more capital over the same strategies

3) Add, adapt and update strategies

Predefining strategies

Currently, there is 40,627 ETH (~$120,459,055) locked in the Spartan Bucket. Working with that kind of capital is challenging. We can start with defining criteria and possible strategies that would fit them.

Please join the discussion, we can start pretty broad and narrow it down when things start to become more clear.

To get a better overview, it might be good if @marh gives us a technical/practical workflow of the bridge and the steps capital takes from and to Enzyme? That will enable us to get a better understanding of its possibilities and limitations.

Furthermore, we can start to compile a list with possible strategies that would be suitable to deploy capital towards. Just post them in the thread, and I’ll compile them in a spreadsheet.

This is a list of assets that can be used within Enzyme

Spreadsheet to compile the strategies

Thank you @Toljona and sure!
From a high level perspective, the Enzyme-Unslashed Bridge does the following:

  1. It receives ETH from the Spartan Bucket and sends it to the Enzyme fund
  2. Then it tracks the value of the ETH that was sent there
  3. The Spartan Bucket can request the ETH to be transferred back at any time and it receives it after it is transformed back into ETH

The main condition I see is to have liquid strategies, meaning strategies that allow us to get ETH out of the fund and back to the bucket quickly whenever it is needed (especially in the case of a claim or a large withdrawal that would have been requested).
Other than that, the Enzyme protocol allows to connect to any other DeFi protocol on Ethereum and deploy funds there, so even if it is not already a ready to use strategy, the Enzyme team could connect it quickly (they have standard connectors).

How is it being tracked? Or do you mean it remembers how much has gone out of the bucket?

How manual is this process?
Does the bridge request a certain amount from Enzyme that is equal to the amount being queued/requested?

It tracks the current value of the assets inside the fund, not how much was sent :slight_smile:

It is the done through one transaction :slight_smile:
The spartan bucket requests an amount of ETH, the bridge receives the request and executes it with the enzyme fund.

Lido x Curve strategy

One potential strategy could be: staking an amount of ETH into Lido’s liquid staking pool in exchange for stETH and putting that with an equal amount of ETH into the Curve stETH pool.

The yield in the Curve stETH pool is around 19.92% apy

If I’m not mistaken (if someone could confirm this), the ETH staked in the Lido staking pool will also gain addition yield because of Ethereum staking rewards deposited into the liquid staking pool. Adding another 7,17% APY on the stETH deposit.

below a simple calculation without considering fees and slippage in both the Curve and Lido pool

1000 ETH Lido apy Curve apy total apy
50% in stETH (500 ETH) 7,17% 19,92% 27,09%
50% ETH (500 ETH) 19,92% 19,92%
total apy 23,51%

This strategy would be relatively scalable, but has a limit because we are also providing insurance for Lido. The risk team (if that’s a good name) might be able to provide a number.

1 Like

I’ve added a couple strats to the spreadsheet, feel free to add more.

1 Like

I think in general we probably want to be pretty conservative with anything that is eth-ish but not actually eth. e.g I’d say lending protocols are actually eth, but staking protocols are eth-ish (ultimately no guarantee that staked eth can be converted back to eth at 1 to 1 or close to it).

This is particularly true with any staked eth other than Lido. I’d say that we probably shouldn’t even touch any staked eth other than ldo because I’ve been messing around with some of these and frankly many of them just don’t have liquidity to hold their staked eth to peg.

The enzyme list of assets isn’t the easiest to parse but from that list, I’d say ibeth from Alpha, Compound, aave, and curve eth/seth are maybe the kind of things we want to look at.

Basically, lending + curve pools of eth or very-close-to-eth things.

Maybe a small allocation to curve eth/steth, with the knowledge that it could go off peg, which means we end up withdrawing at a loss or waiting it out and the eth is stuck.

I agree on being very conservative with the collateral. The strategies I’ve written down in this spreadsheet, are focused on single side ETH (or ETH equivalent) exposure. Picking the right strategies is more than just choosing the ones with the highest yield, and I feel like we are lacking good ways to rate them. How much risk is acceptable, and do we need to hedge that? If so, how?

I’m a Lido staker myself, and I’ve used the curve pool a couple of times, off peg happens from time to time but is arb’d pretty quick. Not counting black swans, I’d say It’s relatively safe to assume we get a very similar amount of ETH back (also yield should compensate for the rest).

But rather than trying to eliminate all risk, we should come up with ways to hedge certain risks.

I’ve heard Marh talk about other buckets coming soon, and that we could use them to hedge other buckets.

Hi @amplice, if you want an easier list of assets to parse try this link Enzyme Asset Universe - Google Sheets

We’re constantly adding new assets to this universe

another thing to consider is lido is covered (heavily?) by unslashed. That in itself is a pretty good reason to not use stEth - if they have validators getting slashed, steth peg could fail at exactly the time we need to pay them the claims.

I’d say, safety wise, lending protocols are safer. We cover compound as well so maybe rule that out.

The straightforward thing to do would be to pick either Aave Eth Lending or Alpha Eth lending. Aave has a longer track record, Alpha has higher yield. Neither are covered by Unslashed at the moment so black swan risk isn’t there.

For stETH, we need to consider our exposure to Lido slashing risk Vs how much we would deploy in a strategy. Taking Curve’s stETH-ETH Pool for example, the stETH exposure would be half the amount deployed. Currently the spartan bucket’s exposure to Lido’s slashing risk is about half of the ETH deposited and that’s the maximum it could be. The amounts of exposure can allow us to pick which percentage of capital we want to deploy there.

In the future, with newly deployed buckets, the deployed capital from one bucket could be protected by buying insurance from another bucket. This is what would allow to mitigate most of the risks we would like to mitigate.

In terms of riskiness or safety of the strategies, that is a real question and the main reason why it would make more sense to have the deployment happen step by step and with a small amount first.
The stETH pool is becoming even more optimized as we speak.

This is all really exciting - we’re happy to see the Enzyme + Unslashed communities working closer together. As an initial starting point, how about deploying a test phase of say, 10% of the capital (around 4k ETH) to run the stETH strategy and test the waters for a couple of weeks before finalising the strategy and portfolio allocation. In the meantime, whilst testing this out - the discussion can continue on whether we should ramp up the stETH strategy or if we want to diversify with other ones too. We have some new integrations coming soon which could open up new possibilities but I agree that the stETH proposal is probably the most interesting (and scalable) atm.

do you guys have alpha’s iBeth? I recall the yields there are pretty good (7%, before alpha rewards) and its not covered by unslashed and there’s no risk of depeg.

They do.

Alpha’s IbETHv2 is currently yielding 5.25% and that’s including the alpha rewards, v1 is even lower.
I do understand your concerns about the peg, and although the risk is probably pretty low we lack a way to hedge it.

@marh How could we hedge this until we get other buckets that can insure that capital?

How about a some sort of backstop mechanism with USF from the DAO?
Absolutely not ideal, because if something goes wrong the price will probably drop, but maybe It could be a temporary measurement just to cover it in case of emergency until we can insure it ourselves or have other funds that could be used to do so.

I am not sure what you exactly have in mind - if we start with 4k ETH as @elisafly suggested it doesn’t impact our ability to cover the Lido slashing risk.

Not the slashing risk, but the off peg risk. There is still a risk we can’t exchange our ETH for the same amount as we put into the Lido or CRV pool, It’s probably a low one (not sure how small, can’t really quantify it). It will only be an issue for the first week or so, after the that the yield will probably cover it.

Might be good to have a back-up plan for any unforeseen events in that time period?

I agree, 10% seems a good amount to start with.

The numbers in the spreadsheet have change a little over the past few days, but the stETH strategy is still yielding the most by far.

From a yield perspective, deploying the capital to the stETH strategy makes the most sense. Although it might be interesting to deploy a part to other strategies just to test out the process and maybe diversify a bit.

Lending on Aave/Compound or Alpha comes to mind.
I personally think that auto farmers are somewhat risky, but in terms of yield they come in second, after the curve pools.

We could incorporate it into the vote and ask the community what they prefer?

Another thing that comes to mind is the day to day management of the Enzyme fund, who is going to run the strategies, move and monitor the funds? I assume this is something done by the core team for now, by you @Marh ?

@elisafly already pushed a proposal here: Scattershot (fork of Snapshot)

seems like ppl are pretty keen on doing steth/eth despite the potential peg issues and stuff so i guess i am ‘outvoted’