What is stopping us from staking the ETH?

In the spirit of What is stopping us from investing the PSM USDC?, I wanted to ask: What is stopping us from staking the ETH that we hold?

Is there significant risk of losing ETH by staking? Would it be too hard to un-stake to return ETH to vault holders if necessary?

Perhaps staking the ETH in existing vaults would be too complicated; maybe a new type (ETH-S?) could be created. Maybe it could have zero or even negative stability fees.

If somehow possible, it seems like the staking rewards from the amount of ETH we hold would be quite substantial. Going wild with this calculator and the amount of ETH that daistats.com currently shows, the upper limit on potential annual revenue is in the hundreds of millions.

MakerDAO could keep it, or pass some or all of it along to vault owners. If the latter, opening a Maker vault could become the easiest way to stake one’s ETH; while also getting to spend it (as DAI), too.

But, I know almost nothing about the technical aspects. So: what is stopping us from staking the ETH that we hold? Thank you!

1 Like

I’ll let others give a more nuanced answer, but if Dai is being generated from ETH, then there is already a potential claim on that Dai, and you do not want two competing claims. Imagine if the ETH used to generate Dai was also staked on Compound, then the price dropped. Both Vaults should be liquidated, but they can’t both be made whole.
It’s different using USDC in the PSM or LP tokens, since that is value that can be recaptured.

2 Likes

Well, for one thing, we don’t hold it. It’s not a deposit. We aren’t custodians. So… not our ETH to stake

7 Likes

It is ETH that is used as collateral, which does not belong to us, therefore we cannot use it.

3 Likes

I think the fundamental misunderstanding here is that the DAO “controls” any individual’s Eth or any other collateral. That is not true. Each vault is under the total and independent control of the user who holds the private key information permitting them access to said vault. This point is incredibly important to remember because it is the non-custodial nature of the software, among other key components, that differentiates it from traditional providers.

Moreover, as a Vault holder, I wouldn’t want anyone being able to stake my Eth unless I specifically agreed to that transfer of funds in the user interface’s terms and conditions.

@someone

1 Like

This sounds right to me; how about the creation of new types that allow users to opt into this in exchange for lower or negative rates?

As a user, I would love to be able to open e.g. an ETH-S-B vault that had the same terms as an ETH-B vault (as far as LR, etc) except that the rate is 5% lower since the DAO will stake it on my behalf.

Why not onboard sETH as collateral for a vault instead?

1 Like

I think that should be done too for people capable of staking ETH themselves, but it feels to me that building it in natively to an ETH vault type would dramatically lower the friction for less technical users like me to access staking rewards, and at the same time make opening Maker vaults / generating DAI much more attractive (we could advertise negative interest rates).

Can you explain a little better how it would work, I’m trying to imagine it and I don’t see it.

@Saludiego_201 Photoshop and copywriting are definitely not my forté, but here’s a quick mockup (differences from current new vault screen outlined in red):

1 Like

The basic idea is that with ETH staking we have the ability to put our users’ collateral to work (if they want us to).

I’m not saying that we shouldn’t onboard staked ETH, for large holders who have the ability to validate themselves. I’m also not saying that we shouldn’t onboard Rocket, Lido, or other pools’ tokens as collateral types.

But by essentially becoming a staking pool ourselves, we can a) make it easier for users by letting them access staking rewards and DAI in one step rather than multiple, b) not have to deal with pools’ lockup periods, c) keep the fees that the pools charge if we want them as a revenue source, d) make opening vaults much more attractive with effectively negative interest rates.

I see d) as the primary challenge in front of us right now: how do we get people to deposit much more non-USDC collateral with us and generate DAI from it?

I imagine an industrious community member might find that to be a viable commercial business opportunity. Perhaps a smart contract proxy that permits the activity you suggest (I think DeFisaver may have something similar but I’m not sure).

hopefully this will come down the pipe for voting soon.

Lido staked ETH is already being onboarded, and recently got its risk assesment. It’s the quickest path towards putting the ETH to work, and making it up to the users if they want it.

1 Like

That is the quickest path, yes; and we should continue down it. But in the long-run, why not:

  • Keep the fees that the pools are making for ourselves, or save them for our users.

    Imagine a future in which we did implement vaults with staking and the new staked vault types grow to equal the non-staked vault types in size; perhaps some of the non-staked vaults are closed in favor of the staked ones.

    As I said in the original post, at current ETH levels, we’d have hundreds of millions in additional annual revenues being generated if they were staked. Whether we keep some of it or pass all of it back to users, why dismiss the possibilities so quickly?

  • Access an untapped market of less technical users who won’t go to Rocket or Lido and then come to Oasis; many of them don’t even know that Ethereum staking exists.

  • Even users who do know that Ethereum staking exists would, I believe, prefer to deposit their ETH with a trusted entity like Maker than with some pool. And then, conveniently, have the ability to withdraw DAI if the need arises.

  • If most of our exposure to staked ETH is coming from our own custom vault type, we won’t have to deal with liquidity and other risks of small tokens from Rocket, Lido, etc; it would just be ETH that we’d be dealing with.

  • Most importantly: we gain a marketing advantage over our competitors with the lower effective rates, hopefully get a bunch more ETH collateral into vaults, and reduce dependency on USDC.

This is definitely an interesting idea, but there are a couple of issues.

  • The protocol engineering team are already fairly stretched, and modifying the vaults system to stake deposited ETH does not sound like a simple task and will likely take a significant amount of time

  • There is a non-zero risk of slashing, and that could theoretically lead to unbacked Dai if it is severe enough - if we are the ones staking the ETH then this is our loss to cover

  • At the moment - we need collateral to be liquid in the event of a market crash and liquidations. At present, you cannot withdraw staked ETH or earnings from the beacon chain so how would we cover our losses in the even of a market crash?

I agree. I usually prefer to stay conservative around introducing big technical changes to the code or protocol, since the fact that Maker has been working for years and remained unhacked despite large incentives to do so is what gives people in my circles confidence in it. And I prefer Maker to be perceived as a stable foundation upon which other things can be built, much like Ethereum.

However in this case I feel that it is warranted for MKR to take a rare, big, evolutionary step forward alongside its substrate ETH.

Good point. Is it known yet how large this risk is? Perhaps we can model it and cover it by retaining a portion of the staking rewards.

Yes, this whole thing will only become possible after “the docking”, whenever withdrawals become possible.

1 Like