Should GnosisDAO manage native bridge funds in a way that improves the Gnosis Chain ecosystem while managing risk properly?

Governance
1

Should GnosisDAO manage native bridge funds in a way that improves the Gnosis Chain ecosystem while managing risk properly?

This post is intended to generate a productive discussion about how to better utilize funds locked in native bridges. If the community signals interest and some degree of alignment in a way forward, this post should result in GIP on phase 2.

There are currently almost 31.4 MM DAI locked in xDai Bridge and $100+MM worth of crypto assets in Omni Bridge which generates 0% revenues, representing an opportunity cost for the Gnosis Chain ecosystem. If a reasonable portion of these assets was deposited in risk-managed strategies, the ecosystem could either get an additional cash-flow which enables further development, or improve the Gnosis Chain liquidity in a sustainable way. An updated number can be found in the xDai Bridge holdings and Omni Bridge holdings. With more than $130MM in crypto assets, even a small amount of capital utilization on low risk strategies would mean significant funding capacity for future projects. As context, one of the last GIPs that passed, GIP-70, represents an expense of $1MM a year.

As a starting point, the community may want to ponder on the following approaches, which could be implemented independently of each other:

  • Deposit a portion of the locked DAI in MakerDAO’s DSR, e.g. one-quarter of locked assets. This strategy yields a 1% APR since the parameters change proposal was voted for on Nov 22, while being an almost risk-free opportunity.
  • Allocate a small portion of the (e.g. 5%) of locked funds in low risk, battle-tested DeFi strategies in mainnet.
  • All funds locked in the bridge are matched by a representation of them in the Gnosis Chain. Another possibility is to mint an extra small portion of these representations, i.e. mint an extra 5%. These funds could be used strategically to increase liquidity in key DeFi pools/protocols (AMMs, bridges, lending markets, etc…).

The following chart is an example of what these numbers could look like:

Bridge TVL Funds to Use (%) APR 1mo Return 1yr Return
xDAI Bridge $31,380,000 25% 1% $6,538 $78,450
Omni Bridge $100,000,000 5% 1.50% $6,250 $75,000
Total $12,788 $153,450

Implementation:

At the very least, xDai Bridge and Omni Bridge would have the capacity to deploy funds in mainnet DeFi protocols (like MakerDAO’s DSR) in a non-custodial way. This solution should also include an efficient way to deploy and withdraw funds when needed. Further definition of requirements should include access controls and the role of the bridge multisig.

Because these funds are key to the Gnosis Chain (and ultimately Gnosis), the solution design and implementation should be properly planned for by a trusted team like Gnosis Guild or Gnosis DevOps and Karpatkey.

Communication:

The voting and implementation of such a GIP need to be communicated in advance to all users and stakeholders of Gnosis Chain, as they might be impacted by the availability of funds. This means sending several communications (prior, and after implementation) through the normal DAO channels: this forum, discord, twitter and community calls. Additional messages could be added to the bridge UI for future users.

I would love to hear what the :owl: community thinks!

9 Likes
2

Many thanks for your post. Like it. Being new to the Gnosis world I wonder how can it be that money doesn’t generate yield? Is your idea almost “risk-free” return? Is there a table of the 100m crypto assets to understand why they are not “used”. Maybe long-term investment into solid projects?

In general, I support your idea.

3 Likes
3

In general, I like the idea to use part of the liquidity to generate income, but as already stated in your post it might come with risks. Another point to consider are the costs, cause management of these strategies don’t come for free (either ppl from the DAO or external parties need to be payed).

Although more liquidity is locked in omnibridge I would prefer to start just with a portion of DAI, the main parameter we need to know is the max liquidity that is needed for a given timeframe to estimate which part (your guess of 25% seems a good start to me) can be put in a bullet proof protocol (besides Maker Aave seems appropriate to me). As I see it, this wouldn’t need much calculation and monitoring and thereby don’t come with much cost.

Using the omnibride funds, due to the may different assets seems more complex and might need a dedicated strategy to manage the risks.

This is something I wouldn’t even think about…although it is possible even from a financial point of view it bears a too large risk of miscommunication…

4 Likes
4

These assets I am refering to in my post are tokens locked in the bridge contracts in mainnet. While being locked, they aren’t allocated to any DeFi strategy and hence don’t generate any yield.

Indeed, I think we should entertain the idea of risk mitigated allocations instead of high APR-high risk. Users wouldn’t be able to bridge all assets back to mainnet if some of them were lost in DeFi. MakerDAO DSR is a good example of low risk

Currently the difference in APRs between DSR and Aave is 0.38%. While it is true that Aave is battletested, it is also true that it has far more risk than Maker’s DSR. Being the Pot smart contracts a core piece of DAI, if it were e.g. exploited, DAI would be impacted wether it is staked in DSR or not. Aave on the other hand has additional risks.
Aave, @refri mentioned, is probably still be a good option for other assets, like USDC or USDT locked in Omni Bridge. Same goes for Compound.

Yes, and in my OP you can find a link to the address holdings as shown in etherscan.

I agree, and we could consider starting with a small portion of a subset of these assets to reduce complexity, like USDT, USDC, WETH and WBTC. Altogether, they represent a bit over 30% of the bridge funds.
Ideally, the solution would be non-custodial and any require methods to deploy or withdraw would be developed in the bridge smart contracts. E.g. a (maybe permiossioned) call to the xDai Bridge would trigger the bridge allocating funds to the DSR within a certain cap of total funds, while another call would withdraw them. As you mentioned, what the decision process is to call for these methods and what indicators will be feeding into it may have to be externally managed. Karpatkey is in a position to help in this space, since monitoring is already going on for the DAO’s managed DeFi positions.

4 Likes
5

What’s an OP? Could you send a link? Tx!

1 Like
6

OP means: original post, link is there included (named omnibridge holdings: https://etherscan.io/tokenholdings?a=0x88ad09518695c6c3712AC10a214bE5109a655671)

3 Likes
7

My first question would be if GnosisDAO has the choice to go yieldfarm with deposits of users? Does the DAO has possession over these funds? Or should there at least be a choice for the user to let their deposits participate in the yield farming (and maybe should they have a part of the profit?)

2 Likes
8

I am in favor of the idea of GnosisDAO managing native bridge funds in a way that improves the Gnosis Chain ecosystem while managing risk properly. Utilizing the funds locked in native bridges is a great opportunity for the ecosystem to generate additional cash-flow, which could be used for:

  1. Gnosis Ecosystem Fund, which could provide funding for various projects and initiatives that contribute to the growth and development of the Gnosis ecosystem.

  2. Buyback and Burn of GNO, which would help to increase the value of GNO and promote sustainable development of the network.

  3. Validator rewards, which would ensure further decentralization and security of the network by incentivizing validators to provide high-quality services.

I think that such a distribution would ensure further decentralization, sustainable development funding, and security of the network, which are all crucial for the long-term success of Gnosis.

3 Likes
9

Great question, indeed. And imho not easy to answer:

  • is bridging something like a change (Dai for xDAI and vice versa)? In this case the bridge owners should be regarded as legitimate owners of the assets.

or

  • does a bridge only holds custody of the users assets to prove he is in possession of the equivalent amount of token/value he uses on either side of the bridge.

I tend to favour the latter view, and this would indeed require some consent of the user. If I remember it right, there was something about this written in the old xDAIbridge docs, but acutally I can’t find it anymore

2 Likes
10

It may be somewhere in between: bridging being an exchange for assets in the destination blockchain with the expectations that these assets can be exchanged back, cause that’s what ensures their value.
I see the risk profile of the strategies being discussed and the caps of what can be invested as dimensions of the mechanisms that should satisfy those expectations.

In that line of thought, if an exploit to the Maker DSR impacts the price of DAI, DAI assets are exposed to it whether they are held in the bridge contract or deposited in the DSR module. Then actual allocation doesn’t seem to be adding more risk, while granting a yield.

1 Like
11

It’s important to mention that the existing smart contracts infrastructure for both xDai and Omni bridges already supports automatic compounding for underlying assets via arbitrary yield bearing protocols.

These contracts has been already fully audited by ChainSecurity (see the audit report) and then successfully used in production for almost a year.

The described functionality was silently switched off prior to Ethereum Merge - (see the Safe tx), as a precaution measure. However, there was no clear statement made regarding presence of absence of future plans for its re-activation.

Over the course of a year prior to Ethereum Merge, the described functionality has beed enabled for the following bridge assets and lending markets:

  • DAI locked in the xDAI bridge - to be invested into Compound V2
  • USDC locked in the Omni bridge - to be invested into AAVE V2
  • USDT locked in the Omni bridge - to be invested into AAVE V2

This has led to the following revenue streams for the DAO, over the course of a year (for a total of 2,744k USD of revenue):

7 Likes
12

@kfedoseev thanks for sharing!
@Karpatkey are those positions within the weekly report? I could not find them.
I guess it would be important to list those income streams in a separate field within the weekly report.

13

Hello @m3rlin5ky! These are not being considered in the Weekly Report as this feature has been paused since the Ethereum Merge, so there are no positions in place at the time. If the community agrees to use these “idle assets” for yield farming, the inflow will definitely be shown in a separate section. I believe this is a great opportunity for creating some extra cash for funding GIPs and growing the Gnosis ecosystem.
As the strategies in mind would be what’s considered risk-free and assuming the community agrees to this, the key question here is how much should we use, and the answer shall come from leveraging expected revenues and exposure.

4 Likes
14

Enabling the management of (parts of) the bridge funds disincentivizes the development of trustless bridges where this is impossible (which I think should be the top priority for the Gnosis Chain). And herby works against the decentralization of the Gnosis Chain. It can be argued that the chain currently, in some ways, is not very decentralized and operates at the mercy of a handful of signers (that can mint GNO and xDAI).

A bridge with managed funds sounds like a different “product” to me and one that should not be allowed to mint xDAI (the chain’s native token…). It should probably mint its own token, “g-yDAI”, or something, and it might be good to give depositors some of the yields in the process.

1 Like
15

I don’t think progressing this initiative will hurt in any way the ongoing development of a trustless bridge.

Would you say more about this? Also noting that xDAI is the token used for fees (exec layer), but different from the token used in the consensus layer (GNO).

1 Like
16

My thinking is that both GNO and xDAI are vitally crucial for the operation and security of the Gnosis Chain. GNO for validation and xDAI because it is used for paying fees.
If xDAI gets compromised in some way, for instance, if somebody was able to get access to the private keys of four of the bridge validators (or there is a bug in the xDAI Bridge), then they could print xDAI and flood the chain for free (denial of service), making it impossible for legitime users to get a tx in (in addition to stealing all the DAI on the Ethereum side of the bridge). A detrimental event, even though there would be no invalid state changes on the Gnosis Chain.

Therefore I think it’s reasonable to say that Gnosis Chain is only as secure as GNO and xDAI.

For the long haul, I think it would be much safer to have no issuance of GNO or fee token on the execution layer, but I don’t think this will be possible as long as the fee token remains xDAI. But at least we can mitigate the risk as much as possible by only allowing trustless minting of xDAI. And even though we don’t have trustless bridges yet, we must protect xDAI in every way we can, and I think this is taking a (small) step in the wrong direction.

Another point is that even though the financial risk of the investments and the size of it would be very small (and even if it goes bad, Gnosis themselves will probably pay up if the value goes below “1 DAI”:“1 xDAI”), it’s still a risk (and dependency upon Gnosis) that gets added to a blockchain. Something we should probably work to minimize. As de-pegging of DAI and xDAI also, would not be good for the chain.

I am not sure it can be called credibly neutral when the GnosisDAO gets this privileged position of being able to utilize the bridged funds, especially since that is the only way of obtaining the gas fee token.

I think there is a tension, though, between the security of the chain and the convenience of using a stable coin for fees. This convenience might diminish as ERC-4337 can make it possible for end users to pay fees in any token, and the fact that the actual gas fees (priority part for GC) fluctuate anyway.

This might be a little off-topic, but I am beginning to realize that making GNO the fee token is the only way to make the Gnosis Chain not dependent on bridges (even though there trustless, it’s still added complex and operational security risks), and able to fully protect itself with its validator set. This would also, in the process, cut its dependency on Maker and make it more equal to Ethereum mainnet and, in the process, probably make the Gnosis Chain-specific changes to the Ethereum clients much smaller.
I am now a “GNO for fee token maxi” :sweat_smile:

17

This is totally incorrect; stop posting this trash, both here and on Telegram.

1 Like