Multi-chain strategy

Over the past year, while the community has governed and grown the Compound protocol, Compound Labs has been been focused on two flagship priorities:

  1. Launching and scaling Compound Treasury, an institutional on-ramp to Compound markets. Treasury is a centrally managed product built on top of the protocol, and the sustainable business model of our company.

  2. Researching new technologies and approaches for Compound markets to include assets and borrowing demand from blockchains (and L2s) beyond Ethereum. This work has centered around Gateway, a stand-alone distributed ledger with “starports” on external chains, used to connect these assets.

Since originally announcing the Gateway project, time has moved quickly. We’ve released multiple versions of a Gateway prototype / testnet; countless teams have begun experimenting with multi-chain approaches; the Compound protocol on Ethereum suffered from a faulty upgrade (unintentionally rewarding COMP to some users); gas costs on Ethereum have risen dramatically; several cross-chain protocols lost assets to security incidents; and multiple blockchain ecosystems evolved rapidly, often with EVM-compatibility at the core of their growth.

Each of these events has required us to adapt what we believe the right multi-chain strategy to be:

Rather than launch Gateway connected to starports on external chains, which will introduce significant new technology & risk, we believe it’s prudent to first launch a Compound presence on EVM-compatible chains as fragmented markets–and if/when the community decides to, upgrade these markets into a multi-chain shared liquidity pool.

At this week’s community developer call, Compound Labs will discuss the steps necessary to enable this strategy, and how we can best equip the community to govern a multi-chain future.

27 Likes

Super hyped for this!

4 Likes

Nice - looking forward to this.

2 Likes

This is an excellent step in expanding Compound protocol outside its homebase of Ethereum. With current market conditions and user profiles firstly being onboarded by L2s and Ethereum scaling solutions like Polygon PoS.

The Polygon community and team was very enthusiastic and welcoming of the Starport deployment - Polygon/Matic Starport - #4 by jared. Compound development team also saw the Polygon Starport integration successful and we would love to replicate that enthusiasm and interest again here.

Happy to put up an extended proposal detailing benefits of a Polygon Compound deployment with more numbers and a detailed view and how it will benefit the Compound DAO.

Just for reference - our team wrote a research article about Stablecoin utilization comparison of Multichain AAVE that might help the community discussions.

6 Likes

This is exciting! Lets go multi-chain

1 Like

Going Multichain would be great! Helps defy centralized blockchain platforms and powers decentralization on a whole new level simultaneously

2 Likes

I believe its high time Compound goes multichain. A welcome move towards various growth opportunities and becoming a one-stop platform.

I will contribute to this initiative by:

  1. Proposing an L2 evaluation framework (started), and
  2. Writing automated deployment and configuration scripts (started)
1 Like

Here’s my L2 evaluation framework so far: https://hackmd.io/@TylerEther/ryu5o1sl5

It’s interesting to see that Polygon has a higher degree of decentralization of active validators/miners (by address).

1 Like

I personally dont care about other L1s. But what I would be interested to see is compound on rollups. Specifically Arbitrum/Optimism and later zk rollups.

Polygon is not an L2, atleast not in its current state. If a chain does its own consensus to validate transactions its an L1. If it relies on another chain to validate transactions its an L2.

I’m not looking to debate the definition of an L2.

What’s important is being able to allow users to use Compound protocol without having to pay ridiculous gas prices relative to the amount of capital they’re working with. This is on top of baseline security, liquidity, on/off-ramps, and other requirements (to be established).

Here are current swapping costs to benchmark transaction costs across different chains:

  • Ethereum: $16.34
  • Arbitrum: $1.16
  • Optimism: $1.13
  • Polygon: $0.005
2 Likes

I understand. I am not saying there shouldnt be compound on polygon. I am saying in its current state Polygon is not a L2.

The trend for transaction prices on L2s goes down, there is still a lot of optimization that can be done plus there is datasharding coming, reducing the prices even more.

Your transactions are actually on the Ethereum blockchain with L2s. Which is way more costly to attack than Polygon.

I am just observing the current trend. L2s will outperform any L1. Polygon will also be a rollup and a true L2 solution. Again I am not saying that Polygon shouldnt get compound I just think that L2 deployment should be also strongly considered because thats the future.

1 Like

Yeah, for sure!

I’m personally very excited for Polygon Hermez 2.0 that offers a zkEVM using plonky2 as an Ethereum rollup.

1 Like

My evaluations for Arbitrum and Optimism are complete. We face a grim reality - decentralization is very questionable.

Arbitrum has one active validator, which should be fine provided there are active defensive validators and watchtower validators. However, I could not find any data on these validators.

Optimism has one active sequencer which is run exclusively by Optimism PBC, and the fraud (“fault”) proof mechanism is temporarily disabled. Optimism is currently completely centralized and users have to fully trust Optimism PBC.

Should the community wish to deploy on these rollups, I’d highly suggest aggressive supply caps until we see better decentralization, openness, and transparency in the block production and validation processes.

1 Like

Decentralization with the sequencers is not the issue with rollups. The thing with optimistic rollups is that you have a long period to submit fraud proofs (optimism is overhauling their system right now). And you are able to interact with the smart contract directly from L1 and flush your funds to L1 incase the Sequencer goes inactive or acts maliciously. The transactions you do are stored on Ethereum so once settled its set in stone.

Optimism/arbitrum are going to open up completly so anyone can run a sequencer and get paid for it. Possibly in combination with the release of their tokens. Aave will also deploy on Arbitrum and Optimism, if the only option is Aave on L2s and they blow up, it was a bad decision in hindsight for compound. Both L2 solutions are still a work in progress.

Based on the fact that there is little liquidity at the moment on both. I dont see a problem with agressive supply caps. The risk reward ratio is in my opinion on the side of deploying compound there. I see an advantage in getting in early.

3 Likes

Are you thinking parachain on Polkadot or maybe using Acala’s EVM+?

I like this approach of prioritizing deployment on chains that are fully EVM-compatible (so the protocol “just works” out of the box) and that minimize centralization risk. Minimizing centralization risk is ultimately about security, which motivates the following suggestion:

Compound’s security is currently underpinned by proof-of-work, but of course this will change to proof-of-stake at the merge. While most users will probably be comfortable with PoS security at the level of decentralization that the beacon chain and certain other large PoS chains enjoy, there’s a case to be made that a security-first money market protocol like Compound ought to offer users an option secured by PoW with a high hash rate for anyone who would prefer to keep their assets on a security model with similar characteristics / risk profile as what they have on ethereum today.

One EVM-compatible chain that will still be secured by proof-of-work after the merge is the Syscoin NEVM. The syscoin PoW base chain dates back about as far as ethereum and is actually mege-mined with bitcoin, so it has an enormous hash rate for the very modest TVL its nascent NEVM smart contract layer currently secures. Being merge-mined with bitcoin also makes syscoin’s PoW security effectively carbon neutral: as long as there is intrinsic demand to mine bitcoin, attempts to ascribe emissions to syscoin’s PoW would be double-counting those emissions against bitcoin’s own carbon intensity.

The syscoin foundation will further scale up the NEVM with a zkrollup and validium this year, but fees on the NEVM layer are already dirt cheap, and the slower-than-average 2.5-minute avg block time of the NEVM isn’t nearly as much of an issue for money markets as it is for most other DeFi apps. Currently, ETH, WBTC, DAI, USDC, and USDT are available on syscoin through a single point-of-entry, Multichain’s router, which is not all that decentralized yet. This seems to be currently the weakest link in terms of potentially deploying Compound on syscoin NEVM. Compound would definitely be the project with the greatest name recognition on syscoin if it were to deploy there.

Summarizing, I could see Compound on syscoin NEVM as our “PoW-secured offering” of the protocol, with ethereum as the default PoS-secured option and then rollups / alt PoS chains as desired. If folks know of other PoW-secured EVMs that should be considered alongside syscoin, feel free to add them here. The only other one that comes to mind for me is ethereum classic, but its hash rate is a tiny, tiny fraction (less than a millionth) of syscoin’s.

Disclosure: it’s probably obvious given the level of detail here, but just so we’re clear, I hold some SYS.

1 Like