0x beyond Ethereum / check-in

On Discord there have been many questions lately about 0x’s plans with regards to the expansion beyond of Ethereum.
The overall strategy was shared back in October, so it’s perhaps useful to have a check-in here on what happened in between.

The strategy

Quoting the blog post:

before investing heavily in productization, we will look at market signals that tell us with a fair amount of confidence that that’s what the ecosystem needs to continue the process of tokenization of value.

We also have informal conversations with 10+ teams building L1/L2/Ethereum Sidechains and learnt more about their progress and roadmaps.
0x being about creating a tokenized world where value flows freely, we obviously made sure there were no obvious blockers, within the progressive decentralizations framework. That is, we are generally more lenient towards early days centralization, as long as a path for decentralization exists.

We formalized rough metrics litmus test for what we would consider a valid candidate for a 0x expansion:

  • Volume: roughly 10% of Ethereum DeFi volumes before making a move
  • Adoption : at least 10K active addresses
    • In absence of that, sustained growth of active addresses - think 20+% W/W over 4 consecutive weeks, or 2X m/m
  • liquidity MVP path: existing liquidity and protocols (better if AMM)

What does the data tell us?

Talking hard data, we have observed a steady growth in adoption by a few key players, across $ volume (when applicable), blockchain wallet # and # transactions.

This data is made of weekly snapshots taken manually by a curated list of data sources, so it could be incomplete or not 100% accurate. However, it paints a pretty clear picture.

MVP Path

We also started formalizing a lean product/engineering path to expand 0x into new blockchain venues. Porting over the entire 0x infrastructure (onchain and offchain) is very costly, so it made sense for us to create the following smaller, sequential milestones:

  1. Plug in Matcha on existing chain+protocol+liquidity
    Assess taker demand (usage, volume). If significant growth, proceed
  2. Deploy 0x protocol (better if EVM compatible)
  3. Update 0x API, aggregate existing AMM liquidity
  4. Update RFQ quote server, work with MMers to build RFQ flow
  5. Update Mesh, unlocks OO
  6. Connect tokeneconomics (how the current ZRX token governs 0x across multiple chains)

The more we go down the list, the more complicated the puzzles around engineering and network dynamics, but also (assuming the first milestones have been cleared), the more visibility we will have on the relevance of the blockchain venue.
In other words, this lean approach will make it possible to invest progressively in an expansion while gaining signals on the prospective ROIs.
Note that the entire process would likely take months of development.

What about the DEX blockchain hub?

As discussed in the blog post, we have researched internally solutions to power trustless, efficient, p2p exchange of assets belonging to different blockchains.
Similar to what expressed in the blog post, we concluded that it’s probably very early to invest in research and implementation before testing the waters with supporting 0x in other blockchains natively.

Taking a 30k feet view, there are two macro approaches:

  • inside-out (left): design and implement a DEX blockchain hub venue, then plug that in into different blockchains where 0x is present
  • outside-in (right): expand 0x in other blockchain venues first, then design a DEX Hub able to orchestrate network liquidity and token economics.

Our current thinking is that the ‘outside-in’ strategy makes the most sense right now. The ‘inside-out’ strategy would likely take us to a research-heavy path with not enough market signals, risking to diverge from real world adoption dynamics.
Also, based on what we have seen out in the wild, there are real challenges in ‘bringing up the stack’ tokenized value where existing network effects are in place (see L2s), and it feels more prudent to collectively research solutions on these hard problems (e.g. bridges) and see how these experiments unfold.
Therefore, we parked that specific research area for now (informally referenced as 0xChain on Discord), while we start paving the way along the MVP path described above.

Conclusions

It would be great to get the community’s take on the above.
What do you think about the MVP path? Any thoughts on what blockchain venue should be prioritized?

13 Likes

Thanks @mintcloud for this detailed report. Many call now for fast, gas-less trading, but we also need decentralization. Therefore, when integrating with other L1/L2s as an additional source of liquidity keep that in mind. The most important thing is to create a network that connects all liquidity and offers it to people to trade in the most decentralized manner at affordable rates. It is also important to keep things simple. To support chains that are not truly decentralized just for growth would go against the 0x ethos of supporting an open, permissionless p2p trading world.

At the end of the day think about how Uniswap launched and the core thing that they offered that catered to what people cared about. That hasn’t changed. It’s just that now gas prices are so expensive. 0x API already has great sources of liquidity. How could we channel that and create a great user experience that would allow people to trade in a decentralized manner for a fraction of the cost across different chains?

Right now, if I am a whale I stay on Ethereum as I don’t care about gas prices nearly as much as normal users. Anyone else at this point is pretty much priced out. Now we go to more centralized versions that use EVM compatible chains, or whatever else other exchanges will use. Every exchange is going to build their own decentralized version. The other option I have is to try L2s that exist today (e.g. Deversifi or Loopring) but their user experience is not the best and onboarding is expensive. 0x Labs will have the best answer as how 0xchain fits into that.

Ultimately 0x needs to become the network that connects all the liquidity sources from different networks and make it for the user so simple that they don’t see all of that on the front end. The user just cares about getting the best prices, tapping into the highest liquidity and execute at reasonable prices in the most decentralized way. At the end of the day the complexity that is powerful needs to be simplified for the user. You have shown that you can do that with Matcha. While you may be the most undervalued project in the space and people do not understand your technology, show them that you can create technology for ordinary people, for anyone, that sources liquidity from many different networks and offers it in such a way that funds are on the most decentralized network and like I said before you can get the best prices with affordable trading costs. If you can do that, everything else will follow from that. You will have network effects and millions of people onboarding onto 0x and from there you will be able to achieve so much more and go after the principled goals to create the world that you envisioned.

6 Likes

I’ve used the Loopring L2, found it to be appealing and the development team responsive from a UI feedback perspective.

Wonder how 0x would deploy on their L2.

3 Likes

Are you saying that the outside-in approach would follow the MVP Path steps for each new chain? If so, can you provide more detail on how the steps would be implemented? For example, say you took milestone 1 (Plug in Matcha on existing chain+protocol+liquidity) and deployed it to BSC (or another chain). What would that entail and can you estimate the time and resources required (for milestone 1 only)?

Have you explored how separate instances of 0x/ZRX on other chains without the hub (0xChain) would impact the staking model ?

I tend to think that the hub option still has merit , especially if done in parallel because it gives you the opportunity to create a standard bus-type architecture and a first-mover advantage, especially for liquidity integration.

2 Likes

Personally I am a big fan of what they have created! And they are growing their volumes lately too.

Still, their excellent solution constitutes ~2-3% of what Ethereum DEX volumes are, as of now.
One big factor is (our assumption) the absence of other DeFi protocols in L2, so you would need to break composability with lending/borrow/LP.
Basically, depositing your funds in L2 creates an opportunity cost (this is true for any L2), and that’s especially high if you don’t find any other DeFi protocols there.

If the user is bound to transfer->swap->widthdraw then the question for the user is
What is better?

  • 2X (medium expensive) transfers + cheap swap
  • 1 expensive swap
2 Likes

can you provide more detail on how the steps would be implemented?

Sure, roughly it would mean to connect a UI to existing liquidity, potentially with no 0x presence. This will allow to gauge user demand before investing further. Milestone 1 challenges are: Wallet management (every chain have their own peculiarities in terms of Wallet support, some are easier than others), data availability (even gas oracles aren’t necessarily present/reliable in other chains), and liquidity interoperability (which depends on the existing protocol).
By the way this is not a necessary step, and depending on the chain, it might be worth deploying contracts and jump to Milestone 2 (deploying 0x contracts). For example, in EVM compatible chains that might be preferable. It would make sense to have aggregation of AMMs right off the bat.

The goal of sequencing these milestones is to maximize the likelihood of catching the right waves of user adoption. Marketplaces are incredibly difficult to bootstrap, therefore it’s preferable to start where portions of the two-sides (liquidity supply and taker demand) are somewhat solved/proven, to then enter the market. We learnt the hard way that it pays off to prioritize market signals while balancing out long term research tech.

Very rough sizing: 80% of engineering investment lays in updating the off-chain infra (0x API, RFQ, Mesh) VS deploying contracts (assuming EVM compatibility). It’s in the “months of dev” order of magnitude. 0x contracts on their own don’t do anything (that is 0x OTC in 2017).

Have you explored how separate instances of 0x/ZRX on other chains without the hub (0xChain) would impact the staking model ?

We did, and there are some general principles we think are sound: ZRX token (ERC20) holders will be able to vote on any local instance of 0x protocol, while liqudity rewards are collected following custom models depending on liquidity type and potentially blockchain venue. Liquidity rewards should flow into a central staking system (today on Ethereum, in the future probably in the Blockchain DEX Hub), and should be redistributed to the 0x stakeholders that have skin in the game when it comes to a good functioning protocol - including ZRX holders who swap a portion of their voting power.

Technical details need to be flashed out and figured out in case by case. The best way to do it is to follow the MVP path imo.

4 Likes

I like the MVP path, and there seems to be benefits of starting with an inside-out strategy. Eventually, it would be helpful to know how we’re measuring success at each level (i.e., what does “if significant growth…” mean?). If you don’t see a ton of user adoption but you learn a ton, is that not successful? Does that mean you wouldn’t deploy elsewhere? Not necessary to answer right away, just something to consider.

High level, it sounds like this approach would allow you to “test and learn” a good deal quickly without over-committing resources and / or deviating too much from the 2021 roadmap. As a community member, I also like that it signals intention for 0x to be chain agnostic, something that most probably don’t know. Even if you don’t get a lot of adoption, it puts 0x into a different conversation and potentially helps you attract more talent, brings more awareness, and increases volume on L1 Ethereum.

A few other questions. Some you may not be able to answer now, but here’s what’s on my mind:

  1. Going off Will’s comments in Discord, I was under the impression 0xChain was launch ready. This post implies that many of the hard research problems are still left to solve. Am I misinterpreting something?

  2. Are there any potential regulatory implications of moving onto other chains?

  3. How long would it take to get Matcha plugged into another existing chain / protocol?

  4. Are there benefits to starting the awareness campaign before Matcha goes live on another chain? I know you don’t want to overpromise and underdeliver, but I’d like to see a little hype generated before this goes live. It also may allow you to gauge market receptivity in advance.

  5. What are some risks? What’s a worse-case scenario…is it just little-to-no adoption and you don’t learn anything?

  6. Any early ideas on where you’d start? (e.g, BSC, Solana)

Thanks for sharing your thoughts @mintcloud, looking forward to helping take 0x to the next level!

2 Likes

Thanks @mintcloud as usual for deep discussion.

I personally would like to see an MVP in action to us start gather experience how to expand behind, and to start get some documentation how this kind of expansion will work and how seamless the integration will be, right now we are on theory field, and we don’t have the experience how it works to expand this whole infra to other chains, even for integrators. We preparing the whole infra to be easy extendable for other chains right way, it will make ZRX and integrators moves a lot easier when it decides to adopt other chains. This expansion should be a matter of adding provider with different chain Id and work right way, I don’t know if the 0x mesh and 0x API are already prepared for this, and would like to know how many months it should take to get at this level of easy expansion.

Additionally, documentation to easy add ZRX tooling to other chains, and as well 0x API and 0x Mesh should be available from MVP experience to get devs from other communities to easy onboard, and at same time align incentives to these new chain communities use the ZRX token (or 0x Hub) instead of deploying their own token without benefiting the ecosystem. For instance, Uniswap Token is not benefiting from PancakeSwap, and they are the same protocol.

From top of head I saw as main candidades Binance Smart Chain or Avalanche for the MVP. As they have AMM’s and are Metamask compatible. I would like to saw an evaluation how close these blockchains are to met the criteria to get ZRX toolchain. Maybe a post with criterias evaluated from 0-10 and blockchains, this will help community understand better how close other chains are to start be looking by ZRX team.

4 Likes

It makes a lot of sense @JoaoCampos89
Updating 0x API /swap with aggregation of would take precedence over 0x native liquidity (including limit orders over Open Orderbook and 0x Mesh), for no other reasons than relative simplicity. We think that it’ll be easier and more justifiable to invest in a fully fledged RFQ+OO setup (+ connected token economics!) once we prove that there is value in supporting markets in these other environments!

I think looking at the chart I shared above (to be updated) it’s pretty obvious that what you say about BSC and Avalanche (Pangolin is their Uniswap fork) makes sense. I would also add Matic (QuickSwap is their Uniswap fork). Serum on Solana is also interesting in volumes (even though the model is different and in a certain sense more similar to 0x

1 Like

Would BSC really be an option though? The lack of true decentralization goes against the core values of 0x’s mission statement. I’m not trying to make this a FUD on Binance, because they obviously have made something that people like and are using. There are plenty of other projects out there that value decentralization and are focused on the long term goals of this space - even though it may be the harder route.

It’s also a very interesting time for layer 2 scaling solutions on Ethereum. As some of the bigger DEX projects port to l2, there will be a renewed demand for 0x’s product. I think it’s hard to even comprehend how much demand is out there right now, sitting idle while being deterred by the high fees to use these products.

Just watched the presentation from 3/11 community call and catching up on this discussion here. Wanted to voice strong support for this proposal here on the forum.

Really like the “outside-in” strategy, seems the more nimble/flexible of the two options.
Seems like there’s going to be a period of extreme fragmentation coming up as ETH mainnet continues to get more expensive and other L1s/ L2s/ Sidechains emerge, which could be a great expansion opportunity for 0x Protocol.

Obviously I’m biased (running a validator) but Polygon Matic is an excellent first candidate due to its EVM compatibility (migrating smart contracts is pretty much plug-n-play), Metamask support, strong ecosystem support, & recent growth.

Alot of NFT and gaming projects are either in the process of migrating over or have already migrated, most of which need DeFi building blocks to work (ERC20/ERC721 trading, AMMs, etc): Decentraland, Atari, The Sandbox Game, Neon District, Aavegotchi to name a few. On the DeFi side, there is a Uniswap clone called Quickswap and I think SushiSwap just ported as well. Additionally, other devs from those projects have been asking when 0x is porting over.

Would love to see 0x ported over in its full form rather than some forked version that may have made tradeoffs/sacrifices (i.e. centralized orders, no upgrades, no governance).

3 Likes

yeah I get your point, but it’s important to stress the fact that an expansion of 0x protocol isn’t a stamp of approval for one blockchain venue vs the other. It’s not a zero-sum move. On the contrary, it’s meant to be additive, in the spirit of 0x’s mission of letting value flows freely. What should be the biggest driving factor is user demand (existing and latent) and the presence of open ecosystem able to flourish. It’s also important to recognize that decentralization exists on a spectrum, the blockchain design space is vast, and all design decisions come with tradeoffs.
The are also real spillover opportunities here, where perhaps other L1/L2 native users (like BSC’s) might be ending up moving to Ethereum ultimately, or demand ‘more decentralization’ out of BSC. What’s important in my opinion is to prioritize the onboarding of millions of users into DEX applications from unbanked/tradFi/CEX. Then the DEX destinations landscape can mutate over time.

While 0x Labs is planning to spearhead the expansion of 0x protocol to multiple L1 and L2 networks in 2021, we believe Ethereum will continue to serve as a nexus for decentralized finance and 0x protocol.

5 Likes