TL;DR; DEXes on Ethereum have hit their stride and are showing strong signs of life. Recent growth has mostly been concentrated in on-chain, curve or reserve based DEX models. With advances in layer 2 such as zkRollup, the next cycle will see scalable & secure orderbook DEXes take root on Ethereum, allowing for true non-custodial trading, not just swapping.
In the middle of a market rout, an Etherean can’t help but put all the pieces together and make sure they are not missing the mark. Luckily, a brief look at your metric or vertical of choice will confirm what we innately know: undeniable progress.
From my perspective — and through the lens of Loopring — that is of course the DEX space. As anyone with a (DeFi) pulse knows, DEXes are having a moment, and there are strong tailwinds to propel them further forward.
All around us, we are seeing the telltale signs of rising volumes, deeper liquidity, consistent usage, and even enjoyable UX.
While ETH locked in DEXes at all time highs is an amazing thing, the very focus on this metric hints at a useful observation: non-lockable, old-fashioned orderbook exchanges have not yet found their footing on Ethereum.
This fact is largely corroborated by volume measures as well. There, too, with new highs and explosive weekly growth, it is concentrated in the pooled, or peer-to-contract DEX structures. Uniswap & Kyber account for almost 2/3 of volume. (Note there is a double counting as Kyber can execute on Uniswap).
Depending on your perspective, the overall DEX-space stats are actually made all the more impressive when you consider they derive primarily from “simple swap” DEXes, such as Uniswap and Kyber. Why is that? Because Ethereum allowed people to invent a whole new type of exchange mechanism that it is uniquely able to underpin. The maxim of ‘don’t play someone else’s game, create your own’, comes to mind (which I cannot find nor attribute :).
If Ethereum’s DEX darling, Uniswap, can invent & implement unkillable, automated market making curves— something that doesn’t exist in the legacy financial world — imagine what will happen when we can port over the parallel world’s familiar frameworks: orderbook exchanges.
That is, after all, how the bulk of trading most users are accustomed to (public equities, in particular) gets done. It’s also still how the bulk of cryptoasset trading gets done — just on sketchy, any-moment-may-be-hacked, exchanges.
Swapping vs Trading
As seen above, Ethereum is home to many great non-custodial token swapping protocols. A user can swap simply and safely between assets on-chain. But they cannot trade fast nor cheap. The difference between swapping and trading is in its ‘purpose’, and price (in)sensitivity.
The swap products are where early signs of DEX success have taken root not only because they’re useful, but also because it’s just about all Ethereum can effectively support at the moment. Trying to build non-custodial orderbook exchanges on Ethereum has proven possible, but not performant. However, the virtue of orderbook exchanges is often performance — or more specifically, high throughput, low cost, and an inviting home to market makers’ liquidity.
If you can’t provide this environment for traders, then you’ll likely be left with the casual swapper, for which the simple swap is a superior experience anyways — no orderbook needed nor wanted. Incidentally, a simple swap platform can rarely be sufficient for ‘true traders’, but a decently liquid orderbook exchange can be sufficient for swappers.
Taking a look at the most important pair on Ethereum, ETH/DAI, we can know the difference between swappers and traders by their venue of choice, intentions, and behaviour.
A swapper can easily go to Uniswap, sell DAI for ETH at whatever rate the curve deterministically dictates, and boom (~15 seconds, $0.05 later), they have ETH exposure at what is likely a reasonable price (because the ETH/DAI liquidity pool is deep). These days you can also go to aggregators like DEX.AG or 1inch.exchange for a similar experience, with best-price-wins guarantees.
A trader — not willing to trust anyone with their ETH money — wants to analyze the chart, the orderbooks, the price action, and set their entry point precisely at, for example, 150.15 DAI…wait no, cancel that order, 150.18 DAI, and boom (~0 seconds, $0.001 later) they have their ETH position. Often, it is a bot executing this logic hundreds of times per minute or more.
The latter situation is not possible today. A trader must basically choose between trusting some centralized exchange with their assets, or giving up some of that performance (high speed, low cost).
The industry is getting closer, but is not there yet. Dolomite launched last month, built atop Loopring v2. You can place/cancel orders for free on Dolomite’s orderbooks and experience fast matching, but you must wait ~15 seconds to be confirmed settled by Ethereum’s next block, and must pay the network’s gas fee (abstracted). This represents the state of the art.
This is not good enough for market makers, HFTs, algos, and other professional traders.
Loopring v3, however, by matching & batching thousands of settlements off-chain and proving/verifying them in a zkSNARK, allows a trader to experience instant finality for fractions of a penny — without sacrificing Ethereum’s security guarantees.
High Performance Orderbook Exchanges — Without Compromising Security
The swap-style DEX success is an ecosystem-wide victory we are thankful for. We are all rooting for the home team in the good fight. However, this team has a glaring hole in its roster. My weird sports analogies can of course mean only one thing: Ethereum does not yet have high-performance, orderbook-based DEXes. As Loopring 3.0 is set to be deployed to mainnet in the coming days, we could not be more thrilled to fill this gap, and complete Ethereum’s emerging exchange infrastructure.
Our past year and a bit has been marked with SNARKs. We’ve gotten up to speed (pun intended) and then some on ZKPs, and its specific application towards scaling DEXes.
With v3, we have built the first zkRollup DEX protocol, complete with on-chain data availability. To be clear, while scalability — high throughput, low settlement cost — has been the design goal, the overarching and sacrosanct principle has been security. That is why we use the most secure layer 2 scaling solution — zkRollup — where every state transition is cryptographically proven. Going one step further, our zkRollup construction enforces on-chain data availability, as opposed to off-chain data availability, which comes in handy during those especially adversarial & scary times. This means the only assumption a user needs to be comfortable with in order to be certain they can always access their funds is that Ethereum will continue to exist. Nothing more.
Loopring builds for the trader (or really, builds for the builders of the traders :). That means fast & cheap trades (1400 tps, <$.002), gasless cancellations, and all the accoutrements traders have come to expect.
v3 is a protocol for high-performance, non-custodial, orderbook exchanges on Ethereum. While that is more of a mouthful than the old way we’d describe it — as just a DEX protocol — the description of the protocol is simultaneously the description of the problem once faced: performance and self-custody couldn’t co-exist for trading before.
Starting in December, when we DeFiers espouse our principles to the world, we will know that there are no more cases where we must cheat/surrender —no more need to go to centralized exchanges to truly trade.