Comparing Order Book Designs: ZKEX + zkLink vs. dYdX + Cosmos
The lack of decentralized order book exchanges has long been one of the missing pieces of the DeFi puzzle.
Although AMM DEXs like Uniswap dominate the market, they suffer from an inherent flaw: they are not suitable for high-frequency trading, automated trading, or large transactions.
Professional traders and institutions prefer order book exchanges, because the exchange rate will deviate significantly and transaction friction become unreasonably high for large value trades on an automated market maker (AMM) exchange.
However, order book exchanges are not easy to build, since the performance (transactions per second and finality) requirements for order book trades are significantly higher than on AMM designs. Decentralization is often sacrificed to increase performance which partially explains why dYdX is leaving Ethereum and StarkEx.
zkLink wants to break this bottleneck, and build a specialized trading network with no compromise on security while delivering scale, and a clear path to full decentralization.
The first ecosystem dApp to be built on top of zkLink, ZKEX, is in some ways conceptually similar to dYdX (which will be built on top of Cosmos), but the underlying architecture, design, and technology are very different.
Multi-Chain Tokens: To Bridge or Not to Bridge?
One of the challenges for dYdX in building a native chain is to “bridge” tokens to and from other chains. dYdX will be able to enjoy chain-interoperability through its IBC protocol, but only within the Cosmos ecosystem.
As the biggest order book DEX on Ethereum, it must find a way to safely connect assets across chains. As dYdX mentioned in its v4 announcement, “most existing bridges suffer issues with user experience and/or security”, so to some extent, dYdX is partially limited by its bridges.
In contrast, zkLink is totally bridgeless - no extra steps, no security risks, and no unnecessary bridging fees.
Instead of a closed ecosystem, zkLink operates as a single, unified network layer that natively connects multiple chains and networks, so tokens on those chains can be easily listed and traded.
Liquidity: Fragmented or Unified?
Order book models rely on deep liquidity for a smooth user experience - buy and sell prices must be matched to fulfill trades.
While dYdX sets a price for each different token (one for USDC on Ethereum and one for USDC on Solana), zkLink eliminates the difference between these tokens and merges them into a single pair.
This unique token merge design increases the depth of liquidity to magnitudes higher, and increases the capital efficiency for traders.
zkLink’s multi-chain interoperability already taps into the native liquidity from connected blockchains (Ethereum, BNB Chain, Polygon, Avalanche, and many others in development), however interoperability is taken a step further - by merging the native stablecoins on these chains into a unified form on zkLink L2 network the liquidity is aggregated, and forms a single giant stablecoin pool.
For example, on ZKEX.com, traders will use an aggregated ‘USD’ token which is merged from fully reserve-backed stables, such as USDC, BUSD, TUSD, and USDP.
Security: Validators or Mathematics?
The dYdX chain will have to give up the security benefits from StarkEx’s ZK-Rollup design and Ethereum’s default security.
As a PoS chain, its security level will be a function of the number/decentralization of validators and the market cap of $DYDX, and it could be a long process for the dYdX chain to build its own decentralized validator network. Moreover, using third-party bridges also adds to its security concern.
The security of zkLink is backed by the underlying connected chains. The computational correctness is guaranteed by mathematical verifications rather than economic assumptions, while maintaining full on-chain data availability on Ethereum and other L1 chains.
Summary: Are Crypto Traders Facing a Compromise?
For now, we can say no one approach today perfectly solves the blockchain trilemma - although both approaches are on the right path.
dYdX on Cosmos balances security in exchange for a higher level of decentralization and scalability.
ZKEX on zkLink puts security and scalability first, because of the belief that in an order book model, user experience (performance) and security are higher priorities than “true decentralization”, which can not be fully achieved yet to the highest degree.
What is certain is that different designs are built for different demands, and they all add to a greater picture for the DeFi-verse.