dYdX’s decision to deploy its own blockchain took much of the crypto world by surprise.
On June 22, the derivatives trading protocol laid out plans to use Cosmos’ SDK and the Tendermint Proof-of-Stake consensus protocol. Both these technologies are fundamental to the broader Cosmos ecosystem and fall outside of the Ethereum-centric DeFi space.
dYdX’s move away from Ethereum comes after its previous iteration, called V3, a flagship deployment on StarkEx. StarkEx is a Layer 2 scaling solution for Ethereum.
Now, with its team planning to build out its next version on Cosmos, experts around the industry are mulling over what dYdX’s move means for the broader crypto industry.
dYdX is no small fish. It’s one of the top exchanges in terms of volume — at $9B in 24-hour volume last September, the exchange passed even the behemoth Uniswap. At $486M in the last 24 hours as of July 6, dYdX has seen volume decrease by almost 95% since that peak.
Despite the drop, dYdX is still competing with major centralized exchanges — dYdX is within $10M of established centralized exchange Kraken and constitutes a third of the Sam Bankman-Fried-led FTX’s volume in the last 24 hours, according to CoinMarketCap.
“The dYdX announcement has surfaced a bunch of things that were latent,” Zaki Manian, the former director of Tendermint Labs who has been helping dYdX develop their appchain, told The Defiant.
Manian said he believes in the rollup-centric vision for Ethereum. He thinks that in five years, rollups will be a compelling decentralized system on par with the Cosmos SDK and Avalanche Subnets. For now though, rollups may not be decentralized enough for dYdX, which announced plans in January to dissolve its centralized foundation by the end of 2022.
Currently, StarkEx has a centralized “operator” according to L2Beat. dYdX used slightly different language, saying that Layer 2s (L2s) as a whole have centralized “sequencers” in their blog post announcing their appchain.
The thrust is the same — L2s may be too centralized for a project aiming to completely decentralize by 2023.
Manian added that dYdX’s design, which keeps its orderbook off-chain while matching the orders on-chain, doesn’t actually demand extremely high transactions per second (TPS) capabilities. Therefore, in terms of performance, StarkEx and other L2s are adequate for dYdX.
David Mihal, who develops the data platform CryptoStats in addition to cryptofees.info, thinks dYdX may have a deeper reason for developing its own chain — to allow the token to capture revenue from the protocol without drawing legal action.
“I believe that having a centralized sequencer means the project is too centralized to capture revenue,” Mihal told The Defiant. Because the sequencer is centralized and can be shut down by one party, according to Mihal, there’s a larger legal attack surface.
The engineer added that if block production is handled by a decentralized chain, rather than a more centralized solution like StarkEx, it could employ token mechanics which push value back into the DYDX token.
Mihal gave the example of MakerDAO’s buyback-and-burn model, whereby fees earned by the protocol go towards buying and then burning Maker’s MKR token.
The DYDX token, which was airdropped to early users last year, has dropped 93% to $1.86 since its all-time high of $27.86 in Sep. 2021.
Traders may think there’s something to Mihal’s idea — the DYDX token has rallied 45.1% to $1.96 since the plan was unveiled on June 22. Over that same stretch, global crypto market capitalization is up only 4.4% to $987B.
Mihal and Manian both agree that dYdX makes a particularly good candidate for the appchain model because of its limited dependence on other projects. DeFi is often lauded for its composability, the ability for different “money legos” to stack together. “It was always more of an app than a protocol,” Mihal said.
As a leveraged trading protocol, dYdX only really depends on one asset that users deposit as collateral — the stablecoin USDC. For all other assets, the app just uses price feeds. That users only need one token to use the app makes the transition to an appchain relatively easy.
“As long as the USDC onboarding experience is easy, I don’t think that the user experience is going to look that different,” Manian said.
Linda Lu, ecosystem director at Oasis, a privacy-focused Layer 1, also pointed out to The Defiant that dYdX has a sizable engineering team. “Building your own chain is not trivial work,” she said. “It’s far more complicated than just deploying dYdX on other Ethereum Virtual Machine compatible chains.”
In all, dYdX’s decision, which blindsided many, appears to be a rational move that many in the industry respect. At the same time, insiders don’t seem to think that the move was as big a blow to rollups as it appeared at first glance.
Even Starkware’s co-founder Eli Ben-Sasson didn’t sound too bothered by the move. “Everyone in crypto constantly discusses the pros and cons of different blockchains,” Ben-Sasson told The Defiant. “When it comes up in my team, we’ve always reached the conclusion that Ethereum is the most secure and developed for general computation.”
The co-founder added that he was proud of the success Starkware achieved with dYdX and will be cheering as the team behind the protocol develops their own chain.
And who knows, if rollups develop the way Manian thinks they might, dYdX V5 or beyond could be deployed on a decentralized Starkware solution.