Every CEX Needs a DEX
-
CEXs and DEXs are often seen as competitors, but can be complementary with a CEX running a decentralized backend to guarantee security, transparency and permissionless transactions
-
Centralized frontends are needed for users that can’t or won’t self-custody. Not everyone will use a DEX
-
On-chain transparency prevents shady issues related to misuse or commingling of funds, including outright fraud
-
A CEX allows frontends to run KYC/AML, filter tokens and take other compliance measures applicable to their jurisdictions and principles
-
A DEX is a backend that is always permissionless: it cannot be halted or censored by governments, whether for benign or malicious reasons
-
Infrastructure improvements mean DEXs can support assets from multiple chains — a key differentiator for CEXs in the past
It’s become very clear that centralized exchanges (CEXs) can no longer support crypto’s growth.
Sergey Gorbunov is the co-founder and CEO of Axelar. He led the cryptography group that build Algorand.
CEXs have lost billions in successive waves of attacks: external hacks (Mt. Gox), internal misuse of funds (FTX) and now, regulatory crackdown (Binance and Coinbase). Their decentralized counterparts, DEXs, defend against all three.
Still, CEXs remain vital for users that don’t want self-custody, and for operators who need jurisdictional compliance. To meet those needs and defend against the threats they face, CEXs now need a hybrid solution that integrates DEX technology. Specifically, they need layers of smart contracts that can be put on-chain as the backend to a centralized frontend.
This need is emerging just as new interoperability infrastructure makes this kind of integration possible. And it’s why I say, every CEX needs a DEX.
Background on DEXs & CEXs
The first DEXs emerged within months of the Mt. Gox hack being made public. They gained popularity as a way to stay safe and clear of CEXs, which were seen as “honey pots:” pools of money that are attractive to hackers.
Indeed, outside attackers lifted over $15 billion from Mt. Gox and generations of CEXs that followed, according to data through 2020. However, the successors matured, and the threat of such CEX “hacks” has greatly diminished. Today, CEXs handle 10 to 100 times the notional volume that DEXs traffic.
Of course, there are other threats besides “hacks.” FTX imploded after a cadre of managers commingled and seemingly lost billions worth of customer funds on an affiliated hedge fund. Binance and Coinbase, the two largest exchanges, face existential threats from the world’s most powerful financial regulator.
Amazingly, DEXs provide strong safeguards against all three of these threats: hacks, fraud and regulatory overreach. And for the first time, they can compete on a feature that until now has set CEXs apart: the ability to trade any token from any chain.
Binance realized this early and built its own decentralized blockchain and DEX. OKX followed soon after. Now, Coinbase has announced it is launching a layer-2 blockchain called Base. The fact that the largest CEXs are developing decentralized systems is telling: DEXs aren’t necessarily competitors to centralized exchanges; they are complements.
Here’s why every CEX needs a DEX (and vice versa):
Decentralization improves robustness against failure and attack. That’s the principle that drove the creation of the internet’s earliest iterations, in an effort to make computer systems resilient against nuclear attack.
The longevity and system reliability demonstrated by Bitcoin and Ethereum are also a testament to the robustness of decentralized approaches.
Decentralized approaches are the best way to protect systems against attack and failure.
In general, DEXs are more robust than CEXs, but historically they haven’t been able to compete on features – like trading tokens issued on separate blockchains. Interchain infrastructure improvements mean DEXs can now do that and handle transactions faster by scaling horizontally.
However, it may be DEXs’ core feature – their decentralization – that keeps many users away. Not everyone wants to manage their own private keys.
Whether measured in users or in dollars, mass adoption won’t be a reality without custodial (centralized) onramps.
Of course, one of the most successful custodial onramps is well-known: FTX. Before its catastrophic failure, FTX was an astonishing success.
The problem was, these users had no way to verify how FTX was using their money. They deposited funds for trading, income, etc.; but funds were gambled and lost.
On a DEX, transactions are published on-chain. Users can verify the integrity of their deposits and see where they are being used. Funds can still be commingled, but it’s harder to hide that fact from users.
Not every transaction needs to be recorded on-chain. But a DEX running on the backend can provide periodic visibility into where funds are stored and how they are being used – whenever the CEX frontend “settles” to the backend blockchain on a regular basis. (This could be similar to the way a roll-up settles to its base chain.) Users can verify their funds are safe using a block explorer.
On a DEX, integrating off-chain information can be difficult, requiring complex and fragile “oracles.” On a centralized exchange, it’s easy.
Centralized frontends can easily handle know-your-customer (KYC) and anti-money laundering (AML) processes, limit what tokens they list and apply other filters required by regulation. applicable to their jurisdictions and filter some tokens if they want. A single CEX, operating with a DEX on the backend, can in fact build multiple frontends to serve various jurisdictions.
It will be up to individual exchanges how compliant user experience (UX) can be integrated seamlessly with permissionless backend: probably different tiers of permissionlessness for different user requirements. This is not unlike the way financial institutions consume customized data feeds via an application programming interfaces (API) while individuals often use one-size-fits-all dashboards.
Since its genesis, cryptocurrency has been identified by its users and creators as a way to protect against governments that abuse or overuse their power. Governments have blacklisted bitcoin addresses, confiscated bitcoin troves and tracked users – but none has yet succeeded at halting the network.
Any backend that inherits bitcoin’s properties can never be stopped. Use of such a system will always be permissionless, i.e., without gatekeepers. This is why bitcoin is used under regimes both benign and malicious, for activities ranging from the dissident to the criminal, including law-abiding individuals simply warding off government incompetence.
For centralized exchanges, a decentralized backend serves the same function: a guarantee to users that no matter what action a government may take against the exchange operator itself, the pipes used to handle transactions are proof against overreach.
Composability, interoperability and horizontal scaling
For many true believers, decentralized exchanges have always been the dream: cryptocurrencies are decentralized, and should run over decentralized systems.
Still, in every wave of enthusiasm for cryptocurrency, CEXs have served as the onramps and primary interaction experiences for the majority of user traffic. There are two primary reasons why:
-
With a database to process transactions, CEXs offer greater and faster throughput; there’s no need to verify transactions over a more robust, decentralized network.
-
Similarly, a database in the middle makes it easy for CEXs to list pairs across various blockchains. DEXs have been limited to trading in pairs on the same blockchain. Something as basic in crypto trading as an ETH-BTC pair has been impossible.
Improvements in blockchain interoperability have changed the game for DEXs on both factors. Interoperability networks like Axelar that handle General Message Passing between blockchains are already allowing DEXs to offer cross-chain swaps.
On the backend, the same cross-chain capabilities mean that Web3 applications of all kinds can now scale horizontally. DEXs are avoiding congestion by building single-application “appchains” dedicated to their own throughput, or by selecting chains with the fastest throughput. These setups can now connect to users, assets and applications hosted on other chains.
Interoperability clears the way for continued horizontal scaling: applications can migrate onto newer and faster blockchain technologies, as they emerge, without requiring their users to move.
Finally, and perhaps most promisingly, a DEX backend enables a CEX to compose with other builders, integrating features and network effects into new “super-app” products. Altogether, blockchain infrastructure is fast approaching a point where building decentralized applications is no longer a matter of principle, it’s a matter of competitive advantage and ultimately, survival.