Blockchains enable frictionless transfers of value, which in turn enable entirely new ways of trading one digital asset for another. This article attempts to categorize and analyze these new exchange systems into four distinct categories:
- Centralized Exchange (“CEX”)
- Smart Contract Custody Exchange (“SCCE”)
- Decentralized Exchange Protocol (“DEP”)
I deliberately avoid the term “Decentralized Exchange” (“DEX”) because it is too generic to adequately describe & distinguish these precise systems.
Centralized Exchanges (“CEX”)
🌉 Bridge Between Finance 1.0 and Finance 2.0
The only legal way to turn fiat currency (“Finance 1.0”) into cryptocurrency (“Finance 2.0”) is by using a regulated Centralized Exchange. This is because exchanges must comply with a variety of regulations on the state and federal levels in order to be able to process deposits via traditional payment methods.
Popular Centralized Exchanges include:
These operations are capital-intensive, both in terms of humans and money. They require lots of employees, offices, budgeted funds, central hardware systems, licenses, and relationships with government regulators & banks.
🐋 Attracting Whales
The benefit of satisfying the above requirements is the ability to attract large liquidity providers (Market Makers, often called “Whales”) and the ability to offer advanced trading features such as margin trading; the combination of these two results in a robust, liquid market.
📒 Order Book Management
A critical differentiator for CEXes is the fact that users can create/modify/fill orders for free because the centralized order book is maintained by traditional database technologies and not on a blockchain. This aspect makes CEXes an ideal place for High-Frequency Trading (“HFT”).
The same is not true for SCCEs, which rely on the blockchain for both order book management and execution. This means that on-chain transactions (which cost time & money) are required in order to create/modify/fill orders. This limitation makes HFT cost-prohibitive on SCCEs.
Relayers, on the other hand, use a hybrid approach where the order book is managed on a central database and trades are executed on the blockchain; this enables users to create/modify/fill orders as quickly and freely as a CEX, although trade settlement is slower and limited by the block times of the underlying blockchain (~15–30 seconds for Ethereum); this latency in trade settlement makes HFT significantly less effective on Relayers than on CEXes.
🌈 Full Asset Support
CEXes make cross-chain trading a seamless experience, enabling users to trade between native assets on different blockchains. Currently, only CEXes make it possible to trade between ETH, BTC, and LTC currency pairs.
Centralized Exchanges store user funds on their platforms, which make them Custodians. This is good for entities that are legally required to store funds with Custodians (such as Hedge Funds) but it does require users to trust the CEX to be a safe Custodian of private keys, which may be difficult given that many reputable exchanges have suffered security breaches in the past.
“You don’t own the coins if you don’t control the private key” is a popular adage among the crypto community. The ability to exchange coins while controlling the private key is a primary value proposition for SCCEs, DEPs, and Relayers — all of which require users to trade directly from their wallets.
🆔 Know Your Customer
CEXes must comply with Know Your Customer (“KYC”) regulations which require users to verify their identity by providing government-issued ID cards and in some jurisdictions to verify SSN. Unfortunately, this excludes a large population of the world— especially those in countries that are hostile towards cryptocurrencies, such as China, Mexico, Bolivia, Colombia, and practically all of Africa.
🎯 Central Points of Failure
Lastly, perhaps the largest existential threat to Centralized Exchanges is the fact that they are large central chokepoints. Specifically, they are at the mercy of government regulators and banks with sufficient risk appetites to bank high-risk businesses such as cryptocurrency exchanges. Furthermore, CEXes rely on centralized infrastructure, making them vulnerable to DDoS attacks and attractive targets for hackers.
Smart Contract Custody Exchange (“SCCE”)
🥇 First Generation
SCCEs are the first generation of decentralized exchange system which allowed users to trade cryptocurrencies without a CEX. A SCCE is a single smart contract that handles all aspects of a trade, including order book management and trade settlement — all performed entirely on-chain. This is different from Relayers, which use a hybrid approach where order books are centrally managed off-chain and trades are settled on-chain.
Popular SCCEs include:
👨💻️ GUI Vulnerability
Users interact with SCCEs via a Graphic User Interface (“GUI”) which is hosted on a central server. GUIs make it easy for users to interact with smart contracts using familiar exchange interfaces rather than code. Centralized GUIs are vulnerable to hackers, as was seen with the EtherDelta DNS hack.
🌍 Global Order Book
CEXes must receive permission (from governments and banks) to operate, which means that their order books are only available in the jurisdictions in which they have received such permission. However, SCCEs and Relayers rely on a borderless blockchain, which enables anyone in the world with an Internet connection to access and interact with their order books. This was simply not feasible prior to the invention of the blockchain.
💧 Isolated Liquidity
Order books give exchanges liquidity; the larger the order book, the more “liquid” the exchange. SCCEs derive liquidity from their own isolated smart contracts and are unable to share liquidity between themselves. Thus, SCCEs contain their own silo’d liquidity pool which are incompatible with Relayers.
🔐 User Anonymity & Control
SCCEs, DEPs, and Relayers all offer the ability to exchange cryptocurrencies anonymously; no account setup or KYC info is required to use them. They are also non-custodial exchanges, meaning that users control their own funds.
SCCEs are inefficient in the sense that on-chain transactions (which cost time and money) are required to create, modify, delete, or fill orders. This is due to the fact that SCCE order books live in a smart contract on the blockchain, and the only way to change/update an order book on a blockchain is to send a transaction containing instructions to either create, modify, delete, or fill a specific order on the order book. An interesting side effect of this efficiency limitation is that it makes high-frequency trading cost-prohibitive on SCCEs.
↔️ Crypto-to-Crypto Only
SCCEs, DEPs, and Relayers are unable to transfer fiat currency for technical and legal reasons. This means that owning cryptocurrency is a prerequisite for their use and users can only trade between various cryptocurrency pairs.
⛓️ Limited Asset Support
SCCEs, DEPs, and Relayers are currently limited to trading between ETH and ERC-20 tokens, making it impossible to trade against BTC or LTC pairs. That said, there are several active projects working on blockchain interoperability, which would mitigate this issue.
😡 Fake Coins
It‘s possible to create “fake” ERC20 tokens which can be listed on SCCEs and used to scam people, as has been reported on the Waves platform. Fake coins have also been reported several times to appear on EtherDelta, as well.
Decentralized Exchange Protocols (“DEP”)
DEPs are open-source protocols which codify the rules for exchanging tokens on SCCEs and Relayers. DEPs are unique in that they are not an exchange platform; rather, they are the rule-layer infrastructure on which SCCEs and Relayers are built on top of; similar to the way that blockchains are the infrastructure layer on which Decentralized Applications (“dApps”) are built.
Popular DEPs include:
💦 Networked Liquidity
DEPs are different from SCCEs in that they have the ability to share liquidity with the Relays built on top of them; they are protocols for pools of decentralized liquidity.
This is a compelling feature for developers building on top of DEPs, as users will naturally migrate to exchanges with the highest liquidity. The result is that we are seeing fewer new SCCEs emerge, with a lot of the developer talent migrating to 0x. Tom Schmidt of 0x recently published a blog post highlighting “18 Ideas for 0x Relayers in 2018” which is worth checking out.
🔌 dApp Engines
DEPs are the first major step towards scalable dApps. The ability to seamlessly move in/out of various tokens is required for a healthy dApp ecosystem to exist; otherwise, dApps will simply be too complex and inefficient for the average user. However, it should be noted that DEP efficiency is entirely dependent on the efficiency of its underlying blockchain. Thus, DEPs built on top of Ethereum are still restricted to roughly 15 transactions per second and are still subject to block confirmation wait times.
The good news is that 0x currently has a list of 16 dApps that will be (or are) using its protocol, including Augur, Maker, and other high-profile projects.
Relayers use a hybrid on/off-chain approach which generally use a centrally-managed order book and/or matching engine, while relying on the blockchain for decentralized trade settlement.
Popular Relayers include:
📲 Interfaces for DEPs
Relayers can best be thought of as interfaces for DEPs; they are the apps built on top of the DEP which enable price discovery and order matching in a way that would simply not be possible with DEPs alone. It is important to note, however, that these interfaces are centralized and thus vulnerable to attack.
Barriers to Mass Adoption for SCCEs & Relayers
In summary, SCCEs & Relayers face the following barriers to mass adoption:
- Chicken/Egg problem: SCCEs & Relayers require liquidity to attract adoption, and adoption requires liquidity. CEXes often solve this problem by making deals with key Market Makers to incentivize them to provide liquidity on their exchanges; these incentives take the shape of a rebate or fee and are provided by CEX to Market Makers in exchange for the Market Makers’ guarantee to maintain a specified level of liquidity on the exchange for a certain period of time.
- Unable to offer Custodial services (which are required by large liquidity providers such as Hedge Funds and Institutional Investors). This disqualifies large liquidity providers from participating, which make it even harder for SCCEs and Relayers to obtain the level of liquidity necessary to compete with CEXes.
- Low liquidity make the order books on SCCEs and Relayers more volatile than their CEX counter-parts.
- No KYC requirements prevents large liquidity providers from participating.
- CEXes have a substantial liquidity lead that will be hard for SCCEs and Relayers to catch up with.
⛓️ Limited Asset Support
- SCCEs and Relayers are unable to support fiat currencies
- SCCEs and Relayers currently do not support blockchain interoperability; this will likely change in the future.
👎 User Experience
- Private key ownership/management (requires technical savvy)
- Creating, modifying, and deleting orders is complicated, expensive and time-consuming.
- Limited to basic features (no margin trading)
- Hosted GUIs can be a security risk.
Nascent technologies present both risks and opportunities and take time to find their place in society. SCCEs, DEPs, and Relayers are no different. Only time will tell whether they will be able to overcome liquidity and usability challenges. In the meantime, the research and development of these exchanges is important to the decentralized ecosystem and could very well lead to some much-needed scaling solutions.
Thanks to Zach Segal for reviewing early drafts of this article.
Citations (in order of appearance)