0x Project vs. AirSwap.io
Prior to AirSwap’s public sale on Oct. 10 (or yesterday), many interested in the decentralized exchange (DEX) space have questioned the redundancy of AirSwap in favor of 0x, which conducted its successful sale back in August. After some research, the two protocols seem to complement each other in the same space because they approach decentralized exchange from opposite ends of the spectrum. Their design differences will initially cater to separate niches in the same market, but their subsequent participation in community efforts is what ultimately determines the leader.
Before we get into the details, here’s a table illustrating their initial strategies to the market:
Order Book vs. Indexer
To facilitate token exchanges, 0x uses order books to keep track of orders, which are left unsigned by makers until takers complete them. The order books can be private or public, and traders can choose the order books they wish to trade on. Because the orders involve no additional step to finalize, 0x’s approach favors speed and can accommodate automated trading. Further, even faster speed can be accomplished if the selected order book supports Centralized Matching, automatically matching orders with identical counterparts.
In contrast, traders on AirSwap use Indexer to solicit token exchanges. Makers and takers post their “intent to trade” for specific tokens, without specifying their desired exchange rate. When a match for a token-pair exchange is identified, AirSwap then pushes both parties to negotiate via its Peer protocol for an agreed-upon exchange rate . If an agreement is unlikely, counterparties can query an Oracle protocol for “a fair price suggestion.” The specifics on how this protocol can deliver “a fair price suggestion” is unclear at this point.
It is worth noting that the AirSwap team advertises its Indexer as a means for counterparty discovery. In a Q&A session, AirSwap’s Michael Oved expanded on the discovery process, and discussed the ability to bypass the discovery process to make subsequent connections with a previously discovered peer. Though this will surely speed up the transaction process, it raises issues with whether the counterparties will simply process future transactions without the AirSwap protocol all together.
Relayer vs. Indexer
0x relies on relayers, an exchange anyone can act as, to maintain order books. To cover the costs of maintaining order books, relayers can require transaction fees from both makers and takers. While 0x does not explicitly require relayers to charge transaction fees, it is in relayers’ interests to do so unless they can find incentives elsewhere. It is also in relayers’ interests to support public, as opposed to a private, order books to maximize the number of transactions.
AirSwap, on the other hand, likely does not depend on third parties to complete transactions. Based on its publicly available information, the AirSwap team has consistently described the Indexer as an AirSwap protocol in its, whitepaper, roadmap release, blog post and Reddit response. As such, AirSwap must be transparent about how the “intents to trade” are fairly listed, sorted, and removed on the Index.
Token Utility: Transaction vs. Membership Fee
Both protocols require fee payments in the form of their tokens, ZRX for 0x and AST for AirSwap, but they differ in their fees.
0x requires fees at the transactional level. Whenever a trade is completed, both maker and taker pay a fee according to the fee schedule of the selected relayer. This allows anyone to make a trade for a fee, effectively increasing liquidity in the market and further lowering the barrier-to-trade. However, if the fees are not insignificant, one may incur considerable costs by trading on 0x.
Instead of a transactional fee, AirSwap requires fees to access its protocol, much like a software licensing fee. To return the favor, AirSwap locks up, rather than extracts, a specific amount of AST for a specific amount of time for each “intent to trade” and returns it back to the holder after the trade is complete or the time expires. Regardless of the price tag on the lock up amount, this model will inevitably price out certain makers/takers at the outset. It does not, however, charge those that complete the “intent to trade.”
Basically, AirSwap is charging for those that wish to become authorized sellers and buyers. Indeed, the team has talked about a reputation system to ensure the quality of the makers and takers in conjunction with the Index.
Token Utility: Voting Rights vs. Voting Rights*
Both protocols use their token for decentralized governance to improve their underlying smart contracts over time.
The 0x “protocol tokens may be used to drive a decentralized update mechanism.” If this is true, it is fairly easy for anyone to contribute to 0x’s update mechanism; anyone can become a ZRX holder by acting as a relayer, and any relayer can grow into a large ZRX holder through various Relayer Strategies functions. With such low barrier-to-entry, the 0x community must carefully balance the interests of existing and prospective ZRX holders to ensure success of the 0x protocol.
For AirSwap, its governance is “weigh[ed] … at least in part, based on each voter’s total AST holdings.” To acquire AST, prospective holders must purchase through its sales or directly from existing AST holders. In addition, because AST holders can vote directly on the lock up amount to post “intent to trade” on the Index, it make sense for competing AirSwap makers and takers to collect more AST tokens to gain voting power. As such, AirSwap’s success not only depends on a careful balancing of the interests of existing and prospective AST holders, it also depends greatly upon the relationship between the haves and the have-littles.
Protocol-Specific Applications
Notwithstanding their differing approach to a decentralized exchange, their proposed ecosystems are still under development and have yet to be tested by the DEX market. That said, their design differences interestingly position themselves on opposite ends of this market.
Specifically, 0x presents an appealing opportunity for those in the business of large-quantity automated trading. Its near frictionless design will win over those that favor speed in the exchange market. This is evident when Augur, a prediction market dApp built on top of the 0x protocol, views “0x as a scaling solution for the speed of transactions.” 0x is also suitable for those that wish to liquidate assets as quickly as possible as it supports orders that any taker can pick up without engaging in an extra step of negotiation.
Catering to the other end of the spectrum, AirSwap is better positioned for OTC (over-the-counter) and everyday retail transactions with its counterparty discovery and negotiation protocols. Imagine buying a car on Craigslist. You’ll review each search result (i.e., Index) and select a few dollar-car pairs that you deem appropriate. You’ll then arrive at one satisfactory pair after some negotiation (i.e., Peer) to finalize the transaction.
As more assets become tokenized, this presents an issue for the 0x ecosystem to trade less popular assets, unless it can efficiently whitelist new tokenized assets. Similarly, scarcity in certain assets will invariably require counterparty discovery and negotiation to achieve liquidity. In other words, until one can trade every available asset at every price point on 0x, AirSwap will continue to fill this void. 0x OTC also does not resolve this issue because it does not support counterparty discovery or negotiation.
Another type of transaction AirSwap is more suitable for involves large quantity of non-automated trades. For example, Starbucks, which sold 671,396,071 cups of coffee in 2016, would logically prefer AirSwap over 0x for its one-time membership fee and freedom for “anyone, including people who do not hold AST, to find and trade with counterparties who are listed on the Index.” This is because each trade requires a unique order, and that 0x depends on relayers to host orders in its order books, the shear number of transactions may result in overwhelming fees with the 0x protocol. Indeed, Starbucks would have to factor in the cost of transaction fees as a taker, as well as absorb or burden its customers with fees as makers. Accordingly, AirSwap’s design may be the more economic alternative for this type of transaction.
Though one could argue that 0x OTC would be the better alternative as it charges no fees, AirSwap’s focus on this type of transaction will likely offer more tailored features for the makers and takers involved.
Summary
If you think about it, 0x and AirSwap are really pushing for the same decentralized exchange from different angles.
For example, 0x would become AirSwap when Starbucks becomes a relayer and acts as the sole taker in its private order book. It would presumably absorb the fees on both ends of the transaction to create a better customer experience. Ultimately, the fees would rack up and eventually function like AirSwap’s membership fee.
Similarly, AirSwap would become 0x when AirSwap supports direct peer-to-peer transaction after a first counterparty discovery via the Indexer, obviating the need to engage in negotiation on future transactions. By doing so, the AirSwap protocol would likely be able to support high frequency trading among peers who have previously discovered each other.
Understandably, there may be scenarios where 0x or AirSwap is always preferred over the other, but because this market is so new and user behavior has yet to be studied, the leader of the two will emerge from responding and adapting to community participation. Now that both exchanges have conducted their ICOs, let’s not dwell on which exchange is better on paper, and instead focus on their execution to become DEXs that we hope to trade with in the near future.
UPDATE: The 0x team explained in a Reddit response that “neither the taker or maker [transaction] fees are mandatory … For instance, https://kinalpha.com/ charges 0 fee for the maker. Hence, the maker doesn’t pay any fees and doesn’t send any transaction, making kinAlpha free of fees for the makers.”
