DEX has no moat.
This spring is the highlight of DeFi. As if the value of people was discovered overnight, the market value of various DeFi projects has started to soar. On June 5, the market value of DeFi’s locked position exceeded US$1 billion again after three months.
In this wave, the forefront of the DEX (decentralized exchange) field has performed well, with Uniswap, Synthetix, Kyber, Bancor, etc. occupying most of the headlines.
The development of the DEX track is fierce and will be the core of the DeFi world in the future, but is DEX really a good track in terms of investment choices?
This article attempts to briefly discuss the spot trading track.
1. Classification of DEX
One notable feature of DeFi is composability. Calling each other between different protocols can produce a “Lego brick”-like magic effect. In the bZx lightning loan event, many people have already seen the magic of DeFi composability.
In the field of DEX, the combination trend of different DEX has become more and more obvious, and DEX can be simply classified according to whether it is combinable.
Composable means that liquidity can be shared. It has become a fact that Uniswap, 0x, Oasis, Kyber, Bancor can be combined to achieve liquidity sharing.
However, decentralized exchanges such as Loopring.io and IDEX are not composable, and there is no way to achieve combination for the time being.
2. Liquidity has no boundaries
With the deepening of the combination of different DEX protocols, an obvious result is that liquidity no longer has boundaries, and all DEXs can share the liquidity of the entire network. No matter which application is used to initiate a transaction, you can enjoy the best liquidity.
As the liquidity of the CEX (centralized exchange) moat, it no longer has moat properties for combinable DEX, but liquidity is still important for value capture.
The logic of Loopring.io and IDEX is more similar to centralized exchanges. Good liquidity will still be their moat, and they must have good liquidity to develop better.
If considered from the perspective of the moat, IDEX and Loopring.io are deeper than the moat of 0x and Kyber.
Third, the price war on the user side
The three major demands of users on DEX are: optimal price; fast transaction; lowest cost.
The best price can be combined with DEX to share liquidity, everyone can achieve;
Fast transactions, the infrastructure that can be combined with DEX are all based on the Ethereum public chain, and everyone can’t achieve it.
The cost of using DEX for users includes handling fee and gas fee.
Gas fee depends on the bottom of Ethereum and the complexity of DEX internal transactions. The cost of DEX Gas with complicated transaction process will be higher, which is not conducive to users. This is very important for small transactions.
Using 1inch as a test, the amount of exchange for small transactions is not much different, but looking at the Gas fee of 0x and Balancer, it will inevitably make people think that it is really too expensive.
Figure 1: https://1inch.exchange
For large transactions, the slippage (the difference between the point where the order is placed and the last transaction) is more important than the Gas fee; for small transactions, the Gas fee is more important than the slippage.
In addition to the gas fee, another important dimension of competition is the handling fee, which is a key step in the value capture of DEX. The competition for handling fees is extremely simple and rude, which is simply a price war. At this point, value capture becomes very difficult, and no one can easily capture value.
4. Whoever aggregates more has advantages
Aggregation attributes are becoming the standard for DEX. Who can aggregate more liquidity and who will have a short-term competitive advantage, but eventually everyone’s aggregated liquidity will reach a similar level.
Some of the polymerization cases that have been implemented:
Kyber aggregated Uniswap, Oasis and Bancor;
0x bridges Kyber, Uniswap, Oasis and Curve;
1inch aggregates most of the combinable DEX, Uniswap, Oasis, Kyber, 0x, MultiSplit, Mooniswap, Balancer, Curve, etc.;
DEX.AG also aggregates most common combinable DEX, Uniswap, Oasis, 0x, Kyber, Curve, etc.
In the short term, 1inch with the most aggregate liquidity is the most convenient to use, but eventually this advantage will fade with the follow-up of other protocols.
From the perspective of liquidity sharing, DEX will be like the outlets of chain companies. Each outlet has a different selling price, and each outlet has the right to adjust goods. Some outlets can only transfer a part of the goods, such as 0x, Kyber, and some outlets can transfer most of the goods, such as 1inch, DEX.AG.
However, the cost of transferring goods within each outlet is not the same, and the cost needs to be borne by the user. Some warehouses are complex, and it takes several hands to transfer goods. The cost of gas transfer (Gas) is high, such as 0x; some warehouses are simple and can be found directly, and the cost of gas transfer (Gas) is low, such as Uniswap.
At 1inch transfer, you can see the transfer cost of each outlet at a glance. The user can choose the lowest cost outlet to buy; if the purchase amount is large, it can also give you the best combination of prices at different outlets. And there is currently no charge for the transfer of goods.
It is also acceptable if 1inch charges a certain transfer fee, but it is obviously not too high and will face competition from free competitors.
V. Comparison of typical models
In the DEX field, only Kyber and 0x are the only ones that can build an ecosystem. Uniswap is suddenly emerging, and it can also be discussed.
0x
0x is a peer-to-peer transaction protocol, used by some protocols.
0x is free before the V3 version. Wallets and DAPPs that use the 0x protocol can be charged separately. For wallets and DAPPs, using the 0x protocol has greater autonomy. The most well-known is Tokenlon under imToken.
But for the 0x protocol itself, it is impossible to capture value. Therefore, on the V3 version, 0x decided to charge a 0.3% handling fee for each transaction. The handling fee will be distributed to the 0x token pledge system, which will be shared by the registered market makers, and the team will also give market makers subsidies.
According to the data, the adoption rate has not been high since the V3 version came online, only about 20%.
Figure 2: First position 0x research report
The application end of the 0x protocol can be charged separately, and the 0x protocol itself has to charge another agreement fee, and the transaction structure of the 0x protocol itself is not simple, and the Gas fee also belongs to a higher level (see Figure 1). Obviously, the user transaction cost will be higher after the upgrade, and the user experience will become worse. Applications should not upgrade the protocol.
At the end of 2019, dYdX announced that it would stop acquiring liquidity from the 0x order book, leaving the 0x ecosystem, precisely because of the agreement fees charged by the 0x V3 version.
The application can always use the V2 version without upgrading, but it also loses the possibility of upgrading to obtain new features, which will damage the competitiveness of the application, and whether the old version of the protocol will be continuously maintained and continuously available is also questionable.
The V3 version of 0x caused a contradiction between application value capture and protocol value capture.
When the application upgrades to the V3 version, the agreement can obtain the fee income, but the user’s experience on the application side will be poor (the fee will be high), and the user will be lost. Or the application lowers its own rate standard and damages its own interests to retain users, but the 0.3% rate standard of the 0x V3 version has made it harder for applications to drop. For example, the maximum rate set by Tokenlon is 0.3%.
If the application does not upgrade the V3 version, the agreement will not be able to obtain fee income.
The problem of application and protocol value capture is difficult to reconcile under V3.
Therefore, there can be a reasonable speculation that 0x may want to abandon the construction of ecology, concentrate on expanding the market maker market, expand its own liquidity, wait for other agreements to call its own liquidity, and obtain fee income.
0x has an advantage in attracting market makers, and operates market makers as core users.
Kyber
Kyber is an on-chain liquidity aggregation protocol that provides flash exchange services.
Kyber is free of charge to users, and charges market makers. Part of the handling fee is allocated to wallets and DAPPs that bring transaction volume, and part of the repurchase and destruction token KNC.
Kyber’s model is very user-friendly. Compared with the multiple charging model of 0x, it is more pleasant to the heart.
Part of Kyber’s Token is that the project party is making its own market. It is logical to charge them like the listing fee charged by a centralized exchange. Because there are many applications that integrate Kyber, it can achieve exposure on many front-ends, which is cost-effective for the project side.
But charging market makers does not feel intuitively long-term, capital is profit-seeking, and it is not difficult to imagine where the proceeds are appropriate. Kyber has no long-term advantage in attracting market makers.
Uniswap
Uniswap is a trading protocol based on an exchange pool and is a representative of automated market makers. Uniswap has also achieved decentralization in market making, which means that ordinary users can also participate in market making. As long as Token is injected into the liquidity pool, they can participate in market making and receive commission dividends.
Uniswap’s market maker model is somewhat similar to Kyber. The liquidity of Uniswap is not built by ordinary users. It is more filled by the project party, or the project party provides incentives to encourage users to provide liquidity.
The fight for market makers is a fight for market share and value capture, but competitors can easily obtain liquidity advantages through more fee sharing and subsidies. Combining the price war of user fees mentioned above, it can be found that whether it is the user side or the liquidity side, DEXs will face extremely fierce, and it is difficult to achieve value capture.
Although DEXs are also looking for ways to increase the liquidity provider’s income outside of the transaction, such as Bancor intends to integrate the loan agreement, users deposit DAI, but the pool can actually become cDAI (Compound) or yDAI (iearn) to improve liquidity Sex provider income.
But Bancor can do it, 0x and aggregator can do the same. This method does not provide long-term competitiveness. In fact, 0x is already doing it.
6. Summary
After the aggregation transaction becomes standard, the protocol with liquidity waits for the call to obtain revenue, and the front-end application uses various methods to obtain traffic, which will extremely test the product design and operation capabilities of the front-end application. Owning traffic can divert its own business or external products to achieve value capture, and 1inch is already experimenting.
To summarize briefly:
DEX is very important in the DeFi world;
The combinable DEX has no moat, the competition between DEX will be extremely fierce, and it will be difficult to capture the value;
Users will be able to enjoy better liquidity depth and more favorable rates;
The possibility of front-end DEX capturing value through transactions is very low, and applications with good product experience and strong operational capabilities can achieve value capture through traffic;
The combinable DEX should develop horizontally, and take advantage of traffic to achieve better value capture.