Discuss the three new trends of DeFi: DEX, lending and stable currency

Written by: Lao Bai, Partner of ABCDE Investment Research

Today, let’s talk about the three major parts of DeFi, DEX, lending, and new trends in stablecoins.

Let’s talk about DEX first

DEX has rarely had major innovations since the appearance of UniV3, Curve War can be counted as one, and the DEX track has officially confirmed the duopoly situation in the past two years. In the past few years, those DEXs that have improved slippage and impermanent losses have basically disappeared.

At present, I have seen four trends on DEX:

1. The suture monster Ve (3, 3) - this is currently happening. AC built a Solidly on FTM before, trying to integrate Uniswap, Curve, and OHM, so there is Ve (3, 3) Things, but it’s not very popular. It’s not popular to open a branch in ETH. On the contrary, Fork Velodrome on OP is popular. I feel that one is related to the decline of Alt L1 and the rise of L2, and the other is that the various rewards of OP are also indispensable.

Now Ve (3, 3) is infiltrating into various chains, Arb has recently been on Chronos, BSC has Thena, and Polygon has lost its name... In short, every project has made some fine-tuning on Solidly, similar to Sushi For Uni, the only fun thing is that Uni is not popular, and the imitation disk Velodrome is very popular, and it directly kills Uni on the OP.

2. Hybrid DEX - It is also a kind of stitching, but it is more of an experience stitching.

One is that after the collapse of FTX, people’s distrust of Cex increased, but they still like the silky experience on Cex, so a combination of CEX experience + DEX self-hosting appeared. The basic front-end is similar to DYDX and Blur. First connect the wallet, then deposit money in, and then the fund operation is completely off-chain, and Settle will go back to the chain when withdrawing money.

The second is the combination of AMM+OrderBook - using the combination of the market maker's Orderbook+ the traditional AMM using LP, the user will automatically select the best price of the two when trading, for long-tail assets (Orberbook may lack MM or lack depth ), you can also use the AMM mechanism to "guarantee the bottom line".

We have talked about more than one project like this, and the most representative one is Vertex, which has just been launched on Arbitrum recently, combining the above one and two, and not only making spot products, but also making derivatives, and the team is also quite reliable. It's just that we think that the DEX track is currently too busy, so we didn't vote. Of course, this is normal in ABCDE. We can talk about forty or fifty projects every month, and in the end we only make one or two shots. What we pursue is the One Shot, One Kill style.

3. UniV3-Fi - A few points, one is that the V3 License protection expired some time ago, which is equivalent to open source, and we will definitely see more V3 forks on various chains, and the other is various Fi based on V3, For example, Paraspace (this is currently on the cusp of a few days) makes loans based on V3 NFT, Panoptic makes options based on V3, Gammaswap makes impermanent loss hedging based on V3, etc. There will definitely be more games based on V3 in the future.

4. Curve's Tricrypto New Generation upgrade - Curve has one foot into the mainstream currency field before, but retail investors still use less, because Gas is much higher than Uni, this time the upgraded Gas is down to the same level as V3, and it is popular among retail investors. And aggregators should have more usage scenarios. Coupled with the gains brought to Atomic Swap by the upgrade of SNX V3, the PK between Curve and Uni in the mainstream spot field should be a highlight for some time to come.

After talking about DEX, let’s look at lending

I have seen the following three trends in lending:

1. Full chain - From the design of Compound V3 and AAVE V3, as well as the popularity of RDNT, it can be seen that full chain lending is an undoubted trend. However, according to Mr. Mindao, the current full-chain lending is still in a "pseudo-full-chain" stage, at most it is a "cross-chain lending". The real full-chain lending should be "allow any chain to deposit, withdraw, borrow, and repay, multiple chains share one liquidity, and multiple chains have one interest rate curve", currently no one can do it.

2. Isolation Pool - The concept of mortgage asset isolation started to emerge from Euler, and now large platforms like AAVE, including some new platforms, are basically equipped as standard. The purpose is to prevent oracle manipulation like XVS or Mango. KO the entire agreement, although some flexibility has been lost, but it must not be allowed to spoil the whole pot of porridge with a single rat droppings.

3. No oracle and no liquidation lending - this is especially suitable for long-tail asset lending, because the depth and liquidation mechanism of mainstream currencies are currently very mature, but many long-tail assets even have isolated pools, and many lending platforms are afraid to use it , because the risks of oracle manipulation and liquidation are high. If these two mechanisms can be avoided, the capital efficiency of long-tail assets can be further released. At present, three different approaches have been seen.

One is Timeswap, which uses an extremely complicated three-variable XYZ=K AMM. The design is a bit 6, but I doubt how many users can really understand it... The other is InfinityPools, which borrows V3 LPs, which is equivalent to automatic liquidation through V3, and finally One is Blur's Blend protocol, which is actually a point-to-point protocol for NFT, but its ingenious design feels that it can be used for long-tail ERC20 loans with a little magic modification. I don't know if there will be a project that does this.

Last look at stablecoins

In fact, there is not much to say about stablecoins. Last year’s Luna crash basically declared that “stable coins are dead.” Some time ago, Frax changed the mortgage rate to 100%, which is a complete solidification of this statement.

Although there are still new mortgage-type stablecoins appearing in the market, using various technologies to ensure that you will not be unanchored, but in fact, apart from pure stability, everyone is not very worried about your unanchoring problem. , the application scenario is king.

To insist on the few trends seen on stablecoins, they should be the following:

1. MakerDAO’s Endgame Plan, AAVE’s GHO, CRV’s crvUSD, SNX V3’s sUSD, and the BTC-based NUSD proposed by Arthur in the concept. I detailed these in a post two months ago However, it will not be repeated here.

2. Stable currency based on RWA - Ondo is currently working on this, and has created an OMMF stable currency, which is supported by MMF (money market fund) and US treasury bonds, which is something that is on the edge of the circle...

3. The development of LSD is in full swing before and after the Shanghai upgrade. I feel that there should be a stable currency based on LSDFI, such as Fork a Liquity, and the collateral is changed to stETH or a basket of LSD version of ETH, and then do some articles on the interest Kind of... I don't know if there are any projects that have already done this. If not, I guess there will definitely be in the future.