Discuss MEV technology trends and privacy track development trends

After the transfer of Ethereum to PoS, MEV has become more and more complex; two new designs have emerged on the privacy track.

Written by: Lao Bai

MEV is a technology-oriented topic, and it has become more and more complicated. Especially after the conversion of Ethereum to PoS, the non-user roles involved in MEV include Searcher, Builder, relayer, Validator, and Proposer. Basically, they do not engage in technology. Looking at the big head, there are various "ten thousand character long essays" series on the principle of MEV on the Internet. I will not repeat them here, but simply talk about the current status and trending technologies.

MEV Status

MEV clip robot (sandwich attack, also known as malicious MEV) often earns more than reasonable MEV (arbitrage and liquidation), and it is also the main income of many MEVs. However, some time ago, the malicious Validator used the relay vulnerability to replace clip robot transactions. The black-eating black incident that caused the robot to lose 25 million US dollars still caused a lot of waves in the circle, and it also made many traps restrained a lot.

In short, most of MEV’s profits go to Searcher and Builder, and those who do MEV protocol or technology can’t make much money. Flashbot is said to have no profit so far for MEV-Boost (but Flashbot is originally a non-profit organization). I saw at least 4 or 5 MEVs in a month, with different technical styles, but we didn't vote for any of them, because we really don't think they can really achieve stable profits (such as the earliest Eden).

Current technical trends of MEV

1. Smart Slippage Management: This is mainly for cross-chain MEV. Seeing that there are projects in progress, it saves users from manually setting slippage + anti-pinch;

2. Threshold Encryption: This Cosmos ecology is relatively good at it. Now Penumbra and Osmosis should be working on it. The transactions entering the memory pool are encrypted. Isn’t MEV basically useless?

3. Delayed Encryption Delayed Encryption: The threshold uses a style similar to multi-signature, and 2/3 of the verifiers decrypt it. If it is still unsafe (because the security assumption of the verifier committee is introduced), then use delayed encryption and let the encryption The information is set to be automatically decrypted after a certain period of time. This is mainly the application of VDF technology. It is still relatively early, and the performance is said to be not good;

4. SGX encryption: similar to the above two, but using trusted hardware, which is mainly done by Flashbot's SUAVE;

5. Fair sorting FSS: Outsource the sorting and hand it over to a trusted subject to prevent MEV, Chainlink is doing it;

6. MEV Auction: MEV auction, the method proposed by the OP, V God is said to like it very much, and it may be used as a solution for the Optimism decentralized sequencer in the future;

7. MEV-Share: Share the profits of MEV with users. This is what Flashbot has just created recently. If it had been earlier, the clip robot that lost 25 million is said to be able to get back about 18 million...

8. Mev-Blocker: This is done by Cowswap. By letting Searcher bid to Backrun your transaction (finally share 90% of your profits), Backrun (mainly arbitrage and liquidation) is relatively gentle in Mev, and users can avoid The larger damage FrontRun (run away) and Sandwich Attack (clip) has.

9. ETH protocol level PBS: The proposer and builder of the Ethereum protocol level are separated. According to the customary style of the Ethereum Foundation, there is a high probability that it will be after 2025...

New design for privacy track

Privacy has always been a track that I am not very optimistic about from an investment point of view, from the earliest Zcash to the later Tornado Cash to the current Aleo, Iron Fish and the like. There is no other reason. Although this thing is politically correct, it is not just needed for 99% of users. Except for hackers, a few institutions and giant whales, who cares about a local dog on Uniswap, a loan on AAVE, and a loan on Lido? Are you afraid that others will see this pledge?

Even if privacy is really needed, it should be a plug-in form like Aztec's Aztec-connect to serve this mainstream Defi protocol, instead of building a "privacy public chain" from scratch. With the closure of Aztec Connect (possibly I am afraid that the SEC will check the water meter, or it may be that it is really not making money), and my bearish sentiment on the privacy track is even worse.

At present, the following two new designs of privacy have been seen in the primary market.

The first category is based on Tornado Cash. The front end of Tornado has been banned. Its design is because it can help hackers launder money and keep the controversy of "political correctness". However, in fact, 80% of the funds in Tornado Cash are clean, and only about 10-20% are clean. Hacking caused money laundering, because there are indeed whales and institutions that need to use similar services to ensure privacy, and V God himself has used it.

So I have seen more than one project that wants to combine Tornado Cash with KYC. Since hackers are afraid of money laundering, wouldn’t it be OK to ensure that the money coming in through the KYC+ whitelist is clean money?

But this actually puts the regulatory risk on the KYC provider, and KYC is theoretically very easy to forge or buy. If hackers use this "Tornado Cash with KYC", once KYC is passed, it may be more convenient for illegal activities.

The second category is similar to the design of Namada in Cosmos, called Multi-Asset Shielded Pool (MASP). Multiple privacy-protected assets share an anonymous set. Coupled with the powerful interoperability of IBC, at least within the Cosmos ecosystem, it can theoretically Provide shielding protection for all mainstream assets on the Cosmos chain. For example, if you want to exchange Osmo and Atom privately, you can implement it like the diagram below.

I have also seen homomorphic encryption for privacy, but homomorphic encryption is too early, and its performance and operability have not reached the stage of "really usable". It is a bit like ZK around 2017 , To develop into a usable high probability, there will be a feeling of 5-10 years.

Let’s talk about this first today, and in the next issue, we will talk about the three major parts of DeFi: DEX, new trends in lending and stable coins.