A weekly recap of the most insightful news, analysis, and capital flows in the wild west we call crypto.

Hello and welcome back to the Web3 Rewind! Per usual in this industry, lots has happened this past week. Here's what we have in store for you:

• Presidential memecoins

• Dencun upgrade

• Coinbase convertible notes

• MakerDAO hikes borrowing rates

• London Stock Exchange makes ETN moves

The Latest

Vote, or buy memecoins

Hard fork dept.

The Dencun upgrade

The Ethereum Dencun upgrade is a go. On March 13th, the Ethereum network upgraded with several key improvements. The most crucial is EIP-4844 which introduced blobs to the network. This allows layer twos to post their data to the network through blobs, instead of competing for blockspace alongside mainnet transactions. Data posting costs make up 99.9% of an L2 transaction’s cost, and thus, this upgrade should significantly decrease the cost of using an L2.

For example, in 2023, Arbitrum spent nearly $47M posting call data to Ethereum. Similarly, Optimism spent ~$27M posting call data to Ethereum. These inordinately high data posting costs make it hard for these L2s to have a high margin. That changes with EIP-4844, which is projected to reduce data posting costs by 5-10x (there is a large range of estimates out there, and we’ll only know after a few days of EIP-4844 live). At this point you may be thinking, are L2s suddenly going to be much more profitable? Not necessarily. L2s will likely pass on the cost savings to users rather than benefiting from the improved margins.

One interesting point of EIP-4844 is that even if it goes live, this doesn’t necessarily mean that builders will include the blobs in their blocks. This is because blobs carry more data, meaning including them in a block carries higher latency. And with higher latency, builders can play less timing games and extract less MEV. Thus, with enough price volatility and MEV opportunities on Ethereum, it may perhaps very well be in a builder’s financial incentives to not include blobs in their blocks, which, well, defeats the whole purpose of blobs in the first place.

Other upgrades include EIP-4788, which allows beacon chain block roots to be readable at the execution layer, and EIP-1153, which introduces transient storage. This new feature allows smart contracts to “store” data within the span of a block or a few transactions. For example, if you needed to execute a bunch of sequential swaps in Uniswap, and each step had its own calculations, now Uniswap doesn’t need to store the transaction call data and output for each of these transactions, they can simply store the data in-between in a “transient” manner and only have to make sure that all balances match up at the end. In fact, this is exactly what Uniswap is doing with flash accounting in its V4.

This brings us to my next question, has DeFi innovation run its course? Something like Uniswap V4 fundamentally relies on the EIP-4844 upgrade to go live, which means that the launch of V4 fully hinges on the fact that the Ethereum Foundation can ship in time. We can call this platform risk, and unfortunately, I don’t think it’s an avoidable one, especially when the bulk of users and liquidity are still on Ethereum. Some of it depends on fundamental improvements to the base layer, which bring about key technological unlocks. On the other hand, you are still seeing new exciting ideas come to market week after week. — JC

Quick Bits

Coinbase announces offering of $1B in convertible notes

• On Tuesday, Coinbase announced a proposed offering of $1B in convertible notes. Someone’s taken a copy out of Michael Saylor’s playbook. The notes will mature in 2030 and be offered at a 1% interest rate.

• Initial purchasers of the notes will have a 30 day option to purchase up to an additional $150M principal amount, and the company will be convertible into cash, Coinbase stock, or a combination of both.

MakerDAO hikes DSR

• In an unscheduled emergency move, MakerDAO has voted to increase borrowing rates and the Dai saving rate (DSR). The DSR will increase to 15% from 5% and nearly triples borrowing rates for most types of collateral.

• This change was a direct response to Ethena, whose stablecoin, sUSDe has been taking crypto and on-chain money markets by storm. Without this, the DAI supply would have likely continued to decrease and the DAI peg would’ve been at risk.

London Stock Exchange to accept physically backed crypto ETN proposals

• The London Stock Exchange is willing to consider proposals for certain crypto investment products. The new guidance came from UK’s Financial Conduct Authority,

• Perhaps this is a mark of FOMO as London watches as US BTC ETFs gobble up an increasing amount of crypto investments, and the exchange will start accepting applications for BTC and ETH traded notes in Q2.

And last but not least

Jito and its mempool

Last week, in a shocking move, Jito decided to shut down its alternate mempool offering. I wouldn’t blame you if you're lost already, but I will break it down for you. Jito is a developer of MEV and transaction related client software on Solana. The mempool is a waiting zone where transactions sit idly while waiting to be picked up by validators and included in a block. Still following? Validators can pick these transactions and include them in their own block, but since the mempool is public, other entities can also view these transactions. That’s where searchers come in. Searchers are entities that aim to extract value from different types of transactions on-chain by inserting/removing/adjusting transactions.

So say you submit a transaction to buy 10 SOL. This transaction flows to the mempool, where a searcher will see that this is likely an opportunity for them to make money. What the searcher does is insert their own buy transaction before this user’s transaction and then insert their own sell transaction after the user’s transaction where the price has been pushed up. This is commonly known as a sandwich attack and is the most common type of MEV that exists today. (In trad fi this is called front running and is totally like illegal — editor). This is enabled because every order on an AMM is a market order, and a user sets the maximum slippage that he is willing to tolerate. If he sets the slippage to 5%, then a searcher will come in with his own sandwich attack, and the user will experience a 5% slippage at the profit of the searcher.

MEV like this is rampant on Ethereum and Solana, but it seems like Solana has a different set of values in mind. Swappers on Solana, and specifically Telegram bot traders, appeared to be sick of getting sandwiched unnecessarily. And thus, with the help of social pressure, Jito decided to suspend its mempool so that unknowing users would not get sandwiched anymore. This is a shocking move in that, ideologically, it stands on the opposite side of the spectrum compared to Ethereum. Mind you, this was a pretty costly business decision. Validator tips from the Jito mempool and block engine added up to $550M, of which Jito takes a 5% cut. Why would a profitable business ever make such a move? And specifically, a move that would be so hard to reverse as the community would push back if they ever decided to do so.

The obvious answer is that Jito is a lot more interested in growing the pie, rather than extracting value from it. But given the economic incentives that are present, it likely won’t take much before a competitor pops up to do the exact same thing. In addition, users now get better “execution” on their swaps, but this comes at the cost of worse transaction inclusion, as searchers have to revert to spamming the network to try to get their transactions and bundles included. Is social pressure the MEV end game? How much time did this really buy Jito? Is this a user problem or a mechanism design problem? Stay tuned, as the Solana community finally finds out how hard MEV is to solve. — JC

Have you read the definitive history of Ethereum? No? Well then get your copy of Out of the Ether while you can.