Can the market resist the seductive narrative of Layer 2 solutions?

Photo by Maria Teneva on Unsplash

Has crypto winter come and gone? Based on price action, one could be forgiven for starting to believe that this has been the case. And as always, we’re open to that possibility. To be sure, recession risks are real, but as we’ve learnt over the past decade, and especially so over the past two years, facts are nothing without context, and context relies almost entirely on stories.

Underestimating the power of stories and narratives in the market can become a fatal mistake: look no further to the gap between the economic reality post covid and the financial reality in markets for lessons on how fundamental analysis would have made for a rather unprofitable strategy in the face of a global pandemic. The meme was “money printer go brrr” leading “numba go up”, and like it or not, the market comprises humans, humans love narratives and narratives are best encapsulated in the memes of their generation.

Nowhere is the impact of memes more pronounced than in the world of crypto, and we are always on the lookout for developing narratives. As it happens, notwithstanding the collapse of hedge funds and liquidity lockouts, crypto has a handful of strong narratives that could go well for it: for starters, a Fed Pivot (actual, perceived or anything in between) fuels “money printer go brrr” = “numba go up” as a meme. To that, add “The Merge”, taking Ethereum to Eth 2.0 and the promise of cheaper transaction fees and higher speeds, potentially heralding in another golden age in crypto (especially with better usability).

On top of those, a new narrative seems to be emerging around scalability solutions on Ethereum, specifically “Layer 2” rollup solutions. As always, this is not “new news” – rollups have been in existence for more than a year now, but as we know, narratives have to do more than just “exist”. They need to be accepted into the mainstream by a sufficient critical mass in order to gain traction.

While the rollups/scaling solutions story has been around for a long time, last year saw the crypto world being obsessed with “alt L1s”, alternatives to Ethereum that echo the “Ethereum Killers” of the previous crypto cycle. Those were the flavours of 2021, just as 2020 saw “DeFi summer”.

This summer, it seems like the L2s are coming to town.

Will it last? Is it another flash in the pan? Maybe, but if DeFi Summer and Alt L1 season were “flashes” in the pan, those are some pretty huge flashes to miss.

There is much that can be written about the technicalities of L2s and we won’t go there, but it suffices to say that as frequent users of the decentralised compute that Ethereum offers, we (personally) enjoyed the L2 experience much more.

That tangible difference in user experience counts, and that could be the difference between a flash in the pan and an enduring step change in technology.

What’s an L2?

While the temptation is strong to get into a long essay about the technicalities of L2s, we’ll try to keep things simple here. The aim is as always to get everyone up to speed at a decent enough level, following which we’d love to hear your thoughts (whether you agree or disagree) on the matter.

Going back to the underlying state of play, crypto’s largest functioning chain is Ethereum, which we’ve written about in some detail over a year ago. While widely accepted to be the godfather of all programmable money chains (Bitcoin, unfortunately despite being the first cryptocurrency in existence, is rather lacking when it comes to functionality and programmability beyond just transferring Bitcoin from one wallet to another), Ethereum is acknowledged to be the exemplar of decentralisation. However, that comes with downsides: transactions at peak volume are slow and expensive, and prohibitively so. While fantastic as a proof of concept and a minimum viable product, if Ethereum were to be the future of the decentralised web, it needed to speed up. A lot.

It was this angle that many of the competing Layer 1s, the successors to the “Ethereum Killers” from the 2017 era, took to jostle for power. From Solana to Avalanche to everything else in between, great technical solutions were offered that could offer cheaper, faster transactions, promising to bring crypto to the masses. Some of these solutions could yet prove to be winners, but one major issue hindered their race for dominance: the bulk of value in crypto, outside of Bitcoin, was sitting on Ethereum. The “whales” were on Eth Layer 1 (the base layer) and accessing that value and using that value to support more value on these new L1s chains required Eth L1 value to be carried over to the new chain, while preventing the double spending problem.

While bridges could be built, each of these bridging contracts constituted a major hack risk that would undermine the backing of the bridged assets on the other side of the chain i.e. being able to compromise the security of the bridge contract and withdraw the assets believed to be “locked up” on one side of the bridge meant that there were assets on the other side which were no longer backed by a deposit.

Furthermore, many of the alt L1s tend to exhibit much greater degrees of centralisation than something like Ethereum: the number of validator nodes is an order of magnitude smaller than Ethereum (Binance smart chain has 21 vs north of 10k for Ethereum). BSC is much cheaper and faster, but by construction it is also easier to hack 51% of 21 than 51% of 10k+ and create a fraudulent transaction.

These risks were further highlighted by Ethereum co-founder Vitalik Buterin in one of his replies to an AMA. And as we are starting to see, these warnings were not unfounded – the most notable bridge contract compromise was that of Axie Infinity’s Ronin bridge. In this case, the devs raised a funding round and made everyone good on hacked funds, but the question is would this ALWAYS be the case?

So, is it the case that Speed/Cost/Security will always be an impossible trinity? Not entirely.

A Layer 2 solution, as its name suggests, is built as a second layer on top of an underlying chain, the Layer 1. In simplified terms, Layer 2s solve the speed, scale and transaction cost problem by batching transactions and settling them all in bulk, as a single cryptographic proof. Put differently, instead of having every individual transaction written onto the chain, requiring the validators to do a large amount of work that is expensive and time consuming, the transactions are batched on the Layer 2 chain, and a summary of these transactions is what gets periodically written onto the main chain.

Sounds complicated? Fret not, for an analogy always helps: between baking one cake in the oven at a time, including a full pre-heat, as opposed to baking 5 cakes at a go in the same oven. Sure, the question is “wouldn’t you need a much bigger oven in that case?” – there’s the beauty of code, for code doesn’t have physical size. Hashing transactions creates a mathematical summary (the hash – yes, grossly simplified, apologies crypto geeks) that is unique for each batch and combination of transactions in a block. It is then this summary hash that gets posted onto the main underlying chain.

The elegant solution here is (as Vitalik suggests) the L2s inherit the security characteristics of its underlying chain: while the L2 itself might get hacked, because the underlying chain is Ethereum, ownership of hacked funds is at least secured by the largely immutable hash on the much more secure L1.

Not a new thing

Writing about L2s now may make L2s seem like a new thing that the crypto world is tinkering with. In reality, it isn’t new at all, at least in the context of what “new” means in crypto land.

On Ethereum, the two most prominent L2 rollups are Optimism and Arbitrum. These are known as “optimistic” rollups, which operate on a system where validators on the L2 rollup “optimistically” accept transactions that are posted as genuine, at the risk of footing the bill and making users good with a reserve of ETH that they put up if transactions are found to be wrongly validated. In return, they are paid fees which, though orders of magnitude smaller than on L1 Ethereum, add up to quite a bit given the large quantity of transactions that can happen.

Arbitrum is a name you might think sounds familiar, because it is: it is the L2 on which TreasureDAO built its projects, from Bridgeworld to Smol Brains. And speaking from personal experience, operating on Arbitrum was an absolute joy.

Low cost and high speed aside, the biggest source of satisfaction was not having to buy another new type of token from a centralised exchange to send in to fund transactions. Unlike something like AVAX, MATIC (Polygon) or SOL (SOL not even being an Ethereum Virtual Machine compatible chain i.e. Ethereum code written in Solidity doesn’t work there), where a new token needs to be purchased to pay for transactions, operating on Arbitrum simply required ETH to be bridged from L1, and then business as usual. Fees are paid in ETH, everything works just like L1, just cheaper and faster.

This actually sets us up for another debate on whether native tokens are always necessary or if they turn into an impediment to user adoption – but we’ll leave that for another day.

If you’ve made it till here and are thinking that maybe the optimistic rollup model is a bit precarious to be economically interesting for a distributed group of validators to run, you wouldn’t be the only one. Running the risk of being slashed for malicious attacks is quite the responsibility to take on in return for fees that hopefully come – perhaps that’s why Arbitrum, for example, runs most of its nodes internally. Centralisation? Maybe. They claim they don’t plan to launch a token – many are skeptical, but perhaps they really don’t need one.

To wrap up the L2 state of play summary, one cannot forget the “other” way.

Next step: ZK

For anyone that’s spent enough time perusing articles about crypto, the two letters “ZK” would seem mildly familiar. Once again, this isn’t a “new” technology. ZK = Zero-knowledge, specifically zero-knowledge proofs.

Zero knowledge proofs first came into existence long ago in crypto’s first iteration, in the form of privacy coins. One of the big points of obsession by the early crypto community was about anonymity, but there was never going to be full anonymity if the entire chain was publicly viewable down to the transaction level.

One solution that was proposed was a zero-knowledge proof. The token ZEC (aka Zcash) was one of the early manifestations of ZK proofs – again, at risk of gross oversimplification, a ZK proof offers a way to mathematically prove that something is true/false without having to know the actual details of what’s inside. ZK proofs come in two forms, ZK Snark and ZK Stark, but the risk once more is that we get sucked into a black hole discussing these so we will leave these for another time.

The key to know here is that ZK rollups that hash back to Eth L1 are way more secure and hack resistant that their optimistic cousins, and that between Stark and Snark, the preference is for Stark because the initial set up of the encryption hash doesn’t require the trust of the party setting it up. And yes, for the REAL ZK geeks out there, all of this goes into the trash if quantum computing happens – but so does all of crypto anyway.

The main reason why ZK rollups, despite all their promise, haven’t caught up in terms of adoption is that the tooling for developing code on them is relatively underdeveloped. Specifically, the deployment of Ethereum’s Virtual Machine (EVM) hasn’t been as widespread on ZK rollup chains – but that is quickly changing.

Businesses like Starkware have been on the move, building developer tools that allow all of the complex calculations and functions that exist on Eth L1 to also happen on ZK rollups. With ZK rollups, it is also the case that the MORE transactions happen per block, the cheaper the cost of each transaction within – economies of scale at its best. Other ZK rollup deployments include the likes of Immutable X (on which their Gods Unchained game is built) and Polygon (aka MATIC) deploying a ZK EVM solution on their chains.

Open season

All things considered, the promise of an extension of Ethereum that has all of the upside and none of the bad stuff, that is delivered cheaper, faster and better than its incumbent version, is definitely one that is tantalising.

How this changes post Ethereum Merge is something to watch – presumably with the canonical chain being accepted to be the new Eth 2.0 hard fork chain, these rollups will follow, but if enough miners hang around on the PoW chain (Eth 1.0, let’s call it), perhaps that too has a future.

As always, things are rather fluid in the world of crypto. This is still a narrative that is gathering steam, but as we know, life in crypto is always “gradually then suddenly” on steroids.

If there’s going to be an L2 summer, we definitely don’t want to be missing it.