This weekend I moderated a panel on “The Interoperable Layers of Bitcoin” at SatsConf where the main discussion points were:
- Why do we need more layers of Bitcoin?
- Are Layer 2s (L2s) a zero-sum game or can we “grow the pie”?
- What is the future of Bitcoin L2s?
Why do we need more layers of Bitcoin?
Bitcoin is a great money, the king of store of value, but it has limitations in terms of scalability, privacy, and programmability.
In terms of scalability, Bitcoin is limited by a block size of 4MB and block time of 10 minutes. This translates to an average throughput of 3 transactions per second (TPS). Whereas Visa has 1,700 TPS on average and can handle 65,000 TPS during peak hours. Scalability is a fundamental limitation of Bitcoin, hence, it was the first driver for the creation of Layer 2 solutions. Lightning Network, for example, was created with the single goal of increasing the throughput of Bitcoin by reducing the on-chain footprint of transactions.
Privacy is another fundamental limitation of Bitcoin. Bitcoin enables pseudonymity, but not anonymity. Bitcoin is a public ledger, meaning that all transactions are visible to everyone. Transactions are linked to an address, but not to an identity. Lots of companies and services use on-chain data to probabilistically identify users, especially at the interface of Bitcoin and traditional finance (TradFi). Yes, you can use coinjoin, but it’s not a perfect solution because it leaks a lot of transaction information, such as inputs, outputs, and amounts.
Finally, in Bitcoin, you are limited by the programmability of Bitcoin Script. It enables you to do simple things such as: verify signatures, hash values, check equality, 32-bit addition, and basic locktimes. You don’t have multiplication or expressive arithmetic operations, global state, transaction introspection, or any other advanced features. This means that you cannot enforce complex business logic. Forget about zero-knowledge proofs1, or any other advanced features. If you want to change Bitcoin, good luck with that. I think the protocol is already ossified2 and the incentives are too strong to keep it that way.
The case of Bitcoin Finance (BiFi) as a counterpoint to DeFi
I want to focus on the third limitation of Bitcoin: programmability3.
One of the main upcoming use cases for Bitcoin that is currently not widely or properly offered in L2s is Bitcoin Finance (BiFi): the ability to do more with your Bitcoin. These things include Bitcoin-backed loans, yield farming, prediction markets, decentralized exchanges, Bitcoin-backed stablecoins, and more.
Yet, most of these solutions exist not only in TradFi, but also in DeFi, spearheaded by Ethereum. You can borrow against your crypto assets or become a liquidity provider in protocols such as Aave (32bi USD Total Value Locked) or Morpho (8bi USD Total Value Locked), stake in protocols such as Lido (30bi USD Total Value Locked), partake in prediction markets such as Polymarket (300mi USD Total Value Locked), or use stablecoins that have a 300bi USD Total Value Locked. It is important to note that these solutions are decentralized, permissionless, and self-custodial most of the time4. This means that you are in control of your funds and you are not relying on third parties to keep them secure.
As a counterpoint to DeFi, Bitcoin has been lacking a proper ecosystem of BiFi solutions. To be fair, some of these solutions have been built on top of Bitcoin, but they are not decentralized, permissionless, and rely on relinquishing custody of funds to third parties. This is far from the ideal situation we want for Bitcoin: uncensorable, permissionless, and self-custodial fuck you money.
You can bash ETH as a token as much as you want, but I’d much prefer to have better security assumptions than “trust me bro”.
What if we could have the best of both worlds? The best money in the world with the best of DeFi protocols and solutions? This is where Bitcoin Finance (BiFi) comes in.
The future of Bitcoin L2s
Given that changing Bitcoin is nearly impossible, even if we could, I think Bitcoin’s value would drop significantly because its value proposition is exactly its ossification and immutability.
With the use case of scalability, already being addressed by L2s such as the Lightning Network and Ark, we have left the privacy and programmability use cases to be addressed.
I think that privacy can be addressed with more expressiveness in Bitcoin, which somewhat unifies the use cases of programmability and privacy. Currently, we have some L2s that are trying to address programmability use cases, such as the Liquid Network and Rootstock. However, these solutions fall short in terms of security assumptions: they are majority federated multisig bridges.
There’s a new wave of L2s that are trying to address programmability use cases with much better security assumptions: 1-of- trust models. As long as you have one live honest operator, you can withdraw on-chain. Spearheaded by companies such as Alpen Labs5 and based on BitVM bridges.
In 2026, we will see a lot of these DeFi protocols coming to Bitcoin, enabling you to do more with your Bitcoin without having to relinquish the best money in the world for other tokens in search of more programmability.