Bitcoin Scaling to the Moon
The Need for Scaling Became Painfully Apparent to the Bitcoin Community in 2017
The scaling discussion took center stage during the 2017 Bitcoin bull market along with many heated debates on the future direction of the Bitcoin protocol. Back then there was just not enough room in a transaction block for everyone to rush their bitcoins onto an exchange to cash in when the price spiked. Since then the protocol has embark on a scalability path starting with the deployment of Segwit along with the publication of the Lightning Network protocol, implemented through multiple open source software projects, and the more recent Bitcoin improvement proposals for the combined implementation of Schnorr signatures and Taproot which will set the stage for further performance and privacy improvements.
Those are great news but there is still work ahead and it doesn’t take much effort to figure it out: the Bitcoin blockchain layer in its current form cannot scale for the masses. It is not an issue now but if we are to push the Bitcoin experience to the limit, the numbers speak for themselves: Approximately 2,500 transactions are validated every ten minutes, and this adds up to approximately 131M transactions per year. It becomes obvious that this volume is far from enough to handle the transaction volume we see in the retail industry, let alone everything else such as salaries, utility bills, mortgages, loans, international remittances and one last one, but not the least, IoT micropayments.
Lightning Network — The First Step for Bitcoin
To begin addressing this limitation, the Lightning Network (LN) was proposed as a network layer adjunct to the Bitcoin blockchain. In a few words, LN works by creating a network of multi-signatures, jointly funded bitcoin transactions, also called channels, between participants (nodes) supported by a communications protocol. Once a channel is created, its participants can send and receive bitcoins and keep a tab of the balance of the jointly signed transaction without the need to interact with the Bitcoin blockchain. The more participants connect to the network the more paths exist to send bitcoins between an ever-growing network of participants. The transactions taking place across the LN remain governed by the protocol implementation and game theory to enforce compliance across all nodes participating in the network. The participants can choose to close a channel, and there are a few variants of how this gets done but this is most frequently done through a jointly agreed-upon settlement transaction that gets broadcasted on the Bitcoin blockchain.
Channel Factories — A Giant Step for Scaling
The current LN implementations are steps in the right direction, and they do a lot to reduce the on-chain traffic back and forth between users, but the LN protocol does not yet address the issue of onboarding the masses onto the network. Let’s run some rough numbers here to illustrate the impact of the current Bitcoin protocol on the number of LN channels that can be added to the blockchain. To keep it simple, let’s assume for a moment that all the on-chain transactions are used up for creating LN channels. This would give room for 131M LN channels per year. Let’s also assume that most of the world population will open only one channel and it will be with one of the major LN hubs. Overall, this is a very optimistic onboarding scenario and even then, it will still take 60 years to open channels for everyone while not closing any of them.
Fortunately, the combination of the upcoming improvement proposals for Schnorr signatures and Taproot together with the concept of Channel Factories presented in a paper published by Conrad Burchert, Christian Decker, and Roger Wattenhofer in 2018, with improvements proposed by Alejandro Ranchal-Pedrosaa, Maria Potop-Butucarub, Sara Tucci-Piergiovannia in 2019 might be enough to scale orders of magnitudes beyond the current blockchain limitations.
A channel factory essentially works by bringing together a large group of users under a single shared account (i.e. an on-chain multi-signature transaction) to open an LN channel. This is a massive scalability improvement over the two-party channel creation model. What is of interest here is that within just one newly created LN channel, sub-channels can be opened for payments, regularly balanced, and closed, etc., all without having to go back on-chain. Participants to a channel factory can also participate in other factories thus acting as gateways between channel factories contributing even more to the scalability and efficiency of the model.
Intuitively, if large numbers of participants can join a channel factory all at once, it directly contributes to reducing the pressure on the blockchain layer by that much. To use my previous back-of-a napkin calculation, while it might take 60 years to onboard everyone using two-party channels, it could take a year using transactions made up of 60 participants per channel factory. This is all theoretical, but my goal here is to illustrate the massive positive impact channel factories would have on Bitcoin scalability and how it would eventually support mass adoption.
Where Do I Join a Channel Factory?
At the time of writing, the Channel Factory concept is still at the white-paper stage, it is debatable whether those factories will be instantiated mostly by business entities that will provide factory setup services and factory overlap capabilities for a fee or if they will be created through a to-be-defined distributed automatic factory setup process. This could turn out to sound like the discussions people are having today about distributed exchanges (DEX) vs custodial exchanges. For now, this is an area that still needs to be explored, even conceptually. I would think that because of the many similarities between sub-channels and LN channels, services and research we see today for LN are likely to be leveraged for channel factories. Below are some areas of development that will probably emerge from the initial concept:
- Third-party channel factories set up/subscription services. Would this trigger AML/KYC rules and other financial regulation compliance?
- Public channel factories created by some form of distributed, permission-less matching engine(s).
- Liquidity and network optimization rules. Are some factories better than others for the network?
- User interface and user experience and how transparent these advancements will be to the end-users.
Some Thoughts on a Possible Adoption Path
Presuming that most on-chain activities are from holders accumulating and lending bitcoins and traders moving their funds on and off exchanges and chasing arbitrage opportunities between exchanges, it then seems plausible that they would be the initial large user group of LN channel factories. It is worth mentioning in this case that the channel factory paper also introduces the concept of LN cold wallets within a factory. This is also likely to resonate well with these early adopters.
Another area of adoption could be catalyzed by the move of large retailers to add Bitcoin to their existing payment options. For as long as Bitcoin remains volatile, most retailers are going to convert bitcoins to fiat at the time of purchase through the services of payment processors and in this scenario, I can see the benefit for setting up channel factories between retailers and payment processors working together with crypto exchanges. Again, exchanges could also play a role here by providing overlapping channel factories.
Beyond these first use cases, my guess is as good as yours but the future of Bitcoin scaling just keeps looking promising. While the LN protocol is gaining traction through maturing implementations, I look forward to seeing the first implementations of channel factories and the associated business models that will come along with it.
References
Nakamoto S. 2008. Bitcoin: a peer-to-peer electronic cash system. See https://bitcoin.org/bitcoin.pdf.
Poon J, Dryja T. 2016. The Bitcoin lightning network: scalable off-chain instant payments. See https://lightning.network/lightning-network-paper.pdf.
Conrad Burchert, Christian Decker, and Roger Wattenhofer. 2018. Scalable funding of Bitcoin micropayment channel networks. See https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6124062/.
Github.com, initial commit by bcongdon. 2017. Implementations of the Lightning Network Protocol. See https://github.com/bcongdon/awesome-lightning-network#implementations
Pascal J. 2018. Lightning Network Economics and business models. See http://www.naturalfinance.net/2018/02/lightning-network-economics-and.html
Alejandro Ranchal-Pedrosaa, Maria Potop-Butucarub, Sara Tucci-Piergiovannia. 2019. Scalable Lightning Factories for Bitcoin. See https://eprint.iacr.org/2018/918.pdf
