Building on Bitcoin

Overview

While the innovation of Bitcoin and its use of blockchain technology is groundbreaking, network scalability challenges persist as interest, usage, and adoption continue to grow. In this brief synopsis, we look at both Bitcoin’s historical and current efforts to scale the network, why these efforts matter, and where they might take Bitcoin in the future.

Setting the Foundation

Before we get into which Bitcoin scaling solutions have been proposed and which are currently being used, it is important to understand what scaling solutions are, why they are important, and what problems they seek to solve. The ability to scale effectively is important to any monetary system to support the demand and scope of that system for all who desire to participate in it.

Gold has been used as a monetary good for thousands of years, but historically, transactions were difficult to settle across space and time as moving substantial amounts of gold was cumbersome, expensive, and challenging to secure.

So how did gold manage to effectively scale? Instead of settling all transactions in physical payments that were difficult to execute, especially across borders, bankers began accepting gold as deposits and producing deposit notes that represented a claim on the gold stored at the bank. These notes could be traded, exchanged, and distributed much more easily and freely than physical gold could previously, and allowed for a much faster and more efficient way of executing payments between willing participants.

Much like the gold monetary system that existed long before it, Bitcoin has faced similar challenges in terms of its own ability to scale as a payment system. While Bitcoin can send payments much more easily, efficiently, and inexpensively in its current state than gold, Bitcoin still faces its own scalability challenges as an increasing number of users look to adopt bitcoin and use the network. Each block can only hold up to a theoretical limit of 4MB of data, or transactions, meaning users must choose between paying higher fees or enduring longer wait times for their transactions to be confirmed in new blocks, especially during periods where demand for block space and transaction volume on the network is higher.

In its current state, Bitcoin can process up to seven transactions per second.¹ If volume on the network is higher than this threshold, then some transactions will have to wait until the next block before being added to the blockchain. As of July 23, 2023, the average confirmation time for a transaction to be included in a newly mined block and added to the blockchain is about 318 minutes² and the average transaction fee is around $1.37 per transaction.³ These are simply averages and users can lower their confirmation times by paying higher fees to have their transactions included in blocks sooner. These numbers are subject to change based on network volume and periods of high network activity can cause longer wait times for transactions to be picked up and included in blocks, on top of higher fees. For comparison, Visa can currently process around 1,700 transactions per second on average.⁴ When compared to existing legacy payment systems, it is easy to see how Bitcoin in its native form may require outside solutions to help make the network more scalable as a payment system. The slower transaction speeds, longer settlement confirmation times, lower throughput capacity, and higher fees at the base layer make Bitcoin difficult to scale in its native form.

Historical Scaling Efforts

A natural question to wonder is why Bitcoin does not increase its block size—would that solve the scalability problem by increasing the amount of data and, therefore, transactions that could be included in each block? Not exactly, but this has been lobbied for and attempted in the past by those who believed in its ability to be a scalability panacea. The reason for this data limit on Bitcoin is intentional and derives its value from two of the core principles of the Bitcoin protocol: security and decentralization. The limit prevents the Bitcoin blockchain from growing too rapidly, a byproduct of which would prevent individual users from running their own node and being able to fully validate the blockchain themselves. In turn, this could harm Bitcoin’s decentralization and create a permissioned environment where only users with copious amounts of capital could run a node.

The Blocksize Wars

Industry participants and Bitcoin community members alike exploded in heated debate between August 2015 and November 2017 over what the maximum amount of data allowed in each Bitcoin block should be, colloquially known as the Blocksize War. This war culminated in a hard fork of the Bitcoin protocol, a change to the Bitcoin code that is not backwards compatible. In other words, users had to choose whether to adopt the new code or stay on the original code. The result of this hard fork was a new blockchain protocol dubbed “Bitcoin Cash,” which is currently only a fraction of the size (in terms of market capitalization) of its originator, Bitcoin or “Bitcoin Core.”

These forks and their proposed changes to the core ruleset never caught on with a majority of Bitcoin community members because they were less secure due to reduced interest from miners and validators. They were also more centralized with a higher degree of permissioned participation as the higher startup costs of running a node with a higher capacity for data storage prevented those with less wherewithal from participating. Most network participants did not see the tradeoff of greater scalability as adequate to give up the coveted and superior characteristics of security and decentralization and, therefore, stayed on the trusted Bitcoin network running the robust Bitcoin Core software. These forks relied on having enough demand and interest from existing Bitcoin users to switch to a new network; however, most chose to stay with the proven network that emphasized security and decentralization over scalability.

Efficiency Improvements

Attempts to increase Bitcoin block sizes have not been the only efforts to natively implement scaling solutions on the Bitcoin base layer. Segregated Witness, colloquially shortened to SegWit, was a Bitcoin upgrade implemented in 2017 that aimed to prevent the problem of transaction malleability. Transaction malleability is an attack that lets someone change the unique ID of a bitcoin transaction before it is confirmed on the Bitcoin network, potentially allowing someone to pretend that a transaction did not happen. Additionally, SegWit enabled the launch of the Lightning Network, a layer 2 scaling solution built on top of Bitcoin that allows for faster transactions and lower fees, which will be explored later in this piece. Before SegWit was implemented, the Lightning Network was not attainable because it relied on unconfirmed bitcoin transactions that were vulnerable to attack if transaction malleability was possible. The SegWit upgrade was only a soft fork of Bitcoin, meaning it updated the ruleset of the protocol in a backwards compatible way. Soft forks are essentially a tighter constraint of the ruleset as opposed to hard forks, which are an expansion or loosening of the ruleset.

SegWit helped the network scale by increasing the number of transactions that could be included in each block, while allowing users who did not update their software to still participate on the network, simply without the added functionality of the SegWit upgrade. This backwards compatibility supported the decentralized nature of the Bitcoin network because it allowed users and node operators to decide whether they wanted to adopt the new software or not, while not being forced to do so by miners or any centralized entity, and did not prevent those unwilling to upgrade from participating on the network. Additionally, the SegWit upgrade was a launchpad for another soft fork known as Taproot, which was introduced in 2021. Taproot introduced several new features, most notably, enabling lower transaction fees.

Current Solutions

Today, there are several prominent off-chain scaling solutions built on Bitcoin. These are made possible through the use of alternatives known as layer 2s and sidechains. A layer 2 is simply an off-chain solution built on top of the base layer (layer 1) that aims to improve upon the shortcomings of the layer 1, such as reducing transaction bottlenecks, while still relying on the base layer for settlement finality. In other words, the work that would be performed by the main chain is outsourced to the second layer. These layer 2 applications then post the data back to the first layer to be confirmed in the blockchain ledger and history when necessary. The most promising current Bitcoin scaling solution using a layer 2 implementation is the Lightning Network, which we will unpack in more detail below.

The Lightning Network

The Lightning Network is a layer 2 payment protocol built on Bitcoin’s base layer. The Lightning Network’s aim is to facilitate faster transactions at lower fees for users and is widely recognized for its potential to effectively scale Bitcoin as a payment system. It achieves this by opening payment channels between two parties off of the main blockchain (layer 1). Lightning transactions only return to layer 1 once these channels are closed. Put simply, one can think of a lightning channel as opening a tab at a restaurant. No transactions are recorded outside of the restaurant. Until you close your tab, Visa and your bank do not yet need to be involved, reducing your “on-chain” transactions to two (open tab and close tab transactions), even if you kept the tab open for 1,000 internal transactions inside the restaurant.

Furthermore, multiple channels can be combined to connect payments between parties who do not have a direct connection. Going off of the restaurant example, this means users can pay for others’ meals by forwarding their connections, creating somewhat of an internal network of connections. Since transactions taking place on Lightning are not recorded on the layer 1 blockchain, and subsequently are not competing for block space, Lightning payments are extremely fast and cheap.

Strike, a Lightning Network payment services provider, has made payment in bitcoin possible for over 400,000 merchants through a partnership with Shopify, NCR, and Blackhawk Network.⁵ Using Bitcoin as a payment rail via the Lightning Network, fiat payments are converted into bitcoin and bitcoin is converted back to fiat for the merchants. This partnership enables merchants to accept bitcoin as an alternative to credit cards, incurring next to no fees compared to the average 1.5% to 3.5% transaction fee associated with most credit cards.⁶ This integration is just one example of how Bitcoin has been able to use the Lightning Network to provide faster and cheaper transactions to everyday users, commercializing bitcoin transactions for both retail and business purposes. With a median fee rate of 0.0029%⁷, the Lightning Network is 1,000 times cheaper than that of Visa or Mastercard and presents the unique ability to revolutionize payments for merchants through the use of this technology. As bitcoin adoption and usage grows, block space will become more constrained and demand for payment alternatives will grow alongside it, underscoring the critical role the Lightning Network may have in the future of bitcoin transactions.

The Lightning Network prioritizes the facilitation of smaller payments on the Bitcoin network and, therefore, does not come without its own limitations. There may not be a limit to how many channels can be opened, but each channel has its own capacity constraints. As of July 23, 2023, there are currently around 16,000 nodes⁸ and ~70,000 channels on the Lightning Network,⁹ with the average channel capacity sitting around 7.2 million satoshis, or around 0.072 bitcoin (roughly $2,085) at current market prices. The total Lightning Network capacity is around 4,900 bitcoin or roughly $144 million at current prices.¹⁰

As more Lightning nodes open payment channels, the overall network capacity increases and is capable of handling more transactions. The Lightning Network has increased by almost 500% in its capacity since 2020¹¹, a strong signal for its potential as a viable layer 2 scaling solution capable of facilitating payments on Bitcoin in a faster and cheaper way than previously possible. Theoretically, the Lightning Network can handle millions of transactions per second (TPS), a vast improvement from the Bitcoin base layer’s seven TPS. This underscores the Lightning Network’s incredible potential as a layer 2 scaling solution for Bitcoin as it achieves faster transaction speeds and a favorable fee market for users with limited or acceptable tradeoffs to decentralization and security.

The Liquid Network

While the Lightning Network chooses to scale through the use of a layer 2, the Liquid Network aims to scale Bitcoin through the use of a sidechain. A sidechain is a separate blockchain network that connects to the main chain via a two-way peg. The two-way peg functions similarly to someone exchanging dollars for tokens at an arcade. In this instance, the participant exchanges their dollars for tokens that are specifically designed for use only within the arcade environment. When the player is finished, they simply exchange their tokens back into dollars that are then usable outside of the arcade. Sidechains rely on the Bitcoin blockchain for data and security, but the transaction execution takes place on a separate blockchain. Like layer 2s, sidechains aim to make the most of Bitcoin’s ultimate security, while allowing for more block throughput and cheaper transactions.

The Liquid Network is a sidechain protocol built on top of the Bitcoin blockchain. It was created by blockchain technology company Blockstream, but is governed by a federation of parties and operated on an open source blockchain platform called Elements.¹² The Liquid Network seeks to improve upon several shortcomings of the Bitcoin base layer, such as providing faster transactions and settlement times with a focus on larger payments. While Bitcoin blocks are added to the blockchain roughly every 10 minutes, Liquid blocks are added every minute.¹³ Because of this feature, Liquid can support a larger quantity of transactions per hour than the Bitcoin base layer. Chain reorganizations are not allowed and two confirmations are enough for settlement finality. The Liquid Network also supports multiple assets, allowing third parties to issue tokens, such as securities and stablecoins, on the sidechain. Transaction confidentiality is another coveted characteristic of the Liquid Network since transactions take place off of the public ledger, supporting privacy for users. While the Liquid Network still supports retail investors, it primarily serves institutional investors, exchanges, and other enterprise clients who desire a higher level of privacy, liquidity, and transaction speed. Like the Lightning Network, Liquid seeks to scale Bitcoin payments by making them faster and cheaper. However, in contrast to the Lightning Network, the Liquid Network focuses on larger institutions and payments rather than smaller payment channels.

Despite the faster transaction speeds and higher degree of privacy, one notable drawback of the Liquid Network is its reliance on a centralized federation to facilitate the two-way peg. This solution is only a viable scaling alternative as long as users trust the federated actors to behave in their best interest. Participation requires users to lock up their bitcoin in a peg-in transaction before they are issued a new synthetic form of bitcoin on the sidechain known as L-BTC. To peg-out and receive their actual bitcoin back, a user must first deposit their L-BTC before the Liquid Network will send back their bitcoin.

To initiate a peg-out, a Liquid Federation member must burn an equivalent amount of L-BTC on the Liquid Network. This burning process is verified by the Liquid Network's consensus mechanism and, once confirmed, an equivalent amount of BTC is released from the Liquid Federation's multisig wallet on the Bitcoin mainchain. The BTC can then be sent to any address on the Bitcoin network. Peg-outs can only be initiated by Liquid Federation members. This is done to ensure that the BTC held on the Liquid Network is always backed by an equivalent amount of BTC on the Bitcoin mainchain. The burning process ensures that the L-BTC being pegged out is not double-spent, and the multisig wallet ensures that the BTC released from the peg-out is always under the control of the Liquid Federation.

This structure may present an opportunity to transact in bitcoin in a faster and more private manner. However, the trust required to rely on a centralized entity to act honestly on behalf of its users and not freeze, steal, or censor payments starkly contrasts to the characteristics of security and decentralization that Bitcoin users have historically been averse to forfeiting.

Rootstock (RSK)

Rootstock, also known as RSK, is another sidechain scaling implementation that, like the previously mentioned solutions, also aims to increase Bitcoin transaction speeds. Unlike these other proposed solutions, RSK promotes programmability that enables developers to create decentralized applications on Bitcoin. Like the Liquid Network, RSK achieves this through a two-way peg. When bitcoin is transferred to RSK, that bitcoin gets converted into RBTC, or smart contract-enabled bitcoin. The RBTC can then be transferred back to the Bitcoin blockchain at any time for associated RSK fees and Bitcoin transaction fees. With Bitcoin as its foundation, RSK seeks to offer enhanced interoperability, making it an option for developers looking to build on top of the Bitcoin blockchain.

Aiming to overcome the high costs and slow settlement of bitcoin transactions on the base layer, block confirmations on RSK take about 33 seconds¹⁴, much less than the time required for the Bitcoin blockchain. This results in significantly faster times for users transacting bitcoin with the bonus of reduced costs. Transactions on RSK can be confirmed, bundled, and sent to Bitcoin’s base layer for final settlement, increasing Bitcoin’s throughput and expanding the network’s capacity to host many more users, applications, and transactions.

RSK’s ability to allow DeFi application developers to create protocols, NFTs, or other digital assets using the Bitcoin blockchain seeks to enable Bitcoin to become a digital asset ecosystem in a similar capacity to Ethereum. As of July 23, 2023, there is $43.04 billion of TVL across all DeFi protocols with over half of the market share (around $25 billion) belonging to Ethereum.¹⁵ TVL, short for Total Value Locked, is a metric that captures the amount of cryptocurrency that is currently staked or deposited into DeFi protocols. Since Rootstock can provide transaction speeds about 10 times faster than that of Ethereum¹⁶ and 50 times cheaper gas fees¹⁷, there could be significant potential for the sidechain to capture some of this market share. This is especially important as fees have continued to rise since Bitcoin and Ethereum were first introduced, and RSK's approach aims to lower these fees for users, while still allowing them to participate in DeFi in the Bitcoin ecosystem. Where RSK differs in its approach to scale the Bitcoin network is its focus on bringing programmability into the Bitcoin ecosystem, which has historically been a tough challenge to solve. Ethereum has the first mover advantage of being the first widely adopted programmable blockchain as well as having stronger network effects from both users and developers. So, while RSK may appear to have a more favorable fee environment and the ability to facilitate faster transaction speeds, its general lack of interest from developers and adoption from users present a challenge for the protocol in capturing market share relative to Ethereum.

The RSK network is secured by a group of validators who are chosen by the RSK community and these validators are required to stake a large amount of RSK tokens to create a financial incentive to act honestly in their efforts to secure the network. So, while RSK is novel in its approach to scale Bitcoin in a developer-friendly way and is more decentralized than a true federated sidechain, it shares similar sidechain centralization risks to that of its Liquid Network counterpart, again forcing users to decide if the tradeoff of security and decentralization is worth the added speed, functionality, and favorable fee market.

Stacks

Most scaling solutions are designed as either sidechains or layer 2s that are built on top of Bitcoin. However, there are other alternatives actively being developed that seek to operate outside of this traditional scope. Stacks is a separate Bitcoin layer 1 blockchain that enables DeFi, NFTs, and other decentralized applications to be built on Bitcoin. Stacks was founded by Muneeb Ali and Ryan Shea under the original name Blockstack in 2017 and transformed into Stacks in 2018.¹⁸

Stacks has its own nodes, network, miners, and native coin. Stacks uses a unique consensus mechanism called proof-of-transfer (PoX) and it is this consensus mechanism that allows Stacks to settle all transactions on Bitcoin. Stacks miners commit bitcoin to the Stacks network by locking up bitcoin to mine stacks blocks and are rewarded in the form of STX tokens. Contrary to the previous iterations of Bitcoin sidechains mentioned in this paper, Stack’s coin (STX) is not pegged to on-chain bitcoin. Instead, the connection to Bitcoin is established by Stacks using the Bitcoin base-layer blockchain as a reliable storage and broadcast medium, meaning that everything that happens on Stacks, such as the transaction history compiled in Stacks’ blocks, is recorded on Bitcoin itself, and Stacks’ entire state is settled on Bitcoin. In other words, to create a Stacks block, a Bitcoin transaction must be initiated on the Bitcoin blockchain itself. This is achieved through hashing a respective Stacks block, ensuring that the Stacks block is anchored within a block on the Bitcoin blockchain. This way, Stacks’ entire transaction and state history is represented on the Bitcoin base layer.

Stacks is like RSK in that it prioritizes programmability on Bitcoin, but differs in that it uniquely uses its own decentralized protocol rather than a sidechain operated by federation members. However, it’s unclear how truly decentralized Stacks is. The more bitcoin a miner is willing to lock up on the Stacks network, the higher the chances are for that miner to win a block and receive the associated STX rewards and transaction fees. This disproportionately favors wealthy bitcoin holders to assume greater control over transactions included in Stacks blocks, leading to concerns over censorship. Despite being decentralized in name, large bitcoin holders and miners can theoretically monopolize the process of building and submitting Stacks blocks in practice because they can simply outbid smaller miners who are competing over the same blocks. Some have pointed out this may already be happening, creating a lot of uncertainty around the viability of the protocol’s current implementation and casting doubt over its ability to effectively improve in future designs given its current state’s ambiguity.¹⁹

Outlook

Looking at the current landscape surrounding Bitcoin scaling solutions, it appears that the Lightning Network will likely continue to gain the most traction and legitimacy based on growing usage and adoption. Integration with Strike has helped to provide more onramps for merchants to accept bitcoin as payment for goods and services and the growing Lightning Network capacity has allowed bitcoin payments to be made faster and cheaper than ever before. We expect that this growth will only continue going forward and expect the Lightning Network to reign supreme as the most legitimate scaling solution built on Bitcoin over the near term.

While the Liquid Network has also shown promise from the standpoint of facilitating speed, low costs, and privacy, its centralization poses a direct contradiction to the core values of trustlessness and security that the Bitcoin base layer promises. The novel approach of Rootstock and Stacks to bring DeFi onto Bitcoin is admirable. However, like the Liquid Network, the ambiguity surrounding their true levels of decentralization forces users to decide on whether they are willing to forgo core values of the Bitcoin base layer for promises of scalability and programmability.

Time and individual preferences will reveal which Bitcoin scalability solutions could garner the greatest usage and demand. Despite each Bitcoin scaling effort having its own native approach and structure, it is encouraging to see solutions being actively developed to address the fundamental problems of slow speeds and high transaction costs in the most established digital asset protocol by market capitalization. This does not have to equate to a winner-take-all end game; rather, a collective effort to scale the network with each proposed solution offering its own unique approach. It is encouraging to see Bitcoin's native limitations around speed, throughput, and programmability being addressed with these proposed solutions and we eagerly look forward to closely following their development and advancements as this robust digital asset ecosystem continues to materialize.