3) “Bitcoin Isn’t Scalable”
A common criticism of Bitcoin is that the number of transactions that the network can handle per 10 minutes is very low compared to, say, Visa (V) datacenters. This limits Bitcoin’s ability to be used for everyday transactions, such as to buy coffee.
In fact, this played a key role in the 2017 hard fork between Bitcoin and Bitcoin Cash. Proponents of Bitcoin Cash wanted to increase the block size, which would allow the network to process more transactions per unit of time.
However, with any payment protocol, there is a trade-off between security, decentralization, and speed. Which variables to maximize is a design choice; it’s currently impossible to maximize all three.
Visa, for example, maximizes speed to handle countless transactions per minute, and has moderate security depending on how you measure it. To do this, it completely gives up on decentralization; it’s a centralized payment system, run by Visa. And it of course relies on the underlying currency, which itself is centralized government fiat currency.
Bitcoin, on the other hand, maximizes security and decentralization, at the cost of speed. By keeping the block size small, it makes it possible for people all over the world to run their own full nodes, which can be used to verify the entire blockchain. Widespread node distribution (over 10,000 nodes) helps ensure decentralization and continual verification of the blockchain.
Bitcoin Cash potentially increases transaction throughput with bigger block sizes, but at the cost of lower security and less decentralization. In addition, it still doesn’t come anywhere close to Visa in terms of transaction throughput, so it doesn’t really maximize any variable.
Basically, the dispute between Bitcoin and Bitcoin Cash is whether Bitcoin should be both a settlement layer and a transaction layer (and thus not be perfect at either of those roles), or whether it should maximize itself as a settlement layer, and allow other networks to build on top of it to optimize for transaction speed and throughput.
The way to think about Bitcoin is that it is an ideal settlement layer. It combines a scarce currency/commodity with transmission and verification features, and has a huge amount of security backing it up from its high global hash rate. In fact, that’s what makes Bitcoin vs Visa an inappropriate comparison; Visa is just a layer on top of deeper settlement layers, with merchant banks and other systems involved under the surface, whereas Bitcoin is foundational.
The global banking system has extremely bad scaling when you go down to the foundation. Wire transfers, for example, generally take days to settle. You don’t pay for everyday things with wire transfers for that reason; they’re mainly for big or important transactions.
However, the banking system builds additional layers of scalability onto those types of settlement layers, so we have things like paper checks, electronic checks, credit cards, PayPal, and so forth. Consumers can use these systems to perform a large number of smaller transactions, and the underlying banks settle with each other with more foundational, larger transactions less frequently. Each form of payment is a trade-off between speed and security; banks and institutions settle with each other with the most secure layers, while consumers use the speedier layers for everyday commerce.
Similarly, there are protocols like the Lightning Network and other smart contract concepts that are built on top of Bitcoin, which increase Bitcoin’s scalability. Lightning can perform tons of quick transactions between counterparties, and reconcile them with Bitcoin’s blockchain in one batch transaction. This reduces the fees and bandwidth limitations per small transaction.
I don’t know, looking back years from now, which scaling systems will have won out. There’s still a lot of development being done. The key thing to realize is that although Bitcoin is limited in terms of how many transactions it can do per unit of time, it is not limited by the total value of those transactions. The amount of value that Bitcoin can settle per unit of time is limitless, depending on its market cap and additional layers.
In other words, suppose that the Bitcoin network is limited to 250 transactions per minute, which is low. Those transactions could average $100 or $1 million, or any number. If they average $100 each, it means only $25,000 in transaction value is performed per minute. If they average $1 million each, it means $250 million in transaction value is performed per minute. If Bitcoin grows in use as a store of value, the transaction fees and inherent limitations prioritize the largest and most important transactions: the major settlement transactions.
Additional layers built on top of Bitcoin can do an arbitrary number of transactions per minute, and settle them with batches on the actual Bitcoin blockchain. This is similar to how consumer layers like Visa or PayPal can process an arbitrary number of transactions per minute, while the banks behind the scenes settle with larger transactions less frequently.
The market has already spoken about which technology it thinks is best, between Bitcoin and others like Bitcoin Cash. Ever since the 2017 hard fork, Bitcoin’s market capitalization and hash rate and number of nodes have greatly outperformed Bitcoin Cash’s. Watching this play out in 2017 was one of my initial risk assessments for the protocol, but three years later, that concern no longer exists.