So, we already have Bitcoin and Ethereum, why do we need another public blockchain? To ask this question is equivalent to ask “We already have horses. Why do we need cars?”. None of the existing infrastructures scale well. Miners are burning more electricity power than Google’s all data centers combined, yet Bitcoin’s speed is only comparable to a computer in 80’s. There are certain benefits of a fixed protocol that once the foundation of Bitcoin is set, it stays unchanged. This restriction makes no room for arbitrary manipulations with the rules. However, the hard-coded 1mb block size has become prohibitive for Bitcoin to serve a wider group of users. The more serious problem with Bitcoin is its incentive mechanism. Mining reward declines while the mining cost increases, miners must charge higher transaction fees to cover their expenses. Bitcoin’s transaction fee has already surpassed $1, resulting its market share dropping to less than 80 percent. As Bitcoin is deflationary, miners will eventually cease to be able to earn block rewards and transaction fees shall become their only source of income. To see how much does a single transaction cost when the block reward drops to zero, let’s make a calculation:
Bitcoin’s current price sits at around $900. Rounding it up, we get that a miner’s block reward in US dollars is 12.5(current reward size)*900 = 11,250. A Bitcoin block size is 1mb, which cost is negligible on most modern computers. Now, let’s zero out the block reward. To make the transaction fees included by a block equal to the current block reward, we get:
11,250 / 4200 (maximum number of transactions in a block) = 2.678
Therefore, miners would need to charge a single transaction at least 2.68 USD on average to make the total transaction fees in a block equal to the current mining reward. As the power required for mining becomes even greater, miners must charge more transaction fees to compensate their spendings. Notice that transaction fees depend on the data size of the transaction, regardless the amount of coins being transferred. It’s too expensive to pay a few bucks to miners when you only want to purchase a train ticket or a sandwich. We can see that Bitcoin price fell sharply after it reached all time high early this year, as the rising transaction fees forced users to switch to altcoins.
Let’s make another calculation. If we want to lower the transaction fee to 1 cent, while still make miners happy by getting the same amount of income, then we get 11,250 * 100 (current block reward in cents) / 600 (block time in seconds), which equals 1,875 transactions per second. Not only this drops the transaction fees significantly, but also saves energy by increasing the number transactions, since most electricity during mining was spent on finding the hash, not processing transactions.
A lot of alternative consensus methods have been proposed so far, i.e. proof-of-stake, claiming to be more energy-saving along with other benefits. However, mining reward is the only fair method to distribute coins so far. Currently, the top 5 most are all proof-of-work based.
Block time Transactions per second
Bitcoin 10 min 7
Ethereum 12 sec 15
Dash 2.5 min 60
Monero 2 min 30
Litecoin 2.5 20
Although Ethereum has the shortest block time, exchanges usually wait until 250 confirmations to finalize a transaction (approx. 2hrs) due to less hash power spent in a single block. Most modern personal computers are able to process hundreds of thousands transactions within a second. Why cannot these cryptocurrencies simply increase their block size and let more transactions pass through? We will explore this mystery in the next article.