In the previous article, we have covered Proof of Work, whereby miners in the network have to put in work to validate transactions and add them to the blockchain. As compensation, they are awarded newly minted coins. In this article, we cover Proof of Stake (PoS), which aims to solve some problems that Proof of Work (PoW) has.
PoW uses up a tremendous amount of electricity to prove that work has been done. A recent study in 2017 demonstrated that Bitcoin’s PoW mining used up more energy than 159 countries. Initially, PoW mining was much cheaper; a small Raspberry Pi could be run 24/7 on minimal electricity and still be profitable, as the total computational power of the PoW network was small. As the network has grown exponentially since 2017, the network computational power has increased exponentially as well. This means that more and more computational power is required to even have a minute chance of ‘winning’ the right to add the next block, a problem that will only grow as the network increases in size.
This also leads to the issue of centralizing the computational power. When mining, it is more efficient to do it in countries with low electricity cost and latency; this leads to large mining pools being set up in China, a region that has both of these attributes. Nodes of the network located outside of these optimal regions are unable to compete as effectively, and might close down to prevent losses.
The equipment used to mine in PoW is also changing. An ordinary individual with commercial hardware (even a good gaming computer) would not be able to compete against specialised hardware dedicated to hashing. As such, more and more of these users would drop out of mining, leaving the mining process to a handful of large companies.
All of these have led to the increasing concern of a ‘51% attack’. In PoW, it is possible for an individual, or group of individuals, to own 51% of the computational power. By doing so, they could severely affect the blockchain in a few ways.
1. Prevent other miners from completing a block. Even if an independent miner were to find the next hash, the group could ignore it and build on blocks found by their own group. By doing so, the independent miner’s block would become orphaned, as it is no longer the longest chain.
2. They could edit information in the blockchain, especially the more recent transactions, which could allow them to spend the money, then reverse the transaction. This would enable them to ‘double spend’. The further back the blocks are, the harder they are to amend (or even impossible), but all new blocks since the 51% takeover could be potentially in jeopardy.
51% attacks have been pulled off in the past, such as Krypton in August 2016.
In response to this, Proof of Stake has been adopted by some blockchains as a consensus mechanism.
In PoS, there are no miners, nor are there new coins created when blocks are added to the chain. Rather, all the coins are created at the start, and distributed to investors as well as the team. Rather than using electricity to prove work done, the coins have a value based on investment through cryptocurrencies like Bitcoin or Ether.
In PoS, there exist ‘validators’ who take transaction fees as a reward for verifying transactions and appending them to the blockchain. Validators have to prove that they have a ‘stake’ by owning some of the cryptocurrency, which they then put up as a stake. Validators are then randomly chosen to verify transactions, and when they do so they receive a fee for their service.
How the validators are chosen is directly related to the percentage of the coins that they stake. For example, an individual who holds 1% of the total coins would have validate 1% of all the transactions.
Validators are rewarded proportionate to how much is staked, much like interest from a bank. This reward is linear, which means that owning 10% would allow the individual to validate 10 times more transactions than an individual who owns 1% of the currency. Compared to PoW, spending 10 times more on hardware would cause an exponential increase in returns, which leads to centralization among large players.
In PoW, each transaction is verified by the work that is required to hash out the next block. In Proof of Stake, there is no such requirement; the validators do not need to sacrifice any computing power or electricity. Rather, they are just guarantors that the transaction is valid, and they receive a reward each time that a block they verified is appended. This means that commercial hardware that regular individuals own are more than enough to be a validator. The protocol only considers their ‘stake’, not how much computational power they can provide.
Reduced ‘Bad Actor’ Incentive
The Proof of Stake model inherently discourages a 51% attack. To own 51% of the network, an individual or group would need to own 51% of the currency, which is a large investment. It is also exponentially more expensive to purchase, as the price increases as the demand goes up. By the time 51% of the currency is under the control of a person or group, they will be strongly discouraged from being a bad actor, as their value is dependent on the stability and reputability of the blockchain.
Also, tokens that are staked will be burnt if the actor acts against the best interest of the protocol. This means that actors are encouraged to continue to make passive income via transaction fees, to ensure that their stake is not risked.
Proof of Stake is a type of consensus mechanism that arose as a solution to problems that Proof of Work faced. Ethereum’s Proof of Stake, Casper, is scheduled to come up sometime this year, and other blockchains such as Qtum (price as of writing QTUM 27.12) have successfully implemented it without any issues.
That being said, Proof of Stake is not a single protocol that is applied to a blockchain. Each blockchain’s Proof of Stake has minor differences, such as how large a stake must be, or what is the penalty for verifying an invalid transaction (the whole stake, or a portion).
Some protocols also have ‘coin-age’, where the coins staked have to undergo a maturing process, and are only eligible for staking after a period of time. Once it has been staked, it has to ‘re-mature’, and will be taken out of consideration for staking for the time being.
All of these are measures to ensure that large investors do not have a disproportionate advantage compared to smaller investors, and the community is still tweaking Proof of Stake protocols to suit their needs.
Stay tuned for the fourth and final part of the Consensus series!