Proof-of-Stake is a method to achieve a distributed consensus. It makes it possible to secure a blockchain by providing the right economic incentives to push for honest transaction validation. It is an alternative to the Proof-of-Work used on Bitcoin.
What are the specific features of the Proof-of-Stake?
The Proof-of-Stake (PoS) replaces a mechanism based on computing power with another based on the active use of its capital. It uses the assets of the blockchain in question, for example Tezos (XTZ) or EOS (EOS) on the smart contract platforms Tezos and EOS.
As a result, it is not possible to simply plug in and start mining as with Proof-of-Work (PoW). You have to buy it or earn it in some way. So PoS blockchains always start with a fund raising in order to make an initial allocation of the native assets.
The holders of the asset in question have the possibility to put into escrow part or all of their capital with the goal of participating directly in the validation of the blocks or, otherwise, voting for an actor who will do it for themselves according to the implementations of this system.
The different implementations of the Proof-of-Stake
There are several implementations of the PoS. Here, we will talk about the well-known ones, i.e. direct PoS systems where you are invited to validate the blocks if you have the necessary capital, and delegated systems where you vote for an actor who will validate the blocks for you). There are other hybrid implementations that will not be covered in this article.
These implementations generally seek to build a robust system by promoting a strong level of decentralisation and allowing anyone, even unidentified actors, to participate in the validation of blocks. From this point of view, we are getting closer to what qualifies the Proof-of-Work in Bitcoin as a “censorship-resistant” system: in order to bring the system down completely by repressive measures, a global attack on all the nodes would be necessary, which is difficult because of their number and the fact that some are not identified.
The mechanism is as follows: the holders of the assets can put part of their funds into escrow. The larger the amount of funds in escrow, the greater the probability of being chosen to create a specific block in the future via an algorithm managed by the protocol.
The selected participant will ensure the legitimacy of the transactions he wants to push on the network and create a block without having to find a hash of the rare block to add it to the blockchain as in PoW.
Indeed, for example, the hash of the Bitcoin block n°571729 is 000000000000000000000000000000110574374f4977cba59e315492589c78451cc2252b894. It is a digital fingerprint of the block that “translates” the information contained in the block, including the transactions added to the block, the hash fingerprint of the previous block, as well as numbers called Nonce that are changed at each iteration of the hash function to create a new hash fingerprint at each test in order to find a valid hash. This one must have a smaller value than the target given by the mining difficulty, which is what makes a valid hash start with a large number of zeros.
Finding this rare hash that requires computing power is the guarantee of the system security, because finding this type of hash for an invalid block requires a lot of electricity to be spent and the block would not be accepted by the grid: a waste of time and money.
Back on the PoS: if the block is valid, the volunteer will receive as a reward the monetary creation planned for the creation of this block as well as the transaction fees of the block, exactly as for the consensus in PoW.
The other participants also have the responsibility to verify that all transactions are correct and that the volunteer is not trying to cheat by spending the same assets twice, for example. If this was the case, depending on the implementations, the cheat will be reported, with several consequences: the block is invalidated, and the funds in escrow can be destroyed, redistributed to the informant, or both. This punitive system is called slashing: it will be present in Ethereum 2.0 and is already active in Tezos.
With slashing, each block validator has a double interest in doing its job properly: a carrot, in the form of a reward obtained when the block was created, and a stick, which is the direct threat of losing the funds in escrow in the event of proven cheating. This replaces a mechanism based on computing power with another based on the risk of losing capital.
PoS blockchains require a minimum number of units of the native asset to be stored to participate in block validation. For example 32 for Ethereum 2.0 or 10,000 for Tezos. Some allow smaller holders to participate by delegating the right to participate in their assets, such as Tezos. Once Ethereum 2.0 will be launched, we will be able to see which of these two models promotes the healthiest distribution of validator nodes.
DPoS, for Delegated Proof-of-Stake, which we find in EOS, Tron, or Ark blockchains, also allows holders of the native assets of these blockchains to store their funds, but this time in order to vote for an identified block validator who will be responsible for validating the blocks for them.
This model isa less decentralised approach with a fixed number of validators: 27 on Tron, 21 on EOS, and 51 on Ark. This limited number of validators theoretically allows to manage more transactions than on classic Pos, because blocks must be shared with fewer people and the conditions to be elected require sustained performance and sophisticated equipment, so the blockchain can afford to be heavier and the time between blocks shorter.
Also, the fact of being able to identify the validating actors is a trade-off that also seeks to fight against certain attacks that can be mounted against Proof-of-Stake systems.
The DPoS trade-offs are important for the investor to understand, as actors holding a very large share of the assets can easily elect the validators of their choice, or even elect between each others. A fairly centralised circle of entities which then monopolises almost all of the monetary creation, which accrues to them for each validated block.
Early assessment of the Proof-of-Stake systems
As of today, the blockchains using the Proof-of-Stake are young. The first implementation was NXT in 2013, but all the protocols we are talking about here started to be developed from 2014 and became operational between 2016 and 2018.
The Proof-of-Work is considered more robust because it has stood the test of time without a hitch for more than 10 years with Bitcoin, and its properties make that even with a colossal 51% of the network’s computing power, the possible attacks are very limited.
The Proof-of-Stake is an interesting and promising model, it is very low in electricity consumption and has some advantages over PoW depending on the implementations. However, these new systems will have to prove their worth by showing that in practice the potential weaknesses can be solved.
A main difference with Proof-of-Work is the ability for an individual to create multiple versions of the blockchain at any height of the blockchain and at zero cost because the protocol does not require this famous valid hash. We talk about “Nothing-at-Stake”. Choosing the longest chain as in PoW is no longer a sufficient condition to recognise the legitimate chain.
This can be problematic in the case of forks because the validator does not have to choose between one of the two versions of the blockchain and his interest is to work on both versions simultaneously to be sure to get his reward. Implementations generally choose slashing methods to force nodes to choose a single version of the blockchain by punishing validators who would push two blocks to the same height.
It can also create problems for new nodes connecting to the network or nodes that have been offline for a long period of time. They must then be able to determine which blockchain is the valid blockchain, without being able to trust the longest chain. This presents attack vectors such as “Long-Range attacks”, which are to be taken into consideration and that we have not yet solved.
Proof-of-Stake is taking a more prominent place in the ecosystem every day and is an important topic, especially with Ethereum, which wants to migrate to this system. This validation method could offer an alternative to PoW in the future, but it will take several years for its robustness to be recognised or for other alternative models to emerge.