Bitcoin relies on a system called Proof of Work to ensure consensus and security on a blockchain. So do other cryptocurrencies such as Peercoin, NXT, Nubits, Qora and Bitshares, But some strongly believe that Proof of Work is wasteful and unreliable, and instead implement an alternative system called Proof of Stake.
How does Proof of Stake work? By offering the chance to validate a block of transactions, and to receive the corresponding reward, to holders of the currency in question.
In Proof of Work, those most likely to validate a block are those with the most computing power. Taking control of the blockchain is, then, a question of churning computations, and would be prohibitively expensive. Work = cost. That is how Proof of Work secures the blockchain, by making it too expensive to retroactively change, and too difficult to control going forward. Consensus is understood to be the chain with the most work behind it, ie. with the greatest number of validated blocks (technically it is possible to have a sequence of blocks with a relatively low level of work behind them, but it is rare).
In Proof of Stake, holders of the underlying currency “deposit”, “pledge” or “bond” an amount, in exchange for the right to validate blocks. Generally, the likelihood that they will successfully validate a block is in proportion to the amount deposited. Security is achieved by the high cost required to control the majority of the network (a validator would have to hold over half of the market capitalization!). Consensus is achieved by the assumption that stakeholders have a strong interest in the health of the network. If trust disappears because of suspected bad behaviour, the value of the currency will crash and the manipulator’s holdings will be worthless. With Proof of Stake, trust becomes a self-fulfilling prophecy.
Also, Proof of Stake in theory is more democratic. With Proof of Work, influence tends to concentrate in the hands of those with the most powerful computers. Not everyone has the wealth to purchase or the skill to maintain that level of equipment. With Proof of Stake, the validation can be done on any computer. The investment required is in the actual currency itself.
Yet Proof of Stake in its simplest form is not conducive to reaching a consensus, since there is no cost associated with mining on a chain. In Proof of Work, if you mine on the wrong chain, you lose the amount that you invested in doing that work (= the cost). In Proof of Stake, it doesn’t matter which chain you try to mine on. You’re depositing an amount of currency, not incurring a cost. If it turns out that you’re trying to mine on the wrong one, you lose very little. In fact, you could theoretically mine on several chains at the same time, since there is no additional cost for doing so. This makes consensus harder to achieve.
And it will obviously lead to increasing concentration, not ideal for a decentralized concept. Why increasing concentration? Because if those that have the highest stakes are more likely to receive the newly issued coins, then their stakes will become even greater, which will make them even likelier to receive the newly issued coins, etc.
So, the currencies that use it have solved these weaknesses by tweaking and adding features, often ending up with a hybrid system that includes some Proof of Work characteristics.
One of the earliest examples of Proof of Stake was PPCoin, subsequently called PeerCoin, in which miners process blocks by submitting a stake. They do this by sending to themselves a chunk of their own coins. Only PeerCoins that have been held for at least 30 days can be used for this, and the longer they have been held without being used (up until 90 days), the higher the chance that block production process has of being successful. Once used, the stake has to sit idle for 520 days. This system ensures that the minting of new coins does not concentrate in the hands of a few participants. The consensus chain is the one with the highest “consumed coin age” behind it. Peercoin also allows for Proof of Work mining as an alternative, but this is being phased out as Proof of Stake becomes more important to the network.
NXT was the first 100% Proof of Stake currency. Block validators are selected at random based on the amount of the currency they hold, and everyone knows who the next miner is going to be. This makes double-spending very difficult, as it the whole network will be able to see if a transaction occurred or not. NXT does not offer fresh coins as a reward for validation – all 1bn coins were created at launch. Block validators focus on maintaining a healthy network, which will increase the value of their stake.
Bitshares uses a derivative called Delegated Proof of Stake, in which wallet holders elect 101 delegates who carry out the voting on which transactions get validated. These delegates take turns producing a block every 10 seconds, in a random manner. This is a less decentralized system than simple Proof of Stake, but more manageable.
Ethereum, the second largest cryptocurrency by market capitalization, currently uses Proof of Work, but plans to move over to a Proof of Stake variation some time in 2017. The twist that Ethereum plans to put on the concept is that validators have a “stake” in the outcome. They stand to lose out if they mine on the wrong chain. To earn the right to try to mine, participants submit a deposit, and are then invited to “bet” on which block will be validated next. Yes, you heard right, you “guess” (presumably in an experienced and insightful way) which block will be the next one to be included in the chain. If you guess right (= if you bet well), you get a reward. If not, you lose your bet. This will make consensus naturally easy to achieve – everyone sees where everyone else is concentrating, and converges on that chain.
As you’ve probably noticed, securing a network and identifying consensus in a decentralized public network that is not controlled by any one entity, is not at all simple. Both systems – Proof of Work and Proof of Stake – are totally ingenious, even though they both have their flaws. Will one turn out to be much better than the others? It’s way too soon to tell. Proof of Work has served Bitcoin well over the past seven years, but the cost and the centralization are becoming serious issues as the profitability of mining falls. Will it withstand the test of time? Proof of Stake has yet to find the magic formula that combines efficiency, security and decentralization. But that doesn’t mean that it won’t happen. We are still in the experimentation phase, launching ideas into the wild and seeing what adaptations and unexpected consequences the users come up with. And the cryptocurrency sector may well end up converging on something totally different. What is most likely, though, is that we will end up with an ecosystem that supports and nurtures combinations of what we have now. And it will be very interesting to see if we can reach a consensus on consensus.