Bitcoin transactions are verified by “miners” using Proof of Work. This research note tells you what the problems are with Proof of Work and explores one alternative called Proof of Stake. I aim for a simple explanation for Fintech entrepreneurs and Bankers. To the 0.1% who have years of hard-core cryptographic experience please excuse my lite knowledge and chip in if I have got something wrong.
For newcomers to Bitcoin: In a proof-of-work blockchain, miners use their CPU power to solve a mathematical challenge and by doing so they “vote” on which transactions came at what time (call it a decentralized timestamp if you like); the more CPU power you have the proportionately larger your influence is. As a miner you are rewarded for this work by being given Bitcoins. The whole “Bitcoin revolution” is based on this Proof of Work.
The problems with Proof of Work is that it is:
- Expensive. Mining rigs are expensive as is the electricity to power them. This leads to the view that Bitcoin is Unsustainable in this Motherboard article, which has this staggering data point:
“a single Bitcoin transaction uses roughly enough electricity to power 1.57 American households for a day.”
- Slow. In the move to real time, Bitcoin is a snail. Blockchain.info has the data. TL:DR it is measured in minutes (which is an eon in consumer facing applications where more than a couple of seconds changes user behavior).
- Subject to some collusion danger.
GHASH.io, the Bitcoin network’s largest mining pool, has briefly directed over 50% of the Bitcoin network’s hashpower, theoretically giving it control over Bitcoin transactions. Translation; some group somewhere could steal your Bitcoins.
Introducing Proof Of Stake
Miners are referred to as an economic set. Proof of Stake does not have “miners” but the objective is the same, which is a group of people who form a consensus that a transaction is valid and make some money by doing so. So lets call both an economic set. A simple way to look at the difference between Proof of Work and Proof of Stake is that:
– In Proof of Work inclusion into the economic set is free & verifying a transaction costs money.
– In Proof of Stake, inclusion into the economic set costs money & verifying a transaction is free.
In Proof Of Stake, people tend to talk about “voting” rather then “mining”. Like all analogies, it goes wrong if taken too literally, but as the aim is to get to consensus that a transaction is correct, the analogy does fit.
In Proof of Stake the economic set vote with their capital (aka their stake) in the form of the internal currency of the platform you use (for example, Ether on Ethereum.
This 2012 paper by Peercoin is credited with inventing Proof of Stake (although as with so much innovation there is probably earlier work that served as inspiration).
Proof Of Stake in action
Proof of Stake is being used by systems such as:
Method: Delegated Proof of Stake
Method: Transparent Forging
Ethereum (Proof Of Stake in development, discussion here)
Different Proof of Stake Methods
- Transparent Forging. Although the node which forges a block is random in the long term, in the immediate future it is highly predictable. This means the network knows where the next block should be forged. If a node does notforge the block it is expected to (perhaps because it is working to build a fraudulent chain instead), it is excluded from the network for a period of time. The likelihood of that node being chosen is instead redistributed across the remaining members of the network.
- Delegated Proof Of Stake. The equivalent of a notary verifies signatures and timestamps transactions.
Lets now look those three problems arising from Proof Of Work and see how Proof of Stake could address them:
- Expensive. Proof of Stake does not waste any more electricity than any other peer to peer internet protocol (eg. BitTorrent).
- Slow. Depending on the precise algorithm in question it can potentially allow for much faster blockchains (eg. NXT has one block every few seconds, compared to one per minute for Ethereum and Bitshares claims 1 second).
- Subject to some collusion danger. In proof of stake, the threshold is 51% of the entire supply of the currency. As these new currencies have a low market cap that is not reassuring but all of the Proof Of Stake methods tackle this problem.
This falls into the wait and watch category (unless you are actively developing these platforms). Proof Of Stake is essential to Blockchain systems but it is not yet proven.