There has been discussion about the motivations and benefits of moving to Proof of Stake. For a risk-benefit analysis of moving to Proof of Stake, what are the major risks?
The biggest risk is, as in Proof-Of-Work, centralisation. Centralisation of staking power in the 'hands of few' with the very deep pockets. To mitigate this risk, Ethereum founders decided not to engage venture capital on the early stages (and hopefully on any stages). Another measure was to hopefully create a diverse enough initial distribution of Ether in a pre-sale. In this blog post, published in the middle of the pre-sale, Vitalik Buterin tried to assess a 'fairness' of Ether distribution: https://blog.ethereum.org/2014/08/08/ether-sale-a-statistical-overview/
Another risk, which the researchers of Casper Proof Of Stake are dealing with, is potential presence of 'stuck' scenarios in the protocol, meaning that in the absence of some other validators and some other conditions, the protocol cannot make any further progress on validating new blocks and the networks 'freezes' (also known as Ice Age). To mitigate (or potentially eliminate) this risk, formal verification of the protocol is required and this is one of the tasks of the researchers.
There are also concerns about the behaviour of the protocol in the case of network partition, potential competition between validators to the point of harming the network.
The biggest risk to proof of stake is "nothing at stake". The "slasher" solution to this isn't tested and may not work. "proof of burn" is similar, possibly safer. But the bigger risk is not switching. Ethereum contracts may be too expensive if a switch isn't made.
More relevant now (from Casper 101):
Adverse Selection – Given the potential draconian penalties in Casper, it may skew the validator pool to people with a high-risk profile (either in or out of protocol). For example, we may observe a higher % of Byzantine actors / malicious validators.
- Response: Casper's parameters are being optimized to handle the tradeoff between draconian penalties vs the loss aversion of honest validators. Also, the increasing total deposit level for any given economic/network activity level will increasingly discourage various attack vectors.
Wealth Concentration – While less regressive than PoW, the best that PoS can offer is proportional (as opposed to progressive) wealth distribution. This does mean larger amount (even if theoretically the same % returns) of the ETH validation returns will accrue to validators with more at stake, which are likely more wealthy individuals/groups.
- Response: Without the reputation and identity problems solved, it is likely impossible to have a safe and fair progressive wealth distribution for miners/validators. In fact, moving from regressive (i.e. higher % returns to larger sums of capital) to proportional ("dollar-is-a-dollar") would be a non-trivial milestone to celebrate for securing a public blockchain. That is what Casper PoS aims to do as the first step towards counteracting increasing wealth inequality.
The size and adoption of Ethereum – Because Ethereum has (1) much more economic value to lose if something goes wrong and (2) many more stakeholders with more serious needs than say a brand new protocol that can start with PoS, Ethereum will necessarily develop more carefully than a new protocol.
- Response: While it is ambitious and daunting to move a large PoW-based network to a PoS-based network, moving to PoS was the intent since the early days, and one of the guiding principles of the project. The community members were very much aware of Ethereum’s plans to switch to PoS (i.e. refer to Ethereum Ice Age — the PoW chain difficult adjustment to encourage migration to PoS — which is commonly known in the ecosystem). That said, the changes will be rolled out in multiple phases (testnet, various clients, hybrid PoW/PoS) to ensure a smooth and successful transition.
- More complex to implement
- The weak subjectivity requirement (search for that phrase here)
- Removes the option of mining as a "one way anonymous decentralized exchange" with which anyone can theoretically get coins to pay for a few transactions
- Can lead to increased stakeholder concentration
- Removes the option of mining as unpredictable leader election, which improves PoW protocols' DoS resistance (that is: you don't know who the next miner is until they make the block, so you can't DoS them)