A quick recap
While there is some debate as to what are the essential properties of sidechains, I usually go by the definition of "a sidechain is a blockchain with a distributed two-way pegged currency from other blockchains". Generally, something like Credits or BitBasket is aiming to do, but not what Liquid is currently offering.
Sidechains are useful as they:
- Move some transaction volume off the main blockchain
- Allow extra functionality on the sidechain not available on the main network
- Allow the transfer of value back onto the main chain without the use of centralized or decentralized third parties
A permissioned ledger is a centralized or decentralized (but not distributed) blockchain ran by one or more parties where the access to the network or various functions on it is gated to only the approved parties. Here an some overview of how the technology compares to traditional, distributed databases.
A permissioned sidechain
If we relax the definition of a sidechain to include any currency (crypto-native, IOU, etc.) on any network (centralized, decentralized, distributed), we can create an interesting sidechain-bankchain combination that would be useful for an international settlement system.
The reason why we'd like to utilize a model like this would be to allow nations and big international organizations to:
- Settle between one another on a global network
- Have autonomy over their national / corporation networks
- Allow for private settlement networks to operate, while still allowing for proof of solvency audits on the main network
- Compartmentalize regional transactions from trans-regional tranasctions for speed and network throughput while still allowing for easy interoperability and global settlement
If the above system was put into place, we probably would see a lot of companies big and small want to get onto the network. If the system would be anything like the current banking system, it would be unlikely that everyone would have the capital or meet other arbitrary requirements to connect directly to the main chain. However, there is nothing stopping us from designing the system with that in mind and perhaps having side-side chains - the biggest companies would connect directly to the main chain, while the smaller companies could connect to them and so on. This way we could have everyone on the same network while separating the peers on the network based on their size and needs (for better or for worse).
Lastly, we could add inter-chain settlement protocols like something Interledger is proposing. This would allow for direct connections between various sidechains without the need of going directly to the top level chain to increase throughput and decrease cost.
Tying it together - how would it work?
Now that we have some overview of how the network might be structured, lets explore a few ways it could work.
The top, global chain would be best served as either a multi-party permissioned ledger (like Eris or Multichain), or a distributed network (like Ethereum or Ripple). This way more participants are likely to join without seeing this as "the USA network" or "the Eurozone network" if it was a more centralized solution developed and controlled by one nation or company.
Ideally, the top chain would be where the various governments and big entities would track their debt / IOUs. This would give a clear insight into who owes who how much and allow lower-tiered chains to use that as base monetary system.
Lower-tiered chains would probably be either permissioned or completely centralized blockchains or other cryptograhy-based networks (like Open Transactions). They would be linked with the main chain through a two-way peg. This would allow for easy settlement between the sidechain and the main chain without completely relying on the chain custodian to forward all of the transactions back and forth by themselves.
The sidechains could also follow some safety mechanisms of the voting pools - being constantly audited for solvency and allowing anyone with a balance on the sidechain to redeem their underlying balance on the main chain according to the protocol.
If you wanted to connect to the network, you could do so by connecting to any of the existing peers on the network - usually some bank or corporation. After that one integration, it would be possible to send money to anyone else on the network easily (and hopefully cheaply).
Is this a good idea?
At the moment, I'm not sure how much of this idea would be useful when implemented in the real world. It seems that a lot of banks and institutions are interested in the blockchain technology, the concept of sidechains is a good way of segmenting the network transactions. Moreover, the entire idea seems similar enough to the way things work nowadays that it might be attractive to the companies from "the old world". That being said, I'm not sure if there are some hidden complexities in the proposed solution that would impair it in some way - a lot of the technologies mentioned are either still in development or are still in the conception phase. So for now I would categorize this as "an idea worth considering" and see where things might go from here.