Hacker News Reader

Comment by Animats | Hacker News Reader

Some big banks are gearing up to use a blockchain to track interbank settlements.[1] But this isn't a coin, nor is there mining. It's simply a shared ledger that can't be altered without cooperation between a majority of the parties. All alterations are clearly visible to all parties, which is good for auditing. Only big banks who have signed up for this have active nodes, and they are not anonymous.

This is an alternative to setting up a neutral party for the purpose, such as the Bank for International Settlements or Depository Trust Corporation. Those organizations are expensive to run. It's a real shared ledger, but it's no more than that.

[1] https://www.reuters.com/article/us-blockchain-banks/six-big-...

View on HN

I disagree that something with those properties is a blockchain. It's simply a distributed append-only ledger.

The exact same functionality could be accomplished with a git repository they each pull/rebase/push between themselves... or with any of a number of append-only databases with WALs and synchronous replication, where each of them runs a replica.

Since only the trustworthy bodies can join the party, they can do the job more efficiently by not using any of the trendy blockchain which may technically prevent the fraud but otherwise inefficient.

And since there are only banks can join the network, I doubt it can technically prevent the fraud among banks because unlike public cryptocurrencies, it does not give incentives to the banks to spend a lot of computational power to the network. So any banks allowed to join can put enough computation power than other banks combined and succeed the attack.

This seems easy to solve though. Couldn’t they just modify the transaction acceptance logic to require 100% agreement across the network?

At risk of breaking NDAs, this is exactly what is being done in deployments I’ve worked on. Regulators also have nodes.

A blockchain system will still work if a minority of the players are down or disconnected. When they come back, they resync to the majority view. That's a useful robustness feature.

I think you misunderstand how private blockchains work:

> "they can do the job more efficiently"

They ARE doing the job more efficiently by not using proof of work, and instead most likely going with a quorum approach based on public keys.

> it does not give incentives to the banks to spend a lot of computational power to the network

Right, because it doesn't use proof of work, it uses a quorum vote. The banks would still need to form a cartel in order to get a (super)majority of the vote.

For one, the use of the word blockchain itself is posing a problem. If I remember correctly there was a case made to use the word - distributed ledger instead to avoid confusion.

And I have never understood one part of it. How are contentious settlements are resolved? Today for any disputes people can approach BIS or DTC and once issue is resolved everything can be rolled back. Who acts as the "watcher" in this case? Surely, going by what majority says is not a solution.

I don't understand how they go around the confidentiality. If you are a trader at big bank A, the last thing you want is traders at big bank B and C to know everything you have in your book.

This probably has to do with payments the banks are making to each other, not between banks and clients. If banks are paying each other, there's going to be a nonzero cost associated with the accounting for those transactions, and this system would remove that cost.

What I didn't understand from the article was the point about membership. If the contributing banks are the only ones using it, why is it giving information about what _I_ could do with one of these USCs - apparently nothing because I won't be able to join the chain

> If banks are paying each other, there's going to be a nonzero cost associated with the accounting for those transactions, and this system would remove that cost

Wouldn’t it just amplify it? Right now, central banks just set up swap facilities between each other. It’s akin to opening a bank account with one another. Crediting, debuting and undoing anything nefarious is cheap and easy.

You are confusing transaction clearing with actual account settlements.

Banks want to use Blockchain as a more reliable alternative to SWIFT it won’t be doing actual account settlements which in your case you mentioned corresponding banking.

That said many of the more progressive institutions stopped calling it Blockchain a while ago they all are talking about distributed ledgers, hyper ledger and the likes are being either used or piloted by many institutions.

The ones that do use Blockchain often mean today actual crypto currency or a settlemen token a good example of a “real world” implementation would be Royal Mint Gold which is a crypto token offered by the Royal Mint which is pegged to gold.

http://rmg.royalmint.com

This captures what is Banks interest in blockchain in a nutshell

> Banks want to use Blockchain as a more reliable alternative to SWIFT it won’t be doing actual account settlements which in your case you mentioned corresponding banking.

> That said many of the more progressive institutions stopped calling it Blockchain a while ago they all are talking about distributed ledgers, hyper ledger and the likes are being either used or piloted by many institutions.

At least one commercial company I've heard of has a sophisticated permissioning model, so that the owner of the chain can publish it and have its causal history and/or integrity be validated by all participants, but with specific permission scopes so that subscribes to the blockchain can only see inside of transactions they have permissions for.

There's also research into cryptocurrencies issued by central banks, both for retail and wholesale between banks [0].

[0] https://www.bis.org/publ/qtrpdf/r_qt1709f.htm

It's interesting too that it will be much easier for them to agree on architectural changes when necessary to compensate for scaling or other issues.

This is a major issue for something like Bitcoin, but a small speed bump for banks utilizing agreed upon blockchain tech.. It's at least interesting to consider the benefits of centralized(exclusive?) guidance coupled with distributed execution.

"simply a ledger" that for some reason didn't exist until Satoshi nakamoto created Bitcoin.

While your comment is in good faith, I don't think the OP was trying to minimise the complexity of a distributed ledger. Instead they were trying to distinguish their mentioned use-case from the currency use-case.

Reply on news.ycombinator.com