I agree with your sentiment. I do not see a scenario where financial institutions would agree to put their transactions into a distributed ledger with copies of the data (even encrypted) held by their competitors. If the financial institutions limit the ledger to computers within their control, what would be the benefit over a clustered database environment (I am assuming that most large financial institutions already have these).

The problem is without blockchain you don't have a meaningless buzzword to sell bullsh*t to banks. Blockchain is a vital technology for this.

Cryptocurrency is the future but I've never talked to more snake oil sales people than when talking to "blockchain" companies. I marvel at people's ability to sell nothingness

I worked in a consulting-type company for a bit and the most valuable lesson was that, generally speaking, nobody knows anything about anything. The most valuable institutions in the world have completely clueless people managing their tech and the ignorance only gets worse as you go up the chain.

It all boils down to "nobody got fired for buying IBM". Your blockchain initiative is only to get you noticed enough to get promoted or hired to a better position somewhere else. By the time anybody realizes how worthless it was you're long gone.

"nobody knows anything about anything". Actually, somebody knows, but they work for someone else.

I don't know much about blockchain, but the salespeople of big consultancy firms like Accenture are master snake oil merchants.

When you combine the two it must be like a perfect storm of bullshit.

> what would be the benefit over a clustered database

With a blockchain, the database and the revision history would be part of the same artifact. Not sure how advantageous this is, however.

Banks spending billions on complex technical processes could save billions if they replace those systems with a new system and you ignore all costs associated with implementing and running that system. Well yes, but..

They also cite "run some processes, like finance reporting".

How would storing this data in a blockchain have less overhead than an http server? Or even a torrent?

It seems like a hammer looking for a nail.

The discussion was around trustworthy data, so I think they want proof that a page served up and stored was in fact the actual page served up and stored, for later audit. HTTP doesn't give you that. Bittorrent doesn't give you that. But blockchain can, if you pay a fractional coin to store the hash of the page in the chain forever. With a reference to the page and to the transaction, you can then later determine whether the page is intact or has been modified. And for bonus points a signature could be stored as well.

This is just passing the overhead cost to whomever is mining the network though.

The trustworthiness of a blockchain comes from it's ability to repel attacks with computing power, with more power meaning more trustiness. This means tremendous overhead for a blockchain that could just as easily be performed with the centralized system we have now.