Hacker News Reader

Comment by retube | Hacker News Reader

With respect i disagree. T+2 is not because of technological limitations of traditional databases or trust issues. Instantaneous settlement is possible (for example in the UK we have instantaneous inter-bank transfers), but it is not desired. Having delayed settlement means you can net offsetting transactions, it means you can handle outages in the infrastructure, it means you can unwind and correct errors.

View on HN

Thank you for this. It's frustrating to hear people making comments who don't have the fundamental understanding of why/how the markets work and are just upset about not being able to join in on the melee and assume it's nefarious behavior. My POV is that it was obviously a risk RH took to not make it more clear what the behavior of people's accounts would be when market volatility is high. But I worked for a Commodity Risk Management Company on their custom derivatives platform and 90% of what we did was around properly valuing customer positions so we could make sure margin requirements were met. This was a huge deal to everyone involved because the counterparty risk was so high for our company.

Help me understand, if the blockchain has 100% uptime, how would there be outages in the infrastructure? To take down a large blockchain, you'd have to take out the entire network, or basically the whole internet.

What errors would occur if all the rules for transfer of assets were encoded on-chain? Crypto is already moving hundreds of billions of dollars worth of tokens 24/7 every single day. There's no central bank entities or clearing houses. The collateral for loans is completely tracked on-chain and fully auditable on the ledger. Tell me why that's not superior.

I think you’re right and the final point of “being able to unwind trades” is most likely a smaller issue than people make it out to be.

Unlike what a concerning number of crypto enthusiasts appear to believe, "sorry for your loss" is often not an acceptable response in the mainstream world of finance.

But just like most players in the space have magically discovered the need for KYC if they want any real mainstream adoption, I expect that they'll also spend the next few years discovering why chargebacks and fraud protection are a thing.

We know, the Fed will step in at any time to back stop the losses. Don't kid yourself into thinking that such a system is without systemic risks. There's a reason why the big wigs spend their time in Davos dreaming and scheming about The Great and Orderly Reset.

As for KYC, crypto is well aware of it. Any centralized exchange you deal with in US has it in place already.

As for fraud protection and chargebacks, that is really just an insurance problem. And that's got solutions in the works.

A hack that wipes out 20% of the market in a single day.

Re outages, it is not about the blockchain itself necessarily, but all the services and systems that interface with it.

Errors occur because humans are humans and fat finger trade details.

That aside, as I mentioned in original comment, instantaneous settlement has other problems, e.g netting is important in managing liquidity.

And to your point about crypto managing billions of dollars of tokens - it's an absolutely trivial amount compared to what's moved in the markets every day. Blockchains are incredibly inefficient and transaction rates orders of magnitude lower than what'd be needed.

You may not be familiar with the layer 2 innovations in crypto enabled by zero knowledge proofs, but blockchains such as Ethereum are about to get several orders of magnitude more efficient. It will absolutely be able to accommodate these transactions in the not too distant future. I get the tech needs to mature and get to a battle tested point first, but blockchain based stock trading is inevitable.

There’s just a single database that ultimately registers trades and ownership. [1] It’s not a technical limitation that creates the “constraints”. That’s what the parent is trying to explain.

[1] https://en.wikipedia.org/wiki/Depository_Trust_%26_Clearing_...

That leaves me with more questions than answers with regards to DTCC's role in the future of finance. To me, it looks like an archaic, centralized institution that was set up to solve a problem they had in the 1960s with paper securities transfers. Smart contracts didn't exist then, so they weren't even an option. This central clearinghouse concept is going the way of the paper ticker. It has no place in the long term future of finance.

Yes, and they could "technically" build a scalable instantaneous system with a usable API. What prevents any movement in that direction is the messy interconnected web around it. A blockchain isn't going to help with it even if it's a technologically superior solution (it may not be).

Reply on news.ycombinator.com