4 ms·

> To me this is like saying 'why do we need a network and a browser? you could just get them to send you a disk with the file you want to look at and install it

by hiq 4y ago

> To me this is like saying 'why do we need a network and a browser? you could just get them to send you a disk with the file you want to look at and install it on your local machine.' I think we disagree of the relative complexity: operating smart contracts is way harder and less proven than what I described, I think the smart contract vulnerabilities described on https://web3isgoinggreat.com/ https://web3isgoinggreat.com/ show that well enough. > A blockchain allows you to atomically transact across multiple instruments and parties. This is not something that can reasonably be done on tradfi I've mostly heard of flash-loans to exploit arbitrage opportunities that shouldn't be there (e.g. oracle bug) or scale up attacks (instead of emptying an address using several transactions, one could be enough). Are you sure there's no equivalent in HFT?

kybernetikos 4y ago
> I think we disagree of the relative complexity: operating smart contracts is way harder and less proven than what I described, I think the smart contract vulnerabilities described on https://web3isgoinggreat.com/ https://web3isgoinggreat.com/ show that well enough. It's not really about the complexity, it's about the fact that finance is about ecosystems. How things integrate. Just like it might not seem that hooking computers together is going to get you that much, in fact it does. By the way, the majority of crypto horror stories at the moment are straightforward fraud and corruption, of exactly the kind of thing that goes on in tradfi the whole time. There are very few that are genuinely about bugs in smart contracts. Flash loans are zero counterparty risk, zero credit risk for the lender - because if the transaction succeeds then the loan was returned, if it wasn't returned then the transaction didn't happen. This would be possible in a closed ecosystem (but much less valuable - limiting what you could do with the borrowed money), but the number of institutions you'd have to get to agree to support a cross institutional transactional API means that it's not practically possible for tradfi to offer an equivalent kind of product.
- hiq 4y ago
  > By the way, the majority of crypto horror stories at the moment are straightforward fraud and corruption, of exactly the kind of thing that goes on in tradfi the whole time. There are very few that are genuinely about bugs in smart contracts. That's focusing too much on the recent FTX stories, you can filter the stories there (select 'bug'), there have been plenty of vulnerabilities this year still. The comparison with traditional finance is also misleading: the proportion of fraud seems to be way higher in the blockchain space relative to "normal" (let's say intended) usage. > it's not practically possible for tradfi to offer an equivalent kind of product. Assuming it would be possible, do you think people would use it? What for? Flash loans are "useful" because they allow anyone to exploit a smart contract without having the required funds, or because they allow one to do arbitrage between decentralized platforms, but do you have such gaping deficiencies in traditional finance for them to be relevant there? A lot of blockchain projects seem to make sense within a blockchain context, but fall apart as soon as you try to link them to the real world. A decentralized exchange sounds fancy, but it's only as interesting as the tokens it allows users to trade. Flash loans look like an innovation until you realize you're still stuck within this context, making concrete comparisons with the real world inadequate.
  - kybernetikos 4y ago
    The modern financial system is built on arbitrage, without it almost nothing works. Arbitrage typically takes advantage of small differences so it needs leverage to make interesting levels of profit. Flash loans are great for arbitrageurs and they're also great for the other side too - imagine being able to make a profit from lending money at interest without having any risk of the party you lend to not being able to return the money. I think perhaps you're only hearing about blockchains in news stories, where of course the shadier aspects are emphasised. As for "stuck within this context", that's actually one of the areas that I think blockchains have a huge advantage over traditional finance. Of course, at the moment there's a lot more traditional finance than there is finance on blockchains, but hooking together different financial institutions and their products is a massive pain, with data sharing (frequently it's just csvs periodically uploaded to an ftp site), and reconcilation taking up lots of time and effort with each hop. Blockchains have the potential to make this much more efficient, standard, secure and transactionally interoperable.
    - hiq 4y ago
      > Flash loans are great for arbitrageurs My point is that traditional finance doesn't need this because there's enough cash available to arbitrageurs already, so if you somehow made flash loans, they wouldn't be used anyway. It's also unclear what legal framework would accompany them, which is again something that Ethereum can ignore. Your last paragraph is an oft-repeated myth that using Ethereum forces transparency and interoperability between systems. But that's not true, it doesn't enforce that, I've seen my share of code trying to implement ERC20 and failing in subtle ways, and that's code trying to follow a common interface. Interoperability is hard, and you don't get by using $MAGIC_PLATFORM. Some projects also restrict access to their smart contract because legally it's really hard to make it fully open, or put a tiny part of their system there mostly for PR reasons ("we're on the blockchain"). Reasons why data sharing etc. are a massive pain include: * no incentive from financial institutions to improve that * legal reasons making it hard and costly to do anyway Ethereum is not solving these problems. In particular: > efficient, standard, secure * So far it's failing at "efficient", and it can only be more efficient by getting rid of the legal aspects of financial transaction, the tech is not the problem. * It's also failing at "secure" as evidenced by the link above. Another counterargument to the potential of this tech is, if indeed it brought such benefits, we'd have seen more than toy projects developed in traditional finance by now, but we're mostly seeing institutions like the ASX scaling down on such experiments.

by hiq 4y ago

kybernetikos 4y ago

> I think we disagree of the relative complexity: operating smart contracts is way harder and less proven than what I described, I think the smart contract vulnerabilities described on https://web3isgoinggreat.com/ https://web3isgoinggreat.com/ show that well enough. It's not really about the complexity, it's about the fact that finance is about ecosystems. How things integrate. Just like it might not seem that hooking computers together is going to get you that much, in fact it does. By the way, the majority of crypto horror stories at the moment are straightforward fraud and corruption, of exactly the kind of thing that goes on in tradfi the whole time. There are very few that are genuinely about bugs in smart contracts. Flash loans are zero counterparty risk, zero credit risk for the lender - because if the transaction succeeds then the loan was returned, if it wasn't returned then the transaction didn't happen. This would be possible in a closed ecosystem (but much less valuable - limiting what you could do with the borrowed money), but the number of institutions you'd have to get to agree to support a cross institutional transactional API means that it's not practically possible for tradfi to offer an equivalent kind of product.

hiq 4y ago
> By the way, the majority of crypto horror stories at the moment are straightforward fraud and corruption, of exactly the kind of thing that goes on in tradfi the whole time. There are very few that are genuinely about bugs in smart contracts. That's focusing too much on the recent FTX stories, you can filter the stories there (select 'bug'), there have been plenty of vulnerabilities this year still. The comparison with traditional finance is also misleading: the proportion of fraud seems to be way higher in the blockchain space relative to "normal" (let's say intended) usage. > it's not practically possible for tradfi to offer an equivalent kind of product. Assuming it would be possible, do you think people would use it? What for? Flash loans are "useful" because they allow anyone to exploit a smart contract without having the required funds, or because they allow one to do arbitrage between decentralized platforms, but do you have such gaping deficiencies in traditional finance for them to be relevant there? A lot of blockchain projects seem to make sense within a blockchain context, but fall apart as soon as you try to link them to the real world. A decentralized exchange sounds fancy, but it's only as interesting as the tokens it allows users to trade. Flash loans look like an innovation until you realize you're still stuck within this context, making concrete comparisons with the real world inadequate.
- kybernetikos 4y ago
  The modern financial system is built on arbitrage, without it almost nothing works. Arbitrage typically takes advantage of small differences so it needs leverage to make interesting levels of profit. Flash loans are great for arbitrageurs and they're also great for the other side too - imagine being able to make a profit from lending money at interest without having any risk of the party you lend to not being able to return the money. I think perhaps you're only hearing about blockchains in news stories, where of course the shadier aspects are emphasised. As for "stuck within this context", that's actually one of the areas that I think blockchains have a huge advantage over traditional finance. Of course, at the moment there's a lot more traditional finance than there is finance on blockchains, but hooking together different financial institutions and their products is a massive pain, with data sharing (frequently it's just csvs periodically uploaded to an ftp site), and reconcilation taking up lots of time and effort with each hop. Blockchains have the potential to make this much more efficient, standard, secure and transactionally interoperable.
  - hiq 4y ago
    > Flash loans are great for arbitrageurs My point is that traditional finance doesn't need this because there's enough cash available to arbitrageurs already, so if you somehow made flash loans, they wouldn't be used anyway. It's also unclear what legal framework would accompany them, which is again something that Ethereum can ignore. Your last paragraph is an oft-repeated myth that using Ethereum forces transparency and interoperability between systems. But that's not true, it doesn't enforce that, I've seen my share of code trying to implement ERC20 and failing in subtle ways, and that's code trying to follow a common interface. Interoperability is hard, and you don't get by using $MAGIC_PLATFORM. Some projects also restrict access to their smart contract because legally it's really hard to make it fully open, or put a tiny part of their system there mostly for PR reasons ("we're on the blockchain"). Reasons why data sharing etc. are a massive pain include: * no incentive from financial institutions to improve that * legal reasons making it hard and costly to do anyway Ethereum is not solving these problems. In particular: > efficient, standard, secure * So far it's failing at "efficient", and it can only be more efficient by getting rid of the legal aspects of financial transaction, the tech is not the problem. * It's also failing at "secure" as evidenced by the link above. Another counterargument to the potential of this tech is, if indeed it brought such benefits, we'd have seen more than toy projects developed in traditional finance by now, but we're mostly seeing institutions like the ASX scaling down on such experiments.