Hmm. I suppose this raises the question of where (and why) it's useful to draw a distinction between "tokens" and "accounts." If all assets on the blockchain are associated with an owner address, that sounds an awful lot like everybody has blockchain accounts.
Is it the token-versus-account distinction that matters here? Or is the biggest change that Citi wouldn't have to manage the accounts using its own database? Or are they using blockchain as an excuse to automate things that, from a technological standpoint, could easily have been automated without blockchain?
The difference is in what you check. With accounts you check the person; with a token you check the thing. I think McLaughlin is saying 1) both approaches are possible and can achieve broadly similar ends, 2) it is not obvious which is better, 3) some people say tokens are better for certain reasons, which I mentioned in the post, and 4) if we suppose that they are right, then we can map out how a tokenized system would have work in order to achieve those benefits.
And I think what I am saying is that I had been asking "why blockchain?" but that "why tokenization?" is a better question.
I'm not following the "check the person" vs. "check the thing" distinction. With both bank accounts and Bitcoin, you can only make a payment if the account or address has enough money in it (ignoring overdrafts).
You've inspired me to go actually read the McLaughlin articles. Here's a quote:
"Bitcoin, for example, is not a liability as there is no central issuer or intermediary."
Is the blockchain itself *not* a central issuer?
Is it wrong to think of the Bitcoin blockchain as a balance sheet consisting entirely of liabilities with nothing on the asset side?
Is Bitcoin *really* a token-based system underneath? Would it be wrong to say that Bitcoin is using accounts to simulate the behavior of tokens?
Conceptually, we could write Bitcoin as a balance sheet with only liabilities. That works, in accounting terms. But banks think in legal terms, and in that sense Bitcoin balances are not accounts, because there is no legal recourse if you can't collect. There is no issuer, because there is no one to sue. Bitcoin is a token and in fact a bearer instrument, because the value is in the token.
Hmm. I suppose this raises the question of where (and why) it's useful to draw a distinction between "tokens" and "accounts." If all assets on the blockchain are associated with an owner address, that sounds an awful lot like everybody has blockchain accounts.
Is it the token-versus-account distinction that matters here? Or is the biggest change that Citi wouldn't have to manage the accounts using its own database? Or are they using blockchain as an excuse to automate things that, from a technological standpoint, could easily have been automated without blockchain?
The difference is in what you check. With accounts you check the person; with a token you check the thing. I think McLaughlin is saying 1) both approaches are possible and can achieve broadly similar ends, 2) it is not obvious which is better, 3) some people say tokens are better for certain reasons, which I mentioned in the post, and 4) if we suppose that they are right, then we can map out how a tokenized system would have work in order to achieve those benefits.
And I think what I am saying is that I had been asking "why blockchain?" but that "why tokenization?" is a better question.
I'm not following the "check the person" vs. "check the thing" distinction. With both bank accounts and Bitcoin, you can only make a payment if the account or address has enough money in it (ignoring overdrafts).
You've inspired me to go actually read the McLaughlin articles. Here's a quote:
"Bitcoin, for example, is not a liability as there is no central issuer or intermediary."
Is the blockchain itself *not* a central issuer?
Is it wrong to think of the Bitcoin blockchain as a balance sheet consisting entirely of liabilities with nothing on the asset side?
Is Bitcoin *really* a token-based system underneath? Would it be wrong to say that Bitcoin is using accounts to simulate the behavior of tokens?
Conceptually, we could write Bitcoin as a balance sheet with only liabilities. That works, in accounting terms. But banks think in legal terms, and in that sense Bitcoin balances are not accounts, because there is no legal recourse if you can't collect. There is no issuer, because there is no one to sue. Bitcoin is a token and in fact a bearer instrument, because the value is in the token.
Ah. That's an interesting point about the legal stuff. I'll have to think about this.