Hacker News Reader

Comment by MatteoFrigo | Hacker News Reader

> I often wonder how money is stored.

The Bank of England published a summary of how modern banks work. See https://www.bankofengland.co.uk/quarterly-bulletin/2014/q1/m...

> It can't be just a number on a computer in a bank, right?

It can and it is, see the BOE paper.

> So I guess USD needs to be recognized by the US somehow?

This is a complicated question that will start a flamewar, but basically, any bank in the world can create dollars out of thin air given adequate collateral. Read about "Eurodollars".

View on HN

> This is a complicated question that will start a flamewar, but basically, any bank in the world can create dollars out of thin air given adequate collateral. Read about "Eurodollars".

This is true, but it's actually sort of a specious point. What a eurodollar deposit "really" is, is just a short contract. Your European bank takes your exchange and promises to give you back dollars when you ask for them. That they may not have them right now is just a technical detail. But the value of the dollar is still set by what the market is willing to pay for it. The bank going "short USD" is a risk it takes, not an inflationary pressure per se.

So... sure. Foreign banks can create dollars just like brokerages can "create stock shares"[1]. Except, they really can't. They're just playing accounting tricks.

[1] Or, more relevantly perhaps, how one crypto exchange can create wrapped tokens representing foreign coins. This capability was part of the exploit chain that led to the DeFi hack last month.

Exactly, they don't give you dollars, they give you a dollar denominated liability.

Just to make sure we are speaking the same language: If I have $1 in a checking account in a US bank, do I have one dollar, or do I have a dollar-denominated asset of $1 face value, backed by a dollar-denominated liability of the bank?

Reserve requirements at US banks are regulated by the US government as one of many mechanisms used to control the money supply. So... yes, sure. It's complicated, but a dollar at a US bank is a "dollar", because that's what "dollar" means.

The contention upthread was that foreign banks could do the same thing to print dollars. And, again, it's complicated. They sorta can, but only in the sense of placing a gargantuan short bet on the dollar. (A ton of short selling on a small security can push the price down for the same reason). So no one does that, because no bank wants to play that kind of game (or is allowed: all those large foreign banks are themselves regulated by their own national regimes).

Oh oh, next do the Depository Trust Company and how people don't actually own their stock shares of most public stocks either :)

If you walk into a US bank and deposit $1 cash, they now owe you $1. To what extent they must have ready cash, fed deposits, liquid assets, illiquid assets, etc to back that debt is the subject of bank regulation

It's amazing to watch people who have never really thought about this before realize the entire banking system is built upon a handful of centralized, permissioned databases which can pretty much be turned on/off at will for certain parties by the central banks.

What is even more amazing is watching these same people (no necessarily you OP, but prevailing opinion on HN...) then go on an argue that a permission-less distributed ledger like Bitcoin is a completely unnecessary waste of resources.

You describe money as if it’s just numbers being moved around: yes, the technology is just numbers moving around, but money is a social construct not a technology, the use of technology is a very small part of what makes money money.

As someone who finds himself thinking “wtf is money” every decade or so, the most recent book I read on this was pretty good - Money: The Unauthorised Biography by Felix Martin. It’s firmly in the “social construct” camp and makes some good arguments (to my lay mind!). There’s some really interesting stuff about the Fei stones https://en.m.wikipedia.org/wiki/Rai_stones

> permission-less distributed ledger like Bitcoin is a completely unnecessary waste of resources.

it certainly is when it immediately de facto centralizes! which is in the event what has happened

I'm not sure how you can say it is de facto centralized. You may have had a decent argument as recently as last summer when most of the mining was done in China. However, we just watched them ban mining there completely, and the network carried on relatively unfazed. Seems pretty distributed to me.

> it immediately de facto centralizes

In what respect? Hash power is well-distributed. Consensus (formal and informal) is working as expected.

>What is even more amazing is watching these same people (no necessarily you OP, but prevailing opinion on HN...) then go on an argue that a permission-less distributed ledger like Bitcoin is a completely unnecessary waste of resources.

Hating the way modern banking works doesn't mean blockchain is a flawless solution. It comes with a ton of its own problems (not the least of which being that it's utterly inscrutable to normies). I personally would prefer we go back to currency backed by gold or some other comparable commodity (silver, etc.)

Sure, I'd agree commodity money is preferable (and certainly trustless in p2p transactions), and in many ways better than blockchain. I just don't see how to map that paradigm onto a world where people who have never met each other regularly transact across continents without introducing some trusted 3rd party. And the moment you add that 3rd party you end up right back where we are now.

How do you transact across continents without a trusted 3rd party, with blockchains? The only transactions you can do trust-less are transactions that happen completely on-chain, i.e. exchanging some cryptocurrency for other cryptocurrency or similar tokens. As soon as you want to exchange currency for good or services, that trusted 3rd party becomes necessary in the exact same way, no matter how "smart" your blockchain money is.

By trust-less I am referring to a trusted 3rd party (e.g bank or escrow service) needed to mediate the transaction between the two parties. Not necessarily a lack of trust between the two parties themselves (obviously it would be foolish to pay someone you didn't trust at all in an irreversible way, but I can order something from Amazon any day with an implicit guarantee I will receive it or a refund, despite not knowing anyone there personally). An if you are referring to the shipping company that is going to move the goods, that would be a separate p2p transaction (and probably insured) between the seller/buyer and the delivery company.

Everyone who transacts in a fiat currency digitally has a defacto third (or more) party in the bank that mediates the transaction, and even non-digital cash transactions inherently include the central issuing bank (who you must trust not to debase the currency...just ask a Turkish citizen how that's going).

So a transaction using cryptocurrency has third parties, who do just the same thing and require trust, but which you don't want to call "defacto" because...?

Honestly curious. What is the dissonance that you see there?

Digital fiat currency is indeed bank databases. Similarly, stock markets and stock ownership are also stored in permissioned databases. Both are highly regulated and enforced by law. For people in countries with a high level of rule-of-law, and a stable financial system, I’d argue that this works reasonably well.

I also think that blockchain and other crypto currencies can be very useful. Especially for people in countries where rule-of-law is not as important and corruption is high.

> For people in countries with a high level of rule-of-law, and a stable financial system, I’d argue that this works reasonably well.

I agree, but that's not always obvious at first glance. For example, I would've said Canada qualifies as a stable rule-of-law country with strong property rights up until a couple weeks ago.

This has to be, by many orders of magnitude, the most baller backdrop of any video ever.

I'm interested in a risk assessment and hardware/software standard for acceptable error rate. How much redundancy is in use and at what layers?

Reply on news.ycombinator.com