Hacker News Reader

Comment by eru | Hacker News Reader

Currency pegs and currency boards can work quite well. You just need sufficient assets to cover them. See https://en.wikipedia.org/wiki/Currency_board

Btw, not only sovereign nations have these issues. Your bank also tries to maintain a peg between one dollar in its account books and one dollar in government money. Similar (but more roundabout) for things like Amazon gift cards.

For a country without an established, credible central bank a currency board can be a big economic improvement over trying to issue their own native fiat currency.

(Even better from an economic point of view is for the country not to issue a government currency at all, but let the market handle that.

When the government is particularly inept, you can see the (black) market trying to fix currency issues even when the government actively tries to prevent that.)

View on HN

No, pegs are nonsense. If you want something USD-denominated, buy USD. The cardinal sin committed was pegging in the first place, which always was doomed to failure because the scale of the effort would attract either a bank run or raid, guarantied.

Let's say you're a hedge fund looking for a one-way bet that will profit massively in a crypto crash (that stablecoins can't survive) and has no downside except carrying costs. Enter Tether, Terra, etc. Risk of loss at initial sale of 1, very low because tokens will not increase in value to 1.5 or 2. Profit potential? 100%. Everyone with a dash of creativity and a fund probably already is short these coins and has been for some time. "The Big Short" of 2022.

There are only two ways to hold USD: physical cash and database entries at the Fed.

The former is really inconvenient, especially if you want to transfer money. The latter is only available to a select few companies.

Everything is else at most assets denominated in (= pegged to) USD.

Does your critique apply to all these other categories as well? Why or why not?

"sure, my house's foundation is made of cardboard, but have you considered that the empire state building could also be destroyed in a large enough earthquake?"

Exactly! Great analogy.

This suggests that we need to look into the engineering details of a building (or peg arrangement) to determine whether we can trust it, instead of just declaring all building (or all pegs) 'nonsense'.

Pegs fail at the worst possible time. You can transfer or store value via oak tokens flown by trained pigeons, because that might be convenient (to you) and doesn't involve the Fed. But if someone tells you the token/pigeon method has the same attributes as a bank-to-bank USD wire, that person would be mistaken. The methods would be no closer to equivalence if one were to stamp "USD" on the oaken rounds and train the birds to collect a receipt.

As long as the guys who issue the token have sufficient assets to redeem them, and regularly redeem them to any comer, I don't see much of a problem.

If there's oak and pigeons involved, that doesn't change much---apart from being might inconvenient.

The problem algorithmic stable coins is that the backing is either non-existent or at best really suspect.

Tether is a bit better in that they at least claim to have real world backing, but they don't tell you what it is, so you can't trust it.

Btw, the backing for a token denominated in USD doesn't have to be USD.

Eg most banks back their deposits with all kinds of assets and bonds, not just physical cash (or balanced at the Fed).

Having a few reserves in USD on hand is important, but in the longer run it's more important that the worth of your backing assets is reliably higher than your liabilities. The difference between the two is your equity cushion.

> You just need sufficient assets to cover them

Oh, is that all?

To have a zero-risk peg you need to have as much reserve currency as you have issued currency; if you issue 1e6 fun-bucks that are pegged at 1:1 with the USD, you need 1e6 dollars.

If your country has banks, and you allow those banks to engage in fractional reserve banking, and you make guarantees to depositors that you will make them whole no matter how terrifyingly bad those banks are at their job, then holding dollars at 1:1 is insufficient. And depending on what kind of reserve requirements you have, the multiplier can get very large very quickly.

If you allow commercial banks to operate, then, as they say, you can fuhgeddaboutit. No amount of assets backing your peg can possibly be sufficient. You can push the tail out further, but such guarantees will only make your commercial banks more aggressive and increase the systemic cost of a failed peg.

If you choose to issue sovereign debt in your pegged currency, or support pension funds for your citizens, or anything like that, then ... don't peg your currency. Issue your own sovereign currency and get yourself some nuclear weapons.

You are making a lot of assumptions here..

Yes, I am suggesting that fractional reserve banking is a good thing. You hold reserves for short-term convenience, so that in the normal course of business, you can redeem immediately on demand. You hold the rest of your balance sheet in different assets, so you can earn a return. (And you make contractual arrangements with your token holders that you have the option of taking your time to redeem, in return for a prohibitive interest rate. At least that's what happened historically in eg Scotland. This was called the 'option clause'.)

You are right that if your liabilities are in thing A, and your assets are in thing B, there's always a non-zero risks that the value of B might drop below the value of A. Even if at the moment, B is much more valuable than A.

You can make that risk pretty small. Small enough that it's comparable to eg the risk of all your computers being hacked, the risk of your CEO running away with bags of cash, the risk of the government just seizing your business, the risk of an asteroid impact hitting your cash vault, etc.

What makes you think this particular risk is special? It's just one of the risks holders of your liabilities take. I don't see how this particular risk should make banks more aggressive?

Btw, commercial banks (and other companies) in most countries are allowed to have liabilities in foreign currencies. Or liabilities in assets that are not currencies at all.

> If you choose to issue sovereign debt in your pegged currency, or support pension funds for your citizens, or anything like that, then ... don't peg your currency. Issue your own sovereign currency and get yourself some nuclear weapons.

Many countries get lower interest rates when issuing debt in a currency they don't control. The creditors weigh both default risk and inflation risk. It's not obvious to me why eliminating all default risk in favour of inflation risk would be an unambiguously good thing?

Before the USD became a full fiat currency (but was instead pegged to gold, silver or a combination of the two at various times), the US issued debt just fine.

Also keep in mind that even if a sovereign issues debt in a pegged currency or a foreign currency they don't control, they can just decide not to pay anyway. That's what Argentina did a few times.

Of course, your creditors won't like that. But they won't like excessive inflation either. Six of one, half a dozen of the other.

Reply on news.ycombinator.com