Hacker News Reader

Yeah, as someone who didn't know about Luna or Terra a week ago I think he did a good job of explaining the system, and the way it would fail, pretty well 2 days ago on Wednesday:

https://www.bloomberg.com/opinion/articles/2022-05-11/terra-...

You can read down through where he details the Death Spiral. I also haven't paid much attention to other stablecoins, but I imagine the systems are similar, based on faith in the system, and only good as long as there's not a run on the system and more than some percentage try to get their money out.

It's absolutely true of other algorithmic stablecoins.

When it comes to "backed" stablecoins like USDT the scam is even simpler: you just say it's backed by cash when it's not.

> When it comes to "backed" stablecoins like USDT the scam is even simpler: you just say it's backed by cash when it's not.

This is such a silly argument. USDT is definitely backed by USD. To claim that it isn’t is just a lie. I think what you meant to imply is that it’s not backed 1:1. This isn’t only an issue with Tether. This is the case with any fractional reserve banking system.

Fractional reserve banking lends out deposits. This creates risk, but the net assets on the books remain the same (actually assets increase due to interest). With USDT, they may have bought something like commercial paper, which would be similar to fractional reserve banking, or they could have spent it on something irreversible (e.g. a dividend) and thus the net assets on the books is lower than the amount of USDT. It's unknown which situation applies to USDT.

It’s not quite that simple, no? If I deposit 1M in the bank, that bank can loan you 900k. You then buy a Bugatti from me for 900k, which I then put into the same bank. Then the bank loans you 810k to buy an NFT from me…

If the bank has 1M deposits, its net assets are 0. It has 1M is cash and has 1M in liabilities to depositors. If the bank loans out 900K, then it still has 0 net assets. It has 1M in liabilities to depositors, 100K in cash, and an "IOU" worth 900K (ignoring interest). Tether is different. When it mints and sells 1M USDT, it gets 1M is cash, but doesn't have any liability to Tether holders other than the honor system. It could easily pay a 500K dividend to its owners and no one would know.

What you describe is called the “money multiplier model” and yes, it’s not how the actual economy or actual banks work.

This document [1] from the Bank of England (UK’s central bank) is the best description I know of about how those ideas (“fractional reserve”, banks lending out deposits, “money multiplier” etc.) are wrong.

1. https://www.bankofengland.co.uk/-/media/boe/files/quarterly-...

That’s not true. It’s not just that real-world banks don’t do that, they just plain aren’t allowed to “lend out deposits”.

I guess you could say “fractional reserve banks” do, because “fractional reserve” is a textbook model and not something that happens in the real world.

The problem is that deposits are a liability. Banks can only lever up assets to lend, so deposits are on the wrong side of the balance sheet to do that. What limits how much banks can lend is capital (and capital adequacy ratios), not deposits.

Banks create new deposits when they lend, as well as creating an equal amount of private debt. The loan is an asset of the bank, which creates a corresponding liability.

Customer deposits coming in as cash or transfers from other banks are useful as liquidity, but can never be lent.

Some details if you’re interested: https://www.bankofengland.co.uk/-/media/boe/files/quarterly-...

> The problem is that deposits are a liability. Banks can only lever up assets to lend, so deposits are on the wrong side of the balance sheet to do that.

When someone makes a deposit in a bank it goes on both sides of the __balance__ sheet. On one side the "deposit" is a liability for the bank - who owes money to the depositor - but on the other side it increases the bank's "reserve" account.

Is it definitely backed? As far as I know there has never been satisfactory proof that USDT has ever been backed by anything.

Yes fractional reserve banking is a real thing, but real banks must publish audited financial statements so that outside observers can asses the banks financial position. From small local banks to national reserve banks, the financial statements are out in the open for depositors and borrowers to scrutinize. This isn't possible with Tether because the fiat side of their books are closed.

If everything is above board at Tether why not release the books? It's such an obvious PR win and would increase adoption of their currency.

They haven't done so because they know the books are cooked.

NYAG, US DOJ and CFTC all found it was backed, from their own subpoenas. NYAG settlement with Tether forces them to keep updating disclosures.

Its 2/3rd backed by dollars and 1/3rd a mixture of commercial paper, which we don't have further information about. Rumor mill always swirls about that paper, but its probably not that controversial and would only be 33% of Tether. Likely would result in a liquidity crisis if more than 70-80% of Tether's were redeemed at once. Although I wonder if that's even possible now given how much USDT is locked in DeFi apps and liquidity pools and burned.

(Projects often lock assets in a liquidity pool share and then destroy their access to that share, to ensure to their community that there is always the ability to trade)

So Tether will probably continue working for partially the dumbest reasons.

Tethers can't be burned.

If anyone ever provably burns a tether, then the tether organisation will help 'recover' the tether to anyone holding keys to a previous address that held it, or just anyone with a reasonable story how they accidentally burned the tethers.

This is their policy since some big players accidentally sent tether to addresses on the wrong Blockchain...

Thats pretty interesting. I don't think that would work for Tether stuck in a liquidity pool share that was burned by being sent to a contract or burn address. Traders can still shift that tether around, but it cannot be unbounded. Be a hard case to prove.

> This is the case with any fractional reserve banking system.

That is insured by GOVERNMENTS...

So if $100m is backed by $1 then it’s backed? Give me a break!

> I think what you meant to imply is that it’s not backed 1:1. This isn’t only an issue with Tether. This is the case with any fractional reserve banking system.

There is a minor difference, though. On a typical fractional reserve bank the deposits are backed by assets worth more than what is the worth of the deposits.[1] Not only that, there are other liabilities that take the possible losses before the deposits are being hit.

On a typical stablecoin, there is a reason to suspect that the coins are backed by assets worth less than the market cap of the stablecoin.

The difference may seem subtle, but it is a difference of a legitimate business and a despicable scam.

[1] In case you are interested, you can go and check the balance sheet of any bank you want. The part where the value of excess assets to back the liabilites of the bank (deposits etc) is listed under "Equity"

Yeah but something happened when USDT was in the unpeg crisis 2 days ago.

See, if it fails then the big exchanges become insolvent, full stop.

Now here's the thing. With some exchange APIs you can see market order statuses, whether they're settled or not.

You can use this to determine market latency. If I want a market buy, how long does it take to settle.

So here's the thing, the latency on USDT buys was on the millisecond scale as best I could determine.

The latency on USDT sells was near minute scale.

It's a giant accusation so I'm not going to name-drop the exchanges I'm talking about and I'm just a single data point but if anyone else saw it contact me.

It looks like they "narrowed the channel" and put their thumb on the scale to prefer buys and float the price, pushing it back up.

Now this was at a time that the gas price was about 250/300gwe because of the massive UST exit and network delays were 15/20 minutes or so.

So even if you wanted to arbitrage it wouldn't have been worth it because the network was going way slower then exchange latency.

This started basically right when Tether tweeted about the billion in Avax and ran for about 12 hours or so

Ultimately it doesn't matter. These are unregulated securities and the exchanges can run their markets as the please and, if what I'm saying is right, good job keeping the thing you need to stay in business afloat. Coinbase stock (which I'm using as proxy for general sentiment) climbed 50% after tether got repegged and contagion appears to have been averted

I'm just wondering if anyone else saw this. It could be a coding defect on my part, a fluke because of the craziness of the market, me projecting and fooling myself, etc.

Please don't go around repeating this as fact, it as of now, needs independent verification

Tether isn't backed by cash but by a combination of short and long term loans and potentially even crypto. Fiat deposits are likely a small allocation. Calling it a scam is still disingenuous, it is backed by reserves (the more pertinent question is how liquid they are). People have been waiting for Tether to "collapse" for over 5 years now.

> I also haven't paid much attention to other stablecoins, but I imagine the systems are similar, based on faith in the system, and only good as long as there's not a run on the system and more than some percentage try to get their money out.

No, that's not true. There are categories of stablecoins. There are some in the category of Luna/Terra, there are some stable in name alone, there are many that work perfectly and some of those would be able to do 100% redemptions of the entire float automatically or in a crisis of confidence.

TerraUSD was a joke. People warned of the death spiral the whole time. Its merely amusing it got that big.

Matt Levine is usually great. But the reason why UST at 0.98 makes the whole thing collapse is not because people “rely on them being worth what they’re worth”. It’s because you can sell UST into Luna for arbitrage minting new Luna in the process, thus inflating Luna.

But Matt knows this. He describes the whole mechanics in his newsletter like 3 days ago. So not sure why the stretched analogy here…

So you sell UST into Luna, buy more UST with Luna, sell UST into even more Luna, rinse and repeat?

Yes, this is the nature of arbitrage opportunities. You are leveling the price of two assets by moving liquidity. You keep doing it until the prices normalize and the arbitrage opportunity dissapears.

In this case "normalizing" was going to zero for luna because that is what became the shared understanding of the value of what it was backed by.

Luna didn't go to zero because the whole crypto market woke up and realized that was no intrinsic value to it. If it did, it never would've got as big or lasted as long as it did.

There was an explicit attack that crippled the peg to the point that their algorithmic stability mechanic couldn't self heal. Which led the Luna project to attempt to assist the peg by deploying their own reserves, which unfortuently didn't work.

Can you send explain some more?

There didn't have to be an attack for this to happen

It may have been, but it equally could have been mass de-risking due to market conditions causing lots of UST withdrawals. Once the peg goes for a minute, confidence soon follows and the whole house of cards unwinds

LFG had $1.5B of reserves to plug a hole in a $18B leak. It never had a realistic chance of working (unless it was fast enough to stop the initial depeg, in which case it could have delayed the inevitable)

but UST burning for luna is capped at $100m/day

Those weren't AAA mortgage securities that went bust in 2008, though some were fraudulently marketed as such. They were liar loans on second-rate properties chopped up into a bunch of tiny pieces and re-sold, based on the notion that real estate would only go up forever, so if people failed to pay the lender would foreclose and own real estate with more value than the original loan.

Seems similar to some methods of stabilizing a "stablecoin": back it by a traditional cryptocoin and assume that this other coin can only go up.

People who had only bank deposits didn't lose anything.

Those chopped-up pieces of crap got lumped together in big collections which got rated AAA, because the people rating them assumed that not too many of the little pieces would fail at the same time.

I think GP covered that with:

> though some were fraudulently marketed as such

I don't think what I described was entirely fraud. I think people honestly underestimated how correlated the failures could become.

(But I won't deny that there was also fraud, and economic incentives for inflated ratings.)

Dead wrong. This is the major take away from 2008 that has been widely reported on in pop media. The junk was repackaged so it could be stamped AAA

off topic, but I wouldn't have understood anything that happened in the last few weeks without Levine. now most of my days are waiting for him to post so that I know wtf just happened

Link?

Edit: Found it, it's https://www.bloomberg.com/opinion/articles/2022-05-13/elon-m...

Surely no one through stablecoins were low risk?

Stable is right there in the name.

By the folks who put “safe” in safemoon.

I am pretty bullish on crypto but Safemoon has to be one of the worst “assets” out there. Bunch of charlatans.

Including overcollaterized ones? They are not risk-free, but DAI has survived much worse market swings that the one from this week.

That’s what the whole crypto industry has been saying for years. If they didn’t think it was stable (or didn’t want you to believe it was), it wouldn’t be called stable. Millions of people at this point have been led to believe a stable coin is stable.

lol

Apparently, you could even get 'depeg' insurance! Multiple levels of Ponzi...