Thursday, May 12, 2022

What, me worry?

 I'll try to get Joseph to come in and explain some of this in greater detail (he knows far more about the subject than I do), but in the Bizarro world of cryptocurrencies, stablecoins actually make a certain amount of sense. There are a number of reasons why someone trading in these markets would rather transact coin to coin rather than cash to coin and back again, particularly if the second coin is pegged to a real currency like the dollar.

The trouble is that all that clever financial engineering has a way of convincing people that there's something there that isn't. In the end, it's just a better engineered perpetual motion machine.





Matt Levine explains the details (from his newsletter):

An “algorithmic stablecoin” sounds complicated, and there are a lot of people with incentives to pretend that it is complicated, but it is not. Here is how an algorithmic stablecoin works:

You wake up one morning and invent two crypto tokens.

One of them is the stablecoin, which I will call “Terra,” for reasons that will become apparent.

The other one is not the stablecoin. I will call it “Luna.”

To be clear, they are both just things you made up, just numbers on a ledger. (Probably the ledger is maintained on a decentralized blockchain, though in theory you could do this on your computer in Excel.) 

You try to find people to buy them.

Luna will trade at some price determined by supply and demand. If you make it up on your computer and keep the list in Excel and smirk when you tell people about this, that price will be zero, and none of this will work.

But if you do a good job of marketing Luna, that price will not be zero. If the price is not zero then you’re in business.

You promise that people can always exchange one Terra for $1 worth of Luna. If Luna trades at $0.10, then one Terra will get you 10 Luna. If Luna trades at $20, then one Terra will get you 0.05 Luna. Doesn’t matter. The price of Luna is arbitrary, but one Terra always gets you $1 worth of Luna. (And vice versa: People can always exchange $1 worth of Luna for one Terra.)

You set up an automated smart contract — the “algorithm” in “algorithmic stablecoin” — to let people exchange their Terras for Lunas and Lunas for Terras. 

Terra should trade at $1. If it trades above $1, people — arbitrageurs — can buy $1 worth of Luna for $1 and exchange them for one Terra worth more than a dollar, for an instant profit. If it trades below $1, people can buy one Terra for less than a dollar and exchange it for $1 worth of Luna, for an instant profit. These arbitrage trades push the price of Terra back to $1 if it ever goes higher or lower.

The price of Luna will fluctuate. Over time, as trust in this ecosystem grows, it will probably mostly go up. But that is not essential to the stablecoin concept. As long as Luna robustly has a non-zero value, you can exchange one Terra for some quantity of Luna that is worth $1, which means Terra should be worth $1, which means that its value should be stable.

All of this is, I think, quite straightforward and correct, except for Point 7, which is insane. If you overcome that — if you can find a way to make Luna worth some nonzero amount of money — then everything works fine. That is the whole ballgame. In theory this seems hard, since you just made up Luna. In practice it seems very easy, as there are dozens and dozens of cryptocurrencies that someone just made up that are now worth billions of dollars. The principal ways to do this are:

Collect some transaction fees from people who exchange Luna for Terra or Terra for Luna, and then pay some of those fees to holders of Luna as, effectively, interest on their Luna holdings. (Or pay interest on Terra, creating demand for Luna that people can exchange into Terra to get the interest.)

Talk about building an ecosystem of smart contracts, programmable money, etc. on top of Terra and Luna, so that people treat Luna as a way to use that ecosystem — as effectively stock in the company that you are building and ascribe a lot of value to it.

These things reinforce each other: The more fees you collect and distribute to Luna holders, the more big and viable your ecosystem looks, so the more highly people value it, so the more Luna they buy, so the more activity you have, so the more fees you collect, etc.

 

But there is no magic here. There is no algorithm to guarantee that Luna is always worth some amount of money. The algorithm just lets people exchange Terra for Luna. Luna is valuable if people think it’s valuable and believe in the long-term value of the system that you are building, and not if they don’t.

Or if they do, then they don't

From the WSJ:

TerraUSD traded as low as 23 cents Wednesday, according to data from CoinDesk. As of about 5 p.m. ET, it had rebounded partially to about 67 cents in volatile trading.

A stablecoin, this breed of cryptocurrencies had gained favor among traders for being the one part of the crypto universe that was known for its stability. While the most popular stablecoins maintain their levels with assets that include dollar-denominated debt and cash, TerraUSD is what is known as an algorithmic stablecoin, which relies on financial engineering to maintain its link to the dollar.

The break in TerraUSD’s peg began over the weekend with a series of large withdrawals of TerraUSD from Anchor Protocol, a sort of decentralized bank for crypto investors.
Here's a characteristically good discussion from Coffeezilla.




While it's easy to mock the celebrities and Silicon Valley visionaries who have been pitching web3, we need to remember that the human toll here is horrifying...

... with the worst yet to come.

As usual, Liz was there first.
But no matter how bad things get, you'll find a rich person with delusions of competence giving bad advice.

Posted by at

No comments:

Post a Comment

Subscribe to: Post Comments (Atom)