Why did Terra’s UST collapse?
Many people have asked for Reserve’s take on UST’s unwinding.
While we will give a deeper analysis in due course, here’s an overview on why this played out like it did.
It was evident from the start that Terra’s UST would eventually collapse.
The idea dates back to 2015 when Robert Sams published a quick paper outlining how one might make a stable version of Bitcoin.
The idea goes like this:
Instead of just one coin like Bitcoin, you have two — call them the “coin” and the “share.”
When the coin is trading above $1.00, you mint more coins out of thin air and sell them into the market for shares, deleting the shares you bought back. As you sell more coins into the market, this drives down the price, and when you get down to $1.00 you stop.
When the coin is trading below $1.00, you do the opposite: mint more shares and use them to buy coins off the market, deleting them as you buy them back. Buying them up drives up the price, until you are back at $1.00, and all you had to do was mint and sell some new shares.
Whoa! It’s actually a very cool idea.
Early share holders get to profit from growth in coin demand, because the more coins are minted, the more shares are bought up and deleted, driving up share price. As long as you anticipate more growth in coin demand, holding shares is attractive.
In Terra’s system, UST is the coin and LUNA is the share. LUNA holders saw incredible appreciation as UST demand grew, fueled by “yield farming” incentives.
(Yield farming is when crypto speculators deposit capital into a decentralized app in order to earn rewards paid out in tokens. Terra had a unique yield farming offering, where the rewards were paid in tokens other than LUNA, because Terra is its own blockchain which is host to its own set of decentralized apps, each with their own token. Those other tokens could be handed out to UST holders to incentivize UST adoption, without having to spend any LUNA.)
But it was obvious upon thinking through the basic design that it could unwind. How?
If confidence in the coin is shaken, it will trade for less than $1.00. Shares will be sold to buy back coins, and this act of selling shares dilutes existing share holders, driving down share price. As a share holder, if you think this will keep happening, you’ll try to sell quickly before more depreciation occurs.
If share holders panic and sell all at once, share price can drop quickly. At some point, coin holders will start to see that the shares they can redeem their coins for may not end up being worth anything. At this point, the coin holders also start to rush for the exit, trading into shares as fast as they can, and dumping those shares immediately.
Among stablecoin geeks, this is called the “death spiral.” Unless some external buyer steps in to buy back all the coins for some reason, it’s probably irrecoverable.
And it was evident to anyone who looked closely at Terra that this could happen.
There have been many variations on that original idea. One was called Basis, and luckily it never launched.
Bizarrely, the Terra team allegedly anonymously launched a version of Basis dubbed “Basis Cash,” perhaps as a learning exercise to see what would happen. As predicted, that project melted down in the same way as UST just did. Surprisingly (if it’s true they were behind it), that didn’t seem to dissuade the Terra team from continuing with their variation on the theme.
But it’s important to keep in mind that not all stablecoins have this issue.
Maker’s DAI stablecoin, which is also a decentralized protocol with elaborate logic, is quite safe in comparison. It over-collateralizes its stablecoin with crypto assets, and has a mechanism for reducing supply before it gets anywhere close to being undercollateralized. That means that unless its backing assets flash crash close to zero in an instant, 1 DAI will always have more than $1.00 in backing.
And fiat-backed coins like TUSD, USDC, and USDT are directly convertible to US dollars. While the issuers sometimes hold assets other than US dollars, they tend to be very low-risk assets, and regulators keep an eye on them to make sure they keep in line.
After UST de-pegging, there has been some fear of USDT following suit, but unless Tether is lying about the assets they hold in reserve, there is little chance of this outcome. Even if the market price dips for a while, big traders can buy up USDT and redeem with Tether for $1.00, turning an easy profit for themselves and bringing USDT back to its intended $1.00 market price.
You can break these down into three categories of stablecoins:
- Fully collateralized, stable asset-backed
- Overcollateralized, volatile crypto-backed
- Algorithmic
Generally speaking, the first two are economically sensible, and Algorithmic stablecoins are not.
Some crypto entrepreneurs have launched variations that blur the lines, holding partial backing. While these may work, they are more dangerous, and should be regarded with caution and skepticism. Terra had started to go down this path, buying billions of dollars worth of Bitcoin to help support the peg, but it was not enough.
It’s incredibly embarrassing for the whole cryptocurrency community that we let an algorithmic stablecoin get so big before failing.
Luckily, the vast majority of its users were probably cryptocurrency speculators, often playing with money they had earned for free from crypto investments, and happy to take some risk on an economic experiment. Still, it was a surprise to most that it unwound so quickly, and many retail investors lost a lot of money.
It is essential that we do not allow Algorithmic stablecoins to be used as a savings and spending tool by the broad population. It’s not practically feasible for every individual to become an expert on monetary dynamics before they decide which financial products to rely on. As an industry, we absolutely must call out any further attempt at popularization of Algorithmic stablecoins among non-sophisticated users, and our exchanges and wallets must not support their mainstream use.
For the other two types of stablecoins, the story is different. While we may need some checks and balances to make sure they are adequately conservative, there’s a lot of value that can come from their replacing outdated money transfer and savings. So while caution is warranted, there’s no need to run for the exit.
