Note: This article is not financial advice. Hubble Protocol does not endorse any tokens or platforms mentioned in this article.
- The vast majority of failed stablecoin projects were algorithmic.
- Other kinds of stablecoins have been downed by poor risk management.
- Stablecoins are a nascent technology carrying inherent risks.
Welcome to the stablecoin cemetery. This is where tokens that attempted to maintain a steady value of $1.00 now trade well under that price.
The stablecoins here represent thousands of hours of work and billions of dollars lost in the quest for creating "the most important development in the crypto asset space since Bitcoin," and there are, unfortunately, a lot of them.
This article will look at why several stablecoins failed to maintain their peg. Hopefully, by examining the points of failure that led to each token's demise, the DeFi community can learn something about building better stablecoins.
The Market Found Ways to Punish Seignorage Token Stablecoin Designs
One factor that unites many of the projects presented here is the use of a seignorage token, also known as a volatility token or shares token. This token helps absorb the price swings of its companion stablecoin, and its price is free to swing accordingly.
Notably, this endogenous relationship (a design that relies on two tokens from the same project working in tandem) has been called into question several times for introducing additional points of fragility to a stablecoin's health. If one token fails, so does the other.
An example of this phenomenon recently played out when TerraUSD (UST) joined the stablecoin cemetery, perishing alongside LUNA. The tokens, which entered a well-documented death spiral, fell victim to panic and were both obliterated.
In fact, these star-crossed pairings are a story as old as the idea of stablecoins. Looking back, the crypto class of 2014 debuted USDT and the earliest algorithmic stablecoins, NuBits (first stablecoin to market) and BitUSD. Today, only USDT, which is backed by fiat, still trades at a dollar.
A post-mortem on BitUSD revealed its design protected the price of the seigniorage token, BitShares, more than the dollar parity of BitUSD. It was later determined that BitUSD's peg stability had been 100% psychological.
As for NuBits, its method for rewarding users for "parking" stablecoins for a set time (removing NuBits from circulation) was criticized for introducing more supply (new NuBits) to the market when its price was already devalued by oversaturation.
Despite having arbitrage bots in place on several markets, NuBits lost its peg in 2016. It regained the peg shortly thereafter, but NuBits drifted from its peg completely in 2018 when users began aggressively swapping out of stables for BTC on the rise.
Users Found Ways to Punish a Stablecoin's TWAP
As pointed out in the article "Built to Fail: The Inherent Fragility of Algorithmic Stablecoins," algo stables are trying to juggle one too many balls:
Algorithmic stablecoins are fundamentally flawed because they rely on three factors which history has shown to be impossible to control. First, they require a support level of demand for operational stability. Second, they rely on independent actors with market incentives to perform price-stabilizing arbitrage. Finally, they require reliable price information at all times. None of these factors are certain, and all of them have proven to be historically tenuous in the context of financial crises or periods of extreme volatility.
Points two and three are an especially dreadful combo. For example, an improper time-weighted average price, or TWAP, crops up again and again when culling through the cause of death for several algorithmic stablecoins.
Iron Finance (IRON), a fork of Frax built on Polygon, was a multi-billion dollar concern when it was raided by users who took advantage of its 10-minute TWAP. The project ran into problems after an extended rise in TITAN's market value was followed by a significant drop in price as users took profits.
The spot price of TITAN, the token that collateralized 25% of IRON (USDC was the other 75%), was less than the TWAP price that minted IRON, and in a post-mortem by the US Federal Reserve:
The peg broke because users were creating IRON that was inherently worth less than $1, as the sum of the spot value of TITAN plus the value of USDC used for the creation of IRON was worth less than the spot price of IRON.
In just a few hours, users had driven the price of TITAN to zero by bringing trillions of tokens to market and squeezing out as much value as possible from the price discrepancy. While some users squeezed value, others were panic selling, and IRON's peg fell well below the missing 25% represented by TITAN.
Other users bought up IRON at a discount as low as $0.58 and brought the peg back to $0.75–its value in USDC backing, and the Iron Finance team redeemed IRON for $0.74 of USDC in the aftermath.
Here, the market acted both rationally and irrationally. Some users saw an opportunity to make a profit and took it, and other users saw two tokens collapsing and jumped ship, swapping stablecoins below their fair value. In the end, IRON realized a proper equilibrium, but that Schelling point was at $0.75, not $1.00.
Iron Finance was joined by Dollar Protocol (USDf) and Empty Set Dollar (ESD). Both projects also blamed less-than-optimal TWAPs for their stablecoins losing peg.
The next generation of ESD, Dynamic Set Dollar (DSD), didn't get to blame its TWAP. Instead, blame was placed on users who didn't play along with the project to help reach a Schelling point of $1.00.
Over the course of one week, the various incentives for user participation led the token's price to highs and lows, from $3.00 to $0.27 to $0.63, but DSD never reached its intended goal of one dollar.
Poor Choices and Hacks Took Down Several Crypto-backed Stablecoins
The crypto-collateralized stablecoins entering this list of dead tokens also suffered at the hands of the market and overzealous profit-seeking behaviors. However, their demise can be attributed to quite different causes of death than algorithmic stables: under-collateralization by various means.
Recipe for Disaster 1: Setting Poor Liquidation Parameters
Some of the ways crypto-backed stablecoins have met their end, in retrospect, were due to specifically poor choices rather than macro-design principles. For example, if a developer sets the TWAP of a token held as collateral to a fixed price, no matter what, they're going to have a bad time.
Kava did this when they hardcoded the price of UST to $1.00 for borrowing USDX. As users watched UST drop in price, they borrowed USDX against UST at a 99% LTV without facing liquidation. Users acquired cheap UST, minted USDX, and dumped enough of it on the market for USDX to depeg.
Another design faux pas that led to a crypto-backed stablecoin's insolvency recently came to light with Fantom's decision to forego liquidations on collateral debt positions (CDPs) for minting FUSD.
Unsurprisingly, without having a liquidation mechanism in place, FUSD never managed to hold a peg from the start.
If a stablecoin is backed by crypto, a highly volatile asset, but the mechanism that mints the stablecoin makes it impossible to liquidate users when crypto prices drop, then the peg becomes impossible to maintain.
Recipe for Disaster 2: Getting Rekt by Hackers
The other ways a few stablecoins have been taken down involve bad actors hacking the platform and taking off with the money. In these cases, in hindsight, the security of the protocol could have been better managed.
BEAN experienced a huge exploit when a user exploited Beanstalk Farm's governance and voted to give themselves $182 million. The attacker flash loaned themselves an amount of Beanstalk Farms governance tokens to gain a majority vote in favor of enriching themselves, and no one noticed until it was too late.
For Cashio, the end came when a user managed to mint $2 billion of CASH and made off with $52 million in collateral. The attacker took advantage of the absence of two missing validation codes and the fact that Cashio's smart contracts had not been audited.
Finally, SafeDollar (SDO), which was categorized as an algorithmic stablecoin with some USDC and USDT backing, experienced a $248,000 attack that threw the token permanently off peg. This attack came a week after the protocol was attacked for $95,000 worth of their SDS tokens waiting to be distributed to users.
Regulators Killed Projects Off Before Launch
There were a couple of stablecoins never had the chance to die, because they were stalled in their early stages due to regulatory scrutiny.
Facebook's Libra, what would have been a fiat-backed stablecoin, was shot down by the US government as Mark Zuckerberg faced off with Congress. Zuckerberg decided to shelve his stablecoin, Facebook was rebranded as Meta, and the project known as Libra was eventually renamed Diem.
Diem was slated to launch in 2021, but by 2022 Meta sold off its stablecoin department, ending the project.
Before Libra, there was Basecoin, also known as Basis. The project aimed to launch an algorithmic stablecoin kept pegged to USD with another token (sound familiar?).
Basis was backed by some huge names in tech, including A16Z and Alphabet, and it raised over $100 million in funding rounds. Eventually, Basis' investors were reimbursed when the founders realized the stablecoin would be impossible to launch due to one of its tokens being considered a security by the SEC.
Another project called Basis Cash copied the idea and launched BAC with an anonymous team. It failed to maintain a peg, twice. It has since been revealed that Do Kwon, the founder behind the recently de-pegged UST, was also allegedly behind BAC.
What Lessons Can We Learn from the Dead?
It's hard work designing a stablecoin that lasts, but fully backed stablecoins seem to survive longer than their algorithmic counterparts. If a stablecoin can ensure that there is collateral behind it, it has a very good chance of survival compared with what has been presented in this article.
As Vitalik Buterin pointed out in his thought experiment on stablecoins, users should be able to safely unwind their stablecoin positions. If a stablecoin is not fully backed by other assets, users will be left holding an empty bag when the music stops, and that's a token doomed to failure.
For fiat-backed stablecoins, survival means guaranteeing there is liquid capital on hand that supports every token minted. For crypto-backed stablecoins, this means ensuring collateral is well protected by running multiple tests and audits.
Any token that tries to maintain $1.00 on its ticker without maintaining as much or more value in reserves seems to be waiting for the grim reaper to come knocking.
Keep in Touch
Website | Twitter | Telegram | Discord | Email | Reddit