Which Is The Best Stable Coin?
A Deep Dive Into the History of Stable (and Unstable) Stables
Why Should You Read This Article?
Truthfully, I don’t know. I don’t know why anyone reads anything I write. Maybe it’s my humor that, for a moment, lets you forget about the bills and the grind and the existential terror and just slow down and enjoy a piece of writing for a second? Or maybe it’s my pervasive (and edgy) use of run-on sentences and complete disregard for English grammatical norms? Yeah…that must be it.
But real talk, I think you should read this because I haven’t found an article like this before. This really could be three different articles, but the problem is that stable coins touch on so many areas. To really understand the implications of them you have to understand how the DeFi markets supporting them work. If you understand this and have a large enough perspective on the innovations in the space, then you can identifying exciting protocols as they emerge. So this is a timeline of stables and a deepdive into their underlying protocols (or at least as deep as I could given the limitations of time and the sheer quantity of protocols to explore).
I also published an article about f(x) Protocol, one of those exciting protocols (at least imo). This article is really necessary to understand what they’re doing and why it’s special.
Even if you don't care about f(x) Protocol or new DeFi protocols, hopefully this article will earn a coveted space in that bookmark file you have for articles that you read once and intend to read again but probably never will.
TLDR: Read this article because you’ll be able to stand on the edge of stable coin innovation and brag to your friends and say, yeah, I get it.
Understanding Stable Coins
There’s a lot of offerings right now. In fact, it feels like every other week you have some new protocol unleashing a stable coin upon us – Aave, Curve, etc.
In some ways, this shows that stable coins might not be one-size-fits-all. As you’ll see below, different design considerations provide different benefit/risk tradeoffs.
What is interesting is how the internal economics of stable coins become what we would call a feature. Usually, these coins are backed by some kind of debt market. I mean, the whole treasury bond and banking market in the U.S. is how they ensure dollar stability. So the value backing the stable coin isn’t only as idle collateral, but often as functional protocols offering access to various financial features.
I like thinking about it like this because protocols really have two potential value offerings:
As a stable stable coin
As a DeFi protocol offering lending, leverage, etc.
Let’s dive into what has turned out to be a very extensive review of stable coins from dawn until today.
Comparison of the Different Types of Stable Coins:
No, we’re not running a review site comparing horse stables. You’ll have to go elsewhere for that.
Among the different stable coins in past and present, we can break them down into three different approaches. You might have read this breakdown elsewhere. It’s not very original.
1. Tokens that use some sort of off-chain collateral.
These stables rely on centralized escrows that manage some sort of off-chain collateral like gold, dollars, or some other centralized currency.
Examples of off-chain collateral backed tokens:
USDC – uses dollars
USDT – dollars
BUSD – dollars
PMGT – gold
GUSD – dollars
USDP – dollars
2. Tokens that don’t have any sort of collateralization –
These try to rely on some sort of internal economic system to maintain stability. Yeah…we know that doesn’t sound very good. These have been called: Seigniorage-style tokens. As you’ll see in our deeper analysis below, it’s not that they have “no” collateral, but rather, they’re collateralized with a token that the protocol can mint at will.
At the heart of them is a lot of truth though. They understand that price fluctuation is just a factor of supply and demand fluctuations. So they try to implement mechanisms to control supply and demand to stabilize price.
Examples of seigniorage-style tokens:
UST
BitUSD
NuBits
3. Tokens that use on-chain collateral
This is where the money is…literally.
DAI
f(x)USD
FRAX
crvUSD
LUSD
GHO
eUSD
We’ll dive more into how these work. But first, let’s all take a moment of silence for some stable coins that are no longer with us:
Some of these will be explored more in depth
Acala USD - dead because of a bug in the iBTC/aUSD liquidity pool which allowed for a huge amount of minting without adequate collateral
UST – the LUNA debacle. Read more below as this is one we focus on.
BeanStalk – an uncollateralized token in which an attacker found a vulnerability in their system and withdrew $180 million. It is in fact still operational. I have serious concerns about their economics.
TITAN – very similar to UST/LUNA. IRON/TITAN collapsed in 2021.
bitUSD – one of the first stable coins. It was an uncollateralized stable tied to Bitshares that collapsed in 2018. Read more below.
Basis Cash – apparently, this was a test project for Do Kwon and some of the Terra team. It collapsed in 2021. Apparently, Kwon wanted to test whether he could successfully develop a stable coin that would eventually collapse before he tried it on the big stage.
NuBits – NuBits/NuShares was one of the first, along with bitUSD. It eventually collapsed for a similar reason as all the other algorithmic coins.
Brief Rant Time:
Ok, skip this is if you want, but can I just say that after reading stacks of stable coin content, there’s nothing that makes me more noxious than hearing people say one of the following things:
“Although it may be complicated under the hood, it’s very simple for the consumer. Just like the average Bitcoin consumer doesn’t know how elliptical curves work…”
Reading this by a team whose product literally burned later. Just because it’s complicated doesn’t mean it works…
“The price is maintained at $1 because everyone trusts that it’s $1, so if it goes above $1 than people will sell, and if it goes below $1, people will buy it…
This feels to me like such a fundamentally backwards way of looking at it. YES, that is an aspect of stable currencies, but I would argue that arbitrage is a result of historically trustworthy currencies, and not the other way around. At least not for the first 20ish years…
Just please, look out for yourself out there…
A Timeline of Stable Coins:
I am putting this on a timeline because I think it’s useful to understand how stable protocols built off each other.
BitUSD
Launched in 2014
How Did BITUSD Work?
BitUSD was the first stable coin, emerging in 2014. It relied on overcollateralization (approx. 200%) with the use of Bitshares as collateral. Simply put, you could mint bitUSD by locking bitshares up as collateral. The system relied on two things:
An assumption that bitshares would be worth something
Arbitrage and other actors would help maintain a $1 price on bitUSD
If the collateral dropped below 200%, the user would have to commit more collateral (as bitshares) or lose some of their bitUSD. They were also always able to trade 1 bitUSD for $1 of bitshares, which they believed would always ensure that if bitUSD ever dropped too much, people would buy bitUSD to trade it in for $1 of bitshares. Naive much?
This whole thing lasted until 2018 when it collapsed.
The real issue was that they had mechanisms to protect against the falling value of bitshares (collateral), but not the falling price of bitUSD. They made the classical assumption that, why would a $1 stable coin trade at any price other than $1?
Read on from BitShares founder, Dan Larimer:
“The market peg works on the premise that all market participants buy and sell based on what they think market participants will be buying and selling in the future. The only rational choice is to assume that it’s going to trade based on the peg in the future. If you don’t believe that, then you have to decide on which way it’s going to go, up or down. And if you don’t have a way of saying you abstain from the market. If you don’t think it works you sell the shares and get out, as the system’s going to fail in the first place. So its a self reinforcing market peg, that causes the asset to always have the purchasing power of the dollar.”
It’s a little hard for me to identify exactly what happened. There doesn’t seem to be so much info regarding the actual decoupling event. This leads me to point out that we really should have some sort of record keeping for this point in on-chain history.
What I could piece together though was that there was a black swan event that created a cascade of liquidation that wrecked bitshares’ price. This happened at the beginning of 2018 when the whole crypto market ate shit. It drove the price down so swiftly, that it depegged bitUSD (how exactly that happened, I don’t know. Write to me if you were there). The system had no mechanism in place to help bitUSD regain its peg since people could only trade 1 bitUSD for $1 worth of bitshares. But there was no demand for bitshares at that moment and no one was able to get $1 worth of bitshares for 1 bitUSD.
TLDR: What Happened to bitUSD?
Ultimately, I think the lessons are that:
Your stable coin is not worth $1 just because you say it is and expecting market forces like arbitrage to protect its value is, at best, naive, and at worst, fraudulent.
Securing a stable coin with on-chain collateral made up of a highly volatile, unproven asset (that you control), is really dangerous. Bitshares was not reliable enough to serve as good collateral.
There was also a serious problem in that there was no reliable price oracle providing data for pricing the exchange at $1. There was no ChainLink back then.
All in all, the biggest problem with this was that there wasn’t any real mechanism to stabilize the currency at $1, other than the expectation that everyone believes it will be $1. Sound familiar? *cough* UST *cough*...
NuBits
Launched in 2014
How Did Nubits Work?
NuBits worked similarly as bitUSD. Anyone who held NuShares (the governance token) could vote to approve the minting of new NuBits which were then sold to increase supply. They could also decrease supply by allowing for a mechanism they called “Parking,” which was essentially an on-chain bond – incentivized through yield. Holders were paid a yield for locking their NuBits, thus choking supply and hopefully stabilizing the price.
As just an example of how sketchy this all feels, take a look at this article that was published at the time, explaining how NuBits works:
https://bitkan.com/learn/what-is-nubit-how-does-it-work-23994
I mean…who the fuck writes this bullshit? How many ways can we say something while saying nothing?
Next, we have the classic line from their whitepaper. Heard this one before?
What ultimately did them in was the great bitcoin bull run of 2016, which my grandaddy tells me was truly a spectacle to behold. Too many people sold their NuBits to chase gold in a digital land and there wasn’t a meaningful mechanism to contract the supply to balance out the falling demand.
I will say that I have read conflicting reports from a different industry report that said NuBits was collateralized with bitcoin. I have found no other evidence of that and my report went off of what I read on the NuBits website.
TLDR: What Happened To NuBits?
The lessons here are very, very similar to BitUSD, so see above.
I will add that:
NuBits had a very poor mechanism for decreasing demand. The bull market exposed this weakness. Ultimately, stable coins need to be able to weather both bull and bear markets. During bear markets, they will have to deal with collapsing collateral prices and skyrocketing demand while in bull season, they will have to contract supply because of falling demand. Parking was a poor approach to this as it’s unlikely that they could offer attractive enough yields to entice people away from rising assets like BTC.
USDT – Tether
Launched in 2015
How Does USDT/Tether Work?
Tether was the first USD-backed stable coin. It’s had a bit of a rocky road since. Back in 2017, I was pretty damn sure USDT was going to implode at any second, but here it is, limping along more than the Simpsons on Season 35.
My real concern was because there was very little transparency and a lack of audits to validate their dollar reserve holdings.
Fast forward seven years and it looks like Tether probably does have the reserves to back up USDT, and the audits show it. The issue though, as highlighted in the S&P Report is that there is not a lot of transparency when it comes to who acts as the custodians of their reserves. There is also not 100% transparency into the total holdings of Tether. While they can prove 104% overcollateralization, according to reports 15% of those holdings are assets with higher volatility – and there is not enough transparency into what exactly those holdings are.
Buuuut, we have reached the oldest, still-surviving stable. 🍻
USDT TLDR:
Dollar backed crypto with decent auditing transparency, but questions regarding who is holding their assets and exactly what assets they’re holding.
MakerDAO – DAI
Established in 2015. Launched in 2017.
How Does DAI Work?
Previous community members of Bitshares went on to establish MakerDAO and the DAI stable coin system. MakerDAO was the first stable collateralized with on-chain assets.
The protocol works by incentivizing people to lockup collateral. This is really a key design feature we’ll highlight later. If you want on-chain collateral with which to secure your stable, you need to offer incentives for people to give it to you.
MakerDAO realized that they could get people to give them collateral if they looked at it like a loan. If you deposited crypto in MakerDAO, you could then borrow in DAI. They required a certain level of over-collateralization to account for asset volatility. There of course is a fee to borrow, kind of like the interest on a loan, and this interest can be manipulated to either increase demand for DAI loans or decrease them. But viola! it’s not really so complicated when you think of it like this.
DAI can also be deposited and locked which entitles the holder to yield derived from the Maker protocol revenues. We’ll explore this more in the section called “Let’s Talk About Treasury Yields.”
Today, MakerDAO has become kind of like a Novo bank and there’s some definite concerns.
First off, let’s take a look at MakerDAO’s collateral list.
Just to give you a sense, there are:
$6.56 billion in on-chain assets, primarily ETH, liquid staking tokens, and WBTC
$384 million of that number above is in stable coins
$7.92 billion in off-chain assets like treasury bonds, loans,
Some of these off-chain assets come from agreements with corporate entities to take USDC from the Peg-Stability Module (PSM) and convert it into treasury bonds and the like. The PSM allows for people to directly convert one DAI for one USDC, which means that MakerDAO has a lot of USDC sitting around.
There is also a Surplus Fund holding approximately $50 million worth of DAI.
One of the issues with DAI is that the collateral is not fungible across all vaults, meaning, if the collateral in one vault crashed, it would impact DAI everywhere. That’s why they have the Surplus Fund. The real question though – is it enough?
I’ll talk more about some of the implications of all of this below, but one of the biggest things is – why do we trust DAI more than USDC? At what point does our decentralized collateral stable become just another dollar-pegged token? And there was significant scare around DAI when USDC depegged because of Silvergate and Silicon Valley Bank (SVB). If I’m hedging against the dollar than I don’t want to use a stable coin that would be caught up in the cascade of collapse.
A good point made by iBetaTesta was that diversification may not be the best thing for stable tokens. If even a little bit of collateral is lost, that could spell disaster for a protocol. It seems to me that we would want to be 100% collateralized with the MOST SECURE ASSET we can find. Not 50% with the most secure and 50% with the second most secure because…diversity is good…
DAI TLDR:
People deposit crypto as collateral and then take their loan in DAI. DAI is overcollateralized, but a large amount of the collateral is now held as USDC, so it doesn’t really protect you from the risks of using centralized fiat. Regardless, it does have a strong track record and is widely used across DeFi. But I don’t love it.
USDC
Launched in 2018
How Does USDC Work?
USDC works the same way as USDT. It’s maintained by a consortium which includes Circle and Coinbase.
Their reserves are secured by U.S. financial institutions and regularly audited, although they don’t indicate directly where ALL of the funds are held. The S&P Report highlights a couple concerns here:
Even though USDC says this won’t happen, there’s not sufficient legal clarity or precedents to guarantee that the USDC reserves would not be fair-game in a bankruptcy case.
Only Circle partners or Type A users like exchanges can go directly to circle to exchange USDC for dollars. Not sure this matters so much, but just know where you are in the pecking order if there ever was a bank run.
And, let’s talk about the elephant in the room…
In March of 2023, USDC almost depegged (13% below the dollar) because they’d stored $3.3 billion with Silver Valley Bank. Coinbase paused redemptions. The peg was restored when good ole Uncle Sam promised that he’d foot the bill for another bank debacle. We’ll talk about this more, but do we really trust this? I mean, it’s not really reasonable to expect the custodian to store tens of billions of dollars in amounts that would be insured by the FDIC.
Just for fun, look at these numbers:
$28,847,880,902 market cap for USDC
Divided by $250,000 which is the maximum amount a deposit is secured by the FDIC
= 115,391
That’s 115,391 banks that Circle would need to deposit into in order for their funds to be secured. Ha. There’s only 4,700 FDIC banks in America. Soooooo…either USDC needs to diversify into land soon, or Uncle Sam is making promises to them that he’s not making to everyone else. Last year, it was reported that USDC even ditched their treasuries because of the U.S. Debt concerns – basically the same concerns that sunk SVB and Silvergate. I’m not sure if this is true, since, today, as you can see below, they do hold most of their debt as treasuries
Regardless of whether that report was true, the fact that USDC needed to get bailed out after the collapse of SVB and Silvergate speaks to a pretty unprecedented situation. When in history has an organization held $28 billion dollars and not been able to diversify into almost any other asset?
Today – looking at the BlackRock balance sheet, it appears that 59% of their holdings are in government repurchase agreements and 41% are held as government treasuries.
USDC TLDR:
Dollar backed stable. More transparency into reserves than Tether, but still almost imploded with Silvergate and SVB went insolvent. Has a serious problem in where to put their money because there literally aren’t enough FDIC insured institutions in America to hold their dollars and it’s not like Treasury Bonds can be fully trusted.
Paxos USD (USDP)
Launched in 2018
How Does USDP Work?
For reference, BinanceUSD (BUSD) is a white-labeled version of Pax Dollar.
Here folks, to our left, we have another boring fiat-backed token grazing in the fields of centralized exchanges. Really though, this stable is owned by Paxos and they publishes a monthly audit of reserves. The S&P Report likes this one, just like it likes USDC.
Paxos solves the issue of FDIC insurance by holding bonds. As of Oct. 2023, only 28% of the reserves were held in such accounts. The additional reserves are largely held as U.S. Treasury Bonds.
USDP TLDR:
All of the concerns I wrote about with USDC. No real red flags here in the sense that I don’t think there’s any actual fraud happening. Just real risks regarding bank insolvency.
Gemini Dollar (GUSD)
Launched in 2018
How does GUSD Work?
Same old, same old. This stable coin is a dollar-pegged, dollar reserve stable managed by Gemini.
Nothing really to note here. No real concerns different from USDC were noted in the S&P 500 Report.
GUSD TLDR:
See USDP and USDC TLDR.
TrueUSD
Launched in 2018
Oh True USD, founded by Justin Sun, which may I add, I trust less than the FDA. Can someone explain to me how Tron has a $12 BILLION MARKETCAP and the “insights'' section on their website literally has three articles, all from 2020 or before, with headlines such as: “Chinese Crypto Pioneer Pays $4.57 Million for Lunch With Warren Buffett,” and “TRON Founder Justin Sun Graduates Hupan University, Forges Ahead.” 🤨
At the time of writing, they have had a miraculous repeg after losing their peg for two months. I wonder how much ETH had to be sold to regain it?
How Does TrueUSD Work?
TrueUSD has realtime reserve updates. Their reserves are held around the world, all denoted in USD. The S&P Report indicates several concerns:
The reserves could be vulnerable to foreign government instability as they are held at institutions in Switzerland, Hong Kong, and the Bahamas.
It’s not 100% clear that all the reserves being reported are legally bound as collateral. It’s possible that the parent company could use them (maybe).
Not much information available regarding Sun’s holdings and business dealings.
To quote:
“TUSD is not regulated, which we see as a weakness. In addition, there is no publicly available information regarding the segregation of its assets from Techteryx. While the attestation report on TUSD reserves states that Techteryx and its agents are not entitled to these assets and that there are no liens, claims, or security interest in the assets, we have not seen any legal opinion to back this statement.”
Yeah…I’d say that’s concerning.
TrueUSD TLDR:
Already depegging. Serious concerns around who is behind TrueUSD and where they’re storing their money.
Ampleforth (AMPL)
Launched in 2019
How Does Ampleforth Work?
Ampleforth is pretty cool. It aims to track the purchasing power of the dollar in 2019. They automatically rebalance the tokens held by each holder to rebalance the price. In other words, if the price goes above the purchasing power of the dollar in 2019, the tokens you hold are automatically decreased, and vice versa.
Since they use a 24h adjusted price feed, they are much less vulnerable to an oracle attack. The price of AMPL has remained within 20% of the $1 target 70% of the time. It’s not perfect, but it’s a pretty sweet mechanism. And honestly, I respect it. Maybe we don’t need a 100% stable asset?
But also, I’m not really sure who would want to hold an asset that has stability with 20% volatility…
AMPL TLDR:
Pretty cool stable that uses a smart contract to automatically reduce and raise the supply of AMPL to adjust for price fluctuations. However, they don’t really hold a strong peg, fluctuating between 20% under and 20% over their $1 peg.
Some Others We Won’t Talk About
In here there were some other tokens – like Facebook’s Libra – that…well…we just don’t want to have to even talk about them.
FRAX
Launched in 2020
How Does Frax Work?
In the beginning, Frax combined collateralization and algorithmic design properties. Previously, to mint a FRAX, you had to deposit 85% collateral of USDC (and other stables) and 15% FXS (the governance token). This ratio was changeable depending on the fluctuations of FRAX. If there is less market trust in FRAX, more USDC would be required, and vice versa. They also have engaged in the Curve Wars – by holding a substantial amount of veCRV – in order to drive liquidity to their stable coin pools. They also hold a substantial amount of Convex’s token: CVX.
As of recently, the Frax team has recognized that they aren’t fully comfortable with their token being undercollateralized, even though there hasn’t been any issues as of yet. They wisely said, and I’m paraphrasing: that the kind of metrics needed to secure $1 billion in value is not the same as necessary to secure $1 trillion in value.
So, now, they’re making a move to be fully collateralized by moving protocol revenues into the collateralization fund.
FRAX also operates AMOs – Algorithmic Market Operations – which automate the movement of FRAX token in and out of various DeFi protocols. This basically automates the process of ensuring liquidity across DeFi, hopefully preventing liquidity crunches. The AMO also deploys treasury funds – such as those held in USDC – across DeFi to earn yield and distributes those to veFXS holders (veFXS is the governance token that has been locked up for 4 years. I.e. Lock up 1 FXS for 4 years, get 1 veFXS).
Frax is currently in the process of developing and implementing a more decentralized governance model.
The S&P Report does not particularly like Frax, citing smart contract risk, questions regarding regulation, and the lack of a guaranteed peg as their main concerns.
FRAX TLDR:
FRAX is slightly undercollateralized but will soon be overcollateralized. They demonstrated success using their hybrid model – maintaining good supply/demand adjustments across liquidity pools with their AMO.
FRAX TLDR:
Honestly, I like FRAX, mostly because of the ingenuity of the team, and the way they really get DeFi. They’re buying a bunch of treasury bonds too, offering that yield to FRAX stakers. They also hold a lot of USDC. Feels a lot like DAI here.
FEI
Founded in 2021
How Does FEI Work?
I hesitate to even discuss this one considering its chart looks like how my heart beat would look like if I ever met Alexandra Daddario.
But sometimes it’s helpful to look at what not to do. The project has also been abandoned by the core developer team at this point, in part due to the $80 million protocol hack in April of 2023.
How Does Fei Work?
FEI offered a novel approach called direct incentives. Essentially, through automated smart contracts, the protocol manipulates trading pairs of the token on exchanges. Supply and demand, baby. They can withdraw and burn tokens and they can flood tokens into the market. This obviously didn’t work, but I kind of like the idea. Also, it’s cool to see the token largely working, albeit with only a $14 million market cap, without any centralized control.
FEI TLDR:
FEI controls supply and demand by manipulating liquidity pools – kind of like FRAX’s AMO. But a protocol hack caused the project to be abandoned. It’s still operational, but not maintaining a tight peg.
UST – LUNA
Ohhhh yes, everyone’s favorite crypto! You either watched UST crash to zero with horror, seeing your money evaporate, or you watched it with that subtle sense of “I told you so.”
It’s questionable whether LUNA was an algorithmic stable coin or a collateralized one, which really highlights how difficult it can be to distinguish between the two. UST was collateralized with LUNA, the Proof-Of-Stake token for the Luna blockchain. $1 in UST could always be traded for $1 in Luna and vice-versa. Obviously, in some regards we could have considered this as collateralized: collateralized by LUNA. However, we get ourselves into a loop here, because, in many ways, the demand for LUNA was a result of demand for UST. This is especially true when you consider the 19.5% APR offered on UST by Anchor, which was clearly unsustainable.
Terra hoped that demand for LUNA would be similar to market demand for ETH. It begs the question whether UST is so different from an ETH collateralized stable coin and truthfully, I don’t think so. However, issuing a multi-billion dollar on a platform whose only real value is the stable coin itself is much much different than issuing the stable-coin on an established protocol hosting 100s of billions of $$$ in collateral with hundreds of projects offering usable protocols. Also, the critical difference is that more LUNA could be minted at will to back UST, meaning a mass exchange of UST to LUNA could potentially hyperinflate LUNA (which happened). That’s obviously not the case with ETH. A good point here is that it’s not a good sign when the developer of the stable coin has the issuance power over the collateral. This, as the youth would say, is highly sus (I’m all about trying to work more slang into my articles to appeal them to the youth…subscribers are subscribers).
The Terra team did recognize the risk of this and acquired over $3 billion in BTC to help balance out the reserves. We know what happened at this point onwards, but just in case you were part of some monastic order during that period, some traders took advantage of the limited liquidity in the 3Pool – a stable-coin liquidity pool on Curve. Within a couple hours, two trades swapped almost $200 million in UST for USDC. This forced the Terra team to withdraw about $100 mil in UST to try to rebalance the pool which further limited liquidity.
The Terra team continued over the next few days to flood the pools with UST, selling off their reserve BTC. They did manage temporarily to repeg UST, but it was short lived. Without more reserves to deploy, they were high and dry.
This liquidity crunch depegged UST. As its price crashed, more and more people began trading UST for Luna which hyperinflated Luna away. It’s not so unsimilar to what happened with bitUSD/Bitshares.
TLDR: What Went Wrong With UST/Terra?
All of the monetary market operations were being conducted manually. They couldn’t react fast enough.
The team is responsible for maintaining liquidity in pools through manual operations.
They were paying 19.5% on UST. This was not native yield, but earned through lending. We’ll discuss this more later.
It had dog-shit backing it
All in all, it was a colossal and extremely costly failure. Let’s at least learn from it.
Curve Stable Coin: crvUSD
Launched in May of 2023
How Does crvUSD Work?
crvUSD works very similarly to MakerDAO where users deposit collateral and then take a loan out in crvUSD. It seems logical that most debt markets would issue such a token of their own and, indeed, it seems like that is happening. Just check out Aave’s GHO.
What Curve’s crvUSD does well is it innovatively does something called soft liquidation – called LLAMMA.
With normal debt positions, there is a set liquidation range. If you fall below that range, you get liquidated, usually with a pretty heavy penalty. The penalty is a result of needing to liquidate large amounts at once (think massive slippage). LLAMA gradually converts collateral into crvUSD as the collateral price drops, and then can un-liquidate as the price climbs again. I like it when a protocol deploys their own stable because they need it – i.e. it unlocks a novel mechanism – rather than just because they can.
One note with Curve is that, yeah, I know it’s a joke that they make it look like Windows98, but eventually, the joke gets old and it just looks like Windows98. I really think, at this point, their UI/UX is limiting their growth.
crvUSD TLDR:
Backed by a basket of collateral, you can deposit either WBTC, WstETH, ETH, or some other ETH LSDs to withdraw crvUSD. They have a novel approach to reducing the risk of liquidity.
AAVE - GHO
Launched in July of 2023
How Does GHO Work?
I don’t know if you’re tired of reading these coin analyses but I sure as hell am tired of writing them. The last one we’ll look at is GHO, a stable launched by Aave in July of 2023.
GHO is pegged to the USD, although they use a basket of tokens as collateral.
Fees earned through GHO lending do not go to liquidity providers but rather go to the DAO. These funds are then distributed to those holders who stake the GHO as native yield. Your GHO stake can be used to cover protocol losses in case of an emergency. That’s the risk you take.
I truly don’t see any major differences between GHO and say crvUSD. However, GHO can be minted by depositing on Aave, which gives the protocol more control over their loan currency.
Also, it hasn’t really been successful maintaining its peg. I think this speaks to why protocols like DAI implemented their 1:1 redemption for USDC policy.
GHO TLDR:
Collateralized similar to crvUSD. Deposit one of their approved currencies, overcollateralized your position, and you can withdraw in GHO. I don’t see anything particularly special here or concerning. They haven’t had amazing success maintaining a tight peg, although I don’t think this indicates a larger risk of total collapse, more just that it’s very hard to have a tight peg to $1 unless you offer redemptions like DAI.
Ethena – USDe
Launched, beginning of 2024. You can read my full review here.
How Does USDe Work?
Ethena focuses on establishing a decentralized stable coin uncoupled from the legacy banking industry. They also cite the need to provide non-Americans with a dollar-denominated, yield-generating savings instrument.
They hold ETH and hold ETH-shorts to create a zero volatility ETH-based stable coin that does not have to be overcollateralized.
I think they share a lot of similarities with f(x) Protocol (read further on). The real difference between the two is that while f(x) Protocol allows other customers to purchase the ETH volatility in the form of xETH which represents leveraged ETH positions, Ethena simply holds short positions on their balance sheet. Fundamentally, it’s the same idea. Both base their stable in ETH and benefit from the ETH yield. The fact that it does not need to be overcollateralized makes it extremely capital efficient.
It is also much more centralized than f(x) protocol which, Ethena argues, is not a problem. They say that the market already has its decentralized currency – Bitcoin – and that the stable coin just needs to be stable, capital efficient, and decoupled from the TradFi banking industry.
I’ll be curious to see how this one grows. I kind of want to find something to hate about it, but I can’t.
Stable Coin Takeaways
Hopefully you got some sleep between reading the massive stable coin article and this one. I, of course, did not have that luxury. So with bloodshot eyes and a kink in my neck, we get to the part I like most: why does this all matters?
To summarize what we’ve covered, we’ve really got fiat-backed stable coins and we have non-fiat backed stables. Fiat backed coins are really a simple matter and if you like them, you like them. They’re reliable and guaranteed 1:1 redemptions do a lot for a good night's sleep. But let’s just quickly run down the problems with them:
The Problems With Fiat-Backed Stable Coins:
You have to rely on a centralized entity – this includes the custodian of the peg
The custodians can be legally ordered to freeze assets
You are exposed to inflationary risks and any risks associated with the U.S. Dollar (or whatever fiat they’re using) – this is true not only for backed stables but any stable that is pegged to the USD, even if it’s not collateralized with USD.
You are exposed to the potential of fraud
Assuming that the reserves are held at banks, it’s very hard to find enough banks to store your money so as to insure it all with FDIC. Do you trust the U.S. Government’s unofficial agreement to bail them out like they did with Silvergate?
You don’t have native yield – (see section below for more)
The Problems With Algorithmic Stable Coins:
I mean, I feel like this is so obvious I don’t even need to write it. They don’t really work and I sure as hell wouldn’t trust them. I mean, I barely trust the U.S. Dollar or as I like to call it the Great American Algorithmic Stable.
Eventually, all algorithmic stables seem to bet on the same process: if we can convince enough people that our economics are sound – basically by having a governance token – arbitrage will keep us stable. This just proves wrong time and again. What I do respect is Ampleforth. I think that when it comes to Algorithmic stables, perhaps they’ll work as long as we don’t expect them to always be stable. If we can allow for like 15% fluctuations, then they can be fine alternatives. But truthfully, in what economic world are we ok holding an asset like that?
The Problem With Collateralized Tokens:
So if you’re with me until now, either a fiat-backed token works for you or you have your doubts and want a token that uses on-chain collateral.
Our models for on-chain collateral are really DAI, crvUSD, GHO, and FRAX.
Even collateralized tokens have trouble maintaining their peg (see GHO). This is why DAI implemented their 1:1 exchange with USDC.
Less Capital Efficient – If we don’t collateralize our stable coin, we end up with a much more capital-efficient model. By capital-efficient we mean that we would be helping make it rain. If I only require 50% collateralization, I’m essentially doubling the money supply in the economy. The dollar is by far the most capital-efficient currency out there because, well, it is basically 0% collateralized. The issue with this is…well obvious. Just look at Terra Luna, BitUSD, Nubits, or the dollar for that matter. Really, the dollar is the only non-collateralized model we would ever trust, because, well, they’re backed by nuclear bombs. Without nuclear bombs, global trade deals, and a too-big-to-fail narrative, uncollateralized stable coins become so capital-efficient that they eventually go to zero. Eventually…always. Even Frax has come to recognize this, which is why they’re transitioning to being fully collateralized.
A pegged $1 stable coin is not protected from inflation, even if it’s not backed by a dollar – Just realize, that even if your token is not backed by dollars, if it’s pegged to a dollar, your purchasing power will be corroded by inflation.
All in all, I am really surprised by how, nearly all of the stables in the market, are tied to dollars one way or another. Either they hold USD as a peg reserve, or they hold USDC or Treasury Bonds. I understand the inclination to do this. However, in the end, I think I would rather hold USDC than DAI. At least with USDC, you’re legally guaranteed a 1:1 redemption (unless it all goes to shit that is).
The Importance of Native Yield for Stable Coin Protocols
A point we want to make is that native yield is important. If it doesn’t have native yield, people get all fancy trying to generate yield. I mean, what kind of person can sit on a treasury of tens of millions in assets and get no yield? Not even the buddha could have taken that temptation. So…they end up lending it out to a bunch of off-chain, centralized entities that really shouldn’t be trusted and they end up collapsing along with the rest of them. Just look at Anchor and UST.
The most obvious native yield is from Proof-of-Stake protocols like Ethereum or Solana. I think this is also the most secure because, well, if the yield collapses, the token also collapses. There is no more risk added by the yield which is really the key here.
So, what about protocol yield? For example, what about with GHO and with DAI where you are entitled to protocol yield by staking your token? First I just want to say that I don’t really have a problem with these protocols doing this. They pay you proceeds from the protocol and, in exchange, they can take your money in an emergency.
The question I want to explore is whether this throws off the economics of the system.
If ever demand for a token is driven by a fundamentally unsustainable model, that’s a huge red flag. That’s essentially a pyramid scheme. UST is a great example of that. You cannot promise 19.5% yields forever. It’s not sustainable. What happens then to the demand for the token when that goes away? Is there more demand for the token outside of that unsustainable yield, or is it totally just people chasing air? In other words, if the yield is only sustainable as long as more and more people deposit, that’s a no go.
Native yield on ETH is great because it isn’t going anywhere. It is sustainable so we can trust that any demand for ETH yield is sustainable demand.
But what about with DAI?
I would say that DAI is a tier below what I would call native yield. I will call it protocol-native-yield. Protocol-native-yield is somewhat less sustainable because it requires the protocol to continue to be profitable. However, unlike with a pyramid scheme, DAI’s yield does not depend on more people depositing DAI. So while some demand for DAI might be a result of the yield, I don’t think it is fundamentally unsustainable. The only downside is that it is not as consistent or reliable as platform PoS yield.
TLDR:
There is a big incentive to generate yield on treasury funds. Yield has to come from somewhere. Native yield from PoS is the most secure. Lending out those funds to others to generate yield is the least secure. Offering protocol revenue as yield is a decently secure model. Just know where that yield is coming from, whether it’s sustainable, and what would happen to the price of the underlying token if that yield disappeared.
Realizations I had While Writing These Articles That Might Be Helpful To You
When it comes down to it, a big part of the economics of a stable is a tight control of the supply of that token. This is true whether we are talking collateralized tokens or uncollateralized tokens.
A huge reserve – like Terra’s $3 billion BTC reserve – are often just a distraction. You’re either have at least 100% collateral or you don’t.
Liquidity across pools is important. With DAI, ultimately, the network is as weak as its weakest debt market.
Diversification may not be a good thing. Would you rather have 100% of the collateral in the most trustworthy asset, or 50% in the most trustworthy asset and 50% in the second most trustworthy asset – especially considering a 50% failure is still total implosion.
It’s a CRAZY new question how these fiat-backed tokens are storing their dollars? I mean – there’s only 4000ish FDIC insured banks in America. USDC itself would require 115,391 such institutions to have proper insurance. There’s never been such a situation before. I think everyone is simply betting on the government to extend that insurance – like they did in 2023. I’m not sure I like that.
Overcollateralization and liquidation plans are how stables address volatility.
The Perfect Stable Coin: A Thought Experiment
I want to run a little thought experiment.
Maybe you read through the analysis of each stable coin or maybe you just skipped ahead to here or maybe you were one of those boomers who learned to speed read in school. Regardless, here’s what we would design if we were designing the next stable coin.
First, we have to consider whether we want to collateralize our token or rely on debt to make it run. If we don’t collateralize our token, essentially we’re running a debt market, right? If I let you take a $1 stable coin for any amount of money under $1, than I am essentially lending you money.
So no uncollateralized stable coins. The next question we have to answer is, what type of collateralization? Back it with the dollar or back it with a token that is backed by the dollar? I mean…if you’re into that kind of thing. And by that I mean, are you into aligning your protocol and digital wealth with the very thing that this whole system is trying to offer an alternative to, than by all means. Go grab yourself some of that delicious dollar-backed-stable-coin at your local exchange. It’s always only $1, but your $1 dollar is now worth .95 cents in goods since I started writing this article.
I don’t mean to disparage the whole dollar-backed stable coin idea. I mean, it works, right? It works as well as the U.S. Dollar works. There’s a purpose for it. Buuuut, it clearly cannot serve all of our needs.
So no dollar. The other criteria is that we would really like for it to have native-yield for reasons discussed above – specifically, that if it doesn’t have native yield, risky loans and questionable collateral management strategies become much more likely.
That leaves us with a handful of choices – ETH, SOL – really any token with proof of stake or proof of something that generates yield for its stakers. Bitcoin isn’t a bad collateral choice other than the fact that it doesn’t have native yield.
Which token you choose really depends on which ecosystem you want to bet on. In the end, the difference between 100% collateralization and 0% collateralization really depends on whether the asset is worth anything. 100% collateralization with SOL is worthless if Solana becomes worthless. Now, this would lead us down a very different rabbit hole, but personally, we wouldn’t want to bet on any ecosystem other than Ethereum right now. I get that this could fail. Bitcoin could fail. Everything could fail. We could be pricing things in corndogs by next summer, but ETH is by far our favorite on-chain, native yield generating asset.
Another point is that, if our stable coin is to be used by on-chain entities like DAOs, crypto-hodlers, DeFi protocols, etc. – then collateralizing the stable coin with the token used by those organizations seems fair. I mean, it’s kind of like trying to invest for a doomsday situation. If everything is destroyed, it doesn’t matter what you chose.
The final issue is volatility. Our on-chain, native-yield asset is volatile AF (as the youth would say). This is where the real fun comes in.
Dealing with volatility is where the ingenuity and additional functionality really come into play. It’s pretty amazing honestly how the creative approaches to mitigating the risks associated with volatility actually unlock a whole range of new functionality.
Generally, we require overcollateralization AND a peg to the USD. The problem is that even pegging it to the USD without collateralizing it with USDs, exposes holders to inflation. I would love to see a model where we could peg the stable to some other price outside of a centralized fiat?
So to summarize, we are looking for a stable with native yield, collateralized by on-chain assets, and not priced in dollars. Do you see the potential of such a token?
Final Thoughts
What I see in the market is a growing recognition that – despite having numerous different projects – we actually don’t have many stable protocol variations. We basically have fiat-backed tokens (USDC, USDT, GUSD, etc) and we have wrapped fiat-backed tokens (FRAX, DAI). At the reserve level, everyone is holding treasuries for their native yield and stability.
The new area of innovation is finding ways to unlock crypto-native yield, like ETH staking rewards, offering that to users, and finding a way to balance out the volatility. The two projects I see doing this right now are Ethena and f(x) Protocol. Ethena – if it works – has the advantage of it being a much more capital efficient model. They are centralized and backed by VCs. f(x) Protocol is bootstrapped. They aren’t only a stable coin but are offering ETH-pegged low volatility and liquidation proof leverage to their customers. We’ll be publishing an article breaking down f(x) Protocol soon. Subscribe for free so you don’t miss it.
This is where we leave you. From here on out, you don’t need us. You now have a solid understanding of the stable coin ecosystem, but perhaps even more importantly, you understand the foundational problems every stable coin must answer and you understand what design techniques have been tried. Now, it becomes about either you building a solution yourself, or identifying early projects that are building one.
If you’re building one, hit us up. We’d like to hear about it. If you want to see what we’ve found, follow us.
As always, none of this is financial advice. I am not telling you to buy or sell anything, just sharing my underlying research and conclusions.
I do hold a portfolio of cryptocurrencies. I was not paid by anyone to write this article. I did it on my own with time that earned me zero dollars.
If you would like to support this writing, please, leave me a comment (I love your feedback), pledge on Substack for when/if this newsletter ever goes paid-mode (you won’t be charged, only if/ever we make this paid:
…or you can buy me a cup of coffee or send some ERC20 tokens my way!
ERC-20 Address: 0x4F9960B2A808e8083C2D12420Aba4496D10e6D37