Originally sent to VIXCONTANGO subscribers on July 16th, 2021
In the last newsletter, I covered Central Bank Digital Currencies (CBDC) and stablecoins (USDT, USDC and BUSD) which is what Wall Street, central banks and governments around the world are talking about right now. In this piece, I want to talk about stablecoins as the crypto industry views them because they are a completely different animal. I will call these “algorithmic stablecoins” to differentiate from the proper stablecoins like USDC. It is important for you to understand what is going on with the algorithmic stablecoins because you can lose a lot of money there and very quickly at that. One thing I want to mention is that algorithmic stablecoins have gotten no market traction and practically nobody is using them. All the real action is in proper stablecoins like USDT and USDC and that is what regulators currently are concerned about and looking to regulate. Algorithmic stablecoins aren’t really on the regulators radar yet because they are not big enough. But still you need to know about them because the crypto media talks about them incessantly and many people have been duped and lost money in them.
According to the ethos of the crypto world stablecoins like Tether’s USDT carry at least 3 different “centralization” risks: the collateral is a risk, the collateral manager/management is a risk and the custody is a risk. What does that mean? In Tether’s USDT, Paolo Ardonio is the fund manager of the assets (commercial paper, treasuries, cash deposits, etc) that back the USDT stablecoin. Paolo Ardonio may decide to sensor some USDT holders and invalidate their holdings of USDT without their knowledge (make the addresses holding USDT invalid). He may do it because some government is forcing him to do it. This is “custodial risk”. Paolo Ardonio makes discretionary decisions about what collateral to back USDT with. He may stop buying commercial paper and instead buy more risky financial instruments without telling anybody. This is “collateral management risk”. And finally, Paolo Ardonio may own commercial paper but then the commercial paper issuers decide to not redeem their paper for US dollars because the government told them to stop doing business with Tether. This is “collateral risk”.
A stablecoin like USDT presents all kinds of problems for crypto people who believe in censorship-resistance. That was the original idea behind Bitcoin – the government can’t take it from you or at the very minimum you can evade government controls quickly. Serious DeFi applications don’t like Bitcoin as a settlement currency because it is very volatile vs US dollar which is what proper businesses pay taxes in. DeFi apps want to use a settlement currency that is more stable. That is what led to the rise of USDT and USDC over the past 2 years. But for some crypto users they are not good enough because they are too centralized along these 3 vectors. So degens have made a number of different attempts to create algorithmic stablecoins that decentralize custodial risk, collateral risk and collateral management risk. Please note that in some white papers/articles about stablecoins, the word “collateral” is replaced with “reserve”. I am going to mostly use the world “collateral” here but “reserve” and “collateral” mean the same thing.
The first type of algo stablecoin is the one that tries to decentralize custody. The custody of the collateral is algorithmic and can’t be influenced by a single individual. Maybe you have a DAO (decentralized autonomous organization) that makes decision about the algorithm’s main parameters but by and large the custody is done programmatically by a decentralized network of computers. In other words, the FBI can’t go and order that protocol to freeze your assets like they could do with Paolo Ardonio. Such algo stablecoin is DAI on Ethereum network. And pretty much all algo stablecoins feature decentralized custody.
Decentralized Collateral/Reserve Management
This type of algo stablecoin looks to take Paolo Ardonio (the stablecoin fund manager) out of the selection process of fiat assets that back the stablecoin. Instead, an algorithm is used that automatically buys and sells the financial instruments backing the stablecoin depending on market conditions. Such a stablecoin is again DAI. DAI is over-collateralized – in other words, they might take commercial paper but it will take $1.20 of commercial paper to issue $1 DAI. Then the algorithm monitors the market conditions of the commercial paper and if the price starts to go down towards $1, it liquidates it (sells it) and gives the commercial paper back and gets 1 US dollar. Then 1 DAI can be redeemed for $1. DAI automates both the custody and collateral management but the collateral is still fiat assets like commercial paper, US treasuries or stablecoins like USDC. Many people consider DAI to be the equivalent of an algorithmic fiat bank. Note that all major algo stablecoins like Empty Set Dollar (ESD), Frax (FRAX), Fei Protocol (FEI) or TerraUSD (UST) also feature decentralized/algorithmic collateral management.
In addition to automating the collateral management process, degens want to decentralize the collateral itself. They don’t want it to be the US dollar. They want a synthetic US dollar – some kind of combination of assets that adds up to 1 US dollar. That combination of assets can be of 3 different kinds:
- Fractional fiat backing – this is equivalent to fractional banking. You don’t have a full $1 of assets as collateral, instead you have some US dollar assets (let’s say 25%) but the rest are crypto assets (75%). Then the algorithms buys and sells the crypto according to market conditions to maintain the peg. Such a coin is Frax (FRAX) or TerraUSD (UST). This approach hasn’t worked very well for Frax at all. For UST, the implementation is considered to be more successful probably because the protocol’s invests in POS (proof-of-stake) coins and their rewards are help to stabilize the crypto portion of the collateral.
- Full crypto backing – in this type of stablecoin there is no fiat backing in USDC or any other fiat asset at all. Crypto coins like Ethereum (ETH) are used to back the coin and then the ETH held in reserve is being bought and sold to keep the peg at $1. Such coin are Empty Set Dollar (ESD). ESD has been quite the debacle. ESD currently trades at 11 cents and has been a non-stop trip down.
- No collateral at all – in this type of stablecoin there is no collateral held in reserve at all. Not even crypto as collateral. Instead there is some buying and selling of some protocol created token that supposedly keeps the price at $1. Such a coin is Fei Protocol (FEI). A normal person would ask – how can that be? How can there be no collateral? I ask the same question myself and I don’t understand it. All you need to know though is that FEI dropped a lot on the day of issuance and many people lost a ton of money. FEI came out around April 1st and after this debacle the crypto market topped. A lot of Ethereum people lost a lot of money in that fiasco and the disappearance of those funds certainly hurt the crypto rally. I think FEI is closer to $1 and seems to maintaining the peg but again, this is a “synthetic” US dollar. There is nothing behind it. FEI usage has remained very low.
Decentralized Peg / Universal Value Stablecoin (UVS)
The final type of algo stablecoin is one with decentralized custody, collateral and collateral management but one which is not pegged to the US dollar or any other sovereign currency. Instead it is its own entity which has been programmed to have low volatility against the US dollar or other sovereign currencies. You can think of it as similar to the Chinese Yuan in that the Yuan is kind of pegged to the US dollar but can move against with some pre-determined daily volatility. Also the Yuan can strengthen or weaken against the US dollar over time but generally speaking within 2-3% annual volatility. There maybe years where the Yuan strengthens or weakens against the US dollar by 5 to 10% but those years are rare. A universal value stablecoin will behave like this – with limited daily and annual volatility. The Facebook’s Libra (currently called Diem) concept is something similar – a stablecoin not pegged directly to any specific currency but is pegged to a basket of sovereign currencies. In the UVS case, the backing will be a basket of cryptocurrencies (Bitcoin, Ethereum, Cardano, etc) and the reserve management mechanism will be swapping in and out of those to keep that universal value stable and not very volatile against the US dollar.
No such UVS exists currently as far as I can tell, but Cardano’s Charles Hoskinson just teased a whitepaper for a stablecoin called the Djed Stablecoin which might attempt to have this design. Hoskinson envisions that stablecoin to be the settlement currency inside the Cardano DeFi ecosystem. Hoskinson does not view ADA as a settlement currency for Cardano – he views it correctly as a token which represents share of the usage of the Cardano network. Nor does he view the US dollar as a settlement currency for Cardano. I haven’t read the Djed paper because it is not out yet, but from preliminary descriptions by Hoskinson, I assume the above description is somewhat representative.
I haven’t covered algo stablecoins over the past 2 years for a reason, even though there has been a lot of action in that space. The simple reason is that I think algo stablecoins are a total scam. While maybe you can technically come up with a formula that prints digital US dollars out of other digital tokens (whether they are fiat or crypto), the LEGAL and ECONOMIC reality is that you are printing unauthorized liabilities of the US government. Which you simply can’t do. You can’t conjure up liabilities of the US government out of thin air (or out of crypto). Only the US government (or more specifically the Fed) can print US dollars. No other private or public entity in the world can print US dollars. So my issue with algo stablecoins is that they are the modern equivalent of alchemy. Before modern times, aristocrats were trying to convert lead to gold by hiring chemists. This is the same thing – trying to convert digital tokens to US dollars. You simply can’t do it. If the digital token is fully backed by dollar denominated stable assets and redeemable to $1 like USDC, yes, then it is a digital dollar. Any other combination simply can’t yield $1 even if the math works out. Fractional backing is a scam. All crypto backing is a scam. No backing is an even bigger scam. At best algo stablecoins create a “synthetic” US dollar. It is not real. If you could print synthetic US dollars, you can devalue the US dollar to zero tomorrow by printing digital tokens. That’s obviously not going to happen. And the US government will not only delegitimize these attempts in law but will also hunt down and incarcerate the people making and trading these synthetic dollars. Creating counterfeit US dollars is a serious crime. The US government doesn’t allow its paper currency to be forged why would it allow the digital representation to be forged? It won’t. And the penalties for forgery are pretty rough. So I have never taken any of these attempts seriously and I wouldn’t touch them with 10 foot pole. FRAX, FEI, ESD, UST, I am not going anywhere near these. These are counterfeit US dollars. They are illegal in every sense of the world.
An algo stablecoin that can’t be pegged to the US dollar is the only viable attempt at this. The peg must be decentralized. Maybe it can have limited volatility versus the US dollar but it most certainly can’t be pegged against it and perceived as a liability of the US government. A universal value stablecoin also can’t have sovereign currencies as a collateral like Diem does. A private entity would be creating liabilities for all the governments involved which it simply can’t do. That’s why I have always been skeptical of Diem. However, a non-pegged stablecoin backed by a basket of digital tokens can make sense.
Long story short – use USDC for US dollar stablecoin and avoid all the unbacked Ethereum experiments. Also avoid Diem.