Stablecoin Basics
Definitions Matter
There has been much sound and fury in the press recently about stablecoins, following the demise of Terra and UST, the wobbles of Tether, and the ongoing questions and confusion about Circle. However, one thing that is consistent across almost all coverage of the stablecoin space is a lack of knowledge and rigor that leads to a riotous mishmash of word salad and unclear meanings (often from issuers themselves, attempting to disguise inconvenient facts).
To that end, I thought I would write this, as someone who actually has done this with my hands, to provide some foundational pieces for the discourse around stablecoins.
What is a Stablecoin?
A stablecoin is a token that attempts to maintain a price peg to a real-world fiat currency.
That’s it. It’s really that simple.
However, like many simple statements, a significant amount of complexity is contained within that definition. First, there are several types of stablecoins, and each of them have a common set of features with regard to what they are attempting.
The types of stablecoins are:
- Fiat-backed stablecoins, which are stablecoins that have a pool of real-world assets behind them to maintain the peg
- Crypto-backed stablecoins, which are stablecoins that attempt to use overcollateralized pools of crypto assets to maintain the peg (sometimes also with real-world assets mixed in)
- Algorithmic stablecoins, which are a class of stablecoin that attempts to use only relationships to other crypto assets to maintain the peg. Notably, none of these attempts have been particularly successful as of yet.
What are the common elements of stablecoins?
All stablecoins have three elements in common, and most (though not necessarily algorithmic coins) have a fourth element.
Fiat Currency Peg
The first choice for all stablecoins is what fiat currency they are going to be pegged to. The most common choice by a significant supermajority is the US Dollar, though fiat-backed stablecoins exist pegged to the Euro, Singaporean Dollar, and others. Certain algorithmic coins have also existed backed by currencies where holding reserves is not a concern, such as the Korean Won or Mongolian tögrög on the now all but defunct Terra blockchain.
Reserve
All stablecoins also need a reserve in order to hold their peg.
For fiat-backed stablecoins, this is usually a pool of conservative, low volatility real-world assets in the appropriate currency (or, theoretically, in a different currency with appropriate FX swaps back to the original currency, but I am not aware of this model being used anywhere).
For crypto-backed stablecoins, this is usually a pool of crypto assets with various haircuts applied to the asset values, able to be liquidated into fiat-backed stablecoins in the case of a significant market decline. For instance, a person might have to deposit $1.5 of BTC in order to mint a single $1 stablecoin.
For algorithmic stablecoins, this is where the fun starts and why they have not been successful so far. Algorithmic stablecoins typically use another crypto asset as the shock absorber / reserve, with the idea that users can mint an algorithmic stablecoin requiring some or all of the contra-asset to be burned in this minting process. The problem, historically, is that the value of this asset tends to be related to the stability of the algorithmic stablecoin, so when it loses its peg, the reserve asset also declines dramatically, and thus a death spiral begins.
Peg Stability Mechanism
A stablecoin is just a coin. The actual mechanics of making sure it is worth a unit of fiat currency at all times in markets driven by supply and demand are what make the coin stable. Thus, to actually be stable, there has to be a mechanism for ensuring that when the value of the coin deviates from the peg, it can be restored rapidly and efficiently.
For fiat-backed stablecoins, this mechanism is usually direct conversion into the reserve that leads to arbitrage profits if the coin trades above or below the peg. Put differently, if a coin can always be minted by giving the reserve $1 or redeemed for $1 by giving the coin back to the reserve, then if the coin trades off peg there is a profit opportunity. If the coin is above $1, mint for $1 and sell for more than $1. If the coin is below $1, buy for less than $1 and burn it for $1.
For crypto-backed stablecoins, the dance is similar. Here, the mechanisms are slightly more divergent across the space, but in general the principle is that a coin is trading above $1, there becomes an incentive to create more until the peg stabilizes, and if a coin is trading below $1, there becomes an incentive to destroy coins until they return to par. Sometimes there is also the ability to explicitly exchange coins for assets worth exactly $1, similar to a fiat reserve (see the PSM for DAI, as an example).
For algorithmic stablecoins, this usually involves an explicit exchange of crypto assets. The ill-fated IRON finance had the model of requiring a certain amount of USDC plus a certain amount of the TITAN token (the project token) to mint a IRON stablecoin. Redeeming the coin meant getting back some combination of USDC and TITAN. The even more ill-fated UST (Terra’s stablecoin) was created by burning $1 of LUNA, the chain token, or destroyed in order to receive $1 of LUNA back.
Transfer Mechanism
Most relevant to fiat-backed stablecoins (though also possibly, depending on the implementation, crypto-backed stablecoins), the transfer mechanism refers to the ability to actually move between the two separate markets of traditional fiat rails and crypto rails.
There are a number of complications that arise in moving between these two worlds: crypto is permissionless (mostly) and always on, while traditional rails are typically highly permissioned and on less than 1/3rd of the time measured by hours for many types of products. Providing 24/7 liquidity for peg stability via mint/burn is non-trivial in this context, and all fiat-backed stablecoins will need some method of dealing with this mismatch in rail capabilities.
Crypto-backed coins have a potentially lesser version of this problem if they have incorporated real-world assets or if their liquidation mechanism relies on any sort of transfer. This is one area where algorithmic coins have one less degree of complexity, though the inability to ultimately rest on more stable forms of real-world assets means that there is a very significant (and so far, always fatal) cost to this tradeoff.
So What?
Given the rapid emergence of the stablecoin space from basically non-existent at the start of 2018 to over $150B of stablecoins outstanding today, increased understanding of the structure, risks, and utility of stablecoins is going to become essential to understanding crypto markets overall.
Therefore, I wanted to lay out the basic structure and language in order to discuss them going forward, and start to arm you, dear reader, with the tools to understand if someone talking about a stablecoin knows what they are and how to approach them, or if it’s more pontificating and screaming from a crypto scammer of fossilized tradfi dinosaur. As a general rule, if someone cannot lay out the characteristics of what these are and how they work for a stablecoin, that person does not know what they are talking about and should likely be disregarded.
I hope this helps.
