An Overview of Stablecoins
Stablecoins have been one of my major areas of interest since I got involved in crypto. I first learned of Bitcoin when I was studying abroad in Argentina in 2014. At the time, Argentina was in the midst of a currency crisis that had resulted in widespread inflation, and Argentinian citizens were still reeling from a 2001 market crash that ended with the government freezing bank accounts for a year.
To combat hyperinflation, Argentinian citizens developed a thriving black market for US dollars. Citizens purchased US dollars legally, but their purchases were capped. People regularly exchanged pesos for dollars from black market vendors, literally stashing their savings in mattresses. When the government currency stopped serving their needs, people widely turned to other currencies. Amidst this crisis, certain features of Bitcoin had serious appeal. It was a currency that wasn’t controlled by any government, could easily be sent or transported internationally without interference, was easier to safeguard than physical dollars, and couldn’t be seized. The only problem was that Bitcoin wasn’t a safe hedge against fiat inflation because Bitcoin itself was too volatile.
The solution is a stablecoin. Stablecoins, in their most ideal form, are simply cryptocurrencies with stable value. They share all the features listed above that make Bitcoin so appealing, but don’t suffer from the same volatility, making them much more usable as a store of value, medium of exchange, and unit of account.
Stablecoins are one of the highest convexity opportunities in crypto. They aim to become global, fiat-free, digital cash, so the total addressable market (TAM) is simply that of all the money in the world: ~$90T. The opportunity for stablecoins is, intrinsically, the largest possible TAM. This vision is larger than that of Bitcoin itself. A fiat-free currency that’s price stable will challenge the legitimacy of weak governments around the world.
You might wonder, how can one profit from a system whose final product is intrinsically a price-stable asset? All of the trustless stablecoins described below have some sort of associated equity-like token, which yields cash flows from the stable functioning of the system.
At a high level, a trustless, fiat-free stablecoin sounds impossible. How can a free-floating currency remain price stable given the natural ebbs and flows of supply and demand? The concept, on the surface, appears to violate basic economic principles. Despite the perceived challenges, many teams are attempting to create stablecoins.
I contend that there are four features that a cryptocurrency needs in order to become global, fiat-free, digital cash:
- Price stability
- Decentralization (i.e. collateral is not held by a single entity, like Tether)
None of the current stablecoin projects have all of these features, but some are aiming to offer all of these. Scalability and privacy are likely further out. But stable, decentralized cryptoassets are possible today.
There are three fundamental approaches to designing stablecoins: centralized IOU issuance, collateral backed, and seigniorage shares. I’ll examine each below.
Stablecoin Model #1: Centralized IOU Issuance
The first is to issue IOUs. This is the model used by tokens like Tether and Digix. Here, a centralized company holds assets in a bank account or vault and issues tokens that represent a claim on the underlying assets. The digital token has value because it represents a claim on another asset with some defined value. The problem with this approach is that it is centralized. These tokens require trust in the issuing party– that they actually own the assets being represented and that they are willing to honor the IOUs. This model imposes serious counterparty risk on holders of the token. Tether is the canonical example given the serious concerns that the public has about their solvency and legitimacy.
Stablecoin Model #2: Collateral Backed
The second approach is to create stablecoins that are backed by other trustless assets on-chain. This model was pioneered by BitShares. It’s also the model used by Maker, Havven, and others (see table below). In this model, the collateral backing the stablecoin is itself a decentralized cryptoasset. In the case of Maker, for example, Maker’s Dai stablecoin is backed by ETH held as collateral in an Ethereum smart contract. This approach has the benefit of being decentralized. The collateral is held trustlessly in a smart contract, so users aren’t relying on any third party to redeem it.
In short, this approach allows users to create stablecoins by locking up collateral in excess of the amount of stablecoins created. For example, a Maker user could generate $100 worth of Dai stablecoins by locking up $150 worth of Ether. The collateral is held in a smart contract, where it can be accessed by paying back the stablecoin debt, or can be automatically sold by the contract software if the collateral falls below a certain threshold. This allows for collateral-backed stablecoins that don’t require trust in a central party.
The problem, of course, is that the collateral backing the stablecoin is often a volatile cryptoasset such as BTS or ETH. If the value of this asset drops too quickly, the stablecoins issued could become undercollateralized. For this reason, most of the projects using this model require that the stablecoins be overcollateralized enough to protect against sharp price movements. While this can provide some degree of certainty, there always exists the possibility of a black swan event that causes collateral prices to drop so quickly that the stablecoins are undercollateralized. Projects using on-chain collateral have different approaches for handling black swan events.
Stablecoin Model #3: Seigniorage Shares
The final approach is the seigniorage shares approach, which algorithmically expands and contracts the supply of the price-stable currency much like a central bank does with fiat currencies. These stablecoins are not actually “backed” by anything other than the expectation that they will retain a certain value.
In this model, some initial allocation of stablecoin tokens is created. They are pegged to some asset such as USD. As total demand for the stablecoin increases or decreases, the supply automatically changes in response. While different projects use different methods to expand and contract the stablecoin supply, the most commonly used is the “bonds and shares” method introduced by Basecoin.
As the network grows, so too does demand for the stablecoins. Given fixed supply, an increase in demand will cause the price to increase. In the seigniorage shares model, however, increased demand causes the system to issue new stablecoins, thus increasing supply, and ultimately lowering price to the target level. This works conversely, using “bonds” to remove coins from circulation (more details below).
The major challenge of seigniorage shares is figuring out how to increase and decrease the monetary supply in a way that is both decentralized, resilient, and un-gameable. Expanding the money supply is easy: print money! Contracting the money supply, on the other hand, is not. Who loses money? Is it forced, or voluntary? If voluntary, what motivation does the person have to part ways with her stablecoin?
When the supply must contract, the system issues bonds with a par value of $1 that are sold at some discount to incentivize holders to remove stablecoins from circulation. Users purchase bonds (which may pay out at some future date) using stablecoins, thus removing some stablecoins from the supply. This creates a mechanism to decrease supply in the event that the price of the stablecoin falls below the target range. At some point in the future, if demand increases such that the system needs to increase the money supply, it first pays out bond holders (in the order that the bonds were purchased). If all of the bond holders have been paid out, then the software pays those who own shares (the equity token of the system). Shares represent a claim on future stablecoin distributions as demand increases. Shares can be thought of much like equity in that both shareholders and equity holders can value their asset as a function of expected dividends of holding the asset. Additionally, in most seigniorage shares implementations, shareholders are offered voting rights.
With the seigniorage shares model, supply never actually contracts with finality. Instead, each contraction involves the promise of a future increase in total supply. We’ve provided a basic overview of these mechanics, with some example estimates, at this link. Basecoin attempts to solve the contraction problem by allowing bonds to expire after five years. These instruments are not actually bonds– they are binary options with an indefinite payout date. This means that buyers will likely demand higher interest rates to account for this risk. One issue this creates is that a rapid decrease in demand can lead to a death spiral in the price of bonds. As the system begins printing new bonds in order to take stablecoins out of the supply, the bond queue becomes increasingly large. This increases the time to payout and decreases the likelihood that each bond is paid. As such, the newly printed bonds must be sold for a cheaper price in order to account for the additional risk. As bond prices fall, the number of stablecoins taken out of circulation for each bond sold also falls. This causes the system to have to print more bonds in order to shrink the supply sufficiently. This creates a recursive feedback loop that could make large-scale supply contraction near impossible unless other measures are put in place to prevent it. The Basecoin FAQ asserts that the system is immune to death spirals and explains their methods for preventing them, which include bond expiration and a bond price floor.
Some projects like Carbon modify the seigniorage shares model. In Carbon, users can elect to freeze portions of their funds to manage contraction and growth cycles. Some projects issue bonds, but simply pay out new stablecoins to all users, pro rata, when all bonds have been paid and supply must increase still. Each approach to the seigniorage shares model has its own set of challenges.
The seigniorage shares model is the most exciting, most experimental, and most “crypto-native” approach to creating a trustless decentralized stablecoin. There are many economists who believe it cannot work. Indeed, it’s fundamentally predicated on perpetual growth of the stablecoin system.
This fluctuating supply concept, while foreign at first, is rooted in a well known theory of economics: the Quantity Theory of Money. It’s also the method used by the Federal Reserve to maintain the stability of the US dollar. The crypto projects adopting the seigniorage shares model are attempting to do what the Federal Reserve does in a decentralized, algorithmic way.
All stablecoins must address the oracle problem. If stablecoins are pegged to the value of some external asset like the US dollar, the system needs some way to get data about the exchange rate between the stablecoin and the asset that it is pegged to. There are three fundamental approaches to this problem.
- Use a trusted data source (aka a trusted oracle).
- This re-centralizes trust in the system on the oracle.
- Data sources can be manipulated.
- Use a set of delegated data feeds and take the median.
- This is the approach used by BitShares. Users use stake-weighted voting to elect delegates to provide price feeds.
- The median of the price feed is used, meaning a majority of the delegates would have to collude to manipulate the price feed.
- The software can set limits on how much the price feed can move in certain time frames.
- Delegates can be voted out for providing faulty data.
- Use a schelling point scheme.
- Users who stake tokens are able to provide price inputs. Votes are weighted by the amount of tokens staked.
- The software sorts the values input by users. Users who provided an answer between the 25th and 75th percentile are rewarded, while users who submitted answers below the 25th percentile and above the 75th percentile are slashed (and their tokens redistributed to those who answered correctly).
- This approaches uses game theory to make the optimal input the one that most accurately reflects reality.
The final two challenges facing all stablecoins (most of which haven’t launched yet) are scalability and privacy. Global digital cash must be fast, cheap, and private. That can only occur if the platform it is built upon can scale. It must also be private, for both philosophical and practical reasons. A decentralized stablecoin could never serve as global, digital cash without some guarantee of privacy. While many people don’t immediately think that they care about privacy, businesses, governments, and financial institutions that transact in the stablecoin would certainly need privacy guarantees to protect their business interests, relationships, and more. A completely transparent ledger like that of Bitcoin is not usable for these purposes. Even though Bitcoin addresses are pseudonymous, simple chain analysis can link addresses with known entities with a fair degree of certainty. This traceability also destroys fungibility, an essential feature of digital cash.
One of the other challenges facing stablecoins is that they all are designed to be “pegged” to some underlying asset, usually USD. The problem is that people generally assume this to mean that the stablecoin is perfectly fungible for USD, when in fact it really means that the stablecoins are designed such that their value generally converges around the price of USD. Even stablecoins that are fully backed by and redeemable for collateral may not always trade at the peg, depending on market dynamics (accounting for counterparty risk). In order for stablecoins to succeed, users must view stablecoins not as fungible to the pegged asset, but as their own free-floating assets that very closely track the value of USD through a combination of redeemable collateral, market incentives, and future expectations. It is entirely possible that stablecoins could provide the desired stability without maintaining a perfect peg. In fact, once economies develop around the stablecoin itself, the peg will begin to matter less and less. If merchants are willing to hold and accept USD-pegged stablecoins, and they in turn pay their suppliers in the same stablecoin, and that stablecoin is widely used as a medium of exchange, then maintaining a perfect peg becomes increasingly less important.
Getting to that future state, however, requires a long process of bootstrapping such a network into existence and getting people to collectively believe that such a stablecoin is sound money. This process will be arduous, and will likely be even more difficult for seigniorage shares-based stablecoins that are not actually “backed” by anything.
While decentralized stablecoins are highly experimental, a successful implementation could be a major catalyst for fundamental long-term changes in the global economy. Lack of price stability prevents cryptocurrencies from displacing most forms of fiat money, and stablecoins can provide the solution. The decoupling of governments and money could provide an end to hyperinflationary policies, economic controls, and other damaging policies that result from government mismanagement of national economies.
Furthermore, stablecoins open up all sorts of possibilities for decentralized applications, especially those that require long-term lockups or escrow mechanisms. Decentralized insurance, prediction markets, savings accounts, decentralized exchange trading pairs, credit and debt markets, remittances, and more are all much more viable with the inclusion of a stablecoin.
While there are different approaches to creating a decentralized stablecoin, ultimately the market will decide which one will emerge the winner.