Stabilising stablecoins: What can we learn from the collapse of TerraUSD?

Co-authored by Jannah Patchay, Claire Conby, and Jonathan Marriott

In May 2022, TerraUSD (UST) – previously the third-largest stablecoin with a market cap of USD $18+ Billion – lost its peg to the US Dollar, falling almost 95 percent within a matter of days. In this article, we look at the circumstances surrounding UST’s spectacular collapse, and the effects of its contagion on the crypto markets. We then look ahead, to understand the tools that central banks and regulators can use to improve consumer protection and systemic stability, and to decrease the potential for market abuse, whilst ensuring that innovation in the still-nascent and evolving digital currency space is not unduly stifled. But first, let’s start with the basics  – what are stablecoins? How do they function, who uses them, and for what purposes are they used?

What are stablecoins?

Stablecoins are digital tokens that are ‘pegged’ to one or more reserve assets. We’ll go through the different ways in which stablecoins can be pegged in more depth; however, fiat currency-backed stablecoins are but one flavour. The two most popular stablecoins, at present, are Tether (USDT) and USD Coin (USDC), both of which are US dollar-denominated, and both of which claim to be backed by reserve assets (not limited to USD),  having a value that approximates the underlying USD notional value. USDT and USDC are mainly used by cryptocurrency investors and traders, both as a means of settlement for crypto transactions on-ledger, and as a means of holding value in a fiat currency without cashing out their positions on a crypto exchange. Stablecoins present key benefits to those operating within the cryptocurrency markets – and beyond – primarily:

• Managing volatility risk – The value of cryptocurrencies like Bitcoin and Ether fluctuates a lot — sometimes by the minute. The ability, in times of volatility, to hold value in an asset that’s pegged to a more stable currency can give buyers and sellers greater certainty that the value of their positions won’t rise or crash unpredictably in the near future.
• Sending or receiving funds and transmitting value – You don’t need a bank account to hold stablecoins, and they’re easy to transfer. Using stablecoins as a means of transmission, value can be sent easily around the globe, including to places where the U.S. dollar may be hard to obtain or where the local currency is unstable. Fast processing times and low transaction fees make some stablecoins a good choice for sending money anywhere in the world. People have sent as much as a million dollars worth of USDC with transfer fees of less than a dollar.

Four types of Stablecoins

The world of stablecoins is a new and diverse one, and one that is subject to constant experimentation and evolution. At present, stablecoins can be roughly grouped into four different types, or ‘flavours’ – based on characteristics such as the nature of the underlying asset or assets, the mechanism by which their value is calculated and ‘pegged’, and the type of reserves or collateral held by their issuer:

• Fiat-backed stablecoins – These derive their value from a peg to one or more underlying fiat currencies. For every coin in circulation, the issuer must hold the equivalent value in cash or other assets in reserve, in order to maintain its peg. Popular examples include Tether (USDT) and USD Coin (USDC), which maintain their peg to the US Dollar by holding cash, cash equivalents, treasury bonds, commercial paper, corporate bonds, or certificates of deposit with foreign banks. These single-currency Fiat-backed stablecoins are currently used as both a means of payment and a store of value, as we’ve seen. 
  • In some jurisdictions, such as the UK and the EU, a single-currency Fiat-backed stablecoin may already fall within the scope of existing regulation (for the UK / EU, this would be the E-Money regime), with multi-currency-backed coins projected to come into scope in the near future. In others, they do not yet fall within the scope of any regulatory regimes beyond AML / KYC requirements. 

• Commodity-backed stablecoins – Less common than other forms of stablecoins, these are backed by commodities such as oil, gold, and other precious metals. The value of the stablecoin will rise and fall, in fiat terms, depending upon how the underlying asset performs in the markets. These stablecoins can therefore be seen as an investment vehicle rather than a means of minimising volatility.
  • In some jurisdictions, and depending on the way in which these are structured, Commodity-backed stablecoins are currently regulated in the same way as securities or other financial instruments. 

• Crypto-backed stablecoins – Each coin is pegged either to the value of a single cryptocurrency, such as Bitcoin with Wrapped Bitcoin (wBTC), or pegged to a fiat currency but collateralised with cryptocurrency. With respect to the latter, the value of the cryptocurrency collateral is typically required to be higher than that of the pegged asset, in order to manage volatility. An example is DAI, which is soft-pegged to USD and collateralised by a range of Ethereum-based cryptoassets. These stablecoins are used for much the same reasons as Fiat-backed stablecoins: hedging, speculation, payments and as a store of value. 
  • At present, Crypto-backed stablecoins do not fall into the regulatory perimeter in any jurisdiction.
• Algorithmic stablecoins – Algorithmic stablecoins, unlike other types of stablecoins, are not backed by assets at all. Instead, their value is controlled by an algorithm that manages the amount of the coin in circulation such that its value remains pegged to an underlying asset. The most popular, and now notorious, algorithmic stablecoin is Terra USD (UST), which is pegged to USD through an algorithmic relationship with the cryptocurrency Terra (LUNA). When the price of UST loses its peg to USD, traders are incentivised to swap between UST and LUNA, burning one for the other (essentially arbitraging the slip in the peg), which either increases or decreases the supply of UST in the market – and thus its value, maintaining the USD peg over time. 
  • At present, Algorithmic stablecoins are also unregulated. 

The risks and pitfalls of stablecoins

Questions have long been asked – not only by regulators, but also by market participants, observers, and investors – about the stability of various stablecoins, and the extent to which they are adequately backed by reserves or collateralised. When a stablecoin is marketed as being pegged to an underlying currency, asset, or basket thereof, it’s important for users to have confidence in how this peg is maintained, and in the issuer’s ability to intervene and to manage and ensure the stability of the stablecoin. 

In periods of market volatility – with respect to either the underlying asset(s) or the reserves / collateral themselves – or of deliberate market abuse, the robustness of the pegging methodology may be compromised. This has a corresponding impact on the value of the stablecoin, and on the consumers or investors who are using it to make payments, to hedge risk or to store value. In essence – the value of the stablecoin is no longer stable, as intended. The extent to which the peg can be maintained, and how it is maintained – through reserves or collateralisation, for example, or through an algorithm – is integral to the stability of a stablecoin, as we’ll see in a closer look at the recent events surrounding the demise of TerraUSD.

May 2022: TerraUSD (UST) drops to $0.05

How did what was only recently the third-largest stablecoin by market cap fall by around 95 percent to $0.05 within a matter of days? Let’s begin to unpack this question by taking a closer look at how Terra maintains the peg between UST and USD. 

In contrast to other popular stablecoins such as USDT and USDC, which rely on their respective issuers holding a reserve of assets as a means of maintaining their peg to USD, UST’s value is maintained by a smart contract-based algorithm. The Terra ecosystem is designed such that LUNA acts as a means of maintaining the UST peg to USD. It’s a complex system, but in short, traders are guaranteed the ability to swap LUNA for UST at a price of 1USD, at all times. This creates arbitrage opportunities; when the market price of UST rises above 1USD, then holders of LUNA can swap 1USD of LUNA for 1UST, thus making a profit (as 1UST is now worth more than 1USD, and therefore more than the LUNA swapped for it). During this process, excess LUNA are removed from circulation, thus driving up the value of LUNA, and more UST are minted, lowering the value of UST in circulation. This process brings the value of both UST and LUNA back into alignment with USD. When the value of UST falls against its USD peg, arbitrage opportunities are created between UST and LUNA, and traders are incentivised to burn UST in exchange for LUNA. Doing so reduces or increases the circulation of UST, thus maintaining its value relative to USD. 

Algorithmic stablecoins such as UST one have been criticised as being too risky, as they rely on traders to maintain the peg to USD, as opposed to maintaining reserves of collateral or assets that continuously support the peg. If traders lose confidence in the system, and aren’t willing to play their part in maintaining the peg by buying or selling LUNA / UST, then the coins can potentially go into a “death spiral”. And this is precisely what occurred in May 2022. The Anchor Protocol is a savings protocol based on the Terra blockchain that offers high yields to traders depositing UST. A series of large withdrawals from the Anchor Protocolled to significant selling pressure on the UST coin, forcing it to lose its USD peg faster than the algorithm could manage to stabilise it. Further compounding this issue was a rapid fall in value of the LUNA token. In an attempt to restore UST’s peg to USD, Luna’s issuer (Luna Foundation Guard) used its reserves to buy up vast amounts of UST in an effort to increase demand on the market and thus increase the price of UST, bringing it closer to its peg. However, the effect was to flood the market with LUNA, resulting in a spectacular price crash. 

The root cause of the crisis – namely, the reasons behind the large withdrawals of UST from the Anchor Protocol in the first instance – remain unclear. Some analysts believe it was a loss of confidence in either the protocol or the wider cryptocurrency market, whilst others believe it was a coordinated attack, a stunning example of market manipulation. Whatever the cause, it begs the question: what is the value in algorithmically-backed stablecoins if the peg can be broken – with such apparent ease – by either market volatility or abuse? 

Some argue that the overall benefit of algorithmically-backed stablecoins lies in their decentralised nature. Without the involvement of regulatory bodies, and without the need for an issuer to maintain reserves or collateral, it is the algorithm’s code that is responsible for monitoring the stablecoin’s price performance against its target value (e.g. 1USD), and to manage supply and demand in such a way that the peg is maintained. However, in the case of UST, the code was not able to function as it was intended, requiring two organisations – Luna Foundation Guard and Terraform Labs – to step in and help alleviate the sharp sell-off, which also called into question the extent to which the stablecoin could truly be regarded as “decentralised”.  

It’s not just the algorithms…

UST’s spectacular implosion, whilst certainly the most recent and most novel in nature, is by no means unique to the woes of existing and would-be stablecoin issuers. In October 2015, the US Commodities Futures Trading Commission (CFTC) fined Tether, the issuer of USDT, 41 million US dollars for misstating its reserves. Tether had consistently claimed that it held cash reserves equivalent to the total USDT in circulation, thus implying that USDT could always be redeemed for the equivalent underlying USD. The CFTC’s investigation found that, between 2016 and 2018, Tether held only 27.6% of the value of its issued USDT in fiat currency reserves. For the  remainder, the CFTC observed, Tether relied on “unregulated entities and certain third-parties to hold funds comprising the reserves”. 

Tether claims that it has since improved its reserve management capabilities – and indeed, during the recent cryptocurrency market volatility following the UST crash, investors were able to withdraw over 10 billion USD, with the value of USDT dropping at times as low as 0.96USD before recovering. Whilst certainly an improvement in performance and resilience over that of UST, this is still by no means what one would expect in terms of ‘stability’ from a stablecoin. 

In June 2019, Facebook (now Meta) announced plans to launch its own digital wallet and stablecoin. The project was called Libra (subsequently Diem). The company claimed that Libra would enable it to reduce the barriers to entry for financial inclusion, whilst introducing fast, low-cost global payments to billions of Facebook’s monthly active users. The stablecoin was to be pegged to the value of a basket of fiat currencies including the US dollar, Pound Sterling, Euro, Yen, and Singapore dollar, and short-term assets (generally considered to be cash equivalents). However, following significant challenges from regulators, the project was abandoned on 31 January 2022. Let’s take a closer look at why this happened.

Unfortunately for the teams at Facebook, trust and privacy aren’t synonymous with the company’s past. In an attempt to bring trust to the Libra project, Facebook launched a not-for-profit organisation called the Libra Association, which was put in place to manage the day-to-day build and operations of the stablecoin project. Facebook invited a slew of trusted, globally-renowned brands to help launch the association, which included the likes of PayPal, eBay, Mastercard, Stripe, Visa, and Mercado Pago. However, this was not enough to stave off the fears of regulators, which included concerns around monopolisation considering Facebook is used by half of the world’s internet users.

Furthermore, the prospect of a global payments network built on a privately issued, unregulated stablecoin, tapping into a base of 2.3 billion users (at the time), did not exactly fill central banks, regulators and governments with joy. Loss of oversight over a significant volume of payments, and the potential for financial crime and tax evasion, as well as a potential loss of monetary sovereignty, were all factors wedged against Libra. It was effectively doomed from the start, with multiple jurisdictions expressing concerns from the moment it was announced, and a general lack of confidence in the company’s ability to address issues such as money laundering, consumer protection and other potential financial risks. Even its subsequent rebranding as Diem, and restructuring as a fiat-backed USD-pegged stablecoin, couldn’t overcome these obstacles. 

Post-Global Financial Crisis, in particular, it is clear that governments, central banks and regulators are increasingly concerned about the potential for commercial entities to create financial products having the potential to destabilise existing global financial structure. In the case of Facebook specifically, how could a company connected with scandals such as Cambridge Analytica be trusted to manage a global payments system? More broadly, how can issuers be trusted to maintain adequate reserves, to prevent and detect market abuse, and to maintain high standards of consumer protection? How can we, as users, trust that our financial data isn’t being sold to other commercial parties, and that the stablecoins we might use will always maintain the ‘stable’ value that we expect from them? 

How do we make stablecoins ‘stable’?

Having explored the specific issues associated with a few stablecoin projects, let’s look ahead to the future of stablecoin regulation, particularly in the UK. We’re talking here about examples such as UST, and USDT, and Libra – all of which could ultimately be regarded as ‘systemically important stablecoins’, those which are deemed to be systemically important to the financial system. Some of those stablecoins, specifically those that are both GBP-pegged and systemically important, will in future constitute the “private digital Pound”, much in the same way that commercial bank money today constitutes “private money” in the UK. 

We recognise that other stablecoins will continue to exist, and new ones will be launched. Not all will be pegged to a single fiat currency, and we may still see some fairly exotic structures emerge. (While regulators may determine what is systemically important, it’s ultimately the degree of uptake by users that makes something systemically important). However, we don’t anticipate that all stablecoins will be systemically important, or that all will have the same risk profile, or require the same degree of regulatory oversight. A stablecoin pegged to the IMF’s Special Drawing Rights (SDR), for example, that is fully asset-backed and used in a closed system of industry participants to transfer value from one to another, may not have the same risk profile as a future Tether-issued GBPT that is used by the general public at scale (this does not exist at present, but it could). There will be other stablecoins that are primarily used as a store of value – commodity-backed stablecoins for example – rather than a means of payment, and the correct approach would be to regulate them in line with similar financial structures today (perhaps subject to certain limitations on their usage as a means of payment). This is typically known as the “same risk, same regulation” approach. 

Regulating to solve for different risks 

Different types of regulation address different risks. 

• Prudential regulation addresses the solvency and liquidity of financial institutions; considered in light of stablecoin issuers, this translates to the type and amount of assets that must be held in reserve by an issuer, and how and where they must be held. Typically, reserve and liquidity requirements are proportionate to the degree of regulatory oversight to which an institution is subject. Banks may have more flexibility with reserves (they can lend more than they hold in deposits and reserves, for example) but they are also subject to an expensive and extensive degree of regulatory scrutiny. In the UK and EU, e-money issuers, on the other hand, are more lightly regulated, but are required to hold 100% of the value of their e-money in circulation on deposit with a bank, giving them less flexibility to engage in higher-risk, leveraged behaviours. 

• Conduct regulation includes requirements around governance, organisational structure and location, remuneration of employees, appropriate treatment of different types of customers, and disclosures about products and services. These requirements are aimed at ensuring that a firm is properly managed, and that it is transparent in its provision of pricing and services. For a stablecoin issuer, conduct requirements could include the requirement to disclose conflicts of interest, transparency around governance arrangements and periodic disclosures of reserves and collateral. 

• Market abuse requirements typically fall under the umbrella of conduct requirements in most jurisdictions. Market participants must implement measures for the prevention and detection of market abuse and market manipulation, by either unscrupulous actors within their own firm, or by third parties. These measures could include transaction reporting, trade and market surveillance, and investigation / escalation of suspicious market activity. In cases where algorithms play a large role – such as algorithmic trading, for example – these must be subject to stringent testing, and have “kill switches” enabling the market to be halted, where there is a risk of a “flash crash” event. 

For systemically important stablecoins, it’s even more important to ensure that the risk of market manipulation is managed with respect to both the coin itself, and the underlying reserve or collateral assets. Even in the case of UST, having provisions in place for detection of market abuse, and having a kill switch for issuance of LUNA, might have enabled more swift action to be taken.

• Safeguarding of client funds – It is important to distinguish between the Financial Services Compensation Scheme (FSCS) that protects individual consumers and eligible organisations from the failure of a financial institution (such as a UK authorised bank) when they are unable to return the user’s funds. FSCS provides compensation up to £85,000 per eligible person, per bank, building society or credit union or up to £170,000 for joint accounts. This is a service funded by the financial services who are covered by it. In theory,  a similar compensation scheme could be created for a systemically important stablecoin; however the downside is that the level of compensation available can be limited and may not cover the full amount owed.

At present, authorised e-money institutions must have appropriate and well-managed safeguarding arrangements so that, if a firm becomes insolvent, customers’ funds are returned in a timely and orderly way. This is achieved through holding client funds in a safeguarded account and kept separate from the issuer’s own money; the funds are designated as belonging to the client and that the e-money issuer has no right over in the event of liquidation. Using the safeguarding method can therefore be considered preferable as all the funds held in an e-money account are safeguarded and the full value (minus administrative costs applied by the insolvency practitioner) will be returned. It may take longer compared to an FSCS claim but the value of funds protected is not limited in the same way as under the FSCS.

Some stablecoins may be structured in such a way that they are backed by a combination of cash-like funds – including client deposits – and “high-quality liquid assets” (HQLAs). In this case the requirements around the HQLAs and liquidity buffers required would likely become more similar to those for banks, so that the issuer is able to ensure that they are able to meet redemptions during a “run” (such as that recently experienced by Tether’s USDT) up to a certain level of expected demand – and potentially necessitating the implementation of an FCSC-like structure to cover higher client losses up to a certain level. In this scenario, all eligible stablecoin issuers would be required to pay into the scheme. 

• Financial crime regulation usually follows global standards set by the Financial Action Task Force (FATF), and involve the implementation of anti-money laundering (AML) / counter-terrorism financing (CTF) measures, such as Know Your Customer (KYC), ensuring that sanctions are adhered to, and procedures for raising and managing suspicious activity. In general, this is the area, globally, which has been applied earliest to participants in the cryptoasset space. 

• Non-financial regulatory requirements include privacy, data protection, information security, and cybersecurity.  Whilst a deceptively short list, these are some of the most fundamental aspects of trust in a digital currency – the extent to which users’ data, and their privacy, are protected, and the extent to which they can be assured that their private data will not be exploited for commercial gain without their informed consent. 

The importance of being proportionate

Fundamentally, a stablecoin must be able to hold its value such that users can be confident that they are holding a stable method of payment that will not fluctuate in value. This is where the challenge with some stablecoin structures lies. For those that are in essence a tradable asset that can be used for payments, there is the potential for the value of these stablecoins to fluctuate as a result of their tradable features. And this means that it does not behave entirely like a currency, but potentially more as a commodity that is used as a means of payment. 

It is also important to distinguish between user types, and between different types of user behaviour. Client classification is a well-established component of EU and UK financial regulation, with clients broadly being classified into Retail or Professional (and Eligible Counterparty) categories, and subject to appropriate protections based on their sophistication. User behaviour can be split across those who are holding stablecoins as a store of value or investment asset and are primarily interested in their tradability (i.e. more speculative behaviour), and those who are more risk averse, and see the benefits of digital currency for payments and settlements, but do not want to expose themselves to price fluctuations.

In the current absence of a regulatory regime for stablecoins (which is soon to change in the EU with the introduction of the MiCA regulations, and in the UK, to some extent, with the broadening of the current e-money and payments regimes to encompass certain types of fiat-backed stablecoins), most stablecoin issuers (with the exception of those that are currently subject to the e-money regime) are not currently subject to any prudential or conduct regulatory requirements. So, whilst they can purport to offer a “stable” coin, there is actually no requirement for them to demonstrate or maintain its stability.  And even with the best intentions, the case studies mentioned above show that the current regulatory infrastructure is not as robust as it needs to be. 

For systemically important stablecoins, in particular, to be truly “stable”, there are some clear risk mitigating steps that can be taken, including lessons that can be learned from the e-money regime. Again, the proportionality of the stablecoin’s usage, and the risk it poses, should always be taken into consideration when determining how this risk can be managed through regulation. Rather than being prescriptive about the types of stablecoins that “should” be allowed, or how all stablecoins should be regulated, it will be important for regulators to consider the types of risks that can arise based on different stablecoin structures – including algorithmic stablecoins – and how these might be addressed in a proportionate manner.

This article was prepared by the authors. The views expressed in this article are the author’s own and do not necessarily reflect the views of the Digital Euro Association.