Research on the Dynamic Model of Stablecoins and the Shadow Banking of Cryptocurrencies: What is the Trap of Instability?

Authors: Ye Li, Simon Mayer, CES ifo Working Papers

Original Title: "Money Creation in Decentralized Finance: A Dynamic Model of Stablecoins and Cryptocurrency Shadow Banking"

Compiled by: Jing Miao, Institute of Financial Technology, Renmin University of China

The rise of stablecoins aims to meet the demand for safe assets in decentralized finance, where stablecoin issuers deploy various strategies to convert risky reserve assets into value-stable tokens. To address issues such as the feasibility of stablecoins, regulations, and initiatives dominated by large platforms, we establish a dynamic model for optimal stablecoin management and describe an unstable trap. The system is binary: stability can persist for a long time, but once a stablecoin breaks its peg under negative shocks, volatility will persist. Since it allows for effective risk-sharing between issuers and stablecoin users, devaluation will trigger an inevitable vicious cycle. The core part of the research has been compiled by the Institute of Financial Technology, Renmin University of China.

1. Introduction

Over a decade ago, Bitcoin heralded a new era of digital payments and decentralized finance. Cryptocurrencies challenge bank-centered payment systems by offering round-the-clock settlement, anonymity, low-cost international remittances, and programmable money through smart contracts. However, the significant volatility exhibited by first-generation cryptocurrencies limits their utility as a means of payment. The purpose of stablecoins is to maintain a stable price relative to a reference currency or a basket of currencies by committing to hold fiat or other assets, while the assets held by stablecoins can be redeemed.

This paper provides a continuous-time dynamic model of stablecoins that rationalizes a rich set of strategies commonly observed in practice, such as open market operations, dynamic collateral requirements for users, transaction fees or subsidies, target price bands, re-pegging, and issuing "secondary units" (or governance tokens) as equity for stablecoin issuers. Our model demonstrates a new instability mechanism and is suitable for evaluating regulatory proposals. It can also be applied to analyze the incentive mechanisms behind stablecoin initiatives led by large platforms (such as Facebook).

The creation of stablecoins is akin to the money creation of shadow banking, which involves the unregulated creation of safe assets to meet the trading needs of agents. Issuers convert risky assets, namely cryptocurrencies in practice, into stablecoins on a one-to-one basis, which can be redeemed for fiat currency on a one-to-one basis. Issuers typically require stablecoin users to provide collateral and stipulate margin requirements. If the value of the user's collateral drops sharply, the issuer will use its reserves (invested in risky assets) to support the value of the stablecoin.

However, the creation of stablecoins differs from traditional shadow banking in a key aspect. Issuers can devalue stablecoins to avoid liquidation, while shadow banks must fulfill debt contracts or face bankruptcy. The devaluation option allows stablecoin issuers to avoid costly liquidation but induces an amplification mechanism that generates a bimodal distribution state. In a high-reserve state, issuers maintain a fixed exchange rate, leading to strong demand for stablecoins and high trading volumes. Through open market operations and transaction fees, issuers generate income, further increasing reserves. In a low-reserve state, issuers provide risk exposure to users through devaluation, which suppresses demand for stablecoins, thereby reducing issuers' income, so issuers can only slowly rebuild their reserves, falling into an unstable trap. Thus, fixed exchange rates can persist for a long time, but once devaluation occurs, recovery is slow.

Our paper discusses key issues regarding the credibility and sustainability of fixed exchange rates. Unlike the speculative attacks and coordination failures on under-collateralized stablecoins described in the studies by Routledge and Zetlin Jones, we describe a new risk amplification mechanism and show that, consistent with evidence, even over-collateralized stablecoins can fail when the issuer's reserves drop below a certain level. When the issuer's reserves fall below a critical threshold, even over-collateralized stablecoins can fail. The reason for devaluation is the trade-off between the issuer's efforts to maintain stable value to stimulate demand and sharing risks with users to avoid liquidation. We focus on fully collateralized stablecoins because recent regulatory proposals and bank run events have raised doubts about the feasibility of under-collateralized stablecoins.

The reserve management of stablecoin issuers resembles dynamic corporate cash management, but unlike corporations, stablecoin issuers can devalue their liabilities (unredeemed stablecoins). This is similar to a country monetizing its debt through inflation. Stablecoins share similarities with contingent convertible bonds (CoCos), which automatically share risks between equity investors and debt holders. However, unlike CoCos, risk-sharing is accomplished through devaluation rather than converting stablecoins into the issuer's equity, and devaluation is determined by the issuer without predetermined trigger events.

After tech giant Facebook and its partners announced their own stablecoin Libra (now "Diem") in June 2019, stablecoins became the subject of intense debate. Leveraging their existing customer networks, global tech or financial companies can rapidly expand the influence of their stablecoins. To understand the advantages of a well-developed network in the stablecoin space, we compare platforms with varying degrees of network effects. Stronger network effects are indeed associated with lower devaluation frequencies, as stronger network effects allow platforms to accumulate fee income more quickly when reserves are depleted, and through higher continuation values, incentivize platforms to maintain larger equity levels to buffer risks.

Stablecoin initiatives sponsored by companies with global customer networks have attracted the attention of regulators, as they not only have the potential for widespread adoption but also raise concerns about monopolistic power. In our model, two countervailing forces determine the share of economic surplus captured by the platform. Under stronger network effects, the platform can extract more rents from users through fees or risk-sharing, but it is also more eager to stimulate token demand by lowering fees and stabilizing tokens, as individual users do not internalize the positive network externalities. In equilibrium, these two forces balance each other. Therefore, the welfare distribution between stablecoin issuers and users is relatively insensitive to the degree of network effects.

The massive transaction data generated by payment systems provides a strong incentive for digital platforms to develop their own stablecoins. Following Parlour (2020), we model data as a byproduct of transactions. With constant velocity of money, the value of stablecoins held by users determines the speed of transaction volume and data accumulation. Data can help platforms improve their productivity, thereby enhancing the flow utility derived by users from their token holdings. A feedback loop emerges: transactions generate more data, which improves the platform and leads to stronger token demand and more transactions. Thus, data grows exponentially over time. The platform balances data acquisition and reserve accumulation.

2. Cryptocurrency Shadow Banking in Decentralized Finance

Blockchain technology supports peer-to-peer transfers of assets on distributed ledgers, potentially eliminating the need for intermediaries in transactions. Decentralization avoids significant intermediary costs. According to blockchain protocols, decentralization can enhance operational resilience by eliminating single points of failure while also achieving scalability. Decentralized finance ("DeFi") provides alternatives to traditional financial services based on blockchain, such as banking, brokerage, and exchanges. It is also built on the foundation of smart contracts (executed through the coding of programmable money), a concept independent of blockchain.

This emerging financial architecture requires blockchain-based money. A viable means of payment should maintain stable value at least during the settlement period (i.e., the time required to achieve decentralized consensus on transactions). However, most cryptocurrencies are highly volatile. They are platform-specific currencies whose values fluctuate without backing, driven by the supply and demand dynamics of the native hosting platform.

Stablecoins are marketed as blockchain-based fiat currency replicas. The total market value exceeds $100 billion. In May 2021 alone, stablecoins worth $76.6 billion were transferred. Issuers can be corporate entities or consortia, such as the consortium led by Facebook or the developers of Diem. It can also be an internet protocol (e.g., MakerDAO, the issuer of Dai), whose rules can be updated based on user consensus. Stablecoins are backed by a combination of the issuer's reserve assets, with value stability maintained through open market operations by the issuer, specifically targeting the trading reserves of stablecoins and meeting redemption requests. The distributed ledger records the ownership and transfer of stablecoins, but verifying reserves still relies on traditional audits.

Stablecoins serve as a link between decentralized finance and the real economy. The volatility of first-generation cryptocurrencies, such as Bitcoin and Ethereum, limits their application in the real world. Stablecoins are designed to have stable exchange rates relative to reference fiat currencies, potentially mediating transactions of blockchain-based goods, services, and real assets. Stablecoins are also significant for the cryptocurrency community. Traders' activities heavily involve rebalancing between stablecoins and more volatile cryptocurrencies. Cryptocurrencies have emerged as a new asset class, with a total market capitalization of approximately $1.5 trillion (with Bitcoin accounting for about $700 billion). It is estimated that 50% to 60% of Bitcoin trading volume involves the cryptocurrency shadow banking of the stablecoin USDT issued by Tether. This illustrates two structures of stablecoins.

Figure 1

As shown in Panel A of Figure 1, a platform issues stablecoins backed by its reserves. The excess reserves, i.e., equity positions, belong to governance token holders who have control (i.e., control over platform policies). When reserves are invested in risky assets, potential losses are absorbed by equity positions. As long as stablecoins are over-collateralized, their value remains unaffected. In Panel B, stablecoins are supported by both user collateral and the platform's reserves. When the value of the collateral declines and users cannot meet margin requirements, the platform liquidates the collateral and uses the proceeds (and its own reserves) to buy back stablecoins in the secondary market.

Despite the importance of stablecoins, there is no clear legal and regulatory framework. Unlike deposit-taking institutions, stablecoin issuers have no obligation to maintain redemption at face value. Given that reserves are often invested in risky assets, many people worry that a major stablecoin could "break the buck," triggering financial turmoil. This concern was sparked by the relatively recent event during the 2007-2008 global financial crisis when money market funds shifted from stable asset net values to floating asset net values. In fact, the creation of stablecoins is essentially a new form of shadow banking, namely the unregulated transformation of safety.

Reserve assets are often risky, and Panel A of Figure 1 illustrates stablecoin creation characterized by over-collateralization. The issuer's excess reserves (equity) buffer the volatility of reserve value. Equity is referred to as governance tokens (or "secondary units"), which carry voting rights over protocol changes (i.e., control) and pay cash flows generated from transaction fees charged to stablecoin users. Governance tokens can be issued to supplement reserves, just as traditional companies can raise cash by issuing equity.

A stablecoin issuer is essentially making a leveraged bet on the value of reserve assets. Issuers can increase their leverage by issuing new stablecoins to fund the purchase of reserve assets, just as banks fund their loans and securities trading with newly issued deposits (i.e., internal money creation). Unlike banks, which commit to redeem deposits at face value, stablecoin issuers can devalue stablecoins.

Panel B of Figure 1 shows a more complex structure, similar to the one adopted by MakerDAO. Users provide their held cryptocurrencies and other assets as collateral for newly created stablecoins but must meet devaluation (margin) requirements. Users can transfer stablecoins and circulate them in the market but must maintain margin requirements. If the value of the collateral declines and the user cannot maintain the margin, they will forfeit the collateral to the stablecoin issuer, who will liquidate the collateral and use the proceeds to buy back (and handle) the stablecoins created for that user. If the liquidation of the collateral does not generate sufficient proceeds, the stablecoin issuer's reserves will be used to cover the costs of repurchasing stablecoins.

The structure in Panel B of Figure 1 is common in traditional shadow banking. A bank establishes a channel (special purpose vehicle) to convert risky investments into debt and equity while simultaneously providing guarantees to debt investors, so that when the channel experiences losses, the bank internalizes the losses. Stablecoins are akin to the debt (priority) portion of the channel, with the stablecoin issuer and users in Panel B corresponding to the bank and users, respectively. The stablecoin issuer commits to buying back stablecoins with its own reserves, similar to a bank's guarantee.

In traditional shadow banking terms, the distinction between the two structures in Figure 1 is that from Panel A to Panel B, the stablecoin issuer transfers risk to an off-balance-sheet entity (i.e., users), so that stablecoins are simultaneously supported by user collateral and the issuer's reserves. Compared to the simple structure in Panel A of Figure 1, this dual collateralization does not necessarily enhance stability, as both the user's collateral and the issuer's reserve assets may be risky and highly correlated in value, especially when both are cryptocurrencies.

3. Conclusion

First-generation cryptocurrencies, such as Bitcoin and Ethereum, were established to serve as mediums of exchange, but price volatility undermines this function. With the rapid development of decentralized finance, various stablecoin initiatives have emerged to meet the demand for stable payment means based on blockchain. Stablecoins are issued by private entities that commit to maintaining price stability by holding collateral assets, redeeming assets held in stablecoins. However, since these issuers maximize their own returns rather than overall welfare, conflicts of interest between issuers and stablecoin users naturally arise, creating space for regulatory interventions to enhance welfare. Additionally, mature network companies (such as Facebook) plan to launch their own stablecoins. Behind these initiatives, the incentive mechanisms are more complex, especially considering that operating a payment system allows platforms to collect transaction data and profit from it.

Despite significant attention from regulators and practitioners, there is currently no unified framework to address these issues. In this paper, we fill this gap and establish a dynamic model for optimal stablecoin management. The model achieves equilibrium characterized by operating two regimes: when the issuer's reserves are sufficiently high, the price of stablecoins is fixed; when reserves fall below a critical value, the price of stablecoins fluctuates with the issuer's reserves, allowing the issuer to share risks with stablecoin users and avoid costly liquidation.

The distribution of states is bimodal; above the devaluation threshold, issuers can credibly maintain a fixed token price, which induces strong token demand, enabling issuers to collect fee income and further increase reserve holdings. When negative shocks deplete the issuer's reserves below the devaluation threshold, this virtuous cycle turns into a vicious cycle. As the token price becomes unstable, user demand for tokens declines, leading to a decrease in the issuer's fee income and revenue from selling tokens in the open market, which then slows the rebuilding of reserves, creating an unstable trap. This vicious cycle can be broken by issuing equity (governance tokens) to supplement reserves. Equity issuance must simultaneously expand the token supply to eliminate arbitrage.

Our model provides a framework for evaluating regulatory proposals. We show that capital requirements could potentially improve overall welfare, but imposing legally binding commitments for perfect price stability would undermine welfare. Our framework can also be applied to analyze the incentives behind stablecoin initiatives led by mature platform companies. We demonstrate that strong network effects can indeed lead to stable token values, making these platform companies natural issuers of stablecoins. Furthermore, the launch of stablecoins can stimulate transactions, and transaction data can be used to enhance the platform's profitability. However, the increase in data productivity can undermine the stability of stablecoins, as platforms become more eager to stimulate transactions, issuing more stablecoins per unit of reserve.